Skip to contentSkip to conversation

arm_mpam: Add basic mpam driver

RFC PATCH In review James Morse 10 participants 117 messages Jul 11, 2025 – Aug 15, 2025 Page 1 of 2

Subsystems (8)
lore
Share & tools
lore

James Morse

author maintainer
#0 lore raw

cover arm_mpam: Add basic mpam driver

Hello,

This is just enough MPAM driver for the ACPI and DT pre-requisites. It doesn't contain any of the resctrl code, meaning you can't actually drive it from user-space yet.

This is the initial group of patches that allows the resctrl code to be built on top. Including that will increase the number of trees that may need to coordinate, so breaking it up make sense.

The locking looks very strange - but is influenced by the 'mpam-fb' firmware interface specification that is still alpha. That thing needs to wait for an interrupt after every system register write, which significantly impacts the driver. Some features just won't work, e.g. reading the monitor registers via perf.
The aim is to not have to make invasive changes to the locking to support the firmware interface, hence it looks strange from day-1.

I've not found a platform that can test all the behaviours around the monitors, so this is where I'd expect the most bugs.

It's unclear where in the tree this should be put. It affects memory bandwidth and cache allocation, but doesn't (yet) interact with perf. The main interaction is with resctrl in fs/resctrl - but there will be no filesystem code in here. Its also likely there will be other in-kernel users. (in-kernel MSC emulation by KVM being an obvious example).
(I'm not a fan of drivers/resctrl or drivers/mpam - its not the sort of thing
that justifies being a 'subsystem'.)

For now, I've put this under drivers/platform/arm64. Other ideas welcome.

The first three patches are currently a series on the list, the PPTT stuff has previously been posted - this is where the users of those helpers appear.

The MPAM spec that describes all the system and MMIO registers can be found here:
https://developer.arm.com/documentation/ddi0598/db/?lang=en
(Ignored the 'RETIRED' warning - that is just arm moving the documentation
around. This document has the best overview)

This series is based on v6.16-rc4, and can be retrieved from:
https://git.kernel.org/pub/scm/linux/kernel/git/morse/linux.git mpam/driver/rfc

The rest of the driver can be found here:
https://git.kernel.org/pub/scm/linux/kernel/git/morse/linux.git mpam/snapshot/v6.16-rc4

What is MPAM? Set your time-machine to 2020:
https://lore.kernel.org/lkml/[email protected]/

Bugs welcome,
Thanks,

James Morse (31):
cacheinfo: Add arch hook to compress CPU h/w id into 32 bits for
cache-id
arm64: cacheinfo: Provide helper to compress MPIDR value into u32
cacheinfo: Expose the code to generate a cache-id from a device_node ACPI / PPTT: Add a helper to fill a cpumask from a processor container ACPI / PPTT: Stop acpi_count_levels() expecting callers to clear
levels
ACPI / PPTT: Find cache level by cache-id
ACPI / PPTT: Add a helper to fill a cpumask from a cache_id
arm64: kconfig: Add Kconfig entry for MPAM
ACPI / MPAM: Parse the MPAM table
platform: arm64: Move ec devices to an ec subdirectory
arm_mpam: Add probe/remove for mpam msc driver and kbuild boiler plate
arm_mpam: Add the class and component structures for ris firmware
described
arm_mpam: Add MPAM MSC register layout definitions
arm_mpam: Add cpuhp callbacks to probe MSC hardware
arm_mpam: Probe MSCs to find the supported partid/pmg values
arm_mpam: Add helpers for managing the locking around the mon_sel
registers
arm_mpam: Probe the hardware features resctrl supports
arm_mpam: Merge supported features during mpam_enable() into
mpam_class
arm_mpam: Reset MSC controls from cpu hp callbacks
arm_mpam: Add a helper to touch an MSC from any CPU
arm_mpam: Extend reset logic to allow devices to be reset any time
arm_mpam: Register and enable IRQs
arm_mpam: Use a static key to indicate when mpam is enabled
arm_mpam: Allow configuration to be applied and restored during cpu
online
arm_mpam: Probe and reset the rest of the features
arm_mpam: Add helpers to allocate monitors
arm_mpam: Add mpam_msmon_read() to read monitor value
arm_mpam: Track bandwidth counter state for overflow and power
management
arm_mpam: Add helper to reset saved mbwu state
arm_mpam: Add kunit test for bitmap reset
arm_mpam: Add kunit tests for props_mismatch()

Rob Herring (2):
cacheinfo: Set cache 'id' based on DT data
dt-bindings: arm: Add MPAM MSC binding

Rohit Mathew (2):
arm_mpam: Probe for long/lwd mbwu counters
arm_mpam: Use long MBWU counters if supported

Shanker Donthineni (1):
arm_mpam: Add support for memory controller MSC on DT platforms

 .../devicetree/bindings/arm/arm,mpam-msc.yaml |  227 ++
 MAINTAINERS                                   |    6 +-
 arch/arm64/Kconfig                            |   19 +
 arch/arm64/include/asm/cache.h                |   17 +
 drivers/acpi/arm64/Kconfig                    |    3 +
 drivers/acpi/arm64/Makefile                   |    1 +
 drivers/acpi/arm64/mpam.c                     |  365 +++
 drivers/acpi/pptt.c                           |  240 +-
 drivers/acpi/tables.c                         |    2 +-
 drivers/base/cacheinfo.c                      |   57 +
 drivers/platform/arm64/Kconfig                |   73 +-
 drivers/platform/arm64/Makefile               |   10 +-
 drivers/platform/arm64/ec/Kconfig             |   73 +
 drivers/platform/arm64/ec/Makefile            |   10 +
 .../platform/arm64/{ => ec}/acer-aspire1-ec.c |    0
 .../arm64/{ => ec}/huawei-gaokun-ec.c         |    0
 .../arm64/{ => ec}/lenovo-yoga-c630.c         |    0
 drivers/platform/arm64/mpam/Kconfig           |   23 +
 drivers/platform/arm64/mpam/Makefile          |    4 +
 drivers/platform/arm64/mpam/mpam_devices.c    | 2910 +++++++++++++++++
 drivers/platform/arm64/mpam/mpam_internal.h   |  697 ++++
 .../platform/arm64/mpam/test_mpam_devices.c   |  390 +++
 include/linux/acpi.h                          |   17 +
 include/linux/arm_mpam.h                      |   56 +
 include/linux/cacheinfo.h                     |    1 +
 25 files changed, 5117 insertions(+), 84 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml
 create mode 100644 drivers/acpi/arm64/mpam.c
 create mode 100644 drivers/platform/arm64/ec/Kconfig
 create mode 100644 drivers/platform/arm64/ec/Makefile
 rename drivers/platform/arm64/{ => ec}/acer-aspire1-ec.c (100%)
 rename drivers/platform/arm64/{ => ec}/huawei-gaokun-ec.c (100%)
 rename drivers/platform/arm64/{ => ec}/lenovo-yoga-c630.c (100%)
 create mode 100644 drivers/platform/arm64/mpam/Kconfig
 create mode 100644 drivers/platform/arm64/mpam/Makefile
 create mode 100644 drivers/platform/arm64/mpam/mpam_devices.c
 create mode 100644 drivers/platform/arm64/mpam/mpam_internal.h
 create mode 100644 drivers/platform/arm64/mpam/test_mpam_devices.c
 create mode 100644 include/linux/arm_mpam.h
Signature · 1 line
2.39.5
author maintainer
#1 lore raw

RFC PATCH 1/36 cacheinfo: Set cache 'id' based on DT data

From: Rob Herring <[email protected]>

Use the minimum CPU h/w id of the CPUs associated with the cache for the cache 'id'. This will provide a stable id value for a given system. As we need to check all possible CPUs, we can't use the shared_cpu_map which is just online CPUs. As there's not a cache to CPUs mapping in DT, we have to walk all CPU nodes and then walk cache levels.

The cache_id exposed to user-space has historically been 32 bits, and is too late to change. This value is parsed into a u32 by user-space libraries such as libvirt:
https://github.com/libvirt/libvirt/blob/master/src/util/virresctrl.c#L1588

Give up on assigning cache-id's if a CPU h/w id greater than 32 bits is found.

match_cache_node() does not make use of the __free() cleanup helpers because of_find_next_cache_node(prev) does not drop a reference to prev, and its too easy to accidentally drop the reference on cpu, which belongs to for_each_of_cpu_node().

Cc Greg Kroah-Hartman <[email protected]>
Cc "Rafael J. Wysocki" <[email protected]>
[ ben: converted to use the __free cleanup idiom ]
[ morse: Add checks to give up if a value larger than 32 bits is seen. ]

---
Use as a 32bit value has also been seen in DPDK patches here:
http://inbox.dpdk.org/dev/[email protected]/

Changes since v2:
* Removed broken use of cleanup in the match helper

Changes since v1:
* Remove the second loop in favour of a helper.

---
 drivers/base/cacheinfo.c | 45 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 45 insertions(+)
drivers/base/cacheinfo.c +45 -0
index cf0d455209d7..4e2f60c85e74 100644
@@ -8,6 +8,7 @@
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
 #include <linux/acpi.h>
+#include <linux/bitfield.h>
 #include <linux/bitops.h>
 #include <linux/cacheinfo.h>
 #include <linux/compiler.h>
@@ -183,6 +184,49 @@ static bool cache_node_is_unified(struct cacheinfo *this_leaf,
 	return of_property_read_bool(np, "cache-unified");
 }
 
+static bool match_cache_node(struct device_node *cpu,
+			     const struct device_node *cache_node)
+{
+	struct device_node *prev, *cache = of_find_next_cache_node(cpu);
+
+	while (cache) {
+		if (cache == cache_node) {
+			of_node_put(cache);
+			return true;
+		}
+
+		prev = cache;
+		cache = of_find_next_cache_node(cache);
+		of_node_put(prev);
+	}
+
+	return false;
+}
+
+static void cache_of_set_id(struct cacheinfo *this_leaf,
+			    struct device_node *cache_node)
+{
+	struct device_node *cpu;
+	u32 min_id = ~0;
+
+	for_each_of_cpu_node(cpu) {
+		u64 id = of_get_cpu_hwid(cpu, 0);
+
+		if (FIELD_GET(GENMASK_ULL(63, 32), id)) {
+			of_node_put(cpu);
+			return;
+		}
+
+		if (match_cache_node(cpu, cache_node))
+			min_id = min(min_id, id);
+	}
+
+	if (min_id != ~0) {
+		this_leaf->id = min_id;
+		this_leaf->attributes |= CACHE_ID;
+	}
+}
+
 static void cache_of_set_props(struct cacheinfo *this_leaf,
 			       struct device_node *np)
 {
@@ -198,6 +242,7 @@ static void cache_of_set_props(struct cacheinfo *this_leaf,
 	cache_get_line_size(this_leaf, np);
 	cache_nr_sets(this_leaf, np);
 	cache_associativity(this_leaf);
+	cache_of_set_id(this_leaf, np);
 }
 
 static int cache_setup_of_node(unsigned int cpu)
Signature · 1 line
2.39.5
author maintainer
#2 lore raw

RFC PATCH 2/36 cacheinfo: Add arch hook to compress CPU h/w id into 32 bits for cache-id

In reply to James Morse

Filesystems like resctrl use the cache-id exposed via sysfs to identify groups of CPUs. The value is also used for PCIe cache steering tags. On DT platforms cache-id is not something that is described in the device-tree, but instead generated from the smallest CPU h/w id of the CPUs associated with that cache.

CPU h/w ids may be larger than 32 bits.

Add a hook to allow architectures to compress the value from the devicetree into 32 bits. Returning the same value is always safe as cache_of_set_id() will stop if a value larger than 32 bits is seen.

For example, on arm64 the value is the MPIDR affinity register, which only has 32 bits of affinity data, but spread accross the 64 bit field. An arch-specific bit swizzle gives a 32 bit value.

---
 drivers/base/cacheinfo.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/base/cacheinfo.c +5 -0
index 4e2f60c85e74..613410705a47 100644
@@ -203,6 +203,10 @@ static bool match_cache_node(struct device_node *cpu,
 	return false;
 }
 
+#ifndef arch_compact_of_hwid
+#define arch_compact_of_hwid(_x)	(_x)
+#endif
+
 static void cache_of_set_id(struct cacheinfo *this_leaf,
 			    struct device_node *cache_node)
 {
@@ -212,6 +216,7 @@ static void cache_of_set_id(struct cacheinfo *this_leaf,
 	for_each_of_cpu_node(cpu) {
 		u64 id = of_get_cpu_hwid(cpu, 0);
 
+		id = arch_compact_of_hwid(id);
 		if (FIELD_GET(GENMASK_ULL(63, 32), id)) {
 			of_node_put(cpu);
 			return;
Signature · 1 line
2.39.5
author maintainer
#3 lore raw

RFC PATCH 3/36 arm64: cacheinfo: Provide helper to compress MPIDR value into u32

In reply to James Morse

Filesystems like resctrl use the cache-id exposed via sysfs to identify groups of CPUs. The value is also used for PCIe cache steering tags. On DT platforms cache-id is not something that is described in the device-tree, but instead generated from the smallest MPIDR of the CPUs associated with that cache. The cache-id exposed to user-space has historically been 32 bits.

MPIDR values may be larger than 32 bits.

MPIDR only has 32 bits worth of affinity data, but the aff3 field lives above 32bits. The corresponding lower bits are masked out by MPIDR_HWID_BITMASK and contain an SMT flag and Uni-Processor flag.

Swizzzle the aff3 field into the bottom 32 bits and using that.

In case more affinity fields are added in the future, the upper RES0 area should be checked. Returning a value greater than 32 bits from this helper will cause the caller to give up on allocating cache-ids.

---
Changes since v1:
* Removal of unrelated changes.
* Added a comment about how the RES0 bit safety net works.

---
 arch/arm64/include/asm/cache.h | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
arch/arm64/include/asm/cache.h +17 -0
index 99cd6546e72e..09963004ceea 100644
@@ -87,6 +87,23 @@ int cache_line_size(void);
 
 #define dma_get_cache_alignment	cache_line_size
 
+/* Compress a u64 MPIDR value into 32 bits. */
+static inline u64 arch_compact_of_hwid(u64 id)
+{
+	u64 aff3 = MPIDR_AFFINITY_LEVEL(id, 3);
+
+	/*
+	 * These bits are expected to be RES0. If not, return a value with
+	 * the upper 32 bits set to force the caller to give up on 32 bit
+	 * cache ids.
+	 */
+	if (FIELD_GET(GENMASK_ULL(63, 40), id))
+		return id;
+
+	return (aff3 << 24) | FIELD_GET(GENMASK_ULL(23, 0), id);
+}
+#define arch_compact_of_hwid	arch_compact_of_hwid
+
 /*
  * Read the effective value of CTR_EL0.
  *
Signature · 1 line
2.39.5
author maintainer
#4 lore raw

RFC PATCH 4/36 cacheinfo: Expose the code to generate a cache-id from a device_node

In reply to James Morse

The MPAM driver identifies caches by id for use with resctrl. It needs to know the cache-id when probe-ing, but the value isn't set in cacheinfo until device_initcall().

Expose the code that generates the cache-id. The parts of the MPAM driver that run early can use this to set up the resctrl structures before cacheinfo is ready in device_initcall().

---
Changes since v1:
* Renamed cache_of_get_id() cache_of_calculate_id().

---
 drivers/base/cacheinfo.c  | 17 ++++++++++++-----
 include/linux/cacheinfo.h |  1 +
 2 files changed, 13 insertions(+), 5 deletions(-)
drivers/base/cacheinfo.c +12 -5
index 613410705a47..0fdd6358ee73 100644
@@ -207,8 +207,7 @@ static bool match_cache_node(struct device_node *cpu,
 #define arch_compact_of_hwid(_x)	(_x)
 #endif
 
-static void cache_of_set_id(struct cacheinfo *this_leaf,
-			    struct device_node *cache_node)
+unsigned long cache_of_calculate_id(struct device_node *cache_node)
 {
 	struct device_node *cpu;
 	u32 min_id = ~0;
@@ -219,15 +218,23 @@ static void cache_of_set_id(struct cacheinfo *this_leaf,
 		id = arch_compact_of_hwid(id);
 		if (FIELD_GET(GENMASK_ULL(63, 32), id)) {
 			of_node_put(cpu);
-			return;
+			return ~0UL;
 		}
 
 		if (match_cache_node(cpu, cache_node))
 			min_id = min(min_id, id);
 	}
 
-	if (min_id != ~0) {
-		this_leaf->id = min_id;
+	return min_id;
+}
+
+static void cache_of_set_id(struct cacheinfo *this_leaf,
+			    struct device_node *cache_node)
+{
+	unsigned long id = cache_of_calculate_id(cache_node);
+
+	if (id != ~0UL) {
+		this_leaf->id = id;
 		this_leaf->attributes |= CACHE_ID;
 	}
 }
include/linux/cacheinfo.h +1 -0
index c8f4f0a0b874..2dcbb69139e9 100644
@@ -112,6 +112,7 @@ int acpi_get_cache_info(unsigned int cpu,
 #endif
 
 const struct attribute_group *cache_get_priv_group(struct cacheinfo *this_leaf);
+unsigned long cache_of_calculate_id(struct device_node *np);
 
 /*
  * Get the cacheinfo structure for the cache associated with @cpu at
Signature · 1 line
2.39.5
maintainer
#5 lore raw

Re: [RFC PATCH 04/36] cacheinfo: Expose the code to generate a cache-id from a device_node

Hi James,

Quoted 36 lines · James Morse

The MPAM driver identifies caches by id for use with resctrl. It needs to know the cache-id when probe-ing, but the value isn't set in cacheinfo until device_initcall().

Expose the code that generates the cache-id. The parts of the MPAM driver that run early can use this to set up the resctrl structures before cacheinfo is ready in device_initcall().

---
Changes since v1:
* Renamed cache_of_get_id() cache_of_calculate_id().

---
  drivers/base/cacheinfo.c  | 17 ++++++++++++-----
  include/linux/cacheinfo.h |  1 +
  2 files changed, 13 insertions(+), 5 deletions(-)
drivers/base/cacheinfo.c +1 -3
index 613410705a47..0fdd6358ee73 100644
@@ -207,8 +207,7 @@ static bool match_cache_node(struct device_node *cpu,
  #define arch_compact_of_hwid(_x)	(_x)
  #endif
 
-static void cache_of_set_id(struct cacheinfo *this_leaf,
-			    struct device_node *cache_node)
+unsigned long cache_of_calculate_id(struct device_node *cache_node)
  {
  	struct device_node *cpu;
  	u32 min_id = ~0;
@@ -219,15 +218,23 @@ static void cache_of_set_id(struct cacheinfo *this_leaf,
  		id = arch_compact_of_hwid(id);
  		if (FIELD_GET(GENMASK_ULL(63, 32), id)) {
  			of_node_put(cpu);
-			return;

drivers/base/cacheinfo.c static void cache_of_set_id(struct cacheinfo *this_leaf, · line 228

Quoted 10 lines
drivers/base/cacheinfo.c +2 -2
+			return ~0UL;
  		}
 
  		if (match_cache_node(cpu, cache_node))
  			min_id = min(min_id, id);
  	}
 
-	if (min_id != ~0) {
-		this_leaf->id = min_id;
+	return min_id;

Looks like some 32bit/64bit confusion. Don't we want to return ~0UL if min_id == ~0?

Quoted 17 lines
drivers/base/cacheinfo.c +9 -0
+}
+
+static void cache_of_set_id(struct cacheinfo *this_leaf,
+			    struct device_node *cache_node)
+{
+	unsigned long id = cache_of_calculate_id(cache_node);
+
+	if (id != ~0UL) {
+		this_leaf->id = id;
  		this_leaf->attributes |= CACHE_ID;
  	}
  }
include/linux/cacheinfo.h +0 -0
index c8f4f0a0b874..2dcbb69139e9 100644

include/linux/cacheinfo.h int acpi_get_cache_info(unsigned int cpu, · line 118

Quoted 7 lines
include/linux/cacheinfo.h +1 -0
@@ -112,6 +112,7 @@ int acpi_get_cache_info(unsigned int cpu,
  #endif
 
  const struct attribute_group *cache_get_priv_group(struct cacheinfo *this_leaf);
+unsigned long cache_of_calculate_id(struct device_node *np);
 
  /*
   * Get the cacheinfo structure for the cache associated with @cpu at

Thanks,

Ben

author maintainer
#6 lore raw

Re: [RFC PATCH 04/36] cacheinfo: Expose the code to generate a cache-id from a device_node

Hi Ben,

Quoted 29 lines · Ben Horgan
Quoted 28 lines · James Morse

The MPAM driver identifies caches by id for use with resctrl. It needs to know the cache-id when probe-ing, but the value isn't set in cacheinfo until device_initcall().

Expose the code that generates the cache-id. The parts of the MPAM driver that run early can use this to set up the resctrl structures before cacheinfo is ready in device_initcall().

drivers/base/cacheinfo.c +2 -2
index 613410705a47..0fdd6358ee73 100644
@@ -207,8 +207,7 @@ static bool match_cache_node(struct device_node *cpu,
   #define arch_compact_of_hwid(_x)    (_x)
   #endif
   -static void cache_of_set_id(struct cacheinfo *this_leaf,
-                struct device_node *cache_node)
+unsigned long cache_of_calculate_id(struct device_node *cache_node)
   {
       struct device_node *cpu;
       u32 min_id = ~0;
@@ -219,15 +218,23 @@ static void cache_of_set_id(struct cacheinfo *this_leaf,
           id = arch_compact_of_hwid(id);
           if (FIELD_GET(GENMASK_ULL(63, 32), id)) {
               of_node_put(cpu);
-            return;
+            return ~0UL;
           }
Quoted 8 lines

drivers/base/cacheinfo.c static void cache_of_set_id(struct cacheinfo *this_leaf, · line 228

Quoted 6 lines
drivers/base/cacheinfo.c +1 -1
             if (match_cache_node(cpu, cache_node))
               min_id = min(min_id, id);
       }
   -    if (min_id != ~0) {
-        this_leaf->id = min_id;
+    return min_id;

Looks like some 32bit/64bit confusion. Don't we want to return ~0UL if min_id == ~0?

Certainly some confusion - yup, because cache_of_calculate_id() needs to return something that is out of range and (u32)-1 might be valid...

I think changing min_id to be defined as:
| unsigned long min_id = ~0UL;

fixes this - any trip round the loop that doesn't match anything will eventually return ~0UL.

Thanks! - I always get the 'UL' suffixes wrong.

James

maintainer
#7 lore raw

Re: [RFC PATCH 04/36] cacheinfo: Expose the code to generate a cache-id from a device_node

Hi James,

Quoted 41 lines · James Morse

Hi Ben,

Quoted 29 lines · Ben Horgan
Quoted 28 lines · James Morse

The MPAM driver identifies caches by id for use with resctrl. It needs to know the cache-id when probe-ing, but the value isn't set in cacheinfo until device_initcall().

Expose the code that generates the cache-id. The parts of the MPAM driver that run early can use this to set up the resctrl structures before cacheinfo is ready in device_initcall().

drivers/base/cacheinfo.c +2 -2
index 613410705a47..0fdd6358ee73 100644
@@ -207,8 +207,7 @@ static bool match_cache_node(struct device_node *cpu,
   #define arch_compact_of_hwid(_x)    (_x)
   #endif
   -static void cache_of_set_id(struct cacheinfo *this_leaf,
-                struct device_node *cache_node)
+unsigned long cache_of_calculate_id(struct device_node *cache_node)
   {
       struct device_node *cpu;
       u32 min_id = ~0;
@@ -219,15 +218,23 @@ static void cache_of_set_id(struct cacheinfo *this_leaf,
           id = arch_compact_of_hwid(id);
           if (FIELD_GET(GENMASK_ULL(63, 32), id)) {
               of_node_put(cpu);
-            return;
+            return ~0UL;
           }
Quoted 8 lines

drivers/base/cacheinfo.c static void cache_of_set_id(struct cacheinfo *this_leaf, · line 228

Quoted 6 lines
drivers/base/cacheinfo.c +1 -1
             if (match_cache_node(cpu, cache_node))
               min_id = min(min_id, id);
       }
   -    if (min_id != ~0) {
-        this_leaf->id = min_id;
+    return min_id;

Looks like some 32bit/64bit confusion. Don't we want to return ~0UL if min_id == ~0?

Quoted 7 lines

Certainly some confusion - yup, because cache_of_calculate_id() needs to return something that is out of range and (u32)-1 might be valid...

I think changing min_id to be defined as:
| unsigned long min_id = ~0UL;

fixes this - any trip round the loop that doesn't match anything will eventually return ~0UL.

Yes, that would work.

Quoted 5 lines

Thanks! - I always get the 'UL' suffixes wrong.

James

Thanks,

Ben

author maintainer
#8 lore raw

RFC PATCH 5/36 ACPI / PPTT: Add a helper to fill a cpumask from a processor container

In reply to James Morse

The PPTT describes CPUs and caches, as well as processor containers. The ACPI table for MPAM describes the set of CPUs that can access an MSC with the UID of a processor container.

Add a helper to find the processor container by its id, then walk the possible CPUs to fill a cpumask with the CPUs that have this processor container as a parent.

---
 drivers/acpi/pptt.c  | 93 ++++++++++++++++++++++++++++++++++++++++++++
 include/linux/acpi.h |  6 +++
 2 files changed, 99 insertions(+)
drivers/acpi/pptt.c +93 -0
index 54676e3d82dd..13619b1b821b 100644
@@ -298,6 +298,99 @@ static struct acpi_pptt_processor *acpi_find_processor_node(struct acpi_table_he
 	return NULL;
 }
 
+/**
+ * acpi_pptt_get_child_cpus() - Find all the CPUs below a PPTT processor node
+ * @table_hdr:		A reference to the PPTT table.
+ * @parent_node:	A pointer to the processor node in the @table_hdr.
+ * @cpus:		A cpumask to fill with the CPUs below @parent_node.
+ *
+ * Walks up the PPTT from every possible CPU to find if the provided
+ * @parent_node is a parent of this CPU.
+ */
+static void acpi_pptt_get_child_cpus(struct acpi_table_header *table_hdr,
+				     struct acpi_pptt_processor *parent_node,
+				     cpumask_t *cpus)
+{
+	struct acpi_pptt_processor *cpu_node;
+	u32 acpi_id;
+	int cpu;
+
+	cpumask_clear(cpus);
+
+	for_each_possible_cpu(cpu) {
+		acpi_id = get_acpi_id_for_cpu(cpu);
+		cpu_node = acpi_find_processor_node(table_hdr, acpi_id);
+
+		while (cpu_node) {
+			if (cpu_node == parent_node) {
+				cpumask_set_cpu(cpu, cpus);
+				break;
+			}
+			cpu_node = fetch_pptt_node(table_hdr, cpu_node->parent);
+		}
+	}
+}
+
+/**
+ * acpi_pptt_get_cpus_from_container() - Populate a cpumask with all CPUs in a
+ *                                       processor containers
+ * @acpi_cpu_id:	The UID of the processor container.
+ * @cpus		The resulting CPU mask.
+ *
+ * Find the specified Processor Container, and fill @cpus with all the cpus
+ * below it.
+ *
+ * Not all 'Processor' entries in the PPTT are either a CPU or a Processor
+ * Container, they may exist purely to describe a Private resource. CPUs
+ * have to be leaves, so a Processor Container is a non-leaf that has the
+ * 'ACPI Processor ID valid' flag set.
+ *
+ * Return: 0 for a complete walk, or an error if the mask is incomplete.
+ */
+int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus)
+{
+	struct acpi_pptt_processor *cpu_node;
+	struct acpi_table_header *table_hdr;
+	struct acpi_subtable_header *entry;
+	bool leaf_flag, has_leaf_flag = false;
+	unsigned long table_end;
+	acpi_status status;
+	u32 proc_sz;
+	int ret = 0;
+
+	cpumask_clear(cpus);
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table_hdr);
+	if (ACPI_FAILURE(status))
+		return 0;
+
+	if (table_hdr->revision > 1)
+		has_leaf_flag = true;
+
+	table_end = (unsigned long)table_hdr + table_hdr->length;
+	entry = ACPI_ADD_PTR(struct acpi_subtable_header, table_hdr,
+			     sizeof(struct acpi_table_pptt));
+	proc_sz = sizeof(struct acpi_pptt_processor);
+	while ((unsigned long)entry + proc_sz <= table_end) {
+		cpu_node = (struct acpi_pptt_processor *)entry;
+		if (entry->type == ACPI_PPTT_TYPE_PROCESSOR &&
+		    cpu_node->flags & ACPI_PPTT_ACPI_PROCESSOR_ID_VALID) {
+			leaf_flag = cpu_node->flags & ACPI_PPTT_ACPI_LEAF_NODE;
+			if ((has_leaf_flag && !leaf_flag) ||
+			    (!has_leaf_flag && !acpi_pptt_leaf_node(table_hdr, cpu_node))) {
+				if (cpu_node->acpi_processor_id == acpi_cpu_id)
+					acpi_pptt_get_child_cpus(table_hdr, cpu_node, cpus);
+			}
+		}
+		entry = ACPI_ADD_PTR(struct acpi_subtable_header, entry,
+				     entry->length);
+	}
+
+	acpi_put_table(table_hdr);
+
+	return ret;
+}
+
 static u8 acpi_cache_type(enum cache_type type)
 {
 	switch (type) {
include/linux/acpi.h +6 -0
index f102c0fe3431..8c3165c2b083 100644
@@ -1541,6 +1541,7 @@ int find_acpi_cpu_topology(unsigned int cpu, int level);
 int find_acpi_cpu_topology_cluster(unsigned int cpu);
 int find_acpi_cpu_topology_package(unsigned int cpu);
 int find_acpi_cpu_topology_hetero_id(unsigned int cpu);
+int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus);
 #else
 static inline int acpi_pptt_cpu_is_thread(unsigned int cpu)
 {
@@ -1562,6 +1563,11 @@ static inline int find_acpi_cpu_topology_hetero_id(unsigned int cpu)
 {
 	return -EINVAL;
 }
+static inline int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id,
+						    cpumask_t *cpus)
+{
+	return -EINVAL;
+}
 #endif
 
 void acpi_arch_init(void);
Signature · 1 line
2.39.5
#9 lore raw

RE: [RFC PATCH 05/36] ACPI / PPTT: Add a helper to fill a cpumask from a processor container

Hello James,

Quoted 89 lines

The PPTT describes CPUs and caches, as well as processor containers. The ACPI table for MPAM describes the set of CPUs that can access an MSC with the UID of a processor container.

Add a helper to find the processor container by its id, then walk the possible CPUs to fill a cpumask with the CPUs that have this processor container as a parent.

---
 drivers/acpi/pptt.c  | 93
++++++++++++++++++++++++++++++++++++++++++++
 include/linux/acpi.h |  6 +++
 2 files changed, 99 insertions(+)
drivers/acpi/pptt.c index renamed from drivers/acpi/pptt.c +0 -0

54676e3d82dd..13619b1b821b 100644

drivers/acpi/pptt.c +2 -0
@@ -298,6 +298,99 @@ static struct acpi_pptt_processor
 *acpi_find_processor_node(struct acpi_table_he
 	return NULL;
 }
 
+/**
+ * acpi_pptt_get_child_cpus() - Find all the CPUs below a PPTT processor

node
+ * @table_hdr: A reference to the PPTT table.
+ * @parent_node: A pointer to the processor node in the @table_hdr.
+ * @cpus: A cpumask to fill with the CPUs below @parent_node.
+ *
+ * Walks up the PPTT from every possible CPU to find if the provided
+ * @parent_node is a parent of this CPU.
+ */

drivers/acpi/pptt.c +2 -0
+static void acpi_pptt_get_child_cpus(struct acpi_table_header *table_hdr,
+				     struct acpi_pptt_processor

*parent_node,

drivers/acpi/pptt.c +18 -0
+				     cpumask_t *cpus)
+{
+	struct acpi_pptt_processor *cpu_node;
+	u32 acpi_id;
+	int cpu;
+
+	cpumask_clear(cpus);
+
+	for_each_possible_cpu(cpu) {
+		acpi_id = get_acpi_id_for_cpu(cpu);
+		cpu_node = acpi_find_processor_node(table_hdr, acpi_id);
+
+		while (cpu_node) {
+			if (cpu_node == parent_node) {
+				cpumask_set_cpu(cpu, cpus);
+				break;
+			}
+			cpu_node = fetch_pptt_node(table_hdr,

cpu_node->parent);

drivers/acpi/pptt.c +6 -0
+		}
+	}
+}
+
+/**
+ * acpi_pptt_get_cpus_from_container() - Populate a cpumask with all CPUs

in a
+ * processor containers
+ * @acpi_cpu_id: The UID of the processor container.
+ * @cpus The resulting CPU mask.
+ *
+ * Find the specified Processor Container, and fill @cpus with all the

drivers/acpi/pptt.c +19 -0
+cpus
+ * below it.
+ *
+ * Not all 'Processor' entries in the PPTT are either a CPU or a
+Processor
+ * Container, they may exist purely to describe a Private resource.
+CPUs
+ * have to be leaves, so a Processor Container is a non-leaf that has
+the
+ * 'ACPI Processor ID valid' flag set.
+ *
+ * Return: 0 for a complete walk, or an error if the mask is incomplete.
+ */
+int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus)
+{
+	struct acpi_pptt_processor *cpu_node;
+	struct acpi_table_header *table_hdr;
+	struct acpi_subtable_header *entry;
+	bool leaf_flag, has_leaf_flag = false;

drivers/acpi/pptt.c static struct acpi_pptt_processor *acpi_find_processor_node(struct acpi_table_he · line 365

Quoted 10 lines
drivers/acpi/pptt.c +10 -0
+	unsigned long table_end;
+	acpi_status status;
+	u32 proc_sz;
+	int ret = 0;
+
+	cpumask_clear(cpus);
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table_hdr);
+	if (ACPI_FAILURE(status))
+		return 0;

If pptt table cannot be got, should -ENODEV be returned?

Quoted 21 lines
drivers/acpi/pptt.c +11 -0
+	if (table_hdr->revision > 1)
+		has_leaf_flag = true;
+
+	table_end = (unsigned long)table_hdr + table_hdr->length;
+	entry = ACPI_ADD_PTR(struct acpi_subtable_header, table_hdr,
+			     sizeof(struct acpi_table_pptt));
+	proc_sz = sizeof(struct acpi_pptt_processor);
+	while ((unsigned long)entry + proc_sz <= table_end) {
+		cpu_node = (struct acpi_pptt_processor *)entry;
+		if (entry->type == ACPI_PPTT_TYPE_PROCESSOR &&
+		    cpu_node->flags &

ACPI_PPTT_ACPI_PROCESSOR_ID_VALID) {

drivers/acpi/pptt.c +1 -0
+			leaf_flag = cpu_node->flags &

ACPI_PPTT_ACPI_LEAF_NODE;

drivers/acpi/pptt.c +2 -0
+			if ((has_leaf_flag && !leaf_flag) ||
+			    (!has_leaf_flag

&& !acpi_pptt_leaf_node(table_hdr, cpu_node))) {

drivers/acpi/pptt.c +1 -0
+				if (cpu_node->acpi_processor_id ==

acpi_cpu_id)

drivers/acpi/pptt.c +1 -0
+					acpi_pptt_get_child_cpus(table_hdr,

cpu_node, cpus);

drivers/acpi/pptt.c static struct acpi_pptt_processor *acpi_find_processor_node(struct acpi_table_he · line 391

Quoted 9 lines
drivers/acpi/pptt.c +9 -0
+			}
+		}
+		entry = ACPI_ADD_PTR(struct acpi_subtable_header, entry,
+				     entry->length);
+	}
+
+	acpi_put_table(table_hdr);
+
+	return ret;

Only 0 is returned here.
There is no action to be taken when the mask is incomplete.

Best regards,
Shaopeng TAN

Quoted 30 lines
drivers/acpi/pptt.c +2 -0
+}
+
 static u8 acpi_cache_type(enum cache_type type)  {
 	switch (type) {
include/linux/acpi.h index renamed from include/linux/acpi.h +0 -0

f102c0fe3431..8c3165c2b083 100644

include/linux/acpi.h +0 -0
@@ -1541,6 +1541,7 @@ int find_acpi_cpu_topology(unsigned int cpu, int

level); int find_acpi_cpu_topology_cluster(unsigned int cpu); int find_acpi_cpu_topology_package(unsigned int cpu); int find_acpi_cpu_topology_hetero_id(unsigned int cpu);

include/linux/acpi.h +3 -0
+int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t
+*cpus);
 #else
 static inline int acpi_pptt_cpu_is_thread(unsigned int cpu)  { @@ -1562,6
+1563,11 @@ static inline int find_acpi_cpu_topology_hetero_id(unsigned int

cpu) {

include/linux/acpi.h +5 -1
 	return -EINVAL;
 }
+static inline int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id,
+						    cpumask_t *cpus)
+{
+	return -EINVAL;
+}
 #endif
 
 void acpi_arch_init(void);
--

2.39.5

author maintainer
#10 lore raw

Re: [RFC PATCH 05/36] ACPI / PPTT: Add a helper to fill a cpumask from a processor container

Hi Shaopeng,

Quoted 47 lines · Shaopeng Tan (Fujitsu)

Hello James,

Quoted 45 lines

The PPTT describes CPUs and caches, as well as processor containers. The ACPI table for MPAM describes the set of CPUs that can access an MSC with the UID of a processor container.

Add a helper to find the processor container by its id, then walk the possible CPUs to fill a cpumask with the CPUs that have this processor container as a parent.

drivers/acpi/pptt.c index renamed from drivers/acpi/pptt.c +0 -0

54676e3d82dd..13619b1b821b 100644

drivers/acpi/pptt.c +2 -0
@@ -298,6 +298,99 @@ static struct acpi_pptt_processor
 
+/**
+ * acpi_pptt_get_cpus_from_container() - Populate a cpumask with all CPUs

in a
+ * processor containers
+ * @acpi_cpu_id: The UID of the processor container.
+ * @cpus The resulting CPU mask.
+ *
+ * Find the specified Processor Container, and fill @cpus with all the

drivers/acpi/pptt.c +23 -0
+cpus
+ * below it.
+ *
+ * Not all 'Processor' entries in the PPTT are either a CPU or a
+Processor
+ * Container, they may exist purely to describe a Private resource.
+CPUs
+ * have to be leaves, so a Processor Container is a non-leaf that has
+the
+ * 'ACPI Processor ID valid' flag set.
+ *
+ * Return: 0 for a complete walk, or an error if the mask is incomplete.
+ */
+int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus)
+{
+	struct acpi_pptt_processor *cpu_node;
+	struct acpi_table_header *table_hdr;
+	struct acpi_subtable_header *entry;
+	bool leaf_flag, has_leaf_flag = false;
+	unsigned long table_end;
+	acpi_status status;
+	u32 proc_sz;
+	int ret = 0;
Quoted 8 lines

drivers/acpi/pptt.c static struct acpi_pptt_processor *acpi_find_processor_node(struct acpi_table_he · line 365

Quoted 6 lines
drivers/acpi/pptt.c +6 -0
+
+	cpumask_clear(cpus);
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table_hdr);
+	if (ACPI_FAILURE(status))
+		return 0;

If pptt table cannot be got, should -ENODEV be returned?

In general its not an error for the PPTT to be missing, there are plenty of platforms where that is the case. I think in this case the caller has to be working with some information that means there has to be a PPTT, so this isn't an error that needs handling.

In MPAM's case, the ACPI table references things in the PPTT, if that table were missing the platform description is unusable. I don't think this is something we need to help debug - just ensure we don't cause a panic() that would make it harder to debug!

Quoted 24 lines
Quoted 24 lines
drivers/acpi/pptt.c +11 -0
+	if (table_hdr->revision > 1)
+		has_leaf_flag = true;
+
+	table_end = (unsigned long)table_hdr + table_hdr->length;
+	entry = ACPI_ADD_PTR(struct acpi_subtable_header, table_hdr,
+			     sizeof(struct acpi_table_pptt));
+	proc_sz = sizeof(struct acpi_pptt_processor);
+	while ((unsigned long)entry + proc_sz <= table_end) {
+		cpu_node = (struct acpi_pptt_processor *)entry;
+		if (entry->type == ACPI_PPTT_TYPE_PROCESSOR &&
+		    cpu_node->flags &

ACPI_PPTT_ACPI_PROCESSOR_ID_VALID) {

drivers/acpi/pptt.c +1 -0
+			leaf_flag = cpu_node->flags &

ACPI_PPTT_ACPI_LEAF_NODE;

drivers/acpi/pptt.c +2 -0
+			if ((has_leaf_flag && !leaf_flag) ||
+			    (!has_leaf_flag

&& !acpi_pptt_leaf_node(table_hdr, cpu_node))) {

drivers/acpi/pptt.c +1 -0
+				if (cpu_node->acpi_processor_id ==

acpi_cpu_id)

drivers/acpi/pptt.c +1 -0
+					acpi_pptt_get_child_cpus(table_hdr,

cpu_node, cpus);

drivers/acpi/pptt.c +3 -0
+			}
+		}
+		entry = ACPI_ADD_PTR(struct acpi_subtable_header, entry,
Quoted 8 lines

drivers/acpi/pptt.c static struct acpi_pptt_processor *acpi_find_processor_node(struct acpi_table_he · line 391

Quoted 6 lines
drivers/acpi/pptt.c +6 -0
+				     entry->length);
+	}
+
+	acpi_put_table(table_hdr);
+
+	return ret;

Only 0 is returned here.

Good spot! I think this allocated memory in the past, it can probably be made 'void', which will make your above point easier too.

Quoted 1 line

There is no action to be taken when the mask is incomplete.

I don't think there needs to be. General callers should be using cacheinfo for this information. This only exists as the MPAM driver needs to know about the topology of the system before 'all' the CPUs are online. (which could be never).

Thanks,

James

#11 lore raw

Re: [RFC PATCH 05/36] ACPI / PPTT: Add a helper to fill a cpumask from a processor container

In reply to James Morse
Quoted 13 lines · On Fri, 11 Jul 2025 18:36:17 +0000 James Morse <[email protected]> wrote:

The PPTT describes CPUs and caches, as well as processor containers. The ACPI table for MPAM describes the set of CPUs that can access an MSC with the UID of a processor container.

Add a helper to find the processor container by its id, then walk the possible CPUs to fill a cpumask with the CPUs that have this processor container as a parent.

Quoted 5 lines

+/**
+ * acpi_pptt_get_cpus_from_container() - Populate a cpumask with all CPUs in a
+ * processor containers
+ * @acpi_cpu_id: The UID of the processor container.
+ * @cpus The resulting CPU mask.

Missing colon.

From a W=1 build (and hence kernel-doc warning).

Quoted 12 lines

+ *
+ * Find the specified Processor Container, and fill @cpus with all the cpus
+ * below it.
+ *
+ * Not all 'Processor' entries in the PPTT are either a CPU or a Processor
+ * Container, they may exist purely to describe a Private resource. CPUs
+ * have to be leaves, so a Processor Container is a non-leaf that has the
+ * 'ACPI Processor ID valid' flag set.
+ *
+ * Return: 0 for a complete walk, or an error if the mask is incomplete.
+ */
+int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus)

author maintainer
#12 lore raw

Re: [RFC PATCH 05/36] ACPI / PPTT: Add a helper to fill a cpumask from a processor container

Hi Jonathan,

Quoted 17 lines · Jonathan Cameron
Quoted 8 lines · On Fri, 11 Jul 2025 18:36:17 +0000 James Morse <[email protected]> wrote:

The PPTT describes CPUs and caches, as well as processor containers. The ACPI table for MPAM describes the set of CPUs that can access an MSC with the UID of a processor container.

Add a helper to find the processor container by its id, then walk the possible CPUs to fill a cpumask with the CPUs that have this processor container as a parent.

Quoted 5 lines

+/**
+ * acpi_pptt_get_cpus_from_container() - Populate a cpumask with all CPUs in a
+ * processor containers
+ * @acpi_cpu_id: The UID of the processor container.
+ * @cpus The resulting CPU mask.

Quoted 3 lines

Missing colon.

From a W=1 build (and hence kernel-doc warning).

Thanks!

W=1 spews so much output I tend to rely on the kbuild robot to report if I'm making it worse!

James

maintainer
#13 lore raw

Re: [RFC PATCH 05/36] ACPI / PPTT: Add a helper to fill a cpumask from a processor container

In reply to James Morse

Hi James,

Quoted 97 lines · James Morse

The PPTT describes CPUs and caches, as well as processor containers. The ACPI table for MPAM describes the set of CPUs that can access an MSC with the UID of a processor container.

Add a helper to find the processor container by its id, then walk the possible CPUs to fill a cpumask with the CPUs that have this processor container as a parent.

---
  drivers/acpi/pptt.c  | 93 ++++++++++++++++++++++++++++++++++++++++++++
  include/linux/acpi.h |  6 +++
  2 files changed, 99 insertions(+)
drivers/acpi/pptt.c +73 -0
index 54676e3d82dd..13619b1b821b 100644
@@ -298,6 +298,99 @@ static struct acpi_pptt_processor *acpi_find_processor_node(struct acpi_table_he
  	return NULL;
  }
 
+/**
+ * acpi_pptt_get_child_cpus() - Find all the CPUs below a PPTT processor node
+ * @table_hdr:		A reference to the PPTT table.
+ * @parent_node:	A pointer to the processor node in the @table_hdr.
+ * @cpus:		A cpumask to fill with the CPUs below @parent_node.
+ *
+ * Walks up the PPTT from every possible CPU to find if the provided
+ * @parent_node is a parent of this CPU.
+ */
+static void acpi_pptt_get_child_cpus(struct acpi_table_header *table_hdr,
+				     struct acpi_pptt_processor *parent_node,
+				     cpumask_t *cpus)
+{
+	struct acpi_pptt_processor *cpu_node;
+	u32 acpi_id;
+	int cpu;
+
+	cpumask_clear(cpus);
+
+	for_each_possible_cpu(cpu) {
+		acpi_id = get_acpi_id_for_cpu(cpu);
+		cpu_node = acpi_find_processor_node(table_hdr, acpi_id);
+
+		while (cpu_node) {
+			if (cpu_node == parent_node) {
+				cpumask_set_cpu(cpu, cpus);
+				break;
+			}
+			cpu_node = fetch_pptt_node(table_hdr, cpu_node->parent);
+		}
+	}
+}
+
+/**
+ * acpi_pptt_get_cpus_from_container() - Populate a cpumask with all CPUs in a
+ *                                       processor containers
+ * @acpi_cpu_id:	The UID of the processor container.
+ * @cpus		The resulting CPU mask.
+ *
+ * Find the specified Processor Container, and fill @cpus with all the cpus
+ * below it.
+ *
+ * Not all 'Processor' entries in the PPTT are either a CPU or a Processor
+ * Container, they may exist purely to describe a Private resource. CPUs
+ * have to be leaves, so a Processor Container is a non-leaf that has the
+ * 'ACPI Processor ID valid' flag set.
+ *
+ * Return: 0 for a complete walk, or an error if the mask is incomplete.
+ */
+int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus)
+{
+	struct acpi_pptt_processor *cpu_node;
+	struct acpi_table_header *table_hdr;
+	struct acpi_subtable_header *entry;
+	bool leaf_flag, has_leaf_flag = false;
+	unsigned long table_end;
+	acpi_status status;
+	u32 proc_sz;
+	int ret = 0;
+
+	cpumask_clear(cpus);
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table_hdr);
+	if (ACPI_FAILURE(status))
+		return 0;
+
+	if (table_hdr->revision > 1)
+		has_leaf_flag = true;
+
+	table_end = (unsigned long)table_hdr + table_hdr->length;
+	entry = ACPI_ADD_PTR(struct acpi_subtable_header, table_hdr,
+			     sizeof(struct acpi_table_pptt));
+	proc_sz = sizeof(struct acpi_pptt_processor);

drivers/acpi/pptt.c static struct acpi_pptt_processor *acpi_find_processor_node(struct acpi_table_he · line 383

Quoted 10 lines
drivers/acpi/pptt.c +10 -0
+	while ((unsigned long)entry + proc_sz <= table_end) {
+		cpu_node = (struct acpi_pptt_processor *)entry;
+		if (entry->type == ACPI_PPTT_TYPE_PROCESSOR &&
+		    cpu_node->flags & ACPI_PPTT_ACPI_PROCESSOR_ID_VALID) {
+			leaf_flag = cpu_node->flags & ACPI_PPTT_ACPI_LEAF_NODE;
+			if ((has_leaf_flag && !leaf_flag) ||
+			    (!has_leaf_flag && !acpi_pptt_leaf_node(table_hdr, cpu_node))) {
+				if (cpu_node->acpi_processor_id == acpi_cpu_id)
+					acpi_pptt_get_child_cpus(table_hdr, cpu_node, cpus);
+			}

acpi_pptt_leaf_node() returns early based on the leaf flag so you can just rely on that here; remove has_leaf_flag and the corresponding extra logic.

Quoted 27 lines
drivers/acpi/pptt.c +10 -0
+		}
+		entry = ACPI_ADD_PTR(struct acpi_subtable_header, entry,
+				     entry->length);
+	}
+
+	acpi_put_table(table_hdr);
+
+	return ret;
+}
+
  static u8 acpi_cache_type(enum cache_type type)
  {
  	switch (type) {
include/linux/acpi.h +1 -0
index f102c0fe3431..8c3165c2b083 100644
@@ -1541,6 +1541,7 @@ int find_acpi_cpu_topology(unsigned int cpu, int level);
  int find_acpi_cpu_topology_cluster(unsigned int cpu);
  int find_acpi_cpu_topology_package(unsigned int cpu);
  int find_acpi_cpu_topology_hetero_id(unsigned int cpu);
+int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus);
  #else
  static inline int acpi_pptt_cpu_is_thread(unsigned int cpu)
  {
@@ -1562,6 +1563,11 @@ static inline int find_acpi_cpu_topology_hetero_id(unsigned int cpu)
  {

include/linux/acpi.h static inline int find_acpi_cpu_topology_hetero_id(unsigned int cpu) · line 1573

Quoted 10 lines
include/linux/acpi.h +5 -0
  	return -EINVAL;
  }
+static inline int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id,
+						    cpumask_t *cpus)
+{
+	return -EINVAL;
+}
  #endif
 
  void acpi_arch_init(void);

Thanks,

Ben

author maintainer
#14 lore raw

Re: [RFC PATCH 05/36] ACPI / PPTT: Add a helper to fill a cpumask from a processor container

Hi Ben,

Quoted 59 lines · Ben Horgan
Quoted 58 lines · James Morse

The PPTT describes CPUs and caches, as well as processor containers. The ACPI table for MPAM describes the set of CPUs that can access an MSC with the UID of a processor container.

Add a helper to find the processor container by its id, then walk the possible CPUs to fill a cpumask with the CPUs that have this processor container as a parent.

drivers/acpi/pptt.c +45 -0
index 54676e3d82dd..13619b1b821b 100644
@@ -298,6 +298,99 @@ static struct acpi_pptt_processor *acpi_find_processor_node(struct
+/**
+ * acpi_pptt_get_cpus_from_container() - Populate a cpumask with all CPUs in a
+ *                                       processor containers
+ * @acpi_cpu_id:    The UID of the processor container.
+ * @cpus        The resulting CPU mask.
+ *
+ * Find the specified Processor Container, and fill @cpus with all the cpus
+ * below it.
+ *
+ * Not all 'Processor' entries in the PPTT are either a CPU or a Processor
+ * Container, they may exist purely to describe a Private resource. CPUs
+ * have to be leaves, so a Processor Container is a non-leaf that has the
+ * 'ACPI Processor ID valid' flag set.
+ *
+ * Return: 0 for a complete walk, or an error if the mask is incomplete.
+ */
+int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus)
+{
+    struct acpi_pptt_processor *cpu_node;
+    struct acpi_table_header *table_hdr;
+    struct acpi_subtable_header *entry;
+    bool leaf_flag, has_leaf_flag = false;
+    unsigned long table_end;
+    acpi_status status;
+    u32 proc_sz;
+    int ret = 0;
+
+    cpumask_clear(cpus);
+
+    status = acpi_get_table(ACPI_SIG_PPTT, 0, &table_hdr);
+    if (ACPI_FAILURE(status))
+        return 0;
+
+    if (table_hdr->revision > 1)
+        has_leaf_flag = true;
+
+    table_end = (unsigned long)table_hdr + table_hdr->length;
+    entry = ACPI_ADD_PTR(struct acpi_subtable_header, table_hdr,
+                 sizeof(struct acpi_table_pptt));
+    proc_sz = sizeof(struct acpi_pptt_processor);
+    while ((unsigned long)entry + proc_sz <= table_end) {
+        cpu_node = (struct acpi_pptt_processor *)entry;
+        if (entry->type == ACPI_PPTT_TYPE_PROCESSOR &&
+            cpu_node->flags & ACPI_PPTT_ACPI_PROCESSOR_ID_VALID) {
+            leaf_flag = cpu_node->flags & ACPI_PPTT_ACPI_LEAF_NODE;
Quoted 8 lines

drivers/acpi/pptt.c static struct acpi_pptt_processor *acpi_find_processor_node(struct acpi_table_he · line 383

Quoted 5 lines
drivers/acpi/pptt.c +5 -0
+            if ((has_leaf_flag && !leaf_flag) ||
+                (!has_leaf_flag && !acpi_pptt_leaf_node(table_hdr, cpu_node))) {
+                if (cpu_node->acpi_processor_id == acpi_cpu_id)
+                    acpi_pptt_get_child_cpus(table_hdr, cpu_node, cpus);
+            }

acpi_pptt_leaf_node() returns early based on the leaf flag so you can just rely on that here; remove has_leaf_flag and the corresponding extra logic.

Aha! I was only doing this to try and avoid that extra descent of the tree. I missed that its already taken into account.

Thanks,

James

author maintainer
#15 lore raw

RFC PATCH 6/36 ACPI / PPTT: Stop acpi_count_levels() expecting callers to clear levels

In reply to James Morse

acpi_count_levels() passes the number of levels back via a pointer argument. It also passes this to acpi_find_cache_level() as the starting_level, and preserves this value as it walks up the cpu_node tree counting the levels.

The only caller acpi_get_cache_info() happens to have already initialised levels to zero, which acpi_count_levels() depends on to get the correct result.

Explicitly zero the levels variable, so the count always starts at zero. This saves any additional callers having to work out they need to do this.

---
 drivers/acpi/pptt.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/acpi/pptt.c +3 -1
index 13619b1b821b..13ca2eee3b98 100644
@@ -183,7 +183,7 @@ acpi_find_cache_level(struct acpi_table_header *table_hdr,
  * @cpu_node: processor node we wish to count caches for
  * @levels: Number of levels if success.
  * @split_levels:	Number of split cache levels (data/instruction) if
- *			success. Can by NULL.
+ *			success. Can be NULL.
  *
  * Given a processor node containing a processing unit, walk into it and count
  * how many levels exist solely for it, and then walk up each level until we hit
@@ -196,6 +196,8 @@ static void acpi_count_levels(struct acpi_table_header *table_hdr,
 			      struct acpi_pptt_processor *cpu_node,
 			      unsigned int *levels, unsigned int *split_levels)
 {
+	*levels = 0;
+
 	do {
 		acpi_find_cache_level(table_hdr, cpu_node, levels, split_levels, 0, 0);
 		cpu_node = fetch_pptt_node(table_hdr, cpu_node->parent);
Signature · 1 line
2.39.5
#16 lore raw

Re: [RFC PATCH 06/36] ACPI / PPTT: Stop acpi_count_levels() expecting callers to clear levels

Quoted 10 lines · On Fri, 11 Jul 2025 18:36:18 +0000 James Morse <[email protected]> wrote:

acpi_count_levels() passes the number of levels back via a pointer argument. It also passes this to acpi_find_cache_level() as the starting_level, and preserves this value as it walks up the cpu_node tree counting the levels.

The only caller acpi_get_cache_info() happens to have already initialised levels to zero, which acpi_count_levels() depends on to get the correct result.

Explicitly zero the levels variable, so the count always starts at zero. This saves any additional callers having to work out they need to do this.

Hi James,

This is all a bit fiddly as we now end up with that initialized in various different places. Perhaps simpler to have acpi_count_levels() return the number of levels rather than void. Then return number of levels rather than 0 on success from acpi_get_cache_info(). Negative error codes used for failure just like now.

That would leave only a local variable in acpi_count_levels being initialized to 0 and passed to acpi_find_cache_level() before being returned when the loop terminates.

I think that sequence then makes it such that we can't fail to initialize it at without the compiler noticing and screaming.

Requires a few changes from if (ret) to if (ret < 0) at callers of acpi_get_cache_info() but looks simple (says the person who hasn't actually coded it!)

Quoted 12 lines
---
 drivers/acpi/pptt.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/acpi/pptt.c +0 -0
index 13619b1b821b..13ca2eee3b98 100644

drivers/acpi/pptt.c acpi_find_cache_level(struct acpi_table_header *table_hdr, · line 186

Quoted 5 lines
drivers/acpi/pptt.c +1 -1
@@ -183,7 +183,7 @@ acpi_find_cache_level(struct acpi_table_header *table_hdr,
  * @cpu_node: processor node we wish to count caches for
  * @levels: Number of levels if success.
  * @split_levels:	Number of split cache levels (data/instruction) if
- *			success. Can by NULL.
+ *			success. Can be NULL.

Grumpy reviewer hat. Unrelated cleanup up - good to have but not in this patch where it's a distraction.

Quoted 12 lines
drivers/acpi/pptt.c +2 -0
  *
  * Given a processor node containing a processing unit, walk into it and count
  * how many levels exist solely for it, and then walk up each level until we hit
@@ -196,6 +196,8 @@ static void acpi_count_levels(struct acpi_table_header *table_hdr,
 			      struct acpi_pptt_processor *cpu_node,
 			      unsigned int *levels, unsigned int *split_levels)
 {
+	*levels = 0;
+
 	do {
 		acpi_find_cache_level(table_hdr, cpu_node, levels, split_levels, 0, 0);
 		cpu_node = fetch_pptt_node(table_hdr, cpu_node->parent);
author maintainer
#17 lore raw

Re: [RFC PATCH 06/36] ACPI / PPTT: Stop acpi_count_levels() expecting callers to clear levels

Hi Jonathan,

Quoted 11 lines · Jonathan Cameron
Quoted 8 lines · On Fri, 11 Jul 2025 18:36:18 +0000 James Morse <[email protected]> wrote:

acpi_count_levels() passes the number of levels back via a pointer argument. It also passes this to acpi_find_cache_level() as the starting_level, and preserves this value as it walks up the cpu_node tree counting the levels.

The only caller acpi_get_cache_info() happens to have already initialised levels to zero, which acpi_count_levels() depends on to get the correct result.

Quoted 5 lines
Quoted 2 lines

Explicitly zero the levels variable, so the count always starts at zero. This saves any additional callers having to work out they need to do this.

This is all a bit fiddly as we now end up with that initialized in various different places.

I've debugged this one a few times, (turns out I'm forgetful) ... I figured doing this was better than adding a comment to warn others.
As its static, I figured it was something the compiler can optimise out if there is a duplicate. (I couldn't find any initialisation I could remove because of this)

Quoted 9 lines

Perhaps simpler to have acpi_count_levels() return the number of levels rather than void. Then return number of levels rather than 0 on success from acpi_get_cache_info(). Negative error codes used for failure just like now.

That would leave only a local variable in acpi_count_levels being initialized to 0 and passed to acpi_find_cache_level() before being returned when the loop terminates.

Quoted 6 lines

I think that sequence then makes it such that we can't fail to initialize it at without the compiler noticing and screaming.

Requires a few changes from if (ret) to if (ret < 0) at callers of acpi_get_cache_info() but looks simple (says the person who hasn't actually coded it!)

Breaking the symmetry between levels and split_levels is an argument against this.

I think within pptt.c this is fine, because 'level's is used internally as 'starting_level', and this expectation it was initialised to zero is a nasty surprise.

But exposing that from acpi_get_cache_info() looks stranger - and would need to touch users in cacheinfo, arm64, riscv.

I've updated acpi_count_levels() to look as you describe - that at least makes it harder to miss this in future. (not sure whether it saves anything)

Quoted 5 lines
Quoted 5 lines
drivers/acpi/pptt.c +0 -0
index 13619b1b821b..13ca2eee3b98 100644
@@ -183,7 +183,7 @@ acpi_find_cache_level(struct acpi_table_header *table_hdr,
Quoted 8 lines

drivers/acpi/pptt.c acpi_find_cache_level(struct acpi_table_header *table_hdr, · line 186

Quoted 5 lines
drivers/acpi/pptt.c +1 -1
  * @cpu_node: processor node we wish to count caches for
  * @levels: Number of levels if success.
  * @split_levels:	Number of split cache levels (data/instruction) if
- *			success. Can by NULL.
+ *			success. Can be NULL.

Grumpy reviewer hat. Unrelated cleanup up - good to have but not in this patch where it's a distraction.

I was hoping diff would keep it as one hunk. Happy to leave it tyopd!

Thanks,

James

author maintainer
#18 lore raw

RFC PATCH 7/36 ACPI / PPTT: Find cache level by cache-id

In reply to James Morse

The MPAM table identifies caches by id. The MPAM driver also wants to know the cache level to determine if the platform is of the shape that can be managed via resctrl. Cacheinfo has this information, but only for CPUs that are online.

Waiting for all CPUs to come online is a problem for platforms where CPUs are brought online late by user-space.

Add a helper that walks every possible cache, until it finds the one identified by cache-id, then return the level.

acpi_count_levels() expects its levels parameter to be initialised to zero as it passes it to acpi_find_cache_level() as starting_level. The existing callers do this. Document it.

---
 drivers/acpi/pptt.c  | 73 ++++++++++++++++++++++++++++++++++++++++++++
 include/linux/acpi.h |  5 +++
 2 files changed, 78 insertions(+)
drivers/acpi/pptt.c +73 -0
index 13ca2eee3b98..f53748a5df19 100644
@@ -912,3 +912,76 @@ int find_acpi_cpu_topology_hetero_id(unsigned int cpu)
 	return find_acpi_cpu_topology_tag(cpu, PPTT_ABORT_PACKAGE,
 					  ACPI_PPTT_ACPI_IDENTICAL);
 }
+
+/**
+ * find_acpi_cache_level_from_id() - Get the level of the specified cache
+ * @cache_id: The id field of the unified cache
+ *
+ * Determine the level relative to any CPU for the unified cache identified by
+ * cache_id. This allows the property to be found even if the CPUs are offline.
+ *
+ * The returned level can be used to group unified caches that are peers.
+ *
+ * The PPTT table must be rev 3 or later,
+ *
+ * If one CPUs L2 is shared with another as L3, this function will return
+ * an unpredictable value.
+ *
+ * Return: -ENOENT if the PPTT doesn't exist, or the cache cannot be found.
+ * Otherwise returns a value which represents the level of the specified cache.
+ */
+int find_acpi_cache_level_from_id(u32 cache_id)
+{
+	u32 acpi_cpu_id;
+	acpi_status status;
+	int level, cpu, num_levels;
+	struct acpi_pptt_cache *cache;
+	struct acpi_table_header *table;
+	struct acpi_pptt_cache_v1 *cache_v1;
+	struct acpi_pptt_processor *cpu_node;
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table);
+	if (ACPI_FAILURE(status)) {
+		acpi_pptt_warn_missing();
+		return -ENOENT;
+	}
+
+	if (table->revision < 3) {
+		acpi_put_table(table);
+		return -ENOENT;
+	}
+
+	/*
+	 * If we found the cache first, we'd still need to walk from each CPU
+	 * to find the level...
+	 */
+	for_each_possible_cpu(cpu) {
+		acpi_cpu_id = get_acpi_id_for_cpu(cpu);
+		cpu_node = acpi_find_processor_node(table, acpi_cpu_id);
+		if (!cpu_node)
+			break;
+		acpi_count_levels(table, cpu_node, &num_levels, NULL);
+
+		/* Start at 1 for L1 */
+		for (level = 1; level <= num_levels; level++) {
+			cache = acpi_find_cache_node(table, acpi_cpu_id,
+						     ACPI_PPTT_CACHE_TYPE_UNIFIED,
+						     level, &cpu_node);
+			if (!cache)
+				continue;
+
+			cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1,
+						cache,
+						sizeof(struct acpi_pptt_cache));
+
+			if (cache->flags & ACPI_PPTT_CACHE_ID_VALID &&
+			    cache_v1->cache_id == cache_id) {
+				acpi_put_table(table);
+				return level;
+			}
+		}
+	}
+
+	acpi_put_table(table);
+	return -ENOENT;
+}
include/linux/acpi.h +5 -0
index 8c3165c2b083..82947f6d2a43 100644
@@ -1542,6 +1542,7 @@ int find_acpi_cpu_topology_cluster(unsigned int cpu);
 int find_acpi_cpu_topology_package(unsigned int cpu);
 int find_acpi_cpu_topology_hetero_id(unsigned int cpu);
 int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus);
+int find_acpi_cache_level_from_id(u32 cache_id);
 #else
 static inline int acpi_pptt_cpu_is_thread(unsigned int cpu)
 {
@@ -1568,6 +1569,10 @@ static inline int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id,
 {
 	return -EINVAL;
 }
+static inline int find_acpi_cache_level_from_id(u32 cache_id)
+{
+	return -EINVAL;
+}
 #endif
 
 void acpi_arch_init(void);
Signature · 1 line
2.39.5
maintainer
#19 lore raw

Re: [RFC PATCH 07/36] ACPI / PPTT: Find cache level by cache-id

Hi James,

Quoted 8 lines · James Morse

The MPAM table identifies caches by id. The MPAM driver also wants to know the cache level to determine if the platform is of the shape that can be managed via resctrl. Cacheinfo has this information, but only for CPUs that are online.

Waiting for all CPUs to come online is a problem for platforms where CPUs are brought online late by user-space.

Quoted 6 lines

Add a helper that walks every possible cache, until it finds the one identified by cache-id, then return the level.

acpi_count_levels() expects its levels parameter to be initialised to zero as it passes it to acpi_find_cache_level() as starting_level. The existing callers do this. Document it.

This paragraph is stale. You dealt with this in the previous commit.

Quoted 102 lines
---
  drivers/acpi/pptt.c  | 73 ++++++++++++++++++++++++++++++++++++++++++++
  include/linux/acpi.h |  5 +++
  2 files changed, 78 insertions(+)
drivers/acpi/pptt.c +73 -0
index 13ca2eee3b98..f53748a5df19 100644
@@ -912,3 +912,76 @@ int find_acpi_cpu_topology_hetero_id(unsigned int cpu)
  	return find_acpi_cpu_topology_tag(cpu, PPTT_ABORT_PACKAGE,
  					  ACPI_PPTT_ACPI_IDENTICAL);
  }
+
+/**
+ * find_acpi_cache_level_from_id() - Get the level of the specified cache
+ * @cache_id: The id field of the unified cache
+ *
+ * Determine the level relative to any CPU for the unified cache identified by
+ * cache_id. This allows the property to be found even if the CPUs are offline.
+ *
+ * The returned level can be used to group unified caches that are peers.
+ *
+ * The PPTT table must be rev 3 or later,
+ *
+ * If one CPUs L2 is shared with another as L3, this function will return
+ * an unpredictable value.
+ *
+ * Return: -ENOENT if the PPTT doesn't exist, or the cache cannot be found.
+ * Otherwise returns a value which represents the level of the specified cache.
+ */
+int find_acpi_cache_level_from_id(u32 cache_id)
+{
+	u32 acpi_cpu_id;
+	acpi_status status;
+	int level, cpu, num_levels;
+	struct acpi_pptt_cache *cache;
+	struct acpi_table_header *table;
+	struct acpi_pptt_cache_v1 *cache_v1;
+	struct acpi_pptt_processor *cpu_node;
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table);
+	if (ACPI_FAILURE(status)) {
+		acpi_pptt_warn_missing();
+		return -ENOENT;
+	}
+
+	if (table->revision < 3) {
+		acpi_put_table(table);
+		return -ENOENT;
+	}
+
+	/*
+	 * If we found the cache first, we'd still need to walk from each CPU
+	 * to find the level...
+	 */
+	for_each_possible_cpu(cpu) {
+		acpi_cpu_id = get_acpi_id_for_cpu(cpu);
+		cpu_node = acpi_find_processor_node(table, acpi_cpu_id);
+		if (!cpu_node)
+			break;
+		acpi_count_levels(table, cpu_node, &num_levels, NULL);
+
+		/* Start at 1 for L1 */
+		for (level = 1; level <= num_levels; level++) {
+			cache = acpi_find_cache_node(table, acpi_cpu_id,
+						     ACPI_PPTT_CACHE_TYPE_UNIFIED,
+						     level, &cpu_node);
+			if (!cache)
+				continue;
+
+			cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1,
+						cache,
+						sizeof(struct acpi_pptt_cache));
+
+			if (cache->flags & ACPI_PPTT_CACHE_ID_VALID &&
+			    cache_v1->cache_id == cache_id) {
+				acpi_put_table(table);
+				return level;
+			}
+		}
+	}
+
+	acpi_put_table(table);
+	return -ENOENT;
+}
include/linux/acpi.h +1 -0
index 8c3165c2b083..82947f6d2a43 100644
@@ -1542,6 +1542,7 @@ int find_acpi_cpu_topology_cluster(unsigned int cpu);
  int find_acpi_cpu_topology_package(unsigned int cpu);
  int find_acpi_cpu_topology_hetero_id(unsigned int cpu);
  int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus);
+int find_acpi_cache_level_from_id(u32 cache_id);
  #else
  static inline int acpi_pptt_cpu_is_thread(unsigned int cpu)
  {

include/linux/acpi.h static inline int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, · line 1578

Quoted 10 lines
include/linux/acpi.h +4 -0
@@ -1568,6 +1569,10 @@ static inline int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id,
  {
  	return -EINVAL;
  }
+static inline int find_acpi_cache_level_from_id(u32 cache_id)
+{
+	return -EINVAL;
+}
  #endif
 
  void acpi_arch_init(void);

Thanks,

Ben

author maintainer
#20 lore raw

Re: [RFC PATCH 07/36] ACPI / PPTT: Find cache level by cache-id

Hi Ben,

Quoted 9 lines · Ben Horgan
Quoted 8 lines · James Morse

The MPAM table identifies caches by id. The MPAM driver also wants to know the cache level to determine if the platform is of the shape that can be managed via resctrl. Cacheinfo has this information, but only for CPUs that are online.

Waiting for all CPUs to come online is a problem for platforms where CPUs are brought online late by user-space.

Quoted 8 lines
Quoted 6 lines

Add a helper that walks every possible cache, until it finds the one identified by cache-id, then return the level.

acpi_count_levels() expects its levels parameter to be initialised to zero as it passes it to acpi_find_cache_level() as starting_level. The existing callers do this. Document it.

This paragraph is stale. You dealt with this in the previous commit.

Fixed, thanks.

James

#21 lore raw

Re: [RFC PATCH 07/36] ACPI / PPTT: Find cache level by cache-idUIRE

In reply to James Morse
Quoted 11 lines · On Fri, 11 Jul 2025 18:36:19 +0000 James Morse <[email protected]> wrote:

The MPAM table identifies caches by id. The MPAM driver also wants to know the cache level to determine if the platform is of the shape that can be managed via resctrl. Cacheinfo has this information, but only for CPUs that are online.

Waiting for all CPUs to come online is a problem for platforms where CPUs are brought online late by user-space.

Add a helper that walks every possible cache, until it finds the one identified by cache-id, then return the level.

Quoted 6 lines

acpi_count_levels() expects its levels parameter to be initialised to zero as it passes it to acpi_find_cache_level() as starting_level. The existing callers do this. Document it.

A few suggestions inline. Mostly driven by the number of missing table puts I've seen in ACPI code. You don't have any missing here but with a bit of restructuring you can make that easy to see.

Quoted 38 lines
---
 drivers/acpi/pptt.c  | 73 ++++++++++++++++++++++++++++++++++++++++++++
 include/linux/acpi.h |  5 +++
 2 files changed, 78 insertions(+)
drivers/acpi/pptt.c +25 -0
index 13ca2eee3b98..f53748a5df19 100644
@@ -912,3 +912,76 @@ int find_acpi_cpu_topology_hetero_id(unsigned int cpu)
 	return find_acpi_cpu_topology_tag(cpu, PPTT_ABORT_PACKAGE,
 					  ACPI_PPTT_ACPI_IDENTICAL);
 }
+
+/**
+ * find_acpi_cache_level_from_id() - Get the level of the specified cache
+ * @cache_id: The id field of the unified cache
+ *
+ * Determine the level relative to any CPU for the unified cache identified by
+ * cache_id. This allows the property to be found even if the CPUs are offline.
+ *
+ * The returned level can be used to group unified caches that are peers.
+ *
+ * The PPTT table must be rev 3 or later,
+ *
+ * If one CPUs L2 is shared with another as L3, this function will return
+ * an unpredictable value.
+ *
+ * Return: -ENOENT if the PPTT doesn't exist, or the cache cannot be found.
+ * Otherwise returns a value which represents the level of the specified cache.
+ */
+int find_acpi_cache_level_from_id(u32 cache_id)
+{
+	u32 acpi_cpu_id;
+	acpi_status status;
+	int level, cpu, num_levels;
+	struct acpi_pptt_cache *cache;
+	struct acpi_table_header *table;

drivers/acpi/pptt.c int find_acpi_cpu_topology_hetero_id(unsigned int cpu) · line 949

Quoted 10 lines
drivers/acpi/pptt.c +10 -0
+	struct acpi_pptt_cache_v1 *cache_v1;
+	struct acpi_pptt_processor *cpu_node;
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table);
+	if (ACPI_FAILURE(status)) {
+		acpi_pptt_warn_missing();
+		return -ENOENT;
+	}
+
+	if (table->revision < 3) {

Maybe a unified exit path given all paths need to do acpi_put_table() and return either error or level.

Or maybe it's time for some cleanup.h magic for acpi tables. I've been thinking about it for a while and mostly stuck on the name ;) (simpler suggestion follows)

static struct acpi_table_header *acpi_get_table_ret(char *signature, u32 instance)

{
	struct acpi_table_header *table;
	int status = acpi_get_table(signature, instance, &table);
	
	if (ACPI_FAILURE(status))
		return ERR_PTR(-ENOENT);
	return table;
}

DEFINE_FREE(acpi_table, struct acpi_table_header *, if (!IS_ERR(_T)) acpi_put_table(_T))

Finally in here and loads of other places we avoid chance of missing an acpi_put_table and generally simplify the code a little.

int find_acpi_cache_level_from_id(u32 cache_id)

{
	u32 acpi_cpu_id;
	acpi_status status;
	int level, cpu, num_levels;
	struct acpi_pptt_cache *cache;
	struct acpi_pptt_cache_v1 *cache_v1;
	struct acpi_pptt_processor *cpu_node;


	struct acpi_table_header *table __free(acpi_table) =
		acpi_get_table_ret(ACPI_SIG_PPTT, 0);

	if (IS_ERR(table)
		return PTR_ERR(table);

	if (table->revision < 3)
		return -ENOENT;

	/*
	 * If we found the cache first, we'd still need to walk from each CPU
	 * to find the level...
	 */
	for_each_possible_cpu(cpu) {
		acpi_cpu_id = get_acpi_id_for_cpu(cpu);
		cpu_node = acpi_find_processor_node(table, acpi_cpu_id);
		if (!cpu_node)
			return -ENOENT;
		acpi_count_levels(table, cpu_node, &num_levels, NULL);

		/* Start at 1 for L1 */
		for (level = 1; level <= num_levels; level++) {
			cache = acpi_find_cache_node(table, acpi_cpu_id,
						     ACPI_PPTT_CACHE_TYPE_UNIFIED,
						     level, &cpu_node);
			if (!cache)
				continue;

			cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1,
						cache,
						sizeof(struct acpi_pptt_cache));

			if (cache->flags & ACPI_PPTT_CACHE_ID_VALID &&
			    cache_v1->cache_id == cache_id) {
				acpi_put_table(table);
				return level;
			}
		}
	}
	return -ENOENT;
}

A less 'fun' alternative is pull some code out as a helper to make put the get and put near each other with no conditionals to confuse things.

static int __find_acpi_cache_level_from_id(u32 cache_id, struct acpi_table_header *head);

{
	u32 acpi_cpu_id;
	int level, cpu, num_levels;
	struct acpi_pptt_cache *cache;
	struct acpi_pptt_cache_v1 *cache_v1;
	struct acpi_pptt_processor *cpu_node;

	if (table->revision < 3)
		return -ENOENT;

	/*
	 * If we found the cache first, we'd still need to walk from each CPU
	 * to find the level...
	 */
	for_each_possible_cpu(cpu) {
		acpi_cpu_id = get_acpi_id_for_cpu(cpu);
		cpu_node = acpi_find_processor_node(table, acpi_cpu_id);
		if (!cpu_node)
			return -ENOENT;
		acpi_count_levels(table, cpu_node, &num_levels, NULL);

		/* Start at 1 for L1 */
		for (level = 1; level <= num_levels; level++) {
			cache = acpi_find_cache_node(table, acpi_cpu_id,
						     ACPI_PPTT_CACHE_TYPE_UNIFIED,
						     level, &cpu_node);
			if (!cache)
				continue;

			cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1,
						cache,
						sizeof(struct acpi_pptt_cache));

			if (cache->flags & ACPI_PPTT_CACHE_ID_VALID &&
			    cache_v1->cache_id == cache_id)
				return level;
		}
	}

	return -ENOENT;
}

int find_acpi_cache_level_from_id(u32 cache_id)

{
	int ret;
	acpi_status status;
	struct acpi_table_header *table;

	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table);
	if (ACPI_FAILURE(status)) {
		acpi_pptt_warn_missing();
		return -ENOENT;
	}

	ret = __find_acpi_cache_level_from_id(cache_id, table)
	acpi_put_table(table);
	return ret;
}
Quoted 61 lines
drivers/acpi/pptt.c +38 -0
+		acpi_put_table(table);
+		return -ENOENT;
+	}
+
+	/*
+	 * If we found the cache first, we'd still need to walk from each CPU
+	 * to find the level...
+	 */
+	for_each_possible_cpu(cpu) {
+		acpi_cpu_id = get_acpi_id_for_cpu(cpu);
+		cpu_node = acpi_find_processor_node(table, acpi_cpu_id);
+		if (!cpu_node)
+			break;
+		acpi_count_levels(table, cpu_node, &num_levels, NULL);
+
+		/* Start at 1 for L1 */
+		for (level = 1; level <= num_levels; level++) {
+			cache = acpi_find_cache_node(table, acpi_cpu_id,
+						     ACPI_PPTT_CACHE_TYPE_UNIFIED,
+						     level, &cpu_node);
+			if (!cache)
+				continue;
+
+			cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1,
+						cache,
+						sizeof(struct acpi_pptt_cache));
+
+			if (cache->flags & ACPI_PPTT_CACHE_ID_VALID &&
+			    cache_v1->cache_id == cache_id) {
+				acpi_put_table(table);
+				return level;
+			}
+		}
+	}
+
+	acpi_put_table(table);
+	return -ENOENT;
+}
include/linux/acpi.h +5 -0
index 8c3165c2b083..82947f6d2a43 100644
@@ -1542,6 +1542,7 @@ int find_acpi_cpu_topology_cluster(unsigned int cpu);
 int find_acpi_cpu_topology_package(unsigned int cpu);
 int find_acpi_cpu_topology_hetero_id(unsigned int cpu);
 int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus);
+int find_acpi_cache_level_from_id(u32 cache_id);
 #else
 static inline int acpi_pptt_cpu_is_thread(unsigned int cpu)
 {
@@ -1568,6 +1569,10 @@ static inline int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id,
 {
 	return -EINVAL;
 }
+static inline int find_acpi_cache_level_from_id(u32 cache_id)
+{
+	return -EINVAL;
+}
 #endif
 
 void acpi_arch_init(void);
author maintainer
#22 lore raw

Re: [RFC PATCH 07/36] ACPI / PPTT: Find cache level by cache-id

Hi Jonathan,

Quoted 14 lines · Jonathan Cameron
Quoted 11 lines · On Fri, 11 Jul 2025 18:36:19 +0000 James Morse <[email protected]> wrote:

The MPAM table identifies caches by id. The MPAM driver also wants to know the cache level to determine if the platform is of the shape that can be managed via resctrl. Cacheinfo has this information, but only for CPUs that are online.

Waiting for all CPUs to come online is a problem for platforms where CPUs are brought online late by user-space.

Add a helper that walks every possible cache, until it finds the one identified by cache-id, then return the level.

Quoted 7 lines
Quoted 3 lines

acpi_count_levels() expects its levels parameter to be initialised to zero as it passes it to acpi_find_cache_level() as starting_level. The existing callers do this. Document it.

A few suggestions inline. Mostly driven by the number of missing table puts I've seen in ACPI code. You don't have any missing here but with a bit of restructuring you can make that easy to see.

Sounds good,

Quoted 48 lines
Quoted 33 lines
drivers/acpi/pptt.c +25 -0
index 13ca2eee3b98..f53748a5df19 100644
@@ -912,3 +912,76 @@ int find_acpi_cpu_topology_hetero_id(unsigned int cpu)
 	return find_acpi_cpu_topology_tag(cpu, PPTT_ABORT_PACKAGE,
 					  ACPI_PPTT_ACPI_IDENTICAL);
 }
+
+/**
+ * find_acpi_cache_level_from_id() - Get the level of the specified cache
+ * @cache_id: The id field of the unified cache
+ *
+ * Determine the level relative to any CPU for the unified cache identified by
+ * cache_id. This allows the property to be found even if the CPUs are offline.
+ *
+ * The returned level can be used to group unified caches that are peers.
+ *
+ * The PPTT table must be rev 3 or later,
+ *
+ * If one CPUs L2 is shared with another as L3, this function will return
+ * an unpredictable value.
+ *
+ * Return: -ENOENT if the PPTT doesn't exist, or the cache cannot be found.
+ * Otherwise returns a value which represents the level of the specified cache.
+ */
+int find_acpi_cache_level_from_id(u32 cache_id)
+{
+	u32 acpi_cpu_id;
+	acpi_status status;
+	int level, cpu, num_levels;
+	struct acpi_pptt_cache *cache;
+	struct acpi_table_header *table;

drivers/acpi/pptt.c int find_acpi_cpu_topology_hetero_id(unsigned int cpu) · line 949

Quoted 10 lines
drivers/acpi/pptt.c +10 -0
+	struct acpi_pptt_cache_v1 *cache_v1;
+	struct acpi_pptt_processor *cpu_node;
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table);
+	if (ACPI_FAILURE(status)) {
+		acpi_pptt_warn_missing();
+		return -ENOENT;
+	}
+
+	if (table->revision < 3) {

Maybe a unified exit path given all paths need to do acpi_put_table() and return either error or level.

Quoted 2 lines

Or maybe it's time for some cleanup.h magic for acpi tables. I've been thinking about it for a while and mostly stuck on the name ;)

(Isn't that the hard bit?)

Quoted 70 lines

(simpler suggestion follows)

static struct acpi_table_header *acpi_get_table_ret(char *signature, u32 instance)

{
	struct acpi_table_header *table;
	int status = acpi_get_table(signature, instance, &table);
	
	if (ACPI_FAILURE(status))
		return ERR_PTR(-ENOENT);
	return table;
}

DEFINE_FREE(acpi_table, struct acpi_table_header *, if (!IS_ERR(_T)) acpi_put_table(_T))

Finally in here and loads of other places we avoid chance of missing an acpi_put_table and generally simplify the code a little.

int find_acpi_cache_level_from_id(u32 cache_id)

{
	u32 acpi_cpu_id;
	acpi_status status;
	int level, cpu, num_levels;
	struct acpi_pptt_cache *cache;
	struct acpi_pptt_cache_v1 *cache_v1;
	struct acpi_pptt_processor *cpu_node;


	struct acpi_table_header *table __free(acpi_table) =
		acpi_get_table_ret(ACPI_SIG_PPTT, 0);

	if (IS_ERR(table)
		return PTR_ERR(table);

	if (table->revision < 3)
		return -ENOENT;

	/*
	 * If we found the cache first, we'd still need to walk from each CPU
	 * to find the level...
	 */
	for_each_possible_cpu(cpu) {
		acpi_cpu_id = get_acpi_id_for_cpu(cpu);
		cpu_node = acpi_find_processor_node(table, acpi_cpu_id);
		if (!cpu_node)
			return -ENOENT;
		acpi_count_levels(table, cpu_node, &num_levels, NULL);

		/* Start at 1 for L1 */
		for (level = 1; level <= num_levels; level++) {
			cache = acpi_find_cache_node(table, acpi_cpu_id,
						     ACPI_PPTT_CACHE_TYPE_UNIFIED,
						     level, &cpu_node);
			if (!cache)
				continue;

			cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1,
						cache,
						sizeof(struct acpi_pptt_cache));

			if (cache->flags & ACPI_PPTT_CACHE_ID_VALID &&
			    cache_v1->cache_id == cache_id) {
				acpi_put_table(table);
				return level;
			}
		}
	}
	return -ENOENT;
}
Quoted 2 lines

A less 'fun' alternative is pull some code out as a helper to make put the get and put near each other with no conditionals to confuse things.

I still find the cleanup stuff slightly sickening ... so lets use it some more.

Added to linux/acpi.h to make it easier to use elsewhere. I think the earlier patches in this series are simple enough in this area its not worth changing them...

Thanks,

James

author maintainer
#23 lore raw

RFC PATCH 8/36 ACPI / PPTT: Add a helper to fill a cpumask from a cache_id

In reply to James Morse

MPAM identifies CPUs by the cache_id in the PPTT cache structure.

The driver needs to know which CPUs are associated with the cache, the CPUs may not all be online, so cacheinfo does not have the information.

Add a helper to pull this information out of the PPTT.

---
 drivers/acpi/pptt.c  | 70 ++++++++++++++++++++++++++++++++++++++++++++
 include/linux/acpi.h |  6 ++++
 2 files changed, 76 insertions(+)
drivers/acpi/pptt.c +70 -0
index f53748a5df19..81f7ac18c023 100644
@@ -985,3 +985,73 @@ int find_acpi_cache_level_from_id(u32 cache_id)
 	acpi_put_table(table);
 	return -ENOENT;
 }
+
+/**
+ * acpi_pptt_get_cpumask_from_cache_id() - Get the cpus associated with the
+ *					   specified cache
+ * @cache_id: The id field of the unified cache
+ * @cpus: Where to build the cpumask
+ *
+ * Determine which CPUs are below this cache in the PPTT. This allows the property
+ * to be found even if the CPUs are offline.
+ *
+ * The PPTT table must be rev 3 or later,
+ *
+ * Return: -ENOENT if the PPTT doesn't exist, or the cache cannot be found.
+ * Otherwise returns 0 and sets the cpus in the provided cpumask.
+ */
+int acpi_pptt_get_cpumask_from_cache_id(u32 cache_id, cpumask_t *cpus)
+{
+	u32 acpi_cpu_id;
+	acpi_status status;
+	int level, cpu, num_levels;
+	struct acpi_pptt_cache *cache;
+	struct acpi_table_header *table;
+	struct acpi_pptt_cache_v1 *cache_v1;
+	struct acpi_pptt_processor *cpu_node;
+
+	cpumask_clear(cpus);
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table);
+	if (ACPI_FAILURE(status)) {
+		acpi_pptt_warn_missing();
+		return -ENOENT;
+	}
+
+	if (table->revision < 3) {
+		acpi_put_table(table);
+		return -ENOENT;
+	}
+
+	/*
+	 * If we found the cache first, we'd still need to walk from each cpu.
+	 */
+	for_each_possible_cpu(cpu) {
+		acpi_cpu_id = get_acpi_id_for_cpu(cpu);
+		cpu_node = acpi_find_processor_node(table, acpi_cpu_id);
+		if (!cpu_node)
+			break;
+		acpi_count_levels(table, cpu_node, &num_levels, NULL);
+
+		/* Start at 1 for L1 */
+		for (level = 1; level <= num_levels; level++) {
+			cache = acpi_find_cache_node(table, acpi_cpu_id,
+						     ACPI_PPTT_CACHE_TYPE_UNIFIED,
+						     level, &cpu_node);
+			if (!cache)
+				continue;
+
+			cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1,
+						cache,
+						sizeof(struct acpi_pptt_cache));
+
+			if (cache->flags & ACPI_PPTT_CACHE_ID_VALID &&
+			    cache_v1->cache_id == cache_id) {
+				cpumask_set_cpu(cpu, cpus);
+			}
+		}
+	}
+
+	acpi_put_table(table);
+	return 0;
+}
include/linux/acpi.h +6 -0
index 82947f6d2a43..61ac3d1de1e8 100644
@@ -1543,6 +1543,7 @@ int find_acpi_cpu_topology_package(unsigned int cpu);
 int find_acpi_cpu_topology_hetero_id(unsigned int cpu);
 int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus);
 int find_acpi_cache_level_from_id(u32 cache_id);
+int acpi_pptt_get_cpumask_from_cache_id(u32 cache_id, cpumask_t *cpus);
 #else
 static inline int acpi_pptt_cpu_is_thread(unsigned int cpu)
 {
@@ -1573,6 +1574,11 @@ static inline int find_acpi_cache_level_from_id(u32 cache_id)
 {
 	return -EINVAL;
 }
+static inline int acpi_pptt_get_cpumask_from_cache_id(u32 cache_id,
+						      cpumask_t *cpus)
+{
+	return -EINVAL;
+}
 #endif
 
 void acpi_arch_init(void);
Signature · 1 line
2.39.5
#24 lore raw

Re: [RFC PATCH 08/36] ACPI / PPTT: Add a helper to fill a cpumask from a cache_id

Quoted 42 lines · On Fri, 11 Jul 2025 18:36:20 +0000 James Morse <[email protected]> wrote:

MPAM identifies CPUs by the cache_id in the PPTT cache structure.

The driver needs to know which CPUs are associated with the cache, the CPUs may not all be online, so cacheinfo does not have the information.

Add a helper to pull this information out of the PPTT.

---
 drivers/acpi/pptt.c  | 70 ++++++++++++++++++++++++++++++++++++++++++++
 include/linux/acpi.h |  6 ++++
 2 files changed, 76 insertions(+)
drivers/acpi/pptt.c +18 -0
index f53748a5df19..81f7ac18c023 100644
@@ -985,3 +985,73 @@ int find_acpi_cache_level_from_id(u32 cache_id)
 	acpi_put_table(table);
 	return -ENOENT;
 }
+
+/**
+ * acpi_pptt_get_cpumask_from_cache_id() - Get the cpus associated with the
+ *					   specified cache
+ * @cache_id: The id field of the unified cache
+ * @cpus: Where to build the cpumask
+ *
+ * Determine which CPUs are below this cache in the PPTT. This allows the property
+ * to be found even if the CPUs are offline.
+ *
+ * The PPTT table must be rev 3 or later,
+ *
+ * Return: -ENOENT if the PPTT doesn't exist, or the cache cannot be found.
+ * Otherwise returns 0 and sets the cpus in the provided cpumask.
+ */
+int acpi_pptt_get_cpumask_from_cache_id(u32 cache_id, cpumask_t *cpus)
+{
+	u32 acpi_cpu_id;

drivers/acpi/pptt.c int find_acpi_cache_level_from_id(u32 cache_id) · line 1015

Quoted 10 lines
drivers/acpi/pptt.c +10 -0
+	acpi_status status;
+	int level, cpu, num_levels;
+	struct acpi_pptt_cache *cache;
+	struct acpi_table_header *table;
+	struct acpi_pptt_cache_v1 *cache_v1;
+	struct acpi_pptt_processor *cpu_node;
+
+	cpumask_clear(cpus);
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table);

Similar suggestions to previous patch apply here as well.

Quoted 26 lines
drivers/acpi/pptt.c +26 -0
+	if (ACPI_FAILURE(status)) {
+		acpi_pptt_warn_missing();
+		return -ENOENT;
+	}
+
+	if (table->revision < 3) {
+		acpi_put_table(table);
+		return -ENOENT;
+	}
+
+	/*
+	 * If we found the cache first, we'd still need to walk from each cpu.
+	 */
+	for_each_possible_cpu(cpu) {
+		acpi_cpu_id = get_acpi_id_for_cpu(cpu);
+		cpu_node = acpi_find_processor_node(table, acpi_cpu_id);
+		if (!cpu_node)
+			break;
+		acpi_count_levels(table, cpu_node, &num_levels, NULL);
+
+		/* Start at 1 for L1 */
+		for (level = 1; level <= num_levels; level++) {
+			cache = acpi_find_cache_node(table, acpi_cpu_id,
+						     ACPI_PPTT_CACHE_TYPE_UNIFIED,
+						     level, &cpu_node);
+			if (!cache)

drivers/acpi/pptt.c int find_acpi_cache_level_from_id(u32 cache_id) · line 1051

Quoted 10 lines
drivers/acpi/pptt.c +10 -0
+				continue;
+
+			cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1,
+						cache,
+						sizeof(struct acpi_pptt_cache));
+
+			if (cache->flags & ACPI_PPTT_CACHE_ID_VALID &&
+			    cache_v1->cache_id == cache_id) {
+				cpumask_set_cpu(cpu, cpus);
+			}

Unnecessary {} Fine to keep them if you add something else here later.

Quoted 30 lines
drivers/acpi/pptt.c +6 -0
+		}
+	}
+
+	acpi_put_table(table);
+	return 0;
+}
include/linux/acpi.h +6 -0
index 82947f6d2a43..61ac3d1de1e8 100644
@@ -1543,6 +1543,7 @@ int find_acpi_cpu_topology_package(unsigned int cpu);
 int find_acpi_cpu_topology_hetero_id(unsigned int cpu);
 int acpi_pptt_get_cpus_from_container(u32 acpi_cpu_id, cpumask_t *cpus);
 int find_acpi_cache_level_from_id(u32 cache_id);
+int acpi_pptt_get_cpumask_from_cache_id(u32 cache_id, cpumask_t *cpus);
 #else
 static inline int acpi_pptt_cpu_is_thread(unsigned int cpu)
 {
@@ -1573,6 +1574,11 @@ static inline int find_acpi_cache_level_from_id(u32 cache_id)
 {
 	return -EINVAL;
 }
+static inline int acpi_pptt_get_cpumask_from_cache_id(u32 cache_id,
+						      cpumask_t *cpus)
+{
+	return -EINVAL;
+}
 #endif
 
 void acpi_arch_init(void);
author maintainer
#25 lore raw

Re: [RFC PATCH 08/36] ACPI / PPTT: Add a helper to fill a cpumask from a cache_id

Hi Jonathan,

Quoted 17 lines · Jonathan Cameron
Quoted 13 lines · On Fri, 11 Jul 2025 18:36:20 +0000 James Morse <[email protected]> wrote:

MPAM identifies CPUs by the cache_id in the PPTT cache structure.

The driver needs to know which CPUs are associated with the cache, the CPUs may not all be online, so cacheinfo does not have the information.

Add a helper to pull this information out of the PPTT.

drivers/acpi/pptt.c +0 -0
index f53748a5df19..81f7ac18c023 100644
@@ -985,3 +985,73 @@ int find_acpi_cache_level_from_id(u32 cache_id)
Quoted 15 lines

drivers/acpi/pptt.c int find_acpi_cache_level_from_id(u32 cache_id) · line 1015

Quoted 13 lines
drivers/acpi/pptt.c +13 -0
+int acpi_pptt_get_cpumask_from_cache_id(u32 cache_id, cpumask_t *cpus)
+{
+	u32 acpi_cpu_id;
+	acpi_status status;
+	int level, cpu, num_levels;
+	struct acpi_pptt_cache *cache;
+	struct acpi_table_header *table;
+	struct acpi_pptt_cache_v1 *cache_v1;
+	struct acpi_pptt_processor *cpu_node;
+
+	cpumask_clear(cpus);
+
+	status = acpi_get_table(ACPI_SIG_PPTT, 0, &table);

Similar suggestions to previous patch apply here as well.

Done!

Quoted 31 lines
Quoted 31 lines
drivers/acpi/pptt.c +30 -0
+	if (ACPI_FAILURE(status)) {
+		acpi_pptt_warn_missing();
+		return -ENOENT;
+	}
+
+	if (table->revision < 3) {
+		acpi_put_table(table);
+		return -ENOENT;
+	}
+
+	/*
+	 * If we found the cache first, we'd still need to walk from each cpu.
+	 */
+	for_each_possible_cpu(cpu) {
+		acpi_cpu_id = get_acpi_id_for_cpu(cpu);
+		cpu_node = acpi_find_processor_node(table, acpi_cpu_id);
+		if (!cpu_node)
+			break;
+		acpi_count_levels(table, cpu_node, &num_levels, NULL);
+
+		/* Start at 1 for L1 */
+		for (level = 1; level <= num_levels; level++) {
+			cache = acpi_find_cache_node(table, acpi_cpu_id,
+						     ACPI_PPTT_CACHE_TYPE_UNIFIED,
+						     level, &cpu_node);
+			if (!cache)
+				continue;
+
+			cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1,
+						cache,
Quoted 8 lines

drivers/acpi/pptt.c int find_acpi_cache_level_from_id(u32 cache_id) · line 1051

Quoted 6 lines
drivers/acpi/pptt.c +6 -0
+						sizeof(struct acpi_pptt_cache));
+
+			if (cache->flags & ACPI_PPTT_CACHE_ID_VALID &&
+			    cache_v1->cache_id == cache_id) {
+				cpumask_set_cpu(cpu, cpus);
+			}

Unnecessary {} Fine to keep them if you add something else here later.

The condition being broken over multiple lines de-rails my C parsing abilities... 'Fixed'.

Thanks,

James

author maintainer
#26 lore raw

RFC PATCH 9/36 arm64: kconfig: Add Kconfig entry for MPAM

In reply to James Morse

The bulk of the MPAM driver lives outside the arch code because it largely manages MMIO devices that generate interrupts. The driver needs a Kconfig symbol to enable it, as MPAM is only found on arm64 platforms, that is where the Kconfig option makes the most sense.

This Kconfig option will later be used by the arch code to enable or disable the MPAM context-switch code, and registering the CPUs properties with the MPAM driver.

---
 arch/arm64/Kconfig | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
arch/arm64/Kconfig +17 -0
index 55fc331af337..5f08214537d0 100644
@@ -2058,6 +2058,23 @@ config ARM64_TLB_RANGE
 	  ARMv8.4-TLBI provides TLBI invalidation instruction that apply to a
 	  range of input addresses.
 
+config ARM64_MPAM
+	bool "Enable support for MPAM"
+	help
+	  Memory Partitioning and Monitoring is an optional extension
+	  that allows the CPUs to mark load and store transactions with
+	  labels for partition-id and performance-monitoring-group.
+	  System components, such as the caches, can use the partition-id
+	  to apply a performance policy. MPAM monitors can use the
+	  partition-id and performance-monitoring-group to measure the
+	  cache occupancy or data throughput.
+
+	  Use of this extension requires CPU support, support in the
+	  memory system components (MSC), and a description from firmware
+	  of where the MSC are in the address space.
+
+	  MPAM is exposed to user-space via the resctrl pseudo filesystem.
+
 endmenu # "ARMv8.4 architectural features"
 
 menu "ARMv8.5 architectural features"
Signature · 1 line
2.39.5
Reviewed-by on v1 9/36
#27 lore raw

Re: [RFC PATCH 09/36] arm64: kconfig: Add Kconfig entry for MPAM

Quoted 10 lines · On Fri, 11 Jul 2025 18:36:21 +0000 James Morse <[email protected]> wrote:

The bulk of the MPAM driver lives outside the arch code because it largely manages MMIO devices that generate interrupts. The driver needs a Kconfig symbol to enable it, as MPAM is only found on arm64 platforms, that is where the Kconfig option makes the most sense.

This Kconfig option will later be used by the arch code to enable or disable the MPAM context-switch code, and registering the CPUs properties with the MPAM driver.

Seems like a reasonable help test so FWIW

Quoted 32 lines
---
 arch/arm64/Kconfig | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
arch/arm64/Kconfig +17 -0
index 55fc331af337..5f08214537d0 100644
@@ -2058,6 +2058,23 @@ config ARM64_TLB_RANGE
 	  ARMv8.4-TLBI provides TLBI invalidation instruction that apply to a
 	  range of input addresses.
 
+config ARM64_MPAM
+	bool "Enable support for MPAM"
+	help
+	  Memory Partitioning and Monitoring is an optional extension
+	  that allows the CPUs to mark load and store transactions with
+	  labels for partition-id and performance-monitoring-group.
+	  System components, such as the caches, can use the partition-id
+	  to apply a performance policy. MPAM monitors can use the
+	  partition-id and performance-monitoring-group to measure the
+	  cache occupancy or data throughput.
+
+	  Use of this extension requires CPU support, support in the
+	  memory system components (MSC), and a description from firmware
+	  of where the MSC are in the address space.
+
+	  MPAM is exposed to user-space via the resctrl pseudo filesystem.
+
 endmenu # "ARMv8.4 architectural features"
 
 menu "ARMv8.5 architectural features"
author maintainer
#28 lore raw

RFC PATCH 10/36 ACPI / MPAM: Parse the MPAM table

In reply to James Morse

Add code to parse the arm64 specific MPAM table, looking up the cache level from the PPTT and feeding the end result into the MPAM driver.

---
 arch/arm64/Kconfig          |   1 +
 drivers/acpi/arm64/Kconfig  |   3 +
 drivers/acpi/arm64/Makefile |   1 +
 drivers/acpi/arm64/mpam.c   | 365 ++++++++++++++++++++++++++++++++++++
 drivers/acpi/tables.c       |   2 +-
 include/linux/arm_mpam.h    |  46 +++++
 6 files changed, 417 insertions(+), 1 deletion(-)
 create mode 100644 drivers/acpi/arm64/mpam.c
 create mode 100644 include/linux/arm_mpam.h
arch/arm64/Kconfig +1 -0
index 5f08214537d0..ad9a49a39e41 100644
@@ -2060,6 +2060,7 @@ config ARM64_TLB_RANGE
 
 config ARM64_MPAM
 	bool "Enable support for MPAM"
+	select ACPI_MPAM if ACPI
 	help
 	  Memory Partitioning and Monitoring is an optional extension
 	  that allows the CPUs to mark load and store transactions with
drivers/acpi/arm64/Kconfig +3 -0
index b3ed6212244c..f2fd79f22e7d 100644
@@ -21,3 +21,6 @@ config ACPI_AGDI
 
 config ACPI_APMT
 	bool
+
+config ACPI_MPAM
+	bool
drivers/acpi/arm64/Makefile +1 -0
index 05ecde9eaabe..27b872249baa 100644
@@ -6,5 +6,6 @@ obj-$(CONFIG_ACPI_GTDT) 	+= gtdt.o
 obj-$(CONFIG_ACPI_IORT) 	+= iort.o
 obj-$(CONFIG_ACPI_PROCESSOR_IDLE) += cpuidle.o
 obj-$(CONFIG_ARM_AMBA)		+= amba.o
+obj-$(CONFIG_ACPI_MPAM) 	+= mpam.o
 obj-y				+= dma.o init.o
 obj-y				+= thermal_cpufreq.o
drivers/acpi/arm64/mpam.c new file +365 -0

365 lines not on this page · raw message

drivers/acpi/tables.c +1 -1
index fa9bb8c8ce95..835e3795ede3 100644
@@ -408,7 +408,7 @@ static const char table_sigs[][ACPI_NAMESEG_SIZE] __nonstring_array __initconst
 	ACPI_SIG_PSDT, ACPI_SIG_RSDT, ACPI_SIG_XSDT, ACPI_SIG_SSDT,
 	ACPI_SIG_IORT, ACPI_SIG_NFIT, ACPI_SIG_HMAT, ACPI_SIG_PPTT,
 	ACPI_SIG_NHLT, ACPI_SIG_AEST, ACPI_SIG_CEDT, ACPI_SIG_AGDI,
-	ACPI_SIG_NBFT };
+	ACPI_SIG_NBFT, ACPI_SIG_MPAM };
 
 #define ACPI_HEADER_SIZE sizeof(struct acpi_table_header)
 
include/linux/arm_mpam.h new file +46 -0

46 lines not on this page · raw message

Signature · 1 line
2.39.5
#29 lore raw

Re: [RFC PATCH 10/36] ACPI / MPAM: Parse the MPAM table

Quoted 2 lines · On Fri, 11 Jul 2025 18:36:22 +0000 James Morse <[email protected]> wrote:

Add code to parse the arm64 specific MPAM table, looking up the cache level from the PPTT and feeding the end result into the MPAM driver.

Throw in a link to the spec perhaps? Particularly useful to know which version this was written against when reviewing it.

Quoted 20 lines
---
 arch/arm64/Kconfig          |   1 +
 drivers/acpi/arm64/Kconfig  |   3 +
 drivers/acpi/arm64/Makefile |   1 +
 drivers/acpi/arm64/mpam.c   | 365 ++++++++++++++++++++++++++++++++++++
 drivers/acpi/tables.c       |   2 +-
 include/linux/arm_mpam.h    |  46 +++++
 6 files changed, 417 insertions(+), 1 deletion(-)
 create mode 100644 drivers/acpi/arm64/mpam.c
 create mode 100644 include/linux/arm_mpam.h
drivers/acpi/arm64/Makefile +0 -0
index 05ecde9eaabe..27b872249baa 100644

drivers/acpi/arm64/Makefile obj-$(CONFIG_ACPI_GTDT) += gtdt.o · line 9

Quoted 4 lines
drivers/acpi/arm64/Makefile +1 -0
@@ -6,5 +6,6 @@ obj-$(CONFIG_ACPI_GTDT) 	+= gtdt.o
 obj-$(CONFIG_ACPI_IORT) 	+= iort.o
 obj-$(CONFIG_ACPI_PROCESSOR_IDLE) += cpuidle.o
 obj-$(CONFIG_ARM_AMBA)		+= amba.o
+obj-$(CONFIG_ACPI_MPAM) 	+= mpam.o

Keep it with the ACPI ones? There doesn't seem to be a lot of order in here though so I guess maybe there is logic behind putting it here I'm missing.

Quoted 14 lines
drivers/acpi/arm64/Makefile +0 -0
 obj-y				+= dma.o init.o
 obj-y				+= thermal_cpufreq.o
drivers/acpi/arm64/mpam.c new file +6 -0
new file mode 100644
index 000000000000..f4791bac9a2a
@@ -0,0 +1,365 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (C) 2025 Arm Ltd.
+
+/* Parse the MPAM ACPI table feeding the discovered nodes into the driver */
+
+#define pr_fmt(fmt) "ACPI MPAM: " fmt

drivers/acpi/arm64/mpam.c line 16

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+
+#include <linux/acpi.h>
+#include <linux/arm_mpam.h>
+#include <linux/cpu.h>
+#include <linux/cpumask.h>
+#include <linux/platform_device.h>
+
+#include <acpi/processor.h>
+
+/* Flags for acpi_table_mpam_msc.*_interrupt_flags */

References.. I'm looking at 3.0-alpha table 5 to check this. I can see why you might be reluctant to point at an alpha if that is what you are using ;)

drivers/acpi/arm64/mpam.c line 18

Quoted 2 lines
drivers/acpi/arm64/mpam.c new file +2 -0
+#define ACPI_MPAM_MSC_IRQ_MODE_EDGE                    1
+#define ACPI_MPAM_MSC_IRQ_TYPE_MASK                    (3 << 1)

GENMASK(3, 2) would be my preference for how to do masks in new code.

Quoted 4 lines
drivers/acpi/arm64/mpam.c new file +4 -0
+#define ACPI_MPAM_MSC_IRQ_TYPE_WIRED                   0
+#define ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER BIT(3)
+#define ACPI_MPAM_MSC_IRQ_AFFINITY_VALID               BIT(4)
+

drivers/acpi/arm64/mpam.c line 32

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+static bool frob_irq(struct platform_device *pdev, int intid, u32 flags,
+		     int *irq, u32 processor_container_uid)
+{
+	int sense;
+
+	if (!intid)
+		return false;
+
+	/* 0 in this field indicates a wired interrupt */
+	if (flags & ACPI_MPAM_MSC_IRQ_TYPE_MASK)

I'd prefer more explicit code (and probably no comment)

	if (FIELD_GET(flags, ACPI_MPAM_MSC_IRQ_TYPE_MASK) !=
	    ACPI_MPAM_MSC_IRQ_TYPE_WIRED)
		return false;

drivers/acpi/arm64/mpam.c line 38

Quoted 6 lines
drivers/acpi/arm64/mpam.c new file +6 -0
+		return false;
+
+	if (flags & ACPI_MPAM_MSC_IRQ_MODE_EDGE)
+		sense = ACPI_EDGE_SENSITIVE;
+	else
+		sense = ACPI_LEVEL_SENSITIVE;

If the spec is supposed to be using standard ACPI_* types for this field (I don't think the connection is explicitly documented though) then

	sense = FIELD_GET(flags, ACPI_MPAM_MSC_IRQ_MODE_MASK); 

Assuming a change to define the mask and rely on the ACPI defs for the values

This one is entirely up to you.

Quoted 21 lines
drivers/acpi/arm64/mpam.c new file +21 -0
+
+	/*
+	 * If the GSI is in the GIC's PPI range, try and create a partitioned
+	 * percpu interrupt.
+	 */
+	if (16 <= intid && intid < 32 && processor_container_uid != ~0) {
+		pr_err_once("Partitioned interrupts not supported\n");
+		return false;
+	}
+
+	*irq = acpi_register_gsi(&pdev->dev, intid, sense, ACPI_ACTIVE_HIGH);
+	if (*irq <= 0) {
+		pr_err_once("Failed to register interrupt 0x%x with ACPI\n",
+			    intid);
+		return false;
+	}
+
+	return true;
+}
+
+static void acpi_mpam_parse_irqs(struct platform_device *pdev,

drivers/acpi/arm64/mpam.c line 69

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+				 struct acpi_mpam_msc_node *tbl_msc,
+				 struct resource *res, int *res_idx)
+{
+	u32 flags, aff = ~0;
+	int irq;
+
+	flags = tbl_msc->overflow_interrupt_flags;
+	if (flags & ACPI_MPAM_MSC_IRQ_AFFINITY_VALID &&
+	    flags & ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER)
+		aff = tbl_msc->overflow_interrupt_affinity;

Just to make the two cases look the same I'd do

	else
		aff = ~0;

here as well and not initialize above. It's not quite worth using a helper function for these two identical blocks but it's close.

drivers/acpi/arm64/mpam.c line 74

Quoted 5 lines
drivers/acpi/arm64/mpam.c new file +5 -0
+	if (frob_irq(pdev, tbl_msc->overflow_interrupt, flags, &irq, aff)) {
+		res[*res_idx].start = irq;
+		res[*res_idx].end = irq;
+		res[*res_idx].flags = IORESOURCE_IRQ;
+		res[*res_idx].name = "overflow";
		res[*res_idx] = DEFINE_RES_IRQ_NAMED(irq, 1, "overflow");

drivers/acpi/arm64/mpam.c line 76

Quoted 2 lines
drivers/acpi/arm64/mpam.c new file +2 -0
+
+		(*res_idx)++;

Can roll this in as well.

drivers/acpi/arm64/mpam.c line 89

Quoted 13 lines
drivers/acpi/arm64/mpam.c new file +13 -0
+	}
+
+	flags = tbl_msc->error_interrupt_flags;
+	if (flags & ACPI_MPAM_MSC_IRQ_AFFINITY_VALID &&
+	    flags & ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER)
+		aff = tbl_msc->error_interrupt_affinity;
+	else
+		aff = ~0;
+	if (frob_irq(pdev, tbl_msc->error_interrupt, flags, &irq, aff)) {
+		res[*res_idx].start = irq;
+		res[*res_idx].end = irq;
+		res[*res_idx].flags = IORESOURCE_IRQ;
+		res[*res_idx].name = "error";

Similar to above.

Quoted 18 lines
drivers/acpi/arm64/mpam.c new file +16 -0
+
+		(*res_idx)++;
+	}
+}
+
 
 
+static bool __init parse_msc_pm_link(struct acpi_mpam_msc_node *tbl_msc,
+				     struct platform_device *pdev,
+				     u32 *acpi_id)
+{
+	bool acpi_id_valid = false;
+	struct acpi_device *buddy;
+	char hid[16], uid[16];
+	int err;
+
+	memset(&hid, 0, sizeof(hid));
+	memcpy(hid, &tbl_msc->hardware_id_linked_device,

drivers/acpi/arm64/mpam.c line 159

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+	       sizeof(tbl_msc->hardware_id_linked_device));
+
+	if (!strcmp(hid, ACPI_PROCESSOR_CONTAINER_HID)) {
+		*acpi_id = tbl_msc->instance_id_linked_device;
+		acpi_id_valid = true;
+	}
+
+	err = snprintf(uid, sizeof(uid), "%u",
+		       tbl_msc->instance_id_linked_device);
+	if (err < 0 || err >= sizeof(uid))

Does snprintf() ever return < 0 ? It's documented as returning number of chars printed (without the NULL) so that can only be 0 or greater.

Can it return >= sizeof(uid) ? Looks odd.

+ return acpi_id_valid;

Quoted 38 lines
drivers/acpi/arm64/mpam.c new file +37 -0
+
+	buddy = acpi_dev_get_first_match_dev(hid, uid, -1);
+	if (buddy)
+		device_link_add(&pdev->dev, &buddy->dev, DL_FLAG_STATELESS);
+
+	return acpi_id_valid;
+}
 
+static int __init _parse_table(struct acpi_table_header *table)
+{
+	char *table_end, *table_offset = (char *)(table + 1);
+	struct property_entry props[4]; /* needs a sentinel */
+	struct acpi_mpam_msc_node *tbl_msc;
+	int next_res, next_prop, err = 0;
+	struct acpi_device *companion;
+	struct platform_device *pdev;
+	enum mpam_msc_iface iface;
+	struct resource res[3];
+	char uid[16];
+	u32 acpi_id;
+
+	table_end = (char *)table + table->length;
+
+	while (table_offset < table_end) {
+		tbl_msc = (struct acpi_mpam_msc_node *)table_offset;
+		table_offset += tbl_msc->length;
+
+		/*
+		 * If any of the reserved fields are set, make no attempt to
+		 * parse the msc structure. This will prevent the driver from
+		 * probing all the MSC, meaning it can't discover the system
+		 * wide supported partid and pmg ranges. This avoids whatever
+		 * this MSC is truncating the partids and creating a screaming
+		 * error interrupt.
+		 */
+		if (tbl_msc->reserved || tbl_msc->reserved1 || tbl_msc->reserved2)
+			continue;
+

drivers/acpi/arm64/mpam.c line 223

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+		if (decode_interface_type(tbl_msc, &iface))
+			continue;
+
+		next_res = 0;
+		next_prop = 0;
+		memset(res, 0, sizeof(res));
+		memset(props, 0, sizeof(props));
+
+		pdev = platform_device_alloc("mpam_msc", tbl_msc->identifier);
+		if (IS_ERR(pdev)) {

returns NULL in at least some error cases (probably all, I'm just to lazy to check)

Quoted 13 lines
drivers/acpi/arm64/mpam.c new file +13 -0
+			err = PTR_ERR(pdev);
+			break;
+		}
+
+		if (tbl_msc->length < sizeof(*tbl_msc)) {
+			err = -EINVAL;
+			break;
+		}
+
+		/* Some power management is described in the namespace: */
+		err = snprintf(uid, sizeof(uid), "%u", tbl_msc->identifier);
+		if (err > 0 && err < sizeof(uid)) {
+			companion = acpi_dev_get_first_match_dev("ARMHAA5C", uid, -1);

drivers/acpi/arm64/mpam.c line 246

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+			if (companion)
+				ACPI_COMPANION_SET(&pdev->dev, companion);
+		}
+
+		if (iface == MPAM_IFACE_MMIO) {
+			res[next_res].name = "MPAM:MSC";
+			res[next_res].start = tbl_msc->base_address;
+			res[next_res].end = tbl_msc->base_address + tbl_msc->mmio_size - 1;
+			res[next_res].flags = IORESOURCE_MEM;
+			next_res++;

DEFINE_RES_MEM_NAMED()?

Quoted 25 lines
drivers/acpi/arm64/mpam.c new file +25 -0
+		} else if (iface == MPAM_IFACE_PCC) {
+			props[next_prop++] = PROPERTY_ENTRY_U32("pcc-channel",
+								tbl_msc->base_address);
+			next_prop++;
+		}
+
+		acpi_mpam_parse_irqs(pdev, tbl_msc, res, &next_res);
+		err = platform_device_add_resources(pdev, res, next_res);
+		if (err)
+			break;
+
+		props[next_prop++] = PROPERTY_ENTRY_U32("arm,not-ready-us",
+							tbl_msc->max_nrdy_usec);
+
+		/*
+		 * The MSC's CPU affinity is described via its linked power
+		 * management device, but only if it points at a Processor or
+		 * Processor Container.
+		 */
+		if (parse_msc_pm_link(tbl_msc, pdev, &acpi_id)) {
+			props[next_prop++] = PROPERTY_ENTRY_U32("cpu_affinity",
+								acpi_id);
+		}
+
+		err = device_create_managed_software_node(&pdev->dev, props,

drivers/acpi/arm64/mpam.c line 281

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+							  NULL);
+		if (err)
+			break;
+
+		/* Come back later if you want the RIS too */
+		err = platform_device_add_data(pdev, tbl_msc, tbl_msc->length);
+		if (err)
+			break;
+
+		platform_device_add(pdev);

Can fail.

Quoted 24 lines
drivers/acpi/arm64/mpam.c new file +24 -0
+	}
+
+	if (err)
+		platform_device_put(pdev);
+
+	return err;
+}
+
+static struct acpi_table_header *get_table(void)
+{
+	struct acpi_table_header *table;
+	acpi_status status;
+
+	if (acpi_disabled || !system_supports_mpam())
+		return NULL;
+
+	status = acpi_get_table(ACPI_SIG_MPAM, 0, &table);
+	if (ACPI_FAILURE(status))
+		return NULL;
+
+	if (table->revision != 1)
+		return NULL;
+
+	return table;

drivers/acpi/arm64/mpam.c line 315

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+}
+
+static int __init acpi_mpam_parse(void)
+{
+	struct acpi_table_header *mpam;
+	int err;
+
+	mpam = get_table();
+	if (!mpam)
+		return 0;

Just what I was suggesting for the PPTT case in early patches. Nice :)

drivers/acpi/arm64/mpam.c line 327

Quoted 12 lines
drivers/acpi/arm64/mpam.c new file +12 -0
+
+	err = _parse_table(mpam);
+	acpi_put_table(mpam);
+
+	return err;
+}
+
+static int _count_msc(struct acpi_table_header *table)
+{
+	char *table_end, *table_offset = (char *)(table + 1);
+	struct acpi_mpam_msc_node *tbl_msc;
+	int ret = 0;

Call it count as it only ever contains the count?

drivers/acpi/arm64/mpam.c line 336

Quoted 9 lines
drivers/acpi/arm64/mpam.c new file +9 -0
+
+	tbl_msc = (struct acpi_mpam_msc_node *)table_offset;
+	table_end = (char *)table + table->length;
+
+	while (table_offset < table_end) {
+		if (tbl_msc->length < sizeof(*tbl_msc))
+			return -EINVAL;
+
+		ret++;

count++ would feel more natural here.

drivers/acpi/arm64/mpam.c line 343

Quoted 7 lines
drivers/acpi/arm64/mpam.c new file +7 -0
+
+		table_offset += tbl_msc->length;
+		tbl_msc = (struct acpi_mpam_msc_node *)table_offset;
+	}
+
+	return ret;
+}

That's all I have time for today. Will get to the rest of the series soonish.

Jonathan

maintainer
#30 lore raw

Re: [RFC PATCH 10/36] ACPI / MPAM: Parse the MPAM table

Hi James, Jonathan,

Quoted 8 lines · Jonathan Cameron
Quoted 2 lines · On Fri, 11 Jul 2025 18:36:22 +0000 James Morse <[email protected]> wrote:

Add code to parse the arm64 specific MPAM table, looking up the cache level from the PPTT and feeding the end result into the MPAM driver.

Throw in a link to the spec perhaps? Particularly useful to know which version this was written against when reviewing it.

As I comment below this code checks the table revision is 1 and so we can assume it was written against version 2 of the spec. As of Monday, there is a new version hot off the press,
https://developer.arm.com/documentation/den0065/3-0bet/?lang=en which introduces an "MMIO size" field to allow for disabled nodes. This should be considered here to avoid advertising msc that aren't present.

Quoted 330 lines
Quoted 21 lines
---
  arch/arm64/Kconfig          |   1 +
  drivers/acpi/arm64/Kconfig  |   3 +
  drivers/acpi/arm64/Makefile |   1 +
  drivers/acpi/arm64/mpam.c   | 365 ++++++++++++++++++++++++++++++++++++
  drivers/acpi/tables.c       |   2 +-
  include/linux/arm_mpam.h    |  46 +++++
  6 files changed, 417 insertions(+), 1 deletion(-)

create mode 100644 drivers/acpi/arm64/mpam.c
create mode 100644 include/linux/arm_mpam.h

drivers/acpi/arm64/Makefile +0 -0
index 05ecde9eaabe..27b872249baa 100644

drivers/acpi/arm64/Makefile obj-$(CONFIG_ACPI_GTDT) += gtdt.o · line 9

Quoted 4 lines
drivers/acpi/arm64/Makefile +1 -0
@@ -6,5 +6,6 @@ obj-$(CONFIG_ACPI_GTDT) 	+= gtdt.o
  obj-$(CONFIG_ACPI_IORT) 	+= iort.o
  obj-$(CONFIG_ACPI_PROCESSOR_IDLE) += cpuidle.o
  obj-$(CONFIG_ARM_AMBA)		+= amba.o
+obj-$(CONFIG_ACPI_MPAM) 	+= mpam.o

Keep it with the ACPI ones? There doesn't seem to be a lot of order in here though so I guess maybe there is logic behind putting it here I'm missing.

Quoted 14 lines
drivers/acpi/arm64/Makefile +0 -0
  obj-y				+= dma.o init.o
  obj-y				+= thermal_cpufreq.o
drivers/acpi/arm64/mpam.c new file +6 -0
new file mode 100644
index 000000000000..f4791bac9a2a
@@ -0,0 +1,365 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (C) 2025 Arm Ltd.
+
+/* Parse the MPAM ACPI table feeding the discovered nodes into the driver */
+
+#define pr_fmt(fmt) "ACPI MPAM: " fmt

drivers/acpi/arm64/mpam.c line 16

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+
+#include <linux/acpi.h>
+#include <linux/arm_mpam.h>
+#include <linux/cpu.h>
+#include <linux/cpumask.h>
+#include <linux/platform_device.h>
+
+#include <acpi/processor.h>
+
+/* Flags for acpi_table_mpam_msc.*_interrupt_flags */

References.. I'm looking at 3.0-alpha table 5 to check this. I can see why you might be reluctant to point at an alpha if that is what you are using ;)

drivers/acpi/arm64/mpam.c line 18

Quoted 2 lines
drivers/acpi/arm64/mpam.c new file +2 -0
+#define ACPI_MPAM_MSC_IRQ_MODE_EDGE                    1
+#define ACPI_MPAM_MSC_IRQ_TYPE_MASK                    (3 << 1)

GENMASK(3, 2) would be my preference for how to do masks in new code.

Quoted 4 lines
drivers/acpi/arm64/mpam.c new file +4 -0
+#define ACPI_MPAM_MSC_IRQ_TYPE_WIRED                   0
+#define ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER BIT(3)
+#define ACPI_MPAM_MSC_IRQ_AFFINITY_VALID               BIT(4)
+

drivers/acpi/arm64/mpam.c line 32

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+static bool frob_irq(struct platform_device *pdev, int intid, u32 flags,
+		     int *irq, u32 processor_container_uid)
+{
+	int sense;
+
+	if (!intid)
+		return false;
+
+	/* 0 in this field indicates a wired interrupt */
+	if (flags & ACPI_MPAM_MSC_IRQ_TYPE_MASK)

I'd prefer more explicit code (and probably no comment)

	if (FIELD_GET(flags, ACPI_MPAM_MSC_IRQ_TYPE_MASK) !=
	    ACPI_MPAM_MSC_IRQ_TYPE_WIRED)
		return false;

drivers/acpi/arm64/mpam.c line 38

Quoted 6 lines
drivers/acpi/arm64/mpam.c new file +6 -0
+		return false;
+
+	if (flags & ACPI_MPAM_MSC_IRQ_MODE_EDGE)
+		sense = ACPI_EDGE_SENSITIVE;
+	else
+		sense = ACPI_LEVEL_SENSITIVE;

If the spec is supposed to be using standard ACPI_* types for this field (I don't think the connection is explicitly documented though) then

	sense = FIELD_GET(flags, ACPI_MPAM_MSC_IRQ_MODE_MASK);

Assuming a change to define the mask and rely on the ACPI defs for the values

This one is entirely up to you.

Quoted 21 lines
drivers/acpi/arm64/mpam.c new file +21 -0
+
+	/*
+	 * If the GSI is in the GIC's PPI range, try and create a partitioned
+	 * percpu interrupt.
+	 */
+	if (16 <= intid && intid < 32 && processor_container_uid != ~0) {
+		pr_err_once("Partitioned interrupts not supported\n");
+		return false;
+	}
+
+	*irq = acpi_register_gsi(&pdev->dev, intid, sense, ACPI_ACTIVE_HIGH);
+	if (*irq <= 0) {
+		pr_err_once("Failed to register interrupt 0x%x with ACPI\n",
+			    intid);
+		return false;
+	}
+
+	return true;
+}
+
+static void acpi_mpam_parse_irqs(struct platform_device *pdev,

drivers/acpi/arm64/mpam.c line 69

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+				 struct acpi_mpam_msc_node *tbl_msc,
+				 struct resource *res, int *res_idx)
+{
+	u32 flags, aff = ~0;
+	int irq;
+
+	flags = tbl_msc->overflow_interrupt_flags;
+	if (flags & ACPI_MPAM_MSC_IRQ_AFFINITY_VALID &&
+	    flags & ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER)
+		aff = tbl_msc->overflow_interrupt_affinity;

Just to make the two cases look the same I'd do

	else
		aff = ~0;

here as well and not initialize above. It's not quite worth using a helper function for these two identical blocks but it's close.

drivers/acpi/arm64/mpam.c line 74

Quoted 5 lines
drivers/acpi/arm64/mpam.c new file +5 -0
+	if (frob_irq(pdev, tbl_msc->overflow_interrupt, flags, &irq, aff)) {
+		res[*res_idx].start = irq;
+		res[*res_idx].end = irq;
+		res[*res_idx].flags = IORESOURCE_IRQ;
+		res[*res_idx].name = "overflow";
		res[*res_idx] = DEFINE_RES_IRQ_NAMED(irq, 1, "overflow");

drivers/acpi/arm64/mpam.c line 76

Quoted 2 lines
drivers/acpi/arm64/mpam.c new file +2 -0
+
+		(*res_idx)++;

Can roll this in as well.

drivers/acpi/arm64/mpam.c line 89

Quoted 13 lines
drivers/acpi/arm64/mpam.c new file +13 -0
+	}
+
+	flags = tbl_msc->error_interrupt_flags;
+	if (flags & ACPI_MPAM_MSC_IRQ_AFFINITY_VALID &&
+	    flags & ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER)
+		aff = tbl_msc->error_interrupt_affinity;
+	else
+		aff = ~0;
+	if (frob_irq(pdev, tbl_msc->error_interrupt, flags, &irq, aff)) {
+		res[*res_idx].start = irq;
+		res[*res_idx].end = irq;
+		res[*res_idx].flags = IORESOURCE_IRQ;
+		res[*res_idx].name = "error";

Similar to above.

Quoted 18 lines
drivers/acpi/arm64/mpam.c new file +16 -0
+
+		(*res_idx)++;
+	}
+}
+
 
 
+static bool __init parse_msc_pm_link(struct acpi_mpam_msc_node *tbl_msc,
+				     struct platform_device *pdev,
+				     u32 *acpi_id)
+{
+	bool acpi_id_valid = false;
+	struct acpi_device *buddy;
+	char hid[16], uid[16];
+	int err;
+
+	memset(&hid, 0, sizeof(hid));
+	memcpy(hid, &tbl_msc->hardware_id_linked_device,

drivers/acpi/arm64/mpam.c line 159

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+	       sizeof(tbl_msc->hardware_id_linked_device));
+
+	if (!strcmp(hid, ACPI_PROCESSOR_CONTAINER_HID)) {
+		*acpi_id = tbl_msc->instance_id_linked_device;
+		acpi_id_valid = true;
+	}
+
+	err = snprintf(uid, sizeof(uid), "%u",
+		       tbl_msc->instance_id_linked_device);
+	if (err < 0 || err >= sizeof(uid))

Does snprintf() ever return < 0 ? It's documented as returning number of chars printed (without the NULL) so that can only be 0 or greater.

Can it return >= sizeof(uid) ? Looks odd.

(quoted diff) +1 -0
+		return acpi_id_valid;
Quoted 38 lines
drivers/acpi/arm64/mpam.c new file +37 -0
+
+	buddy = acpi_dev_get_first_match_dev(hid, uid, -1);
+	if (buddy)
+		device_link_add(&pdev->dev, &buddy->dev, DL_FLAG_STATELESS);
+
+	return acpi_id_valid;
+}
 
+static int __init _parse_table(struct acpi_table_header *table)
+{
+	char *table_end, *table_offset = (char *)(table + 1);
+	struct property_entry props[4]; /* needs a sentinel */
+	struct acpi_mpam_msc_node *tbl_msc;
+	int next_res, next_prop, err = 0;
+	struct acpi_device *companion;
+	struct platform_device *pdev;
+	enum mpam_msc_iface iface;
+	struct resource res[3];
+	char uid[16];
+	u32 acpi_id;
+
+	table_end = (char *)table + table->length;
+
+	while (table_offset < table_end) {
+		tbl_msc = (struct acpi_mpam_msc_node *)table_offset;
+		table_offset += tbl_msc->length;
+
+		/*
+		 * If any of the reserved fields are set, make no attempt to
+		 * parse the msc structure. This will prevent the driver from
+		 * probing all the MSC, meaning it can't discover the system
+		 * wide supported partid and pmg ranges. This avoids whatever
+		 * this MSC is truncating the partids and creating a screaming
+		 * error interrupt.
+		 */
+		if (tbl_msc->reserved || tbl_msc->reserved1 || tbl_msc->reserved2)
+			continue;
+

drivers/acpi/arm64/mpam.c line 223

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+		if (decode_interface_type(tbl_msc, &iface))
+			continue;
+
+		next_res = 0;
+		next_prop = 0;
+		memset(res, 0, sizeof(res));
+		memset(props, 0, sizeof(props));
+
+		pdev = platform_device_alloc("mpam_msc", tbl_msc->identifier);
+		if (IS_ERR(pdev)) {

returns NULL in at least some error cases (probably all, I'm just to lazy to check)

Quoted 13 lines
drivers/acpi/arm64/mpam.c new file +13 -0
+			err = PTR_ERR(pdev);
+			break;
+		}
+
+		if (tbl_msc->length < sizeof(*tbl_msc)) {
+			err = -EINVAL;
+			break;
+		}
+
+		/* Some power management is described in the namespace: */
+		err = snprintf(uid, sizeof(uid), "%u", tbl_msc->identifier);
+		if (err > 0 && err < sizeof(uid)) {
+			companion = acpi_dev_get_first_match_dev("ARMHAA5C", uid, -1);

drivers/acpi/arm64/mpam.c line 246

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+			if (companion)
+				ACPI_COMPANION_SET(&pdev->dev, companion);
+		}
+
+		if (iface == MPAM_IFACE_MMIO) {
+			res[next_res].name = "MPAM:MSC";
+			res[next_res].start = tbl_msc->base_address;
+			res[next_res].end = tbl_msc->base_address + tbl_msc->mmio_size - 1;
+			res[next_res].flags = IORESOURCE_MEM;
+			next_res++;

DEFINE_RES_MEM_NAMED()?

Quoted 25 lines
drivers/acpi/arm64/mpam.c new file +25 -0
+		} else if (iface == MPAM_IFACE_PCC) {
+			props[next_prop++] = PROPERTY_ENTRY_U32("pcc-channel",
+								tbl_msc->base_address);
+			next_prop++;
+		}
+
+		acpi_mpam_parse_irqs(pdev, tbl_msc, res, &next_res);
+		err = platform_device_add_resources(pdev, res, next_res);
+		if (err)
+			break;
+
+		props[next_prop++] = PROPERTY_ENTRY_U32("arm,not-ready-us",
+							tbl_msc->max_nrdy_usec);
+
+		/*
+		 * The MSC's CPU affinity is described via its linked power
+		 * management device, but only if it points at a Processor or
+		 * Processor Container.
+		 */
+		if (parse_msc_pm_link(tbl_msc, pdev, &acpi_id)) {
+			props[next_prop++] = PROPERTY_ENTRY_U32("cpu_affinity",
+								acpi_id);
+		}
+
+		err = device_create_managed_software_node(&pdev->dev, props,

drivers/acpi/arm64/mpam.c line 281

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+							  NULL);
+		if (err)
+			break;
+
+		/* Come back later if you want the RIS too */
+		err = platform_device_add_data(pdev, tbl_msc, tbl_msc->length);
+		if (err)
+			break;
+
+		platform_device_add(pdev);

Can fail.

Quoted 14 lines
drivers/acpi/arm64/mpam.c new file +14 -0
+	}
+
+	if (err)
+		platform_device_put(pdev);
+
+	return err;
+}
+
+static struct acpi_table_header *get_table(void)
+{
+	struct acpi_table_header *table;
+	acpi_status status;
+
+	if (acpi_disabled || !system_supports_mpam())
Quoted 8 lines

drivers/acpi/arm64/mpam.c line 303

Quoted 8 lines
drivers/acpi/arm64/mpam.c new file +8 -0
+		return NULL;
+
+	status = acpi_get_table(ACPI_SIG_MPAM, 0, &table);
+	if (ACPI_FAILURE(status))
+		return NULL;
+
+	if (table->revision != 1)
+		return NULL;

Indicates that this was written against version 2 of the spec.

Quoted 52 lines

drivers/acpi/arm64/mpam.c line 315

Quoted 12 lines
drivers/acpi/arm64/mpam.c new file +12 -0
+
+	return table;
+}
+
+static int __init acpi_mpam_parse(void)
+{
+	struct acpi_table_header *mpam;
+	int err;
+
+	mpam = get_table();
+	if (!mpam)
+		return 0;

Just what I was suggesting for the PPTT case in early patches. Nice :)

drivers/acpi/arm64/mpam.c line 327

Quoted 12 lines
drivers/acpi/arm64/mpam.c new file +12 -0
+
+	err = _parse_table(mpam);
+	acpi_put_table(mpam);
+
+	return err;
+}
+
+static int _count_msc(struct acpi_table_header *table)
+{
+	char *table_end, *table_offset = (char *)(table + 1);
+	struct acpi_mpam_msc_node *tbl_msc;
+	int ret = 0;

Call it count as it only ever contains the count?

drivers/acpi/arm64/mpam.c line 336

Quoted 9 lines
drivers/acpi/arm64/mpam.c new file +9 -0
+
+	tbl_msc = (struct acpi_mpam_msc_node *)table_offset;
+	table_end = (char *)table + table->length;
+
+	while (table_offset < table_end) {
+		if (tbl_msc->length < sizeof(*tbl_msc))
+			return -EINVAL;
+
+		ret++;

count++ would feel more natural here.

drivers/acpi/arm64/mpam.c line 343

Quoted 7 lines
drivers/acpi/arm64/mpam.c new file +7 -0
+
+		table_offset += tbl_msc->length;
+		tbl_msc = (struct acpi_mpam_msc_node *)table_offset;
+	}
+
+	return ret;
+}

That's all I have time for today. Will get to the rest of the series soonish.

Quoted 3 lines

Jonathan

Thanks,

Ben

author maintainer
#31 lore raw

Re: [RFC PATCH 10/36] ACPI / MPAM: Parse the MPAM table

Hi Ben,

Quoted 10 lines · Ben Horgan
Quoted 8 lines · Jonathan Cameron
Quoted 2 lines · On Fri, 11 Jul 2025 18:36:22 +0000 James Morse <[email protected]> wrote:

Add code to parse the arm64 specific MPAM table, looking up the cache level from the PPTT and feeding the end result into the MPAM driver.

Throw in a link to the spec perhaps?  Particularly useful to know which version this was written against when reviewing it.

Quoted 6 lines

As I comment below this code checks the table revision is 1 and so we can assume it was written against version 2 of the spec. As of Monday, there is a new version hot off the press,
https://developer.arm.com/documentation/den0065/3-0bet/?lang=en which introduces an "MMIO size" field to allow for disabled nodes. This should be considered here to avoid advertising msc that aren't present.

Sure. Bit of an unfortunate race with the spec people there!

Added as:
--------------------%<--------------------

drivers/acpi/arm64/mpam.c +7 -2
index 9ff5a6df9f1b..d8c6224a76f8 100644
@@ -202,6 +202,9 @@ static int __init _parse_table(struct acpi_table_header *table)
                if (tbl_msc->reserved || tbl_msc->reserved1 || tbl_msc->reserved2)
                        continue;
 
+               if (!tbl_msc->mmio_size)
+                       continue;
+
                if (decode_interface_type(tbl_msc, &iface))
                        continue;
 
@@ -290,7 +293,7 @@ static struct acpi_table_header *get_table(void)
        if (ACPI_FAILURE(status))
                return NULL;
 
-       if (table->revision != 1)
+       if (table->revision < 1)
                return NULL;
 
        return table;
@@ -321,6 +324,9 @@ static int _count_msc(struct acpi_table_header *table)
        table_end = (char *)table + table->length;
 
        while (table_offset < table_end) {
+               if (!tbl_msc->mmio_size)
+                       continue;
+
                if (tbl_msc->length < sizeof(*tbl_msc))
                        return -EINVAL;
--------------------%<--------------------

Amusingly, PCC also defines mmio_size==0 as disabled, so _count_msc() doesn't need to know what kind of thing this is. In principle they could change this as its beta, but a zero sized MSC should probably be treated as an error anyway.

Thanks,

James

maintainer
#32 lore raw

Re: [RFC PATCH 10/36] ACPI / MPAM: Parse the MPAM table

Hi James,

Quoted 48 lines · James Morse

Hi Ben,

Quoted 10 lines · Ben Horgan
Quoted 8 lines · Jonathan Cameron
Quoted 2 lines · On Fri, 11 Jul 2025 18:36:22 +0000 James Morse <[email protected]> wrote:

Add code to parse the arm64 specific MPAM table, looking up the cache level from the PPTT and feeding the end result into the MPAM driver.

Throw in a link to the spec perhaps?  Particularly useful to know which version this was written against when reviewing it.

Quoted 6 lines

As I comment below this code checks the table revision is 1 and so we can assume it was written against version 2 of the spec. As of Monday, there is a new version hot off the press,
https://developer.arm.com/documentation/den0065/3-0bet/?lang=en which introduces an "MMIO size" field to allow for disabled nodes. This should be considered here to avoid advertising msc that aren't present.

Sure. Bit of an unfortunate race with the spec people there!

Added as:
--------------------%<--------------------

drivers/acpi/arm64/mpam.c +4 -1
index 9ff5a6df9f1b..d8c6224a76f8 100644
@@ -202,6 +202,9 @@ static int __init _parse_table(struct acpi_table_header *table)
                 if (tbl_msc->reserved || tbl_msc->reserved1 || tbl_msc->reserved2)
                         continue;
 
+               if (!tbl_msc->mmio_size)
+                       continue;
+
                 if (decode_interface_type(tbl_msc, &iface))
                         continue;
 
@@ -290,7 +293,7 @@ static struct acpi_table_header *get_table(void)
         if (ACPI_FAILURE(status))
                 return NULL;
 
-       if (table->revision != 1)
+       if (table->revision < 1)
                 return NULL;
 
         return table;

drivers/acpi/arm64/mpam.c static int _count_msc(struct acpi_table_header *table) · line 326

Quoted 9 lines
drivers/acpi/arm64/mpam.c +3 -1
@@ -321,6 +324,9 @@ static int _count_msc(struct acpi_table_header *table)
         table_end = (char *)table + table->length;
 
         while (table_offset < table_end) {
+               if (!tbl_msc->mmio_size)
+                       continue;
+
                 if (tbl_msc->length < sizeof(*tbl_msc))
                         return -EINVAL;
--------------------%<--------------------

This seems fine as long as any later table revisions are guaranteed to be backwards compatible.

Quoted 9 lines

Amusingly, PCC also defines mmio_size==0 as disabled, so _count_msc() doesn't need to know what kind of thing this is. In principle they could change this as its beta, but a zero sized MSC should probably be treated as an error anyway.

Thanks,

James

Thanks,

Ben

#33 lore raw

Re: [RFC PATCH 10/36] ACPI / MPAM: Parse the MPAM table

In reply to Jonathan Cameron
Quoted 14 lines
Quoted 4 lines
drivers/acpi/arm64/mpam.c new file +3 -0
+static struct acpi_table_header *get_table(void)
+{
+	struct acpi_table_header *table;

drivers/acpi/arm64/mpam.c line 302

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+	acpi_status status;
+
+	if (acpi_disabled || !system_supports_mpam())
+		return NULL;
+
+	status = acpi_get_table(ACPI_SIG_MPAM, 0, &table);
+	if (ACPI_FAILURE(status))
+		return NULL;
+
+	if (table->revision != 1)

Missing an acpi_put_table()

I'm messing around with ACQUIRE() that is queued in the CXL tree for the coming merge window and noticed this.
https://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl.git/log/?h=for-6.17/cleanup-acquire

Interestingly this is a new corner case where we want conditional locking style handling but with return_ptr() style handling. Maybe too much of a niche to bother with infrastructure.

Worth noting though that one layer up it is probably worth something like:

DEFINE_FREE(acpi_table_mpam, struct acpi_table_header *, if (_T) acpi_put_table(_T));

That enables nice clean code like:

static int __init acpi_mpam_parse(void)

{
	struct acpi_table_header *mpam = __free(acpi_table_mpam) = get_table();

	if (!mpam)
		return 0;

	return _parse_table;
}

This series was big enough that I'm spinning a single 'suggested changes' patch on top of it that includes stuff like this. Might take another day or so.

Jonathan

Quoted 4 lines
Quoted 4 lines
drivers/acpi/arm64/mpam.c new file +4 -0
+		return NULL;
+
+	return table;
+}
author maintainer
#34 lore raw

Re: [RFC PATCH 10/36] ACPI / MPAM: Parse the MPAM table

Hi Jonathan,

Quoted 7 lines · Jonathan Cameron
Quoted 7 lines
Quoted 7 lines
drivers/acpi/arm64/mpam.c new file +7 -0
+static struct acpi_table_header *get_table(void)
+{
+	struct acpi_table_header *table;
+	acpi_status status;
+
+	if (acpi_disabled || !system_supports_mpam())
+		return NULL;
Quoted 8 lines
Quoted 6 lines

drivers/acpi/arm64/mpam.c line 302

Quoted 6 lines
drivers/acpi/arm64/mpam.c new file +6 -0
+
+	status = acpi_get_table(ACPI_SIG_MPAM, 0, &table);
+	if (ACPI_FAILURE(status))
+		return NULL;
+
+	if (table->revision != 1)

Missing an acpi_put_table()

Oops,

Quoted 3 lines

I'm messing around with ACQUIRE() that is queued in the CXL tree for the coming merge window and noticed this.
https://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl.git/log/?h=for-6.17/cleanup-acquire

(not more C++!)

Quoted 12 lines

Interestingly this is a new corner case where we want conditional locking style handling but with return_ptr() style handling. Maybe too much of a niche to bother with infrastructure.

Worth noting though that one layer up it is probably worth something like:

DEFINE_FREE(acpi_table_mpam, struct acpi_table_header *, if (_T) acpi_put_table(_T));

That enables nice clean code like:

static int __init acpi_mpam_parse(void)

Quoted 8 lines
{
	struct acpi_table_header *mpam = __free(acpi_table_mpam) = get_table();

	if (!mpam)
		return 0;

	return _parse_table;
}

I've got bits of that from your PPTT suggestions. I ended up folding the get_table() helper in here.

count_msc() gets the same treatment and the cleanup thing lets _count_msc() be folded into it.

Thanks,

James

Quoted 4 lines

This series was big enough that I'm spinning a single 'suggested changes' patch on top of it that includes stuff like this. Might take another day or so.

Jonathan

author maintainer
#35 lore raw

Re: [RFC PATCH 10/36] ACPI / MPAM: Parse the MPAM table

In reply to Jonathan Cameron

Hi Jonathan,

Quoted 8 lines · Jonathan Cameron
Quoted 2 lines · On Fri, 11 Jul 2025 18:36:22 +0000 James Morse <[email protected]> wrote:

Add code to parse the arm64 specific MPAM table, looking up the cache level from the PPTT and feeding the end result into the MPAM driver.

Throw in a link to the spec perhaps? Particularly useful to know which version this was written against when reviewing it.

Will do. Ben has already pointed out it wasn't written against the latest version...

Quoted 11 lines

drivers/acpi/arm64/Makefile obj-$(CONFIG_ACPI_GTDT) += gtdt.o · line 9

Quoted 9 lines
drivers/acpi/arm64/Makefile +1 -0
index 05ecde9eaabe..27b872249baa 100644
@@ -6,5 +6,6 @@ obj-$(CONFIG_ACPI_GTDT) 	+= gtdt.o
 obj-$(CONFIG_ACPI_IORT) 	+= iort.o
 obj-$(CONFIG_ACPI_PROCESSOR_IDLE) += cpuidle.o
 obj-$(CONFIG_ARM_AMBA)		+= amba.o
+obj-$(CONFIG_ACPI_MPAM) 	+= mpam.o

Keep it with the ACPI ones?

Sure,

Quoted 2 lines

There doesn't seem to be a lot of order in here
though so I guess maybe there is logic behind putting it here I'm missing.

merge conflicts over many years always put it at the bottom of the file. I at least kept the conditional ones together.

Moving it up lets the table 'drivers' appear together in alphabetical order.

Quoted 18 lines
Quoted 18 lines
drivers/acpi/arm64/mpam.c new file +12 -0
new file mode 100644
index 000000000000..f4791bac9a2a
@@ -0,0 +1,365 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (C) 2025 Arm Ltd.
+
+/* Parse the MPAM ACPI table feeding the discovered nodes into the driver */
+
+#define pr_fmt(fmt) "ACPI MPAM: " fmt
+
+#include <linux/acpi.h>
+#include <linux/arm_mpam.h>
+#include <linux/cpu.h>
+#include <linux/cpumask.h>
+#include <linux/platform_device.h>
Quoted 8 lines

drivers/acpi/arm64/mpam.c line 16

Quoted 4 lines
drivers/acpi/arm64/mpam.c new file +4 -0
+
+#include <acpi/processor.h>
+
+/* Flags for acpi_table_mpam_msc.*_interrupt_flags */

References.. I'm looking at 3.0-alpha table 5 to check this. I can see why you might be reluctant to point at an alpha if that is what you are using ;)

I did this against the released(?) version 2.0. (aka table revision 1). I'll add references based on the v3 beta ... it looks like that defines the mmio-size=0 behaviour and the pcc stuff. The mmio-size is harmless - we'd need to handle that as an error anyay. I don't want to touch the pcc thing until there is a real platform that needs it, and the spec is finished...

e.g.
| * See 2.1.1 Interrupt Flags, Table 5, of DEN0065B_MPAM_ACPI_3.0-bet.

Quoted 4 lines

drivers/acpi/arm64/mpam.c line 18

Quoted 2 lines
drivers/acpi/arm64/mpam.c new file +2 -0
+#define ACPI_MPAM_MSC_IRQ_MODE_EDGE                    1
+#define ACPI_MPAM_MSC_IRQ_TYPE_MASK                    (3 << 1)

GENMASK(3, 2) would be my preference for how to do masks in new code.

GENMASK(2, 1), but yes.

Quoted 17 lines
Quoted 4 lines
drivers/acpi/arm64/mpam.c new file +4 -0
+#define ACPI_MPAM_MSC_IRQ_TYPE_WIRED                   0
+#define ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER BIT(3)
+#define ACPI_MPAM_MSC_IRQ_AFFINITY_VALID               BIT(4)
+

drivers/acpi/arm64/mpam.c line 32

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+static bool frob_irq(struct platform_device *pdev, int intid, u32 flags,
+		     int *irq, u32 processor_container_uid)
+{
+	int sense;
+
+	if (!intid)
+		return false;
+
+	/* 0 in this field indicates a wired interrupt */
+	if (flags & ACPI_MPAM_MSC_IRQ_TYPE_MASK)

I'd prefer more explicit code (and probably no comment)

Quoted 3 lines
	if (FIELD_GET(flags, ACPI_MPAM_MSC_IRQ_TYPE_MASK) !=
	    ACPI_MPAM_MSC_IRQ_TYPE_WIRED)
		return false;

Sure,

Quoted 9 lines

drivers/acpi/arm64/mpam.c line 38

Quoted 6 lines
drivers/acpi/arm64/mpam.c new file +6 -0
+		return false;
+
+	if (flags & ACPI_MPAM_MSC_IRQ_MODE_EDGE)
+		sense = ACPI_EDGE_SENSITIVE;
+	else
+		sense = ACPI_LEVEL_SENSITIVE;

If the spec is supposed to be using standard ACPI_* types for this field (I don't think the connection is explicitly documented though) then

Sent as feedback on the spec. (I didn't realise those were standard!)

Quoted 45 lines
	sense = FIELD_GET(flags, ACPI_MPAM_MSC_IRQ_MODE_MASK); 

Assuming a change to define the mask and rely on the ACPI defs for the values

This one is entirely up to you.

Quoted 21 lines
drivers/acpi/arm64/mpam.c new file +21 -0
+
+	/*
+	 * If the GSI is in the GIC's PPI range, try and create a partitioned
+	 * percpu interrupt.
+	 */
+	if (16 <= intid && intid < 32 && processor_container_uid != ~0) {
+		pr_err_once("Partitioned interrupts not supported\n");
+		return false;
+	}
+
+	*irq = acpi_register_gsi(&pdev->dev, intid, sense, ACPI_ACTIVE_HIGH);
+	if (*irq <= 0) {
+		pr_err_once("Failed to register interrupt 0x%x with ACPI\n",
+			    intid);
+		return false;
+	}
+
+	return true;
+}
+
+static void acpi_mpam_parse_irqs(struct platform_device *pdev,

drivers/acpi/arm64/mpam.c line 69

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+				 struct acpi_mpam_msc_node *tbl_msc,
+				 struct resource *res, int *res_idx)
+{
+	u32 flags, aff = ~0;
+	int irq;
+
+	flags = tbl_msc->overflow_interrupt_flags;
+	if (flags & ACPI_MPAM_MSC_IRQ_AFFINITY_VALID &&
+	    flags & ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER)
+		aff = tbl_msc->overflow_interrupt_affinity;

Just to make the two cases look the same I'd do

	else
		aff = ~0;

here as well and not initialize above. It's not quite worth using a helper function for these two identical blocks but it's close.

Quoted 7 lines

drivers/acpi/arm64/mpam.c line 74

Quoted 5 lines
drivers/acpi/arm64/mpam.c new file +5 -0
+	if (frob_irq(pdev, tbl_msc->overflow_interrupt, flags, &irq, aff)) {
+		res[*res_idx].start = irq;
+		res[*res_idx].end = irq;
+		res[*res_idx].flags = IORESOURCE_IRQ;
+		res[*res_idx].name = "overflow";
		res[*res_idx] = DEFINE_RES_IRQ_NAMED(irq, 1, "overflow");

Handy, not seen that before.

Quoted 5 lines

drivers/acpi/arm64/mpam.c line 76

Quoted 2 lines
drivers/acpi/arm64/mpam.c new file +2 -0
+
+		(*res_idx)++;

Can roll this in as well.

Quoted 15 lines

drivers/acpi/arm64/mpam.c line 89

Quoted 13 lines
drivers/acpi/arm64/mpam.c new file +13 -0
+	}
+
+	flags = tbl_msc->error_interrupt_flags;
+	if (flags & ACPI_MPAM_MSC_IRQ_AFFINITY_VALID &&
+	    flags & ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER)
+		aff = tbl_msc->error_interrupt_affinity;
+	else
+		aff = ~0;
+	if (frob_irq(pdev, tbl_msc->error_interrupt, flags, &irq, aff)) {
+		res[*res_idx].start = irq;
+		res[*res_idx].end = irq;
+		res[*res_idx].flags = IORESOURCE_IRQ;
+		res[*res_idx].name = "error";

Similar to above.

Yup,

Quoted 24 lines
Quoted 24 lines
drivers/acpi/arm64/mpam.c new file +22 -0
+
+		(*res_idx)++;
+	}
+}
+
 
 
+static bool __init parse_msc_pm_link(struct acpi_mpam_msc_node *tbl_msc,
+				     struct platform_device *pdev,
+				     u32 *acpi_id)
+{
+	bool acpi_id_valid = false;
+	struct acpi_device *buddy;
+	char hid[16], uid[16];
+	int err;
+
+	memset(&hid, 0, sizeof(hid));
+	memcpy(hid, &tbl_msc->hardware_id_linked_device,
+	       sizeof(tbl_msc->hardware_id_linked_device));
+
+	if (!strcmp(hid, ACPI_PROCESSOR_CONTAINER_HID)) {
+		*acpi_id = tbl_msc->instance_id_linked_device;
+		acpi_id_valid = true;
+	}
Quoted 8 lines

drivers/acpi/arm64/mpam.c line 159

Quoted 4 lines
drivers/acpi/arm64/mpam.c new file +4 -0
+
+	err = snprintf(uid, sizeof(uid), "%u",
+		       tbl_msc->instance_id_linked_device);
+	if (err < 0 || err >= sizeof(uid))

Does snprintf() ever return < 0 ? It's documented as returning number of chars printed (without the NULL) so that can only be 0 or greater.

That looks like paranoia around string parsing in C, and snprintf() returning an int. I've removed the first half,

Quoted 1 line

Can it return >= sizeof(uid) ? Looks odd.

More paranoia, it should be impossible given the arguments, but the documentation has:
| If the return is greater than or equal to @size, the resulting string is truncated.

If the string is truncated, there is no reason to feed it into acpi_dev_get_first_match_dev().

Quoted 43 lines
(quoted diff) +1 -0
+		return acpi_id_valid;
Quoted 42 lines
drivers/acpi/arm64/mpam.c new file +41 -0
+
+	buddy = acpi_dev_get_first_match_dev(hid, uid, -1);
+	if (buddy)
+		device_link_add(&pdev->dev, &buddy->dev, DL_FLAG_STATELESS);
+
+	return acpi_id_valid;
+}
 
+static int __init _parse_table(struct acpi_table_header *table)
+{
+	char *table_end, *table_offset = (char *)(table + 1);
+	struct property_entry props[4]; /* needs a sentinel */
+	struct acpi_mpam_msc_node *tbl_msc;
+	int next_res, next_prop, err = 0;
+	struct acpi_device *companion;
+	struct platform_device *pdev;
+	enum mpam_msc_iface iface;
+	struct resource res[3];
+	char uid[16];
+	u32 acpi_id;
+
+	table_end = (char *)table + table->length;
+
+	while (table_offset < table_end) {
+		tbl_msc = (struct acpi_mpam_msc_node *)table_offset;
+		table_offset += tbl_msc->length;
+
+		/*
+		 * If any of the reserved fields are set, make no attempt to
+		 * parse the msc structure. This will prevent the driver from
+		 * probing all the MSC, meaning it can't discover the system
+		 * wide supported partid and pmg ranges. This avoids whatever
+		 * this MSC is truncating the partids and creating a screaming
+		 * error interrupt.
+		 */
+		if (tbl_msc->reserved || tbl_msc->reserved1 || tbl_msc->reserved2)
+			continue;
+
+		if (decode_interface_type(tbl_msc, &iface))
+			continue;
+
+		next_res = 0;
Quoted 8 lines

drivers/acpi/arm64/mpam.c line 223

Quoted 6 lines
drivers/acpi/arm64/mpam.c new file +6 -0
+		next_prop = 0;
+		memset(res, 0, sizeof(res));
+		memset(props, 0, sizeof(props));
+
+		pdev = platform_device_alloc("mpam_msc", tbl_msc->identifier);
+		if (IS_ERR(pdev)) {

returns NULL in at least some error cases (probably all, I'm just to lazy to check)

So it does ... Fixed.

Quoted 17 lines
Quoted 17 lines
drivers/acpi/arm64/mpam.c new file +17 -0
+			err = PTR_ERR(pdev);
+			break;
+		}
+
+		if (tbl_msc->length < sizeof(*tbl_msc)) {
+			err = -EINVAL;
+			break;
+		}
+
+		/* Some power management is described in the namespace: */
+		err = snprintf(uid, sizeof(uid), "%u", tbl_msc->identifier);
+		if (err > 0 && err < sizeof(uid)) {
+			companion = acpi_dev_get_first_match_dev("ARMHAA5C", uid, -1);
+			if (companion)
+				ACPI_COMPANION_SET(&pdev->dev, companion);
+		}
+
Quoted 8 lines

drivers/acpi/arm64/mpam.c line 246

Quoted 6 lines
drivers/acpi/arm64/mpam.c new file +6 -0
+		if (iface == MPAM_IFACE_MMIO) {
+			res[next_res].name = "MPAM:MSC";
+			res[next_res].start = tbl_msc->base_address;
+			res[next_res].end = tbl_msc->base_address + tbl_msc->mmio_size - 1;
+			res[next_res].flags = IORESOURCE_MEM;
+			next_res++;

DEFINE_RES_MEM_NAMED()?

Done,

Quoted 29 lines
Quoted 29 lines
drivers/acpi/arm64/mpam.c new file +29 -0
+		} else if (iface == MPAM_IFACE_PCC) {
+			props[next_prop++] = PROPERTY_ENTRY_U32("pcc-channel",
+								tbl_msc->base_address);
+			next_prop++;
+		}
+
+		acpi_mpam_parse_irqs(pdev, tbl_msc, res, &next_res);
+		err = platform_device_add_resources(pdev, res, next_res);
+		if (err)
+			break;
+
+		props[next_prop++] = PROPERTY_ENTRY_U32("arm,not-ready-us",
+							tbl_msc->max_nrdy_usec);
+
+		/*
+		 * The MSC's CPU affinity is described via its linked power
+		 * management device, but only if it points at a Processor or
+		 * Processor Container.
+		 */
+		if (parse_msc_pm_link(tbl_msc, pdev, &acpi_id)) {
+			props[next_prop++] = PROPERTY_ENTRY_U32("cpu_affinity",
+								acpi_id);
+		}
+
+		err = device_create_managed_software_node(&pdev->dev, props,
+							  NULL);
+		if (err)
+			break;
+
Quoted 8 lines

drivers/acpi/arm64/mpam.c line 281

Quoted 6 lines
drivers/acpi/arm64/mpam.c new file +6 -0
+		/* Come back later if you want the RIS too */
+		err = platform_device_add_data(pdev, tbl_msc, tbl_msc->length);
+		if (err)
+			break;
+
+		platform_device_add(pdev);

Can fail.

Fixed,

Quoted 8 lines
Quoted 7 lines
drivers/acpi/arm64/mpam.c new file +7 -0
+	}
+
+	if (err)
+		platform_device_put(pdev);
+
+	return err;
+}
Quoted 7 lines

drivers/acpi/arm64/mpam.c line 327

Quoted 5 lines
drivers/acpi/arm64/mpam.c new file +5 -0
+static int _count_msc(struct acpi_table_header *table)
+{
+	char *table_end, *table_offset = (char *)(table + 1);
+	struct acpi_mpam_msc_node *tbl_msc;
+	int ret = 0;

Call it count as it only ever contains the count?

Sure,

Quoted 12 lines

drivers/acpi/arm64/mpam.c line 336

Quoted 9 lines
drivers/acpi/arm64/mpam.c new file +9 -0
+
+	tbl_msc = (struct acpi_mpam_msc_node *)table_offset;
+	table_end = (char *)table + table->length;
+
+	while (table_offset < table_end) {
+		if (tbl_msc->length < sizeof(*tbl_msc))
+			return -EINVAL;
+
+		ret++;

count++ would feel more natural here.

Quoted 9 lines

drivers/acpi/arm64/mpam.c line 343

Quoted 7 lines
drivers/acpi/arm64/mpam.c new file +7 -0
+
+		table_offset += tbl_msc->length;
+		tbl_msc = (struct acpi_mpam_msc_node *)table_offset;
+	}
+
+	return ret;
+}

That's all I have time for today. Will get to the rest of the series soonish.

Thanks for taking a look!

James

maintainer
#36 lore raw

Re: [RFC PATCH 10/36] ACPI / MPAM: Parse the MPAM table

In reply to James Morse

Hi James,

Quoted 15 lines · James Morse

Add code to parse the arm64 specific MPAM table, looking up the cache level from the PPTT and feeding the end result into the MPAM driver.

---
  arch/arm64/Kconfig          |   1 +
  drivers/acpi/arm64/Kconfig  |   3 +
  drivers/acpi/arm64/Makefile |   1 +
  drivers/acpi/arm64/mpam.c   | 365 ++++++++++++++++++++++++++++++++++++
  drivers/acpi/tables.c       |   2 +-
  include/linux/arm_mpam.h    |  46 +++++
  6 files changed, 417 insertions(+), 1 deletion(-)

create mode 100644 drivers/acpi/arm64/mpam.c
create mode 100644 include/linux/arm_mpam.h

[snip]

Quoted 103 lines
drivers/acpi/arm64/mpam.c new file +97 -0
new file mode 100644
index 000000000000..f4791bac9a2a
@@ -0,0 +1,365 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (C) 2025 Arm Ltd.
+
+/* Parse the MPAM ACPI table feeding the discovered nodes into the driver */
+
+#define pr_fmt(fmt) "ACPI MPAM: " fmt
+
+#include <linux/acpi.h>
+#include <linux/arm_mpam.h>
+#include <linux/cpu.h>
+#include <linux/cpumask.h>
+#include <linux/platform_device.h>
+
+#include <acpi/processor.h>
+
+/* Flags for acpi_table_mpam_msc.*_interrupt_flags */
+#define ACPI_MPAM_MSC_IRQ_MODE_EDGE                    1
+#define ACPI_MPAM_MSC_IRQ_TYPE_MASK                    (3 << 1)
+#define ACPI_MPAM_MSC_IRQ_TYPE_WIRED                   0
+#define ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER BIT(3)
+#define ACPI_MPAM_MSC_IRQ_AFFINITY_VALID               BIT(4)
+
+static bool frob_irq(struct platform_device *pdev, int intid, u32 flags,
+		     int *irq, u32 processor_container_uid)
+{
+	int sense;
+
+	if (!intid)
+		return false;
+
+	/* 0 in this field indicates a wired interrupt */
+	if (flags & ACPI_MPAM_MSC_IRQ_TYPE_MASK)
+		return false;
+
+	if (flags & ACPI_MPAM_MSC_IRQ_MODE_EDGE)
+		sense = ACPI_EDGE_SENSITIVE;
+	else
+		sense = ACPI_LEVEL_SENSITIVE;
+
+	/*
+	 * If the GSI is in the GIC's PPI range, try and create a partitioned
+	 * percpu interrupt.
+	 */
+	if (16 <= intid && intid < 32 && processor_container_uid != ~0) {
+		pr_err_once("Partitioned interrupts not supported\n");
+		return false;
+	}
+
+	*irq = acpi_register_gsi(&pdev->dev, intid, sense, ACPI_ACTIVE_HIGH);
+	if (*irq <= 0) {
+		pr_err_once("Failed to register interrupt 0x%x with ACPI\n",
+			    intid);
+		return false;
+	}
+
+	return true;
+}
+
+static void acpi_mpam_parse_irqs(struct platform_device *pdev,
+				 struct acpi_mpam_msc_node *tbl_msc,
+				 struct resource *res, int *res_idx)
+{
+	u32 flags, aff = ~0;
+	int irq;
+
+	flags = tbl_msc->overflow_interrupt_flags;
+	if (flags & ACPI_MPAM_MSC_IRQ_AFFINITY_VALID &&
+	    flags & ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER)
+		aff = tbl_msc->overflow_interrupt_affinity;
+	if (frob_irq(pdev, tbl_msc->overflow_interrupt, flags, &irq, aff)) {
+		res[*res_idx].start = irq;
+		res[*res_idx].end = irq;
+		res[*res_idx].flags = IORESOURCE_IRQ;
+		res[*res_idx].name = "overflow";
+
+		(*res_idx)++;
+	}
+
+	flags = tbl_msc->error_interrupt_flags;
+	if (flags & ACPI_MPAM_MSC_IRQ_AFFINITY_VALID &&
+	    flags & ACPI_MPAM_MSC_IRQ_AFFINITY_PROCESSOR_CONTAINER)
+		aff = tbl_msc->error_interrupt_affinity;
+	else
+		aff = ~0;
+	if (frob_irq(pdev, tbl_msc->error_interrupt, flags, &irq, aff)) {
+		res[*res_idx].start = irq;
+		res[*res_idx].end = irq;
+		res[*res_idx].flags = IORESOURCE_IRQ;
+		res[*res_idx].name = "error";
+
+		(*res_idx)++;
+	}
+}
+
+static int acpi_mpam_parse_resource(struct mpam_msc *msc,
+				    struct acpi_mpam_resource_node *res)
+{

drivers/acpi/arm64/mpam.c line 107

Quoted 10 lines
drivers/acpi/arm64/mpam.c new file +10 -0
+	int level, nid;
+	u32 cache_id;
+
+	switch (res->locator_type) {
+	case ACPI_MPAM_LOCATION_TYPE_PROCESSOR_CACHE:
+		cache_id = res->locator.cache_locator.cache_reference;
+		level = find_acpi_cache_level_from_id(cache_id);
+		if (level < 0) {
+			pr_err_once("Bad level (%u) for cache with id %u\n", level, cache_id);
+			return -EINVAL;

Nit: More robust to check for level <= 0.

Thanks,

Ben

author maintainer
#37 lore raw

Re: [RFC PATCH 10/36] ACPI / MPAM: Parse the MPAM table

Hi Ben,

Quoted 11 lines · Ben Horgan
Quoted 9 lines · James Morse

Add code to parse the arm64 specific MPAM table, looking up the cache level from the PPTT and feeding the end result into the MPAM driver.

drivers/acpi/arm64/mpam.c new file +0 -0
new file mode 100644
index 000000000000..f4791bac9a2a
@@ -0,0 +1,365 @@
Quoted 15 lines

drivers/acpi/arm64/mpam.c line 107

Quoted 13 lines
drivers/acpi/arm64/mpam.c new file +13 -0
+static int acpi_mpam_parse_resource(struct mpam_msc *msc,
+                    struct acpi_mpam_resource_node *res)
+{
+    int level, nid;
+    u32 cache_id;
+
+    switch (res->locator_type) {
+    case ACPI_MPAM_LOCATION_TYPE_PROCESSOR_CACHE:
+        cache_id = res->locator.cache_locator.cache_reference;
+        level = find_acpi_cache_level_from_id(cache_id);
+        if (level < 0) {
+            pr_err_once("Bad level (%u) for cache with id %u\n", level, cache_id);
+            return -EINVAL;

Nit: More robust to check for level <= 0.

Sure, that can probably happen!

Thanks,

James

author maintainer
#38 lore raw

RFC PATCH 11/36 dt-bindings: arm: Add MPAM MSC binding

In reply to James Morse

From: Rob Herring <[email protected]>

The binding is designed around the assumption that an MSC will be a sub-block of something else such as a memory controller, cache controller, or IOMMU. However, it's certainly possible a design does not have that association or has a mixture of both, so the binding illustrates how we can support that with RIS child nodes.

A key part of MPAM is we need to know about all of the MSCs in the system before it can be enabled. This drives the need for the genericish 'arm,mpam-msc' compatible. Though we can't assume an MSC is accessible until a h/w specific driver potentially enables the h/w.

---
 .../devicetree/bindings/arm/arm,mpam-msc.yaml | 227 ++++++++++++++++++
 1 file changed, 227 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +227 -0
new file mode 100644
index 000000000000..9d542ecb1a7d
@@ -0,0 +1,227 @@
+# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/arm/arm,mpam-msc.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Arm Memory System Resource Partitioning and Monitoring (MPAM)
+
+description: |
+  The Arm MPAM specification can be found here:
+
+  https://developer.arm.com/documentation/ddi0598/latest
+
+maintainers:
+  - Rob Herring <[email protected]>
+
+properties:
+  compatible:
+    items:
+      - const: arm,mpam-msc                   # Further details are discoverable
+      - const: arm,mpam-memory-controller-msc
+
+  reg:
+    maxItems: 1
+    description: A memory region containing registers as defined in the MPAM
+      specification.
+
+  interrupts:
+    minItems: 1
+    items:
+      - description: error (optional)
+      - description: overflow (optional, only for monitoring)
+
+  interrupt-names:
+    oneOf:
+      - items:
+          - enum: [ error, overflow ]
+      - items:
+          - const: error
+          - const: overflow
+
+  arm,not-ready-us:
+    description: The maximum time in microseconds for monitoring data to be
+      accurate after a settings change. For more information, see the
+      Not-Ready (NRDY) bit description in the MPAM specification.
+
+  numa-node-id: true # see NUMA binding
+
+  '#address-cells':
+    const: 1
+
+  '#size-cells':
+    const: 0
+
+patternProperties:
+  '^ris@[0-9a-f]$':
+    type: object
+    additionalProperties: false
+    description: |
+      RIS nodes for each RIS in an MSC. These nodes are required for each RIS
+      implementing known MPAM controls
+
+    properties:
+      compatible:
+        enum:
+            # Bulk storage for cache
+          - arm,mpam-cache
+            # Memory bandwidth
+          - arm,mpam-memory
+
+      reg:
+        minimum: 0
+        maximum: 0xf
+
+      cpus:
+        $ref: '/schemas/types.yaml#/definitions/phandle-array'
+        description:
+          Phandle(s) to the CPU node(s) this RIS belongs to. By default, the parent
+          device's affinity is used.
+
+      arm,mpam-device:
+        $ref: '/schemas/types.yaml#/definitions/phandle'
+        description:
+          By default, the MPAM enabled device associated with a RIS is the MSC's
+          parent node. It is possible for each RIS to be associated with different
+          devices in which case 'arm,mpam-device' should be used.
+
+    required:
+      - compatible
+      - reg
+
+required:
+  - compatible
+  - reg
+
+dependencies:
+  interrupts: [ interrupt-names ]
+
+additionalProperties: false
+
+examples:
+  - |
+    /*
+    cpus {
+        cpu@0 {
+            next-level-cache = <&L2_0>;
+        };
+        cpu@100 {
+            next-level-cache = <&L2_1>;
+        };
+    };
+    */
+    L2_0: cache-controller-0 {
+        compatible = "cache";
+        cache-level = <2>;
+        cache-unified;
+        next-level-cache = <&L3>;
+
+    };
+
+    L2_1: cache-controller-1 {
+        compatible = "cache";
+        cache-level = <2>;
+        cache-unified;
+        next-level-cache = <&L3>;
+
+    };
+
+    L3: cache-controller@30000000 {
+        compatible = "arm,dsu-l3-cache", "cache";
+        cache-level = <3>;
+        cache-unified;
+
+        ranges = <0x0 0x30000000 0x800000>;
+        #address-cells = <1>;
+        #size-cells = <1>;
+
+        msc@10000 {
+            compatible = "arm,mpam-msc";
+
+            /* CPU affinity implied by parent cache node's  */
+            reg = <0x10000 0x2000>;
+            interrupts = <1>, <2>;
+            interrupt-names = "error", "overflow";
+            arm,not-ready-us = <1>;
+        };
+    };
+
+    mem: memory-controller@20000 {
+        compatible = "foo,a-memory-controller";
+        reg = <0x20000 0x1000>;
+
+        #address-cells = <1>;
+        #size-cells = <1>;
+        ranges;
+
+        msc@21000 {
+            compatible = "arm,mpam-memory-controller-msc", "arm,mpam-msc";
+            reg = <0x21000 0x1000>;
+            interrupts = <3>;
+            interrupt-names = "error";
+            arm,not-ready-us = <1>;
+            numa-node-id = <1>;
+        };
+    };
+
+    iommu@40000 {
+        reg = <0x40000 0x1000>;
+
+        ranges;
+        #address-cells = <1>;
+        #size-cells = <1>;
+
+        msc@41000 {
+            compatible = "arm,mpam-msc";
+            reg = <0 0x1000>;
+            interrupts = <5>, <6>;
+            interrupt-names = "error", "overflow";
+            arm,not-ready-us = <1>;
+
+            #address-cells = <1>;
+            #size-cells = <0>;
+
+            ris@2 {
+                compatible = "arm,mpam-cache";
+                reg = <0>;
+                // TODO: How to map to device(s)?
+            };
+        };
+    };
+
+    msc@80000 {
+        compatible = "foo,a-standalone-msc";
+        reg = <0x80000 0x1000>;
+
+        clocks = <&clks 123>;
+
+        ranges;
+        #address-cells = <1>;
+        #size-cells = <1>;
+
+        msc@10000 {
+            compatible = "arm,mpam-msc";
+
+            reg = <0x10000 0x2000>;
+            interrupts = <7>;
+            interrupt-names = "overflow";
+            arm,not-ready-us = <1>;
+
+            #address-cells = <1>;
+            #size-cells = <0>;
+
+            ris@0 {
+                compatible = "arm,mpam-cache";
+                reg = <0>;
+                arm,mpam-device = <&L2_0>;
+            };
+
+            ris@1 {
+                compatible = "arm,mpam-memory";
+                reg = <1>;
+                arm,mpam-device = <&mem>;
+            };
+        };
+    };
+
+...
Signature · 1 line
2.39.5
maintainer
#39 lore raw

Re: [RFC PATCH 11/36] dt-bindings: arm: Add MPAM MSC binding

Quoted 13 lines · James Morse

From: Rob Herring <[email protected]>

The binding is designed around the assumption that an MSC will be a sub-block of something else such as a memory controller, cache controller, or IOMMU. However, it's certainly possible a design does not have that association or has a mixture of both, so the binding illustrates how we can support that with RIS child nodes.

A key part of MPAM is we need to know about all of the MSCs in the system before it can be enabled. This drives the need for the genericish 'arm,mpam-msc' compatible. Though we can't assume an MSC is accessible until a h/w specific driver potentially enables the h/w.

Quoted 7 lines
---
 .../devicetree/bindings/arm/arm,mpam-msc.yaml | 227 ++++++++++++++++++
 1 file changed, 227 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml

Is there any DT based h/w using this? I'm not aware of any. I would prefer not merging this until there is. I have little insight whether these genericish compatibles will be sufficient, but I have lots of experience to say they won't be. I would also suspect that if anyone has started using this, they've just extended/modified it however they wanted and no feedback got to me.

Quoted 55 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +49 -0
new file mode 100644
index 000000000000..9d542ecb1a7d
@@ -0,0 +1,227 @@
+# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/arm/arm,mpam-msc.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Arm Memory System Resource Partitioning and Monitoring (MPAM)
+
+description: |
+  The Arm MPAM specification can be found here:
+
+  https://developer.arm.com/documentation/ddi0598/latest
+
+maintainers:
+  - Rob Herring <[email protected]>
+
+properties:
+  compatible:
+    items:
+      - const: arm,mpam-msc                   # Further details are discoverable
+      - const: arm,mpam-memory-controller-msc
+
+  reg:
+    maxItems: 1
+    description: A memory region containing registers as defined in the MPAM
+      specification.
+
+  interrupts:
+    minItems: 1
+    items:
+      - description: error (optional)
+      - description: overflow (optional, only for monitoring)
+
+  interrupt-names:
+    oneOf:
+      - items:
+          - enum: [ error, overflow ]
+      - items:
+          - const: error
+          - const: overflow
+
+  arm,not-ready-us:
+    description: The maximum time in microseconds for monitoring data to be
+      accurate after a settings change. For more information, see the
+      Not-Ready (NRDY) bit description in the MPAM specification.
+
+  numa-node-id: true # see NUMA binding
+
+  '#address-cells':

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 59

Quoted 10 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +10 -0
+    const: 1
+
+  '#size-cells':
+    const: 0
+
+patternProperties:
+  '^ris@[0-9a-f]$':
+    type: object
+    additionalProperties: false
+    description: |

'|' can be dropped.

Quoted 7 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +7 -0
+      RIS nodes for each RIS in an MSC. These nodes are required for each RIS
+      implementing known MPAM controls
+
+    properties:
+      compatible:
+        enum:
+            # Bulk storage for cache

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 76

Quoted 10 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +10 -0
+          - arm,mpam-cache
+            # Memory bandwidth
+          - arm,mpam-memory
+
+      reg:
+        minimum: 0
+        maximum: 0xf
+
+      cpus:
+        $ref: '/schemas/types.yaml#/definitions/phandle-array'

Don't need the type. It's in the core schemas now.

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 82

Quoted 6 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +6 -0
+        description:
+          Phandle(s) to the CPU node(s) this RIS belongs to. By default, the parent
+          device's affinity is used.
+
+      arm,mpam-device:
+        $ref: '/schemas/types.yaml#/definitions/phandle'

Don't need quotes. This should be a warning, but no testing happened because the DT list and maintainers weren't CCed.

Quoted 35 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +35 -0
+        description:
+          By default, the MPAM enabled device associated with a RIS is the MSC's
+          parent node. It is possible for each RIS to be associated with different
+          devices in which case 'arm,mpam-device' should be used.
+
+    required:
+      - compatible
+      - reg
+
+required:
+  - compatible
+  - reg
+
+dependencies:
+  interrupts: [ interrupt-names ]
+
+additionalProperties: false
+
+examples:
+  - |
+    /*
+    cpus {
+        cpu@0 {
+            next-level-cache = <&L2_0>;
+        };
+        cpu@100 {
+            next-level-cache = <&L2_1>;
+        };
+    };
+    */
+    L2_0: cache-controller-0 {
+        compatible = "cache";
+        cache-level = <2>;
+        cache-unified;
+        next-level-cache = <&L3>;

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 127

Quoted 10 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +10 -0
+
+    };
+
+    L2_1: cache-controller-1 {
+        compatible = "cache";
+        cache-level = <2>;
+        cache-unified;
+        next-level-cache = <&L3>;
+
+    };

All the above should be dropped. Not part of this binding.

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 130

Quoted 3 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +3 -0
+
+    L3: cache-controller@30000000 {
+        compatible = "arm,dsu-l3-cache", "cache";

Pretty sure this is a warning because that compatible doesn't exist.

Quoted 100 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +97 -0
+        cache-level = <3>;
+        cache-unified;
+
+        ranges = <0x0 0x30000000 0x800000>;
+        #address-cells = <1>;
+        #size-cells = <1>;
+
+        msc@10000 {
+            compatible = "arm,mpam-msc";
+
+            /* CPU affinity implied by parent cache node's  */
+            reg = <0x10000 0x2000>;
+            interrupts = <1>, <2>;
+            interrupt-names = "error", "overflow";
+            arm,not-ready-us = <1>;
+        };
+    };
+
+    mem: memory-controller@20000 {
+        compatible = "foo,a-memory-controller";
+        reg = <0x20000 0x1000>;
+
+        #address-cells = <1>;
+        #size-cells = <1>;
+        ranges;
+
+        msc@21000 {
+            compatible = "arm,mpam-memory-controller-msc", "arm,mpam-msc";
+            reg = <0x21000 0x1000>;
+            interrupts = <3>;
+            interrupt-names = "error";
+            arm,not-ready-us = <1>;
+            numa-node-id = <1>;
+        };
+    };
+
+    iommu@40000 {
+        reg = <0x40000 0x1000>;
+
+        ranges;
+        #address-cells = <1>;
+        #size-cells = <1>;
+
+        msc@41000 {
+            compatible = "arm,mpam-msc";
+            reg = <0 0x1000>;
+            interrupts = <5>, <6>;
+            interrupt-names = "error", "overflow";
+            arm,not-ready-us = <1>;
+
+            #address-cells = <1>;
+            #size-cells = <0>;
+
+            ris@2 {
+                compatible = "arm,mpam-cache";
+                reg = <0>;
+                // TODO: How to map to device(s)?
+            };
+        };
+    };
+
+    msc@80000 {
+        compatible = "foo,a-standalone-msc";
+        reg = <0x80000 0x1000>;
+
+        clocks = <&clks 123>;
+
+        ranges;
+        #address-cells = <1>;
+        #size-cells = <1>;
+
+        msc@10000 {
+            compatible = "arm,mpam-msc";
+
+            reg = <0x10000 0x2000>;
+            interrupts = <7>;
+            interrupt-names = "overflow";
+            arm,not-ready-us = <1>;
+
+            #address-cells = <1>;
+            #size-cells = <0>;
+
+            ris@0 {
+                compatible = "arm,mpam-cache";
+                reg = <0>;
+                arm,mpam-device = <&L2_0>;
+            };
+
+            ris@1 {
+                compatible = "arm,mpam-memory";
+                reg = <1>;
+                arm,mpam-device = <&mem>;
+            };
+        };
+    };
+
+...
Signature · 1 line
2.39.5
author maintainer
#40 lore raw

Re: [RFC PATCH 11/36] dt-bindings: arm: Add MPAM MSC binding

Hi Rob,

Quoted 9 lines · Rob Herring
Quoted 8 lines · James Morse

From: Rob Herring <[email protected]>

The binding is designed around the assumption that an MSC will be a sub-block of something else such as a memory controller, cache controller, or IOMMU. However, it's certainly possible a design does not have that association or has a mixture of both, so the binding illustrates how we can support that with RIS child nodes.

Quoted 6 lines
Quoted 4 lines

A key part of MPAM is we need to know about all of the MSCs in the system before it can be enabled. This drives the need for the genericish 'arm,mpam-msc' compatible. Though we can't assume an MSC is accessible until a h/w specific driver potentially enables the h/w.

Is there any DT based h/w using this? I'm not aware of any.

I'm told there is. I'll let them come out of the wood work to confirm it ...

Quoted 6 lines

I would
prefer not merging this until there is. I have little insight whether these genericish compatibles will be sufficient, but I have lots of experience to say they won't be. I would also suspect that if anyone has started using this, they've just extended/modified it however they wanted and no feedback got to me.

Sure, what are you looking for here, Reviewed-by tags from someone with a hardware platform that is going to ship with DT?

Quoted 79 lines
Quoted 55 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +49 -0
new file mode 100644
index 000000000000..9d542ecb1a7d
@@ -0,0 +1,227 @@
+# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/arm/arm,mpam-msc.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Arm Memory System Resource Partitioning and Monitoring (MPAM)
+
+description: |
+  The Arm MPAM specification can be found here:
+
+  https://developer.arm.com/documentation/ddi0598/latest
+
+maintainers:
+  - Rob Herring <[email protected]>
+
+properties:
+  compatible:
+    items:
+      - const: arm,mpam-msc                   # Further details are discoverable
+      - const: arm,mpam-memory-controller-msc
+
+  reg:
+    maxItems: 1
+    description: A memory region containing registers as defined in the MPAM
+      specification.
+
+  interrupts:
+    minItems: 1
+    items:
+      - description: error (optional)
+      - description: overflow (optional, only for monitoring)
+
+  interrupt-names:
+    oneOf:
+      - items:
+          - enum: [ error, overflow ]
+      - items:
+          - const: error
+          - const: overflow
+
+  arm,not-ready-us:
+    description: The maximum time in microseconds for monitoring data to be
+      accurate after a settings change. For more information, see the
+      Not-Ready (NRDY) bit description in the MPAM specification.
+
+  numa-node-id: true # see NUMA binding
+
+  '#address-cells':

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 59

Quoted 10 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +10 -0
+    const: 1
+
+  '#size-cells':
+    const: 0
+
+patternProperties:
+  '^ris@[0-9a-f]$':
+    type: object
+    additionalProperties: false
+    description: |

'|' can be dropped.

Quoted 11 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +11 -0
+      RIS nodes for each RIS in an MSC. These nodes are required for each RIS
+      implementing known MPAM controls
+
+    properties:
+      compatible:
+        enum:
+            # Bulk storage for cache
+          - arm,mpam-cache
+            # Memory bandwidth
+          - arm,mpam-memory
+
Quoted 8 lines

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 76

Quoted 6 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +6 -0
+      reg:
+        minimum: 0
+        maximum: 0xf
+
+      cpus:
+        $ref: '/schemas/types.yaml#/definitions/phandle-array'

Don't need the type. It's in the core schemas now.

(I assume the type is that '$ref' line)

Quoted 9 lines

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 82

Quoted 6 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +6 -0
+        description:
+          Phandle(s) to the CPU node(s) this RIS belongs to. By default, the parent
+          device's affinity is used.
+
+      arm,mpam-device:
+        $ref: '/schemas/types.yaml#/definitions/phandle'

Don't need quotes. This should be a warning, but no testing happened because the DT list and maintainers weren't CCed.

Yup, its an RFC - I only CC'd the folk that have previously expressed an interest for the first pass. (git send-email put you on CC!)

Quoted 48 lines
Quoted 35 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +35 -0
+        description:
+          By default, the MPAM enabled device associated with a RIS is the MSC's
+          parent node. It is possible for each RIS to be associated with different
+          devices in which case 'arm,mpam-device' should be used.
+
+    required:
+      - compatible
+      - reg
+
+required:
+  - compatible
+  - reg
+
+dependencies:
+  interrupts: [ interrupt-names ]
+
+additionalProperties: false
+
+examples:
+  - |
+    /*
+    cpus {
+        cpu@0 {
+            next-level-cache = <&L2_0>;
+        };
+        cpu@100 {
+            next-level-cache = <&L2_1>;
+        };
+    };
+    */
+    L2_0: cache-controller-0 {
+        compatible = "cache";
+        cache-level = <2>;
+        cache-unified;
+        next-level-cache = <&L3>;

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 127

Quoted 10 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +10 -0
+
+    };
+
+    L2_1: cache-controller-1 {
+        compatible = "cache";
+        cache-level = <2>;
+        cache-unified;
+        next-level-cache = <&L3>;
+
+    };

All the above should be dropped. Not part of this binding.

Quoted 5 lines

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 130

Quoted 3 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +3 -0
+
+    L3: cache-controller@30000000 {
+        compatible = "arm,dsu-l3-cache", "cache";

Pretty sure this is a warning because that compatible doesn't exist.

Not sure what to do with this. I see plenty of DT in the tree with 'cache', and you've got 'foo,a-memory-controller' below ...

Quoted 27 lines
Quoted 27 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +27 -0
+        cache-level = <3>;
+        cache-unified;
+
+        ranges = <0x0 0x30000000 0x800000>;
+        #address-cells = <1>;
+        #size-cells = <1>;
+
+        msc@10000 {
+            compatible = "arm,mpam-msc";
+
+            /* CPU affinity implied by parent cache node's  */
+            reg = <0x10000 0x2000>;
+            interrupts = <1>, <2>;
+            interrupt-names = "error", "overflow";
+            arm,not-ready-us = <1>;
+        };
+    };
+
+    mem: memory-controller@20000 {
+        compatible = "foo,a-memory-controller";
+        reg = <0x20000 0x1000>;
+
+        #address-cells = <1>;
+        #size-cells = <1>;
+        ranges;
+
+        msc@21000 {
Quoted 8 lines

Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml line 165

Quoted 8 lines
Documentation/devicetree/bindings/arm/arm,mpam-msc.yaml new file +8 -0
+            compatible = "arm,mpam-memory-controller-msc", "arm,mpam-msc";
+            reg = <0x21000 0x1000>;
+            interrupts = <3>;
+            interrupt-names = "error";
+            arm,not-ready-us = <1>;
+            numa-node-id = <1>;
+        };
+    };

James

author maintainer
#41 lore raw

RFC PATCH 12/36 platform: arm64: Move ec devices to an ec subdirectory

In reply to James Morse

commit 363c8aea257 "platform: Add ARM64 platform directory" added a subdirectory for arm64 platform devices, but claims that all such devices must be 'EC like'.

The arm64 MPAM driver manages an MMIO interface that appears in memory controllers, caches, IOMMU and connection points on the interconnect. It doesn't fit into any existing subsystem.

It would be convenient to use this subdirectory for drivers for other arm64 platform devices which aren't closely coupled to the architecture code and don't fit into any existing subsystem.

Move the existing code and maintainer entries to be under drivers/platform/arm64/ec. The MPAM driver will be added under drivers/platform/arm64/mpam.

---
 MAINTAINERS                                   |  6 +-
 drivers/platform/arm64/Kconfig                | 72 +-----------------
 drivers/platform/arm64/Makefile               |  9 +--
 drivers/platform/arm64/ec/Kconfig             | 73 +++++++++++++++++++
 drivers/platform/arm64/ec/Makefile            | 10 +++
 .../platform/arm64/{ => ec}/acer-aspire1-ec.c |  0
 .../arm64/{ => ec}/huawei-gaokun-ec.c         |  0
 .../arm64/{ => ec}/lenovo-yoga-c630.c         |  0
 8 files changed, 88 insertions(+), 82 deletions(-)
 create mode 100644 drivers/platform/arm64/ec/Kconfig
 create mode 100644 drivers/platform/arm64/ec/Makefile
 rename drivers/platform/arm64/{ => ec}/acer-aspire1-ec.c (100%)
 rename drivers/platform/arm64/{ => ec}/huawei-gaokun-ec.c (100%)
 rename drivers/platform/arm64/{ => ec}/lenovo-yoga-c630.c (100%)
MAINTAINERS +3 -3
index 4bac4ea21b64..bea01d413666 100644
@@ -3549,15 +3549,15 @@ S:	Maintained
 F:	arch/arm64/boot/Makefile
 F:	scripts/make_fit.py
 
-ARM64 PLATFORM DRIVERS
-M:	Hans de Goede <[email protected]>
+ARM64 EC PLATFORM DRIVERS
+M:	Hans de Goede <[email protected]>
 M:	Ilpo Järvinen <[email protected]>
 R:	Bryan O'Donoghue <[email protected]>
 L:	[email protected]
 S:	Maintained
 Q:	https://patchwork.kernel.org/project/platform-driver-x86/list/
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86.git
-F:	drivers/platform/arm64/
+F:	drivers/platform/arm64/ec
 
 ARM64 PORT (AARCH64 ARCHITECTURE)
 M:	Catalin Marinas <[email protected]>
drivers/platform/arm64/Kconfig +1 -71

74 lines not on this page · raw message

drivers/platform/arm64/Makefile +1 -8
index 46a99eba3264..ce840a8cf8cc 100644
@@ -1,10 +1,3 @@
 # SPDX-License-Identifier: GPL-2.0-only
-#
-# Makefile for linux/drivers/platform/arm64
-#
-# This dir should only include drivers for EC-like devices.
-#
 
-obj-$(CONFIG_EC_ACER_ASPIRE1)	+= acer-aspire1-ec.o
-obj-$(CONFIG_EC_HUAWEI_GAOKUN)	+= huawei-gaokun-ec.o
-obj-$(CONFIG_EC_LENOVO_YOGA_C630) += lenovo-yoga-c630.o
+obj-y             += ec/
drivers/platform/arm64/ec/Kconfig new file +73 -0

73 lines not on this page · raw message

drivers/platform/arm64/ec/Makefile new file +10 -0
new file mode 100644
index 000000000000..b3a7c4096f08
@@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Makefile for linux/drivers/platform/arm64/ec
+#
+# This dir should only include drivers for EC-like devices.
+#
+
+obj-$(CONFIG_EC_ACER_ASPIRE1)	+= acer-aspire1-ec.o
+obj-$(CONFIG_EC_HUAWEI_GAOKUN)	+= huawei-gaokun-ec.o
+obj-$(CONFIG_EC_LENOVO_YOGA_C630) += lenovo-yoga-c630.o
drivers/platform/arm64/ec/acer-aspire1-ec.c renamed from drivers/platform/arm64/acer-aspire1-ec.c +0 -0
similarity index 100%
rename from drivers/platform/arm64/acer-aspire1-ec.c
rename to drivers/platform/arm64/ec/acer-aspire1-ec.c
drivers/platform/arm64/ec/huawei-gaokun-ec.c renamed from drivers/platform/arm64/huawei-gaokun-ec.c +0 -0
similarity index 100%
rename from drivers/platform/arm64/huawei-gaokun-ec.c
rename to drivers/platform/arm64/ec/huawei-gaokun-ec.c
drivers/platform/arm64/ec/lenovo-yoga-c630.c renamed from drivers/platform/arm64/lenovo-yoga-c630.c +0 -0
similarity index 100%
rename from drivers/platform/arm64/lenovo-yoga-c630.c
rename to drivers/platform/arm64/ec/lenovo-yoga-c630.c
Signature · 1 line
2.39.5
#42 lore raw

Re: [RFC PATCH 12/36] platform: arm64: Move ec devices to an ec subdirectory

Quoted 25 lines · On Fri, 11 Jul 2025 18:36:24 +0000 James Morse <[email protected]> wrote:

commit 363c8aea257 "platform: Add ARM64 platform directory" added a subdirectory for arm64 platform devices, but claims that all such devices must be 'EC like'.

The arm64 MPAM driver manages an MMIO interface that appears in memory controllers, caches, IOMMU and connection points on the interconnect. It doesn't fit into any existing subsystem.

It would be convenient to use this subdirectory for drivers for other arm64 platform devices which aren't closely coupled to the architecture code and don't fit into any existing subsystem.

Move the existing code and maintainer entries to be under drivers/platform/arm64/ec. The MPAM driver will be added under drivers/platform/arm64/mpam.

MAINTAINERS +0 -0
index 4bac4ea21b64..bea01d413666 100644

MAINTAINERS S: Maintained · line 3553

Quoted 7 lines
MAINTAINERS +2 -2
@@ -3549,15 +3549,15 @@ S:	Maintained
 F:	arch/arm64/boot/Makefile
 F:	scripts/make_fit.py
 
-ARM64 PLATFORM DRIVERS
-M:	Hans de Goede <[email protected]>
+ARM64 EC PLATFORM DRIVERS
+M:	Hans de Goede <[email protected]>

Smells like a rebase error as Hans' email address chagned to the kernel.org one in the 6.16 cycle.

MAINTAINERS S: Maintained · line 3560

Quoted 8 lines
MAINTAINERS +1 -1
 M:	Ilpo Järvinen <[email protected]>
 R:	Bryan O'Donoghue <[email protected]>
 L:	[email protected]
 S:	Maintained
 Q:	https://patchwork.kernel.org/project/platform-driver-x86/list/
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86.git
-F:	drivers/platform/arm64/
+F:	drivers/platform/arm64/ec

Other than that looks sensible to me but obviously needs tags from Hans or Ilpo.

Jonathan

author maintainer
#43 lore raw

Re: [RFC PATCH 12/36] platform: arm64: Move ec devices to an ec subdirectory

Hi Jonathan,

Quoted 27 lines · Jonathan Cameron
Quoted 24 lines · On Fri, 11 Jul 2025 18:36:24 +0000 James Morse <[email protected]> wrote:

commit 363c8aea257 "platform: Add ARM64 platform directory" added a subdirectory for arm64 platform devices, but claims that all such devices must be 'EC like'.

The arm64 MPAM driver manages an MMIO interface that appears in memory controllers, caches, IOMMU and connection points on the interconnect. It doesn't fit into any existing subsystem.

It would be convenient to use this subdirectory for drivers for other arm64 platform devices which aren't closely coupled to the architecture code and don't fit into any existing subsystem.

Move the existing code and maintainer entries to be under drivers/platform/arm64/ec. The MPAM driver will be added under drivers/platform/arm64/mpam.

MAINTAINERS +0 -0
index 4bac4ea21b64..bea01d413666 100644
@@ -3549,15 +3549,15 @@ S:	Maintained
 F:	arch/arm64/boot/Makefile
 F:	scripts/make_fit.py
 
Quoted 7 lines

MAINTAINERS S: Maintained · line 3553

Quoted 4 lines
MAINTAINERS +2 -2
-ARM64 PLATFORM DRIVERS
-M:	Hans de Goede <[email protected]>
+ARM64 EC PLATFORM DRIVERS
+M:	Hans de Goede <[email protected]>

Smells like a rebase error as Hans' email address chagned to the kernel.org one in the 6.16 cycle.

Bother - yes, that's exactly what happened.

Quoted 10 lines

MAINTAINERS S: Maintained · line 3560

Quoted 8 lines
MAINTAINERS +1 -1
 M:	Ilpo Järvinen <[email protected]>
 R:	Bryan O'Donoghue <[email protected]>
 L:	[email protected]
 S:	Maintained
 Q:	https://patchwork.kernel.org/project/platform-driver-x86/list/
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86.git
-F:	drivers/platform/arm64/
+F:	drivers/platform/arm64/ec

Other than that looks sensible to me but obviously needs tags from Hans or Ilpo.

Thanks,

James

maintainer
#44 lore raw

Re: [RFC PATCH 12/36] platform: arm64: Move ec devices to an ec subdirectory

In reply to James Morse

Hi James,

Quoted 165 lines · James Morse

commit 363c8aea257 "platform: Add ARM64 platform directory" added a subdirectory for arm64 platform devices, but claims that all such devices must be 'EC like'.

The arm64 MPAM driver manages an MMIO interface that appears in memory controllers, caches, IOMMU and connection points on the interconnect. It doesn't fit into any existing subsystem.

It would be convenient to use this subdirectory for drivers for other arm64 platform devices which aren't closely coupled to the architecture code and don't fit into any existing subsystem.

Move the existing code and maintainer entries to be under drivers/platform/arm64/ec. The MPAM driver will be added under drivers/platform/arm64/mpam.

---
  MAINTAINERS                                   |  6 +-
  drivers/platform/arm64/Kconfig                | 72 +-----------------
  drivers/platform/arm64/Makefile               |  9 +--
  drivers/platform/arm64/ec/Kconfig             | 73 +++++++++++++++++++
  drivers/platform/arm64/ec/Makefile            | 10 +++
  .../platform/arm64/{ => ec}/acer-aspire1-ec.c |  0
  .../arm64/{ => ec}/huawei-gaokun-ec.c         |  0
  .../arm64/{ => ec}/lenovo-yoga-c630.c         |  0
  8 files changed, 88 insertions(+), 82 deletions(-)

create mode 100644 drivers/platform/arm64/ec/Kconfig create mode 100644 drivers/platform/arm64/ec/Makefile rename drivers/platform/arm64/{ => ec}/acer-aspire1-ec.c (100%) rename drivers/platform/arm64/{ => ec}/huawei-gaokun-ec.c (100%) rename drivers/platform/arm64/{ => ec}/lenovo-yoga-c630.c (100%)

MAINTAINERS +3 -3
index 4bac4ea21b64..bea01d413666 100644
@@ -3549,15 +3549,15 @@ S:	Maintained
  F:	arch/arm64/boot/Makefile
  F:	scripts/make_fit.py
 
-ARM64 PLATFORM DRIVERS
-M:	Hans de Goede <[email protected]>
+ARM64 EC PLATFORM DRIVERS
+M:	Hans de Goede <[email protected]>
  M:	Ilpo Järvinen <[email protected]>
  R:	Bryan O'Donoghue <[email protected]>
  L:	[email protected]
  S:	Maintained
  Q:	https://patchwork.kernel.org/project/platform-driver-x86/list/
  T:	git git://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86.git
-F:	drivers/platform/arm64/
+F:	drivers/platform/arm64/ec
 
  ARM64 PORT (AARCH64 ARCHITECTURE)
  M:	Catalin Marinas <[email protected]>
drivers/platform/arm64/Kconfig +1 -71
index 06288aebc559..1eb8ab0855e5 100644
@@ -1,73 +1,3 @@
  # SPDX-License-Identifier: GPL-2.0-only
-#
-# EC-like Drivers for aarch64 based devices.
-#
 
-menuconfig ARM64_PLATFORM_DEVICES
-	bool "ARM64 Platform-Specific Device Drivers"
-	depends on ARM64 || COMPILE_TEST
-	default ARM64
-	help
-	  Say Y here to get to see options for platform-specific device drivers
-	  for arm64 based devices, primarily EC-like device drivers.
-	  This option alone does not add any kernel code.
-
-	  If you say N, all options in this submenu will be skipped and disabled.
-
-if ARM64_PLATFORM_DEVICES
 
-
-config EC_ACER_ASPIRE1
-	tristate "Acer Aspire 1 Embedded Controller driver"
-	depends on ARCH_QCOM || COMPILE_TEST
-	depends on I2C
-	depends on DRM
-	depends on POWER_SUPPLY
-	depends on INPUT
-	help
-	  Say Y here to enable the EC driver for the (Snapdragon-based)
-	  Acer Aspire 1 laptop. The EC handles battery and charging
-	  monitoring as well as some misc functions like the lid sensor
-	  and USB Type-C DP HPD events.
-
-	  This driver provides battery and AC status support for the mentioned
-	  laptop where this information is not properly exposed via the
-	  standard ACPI devices.
-
-config EC_HUAWEI_GAOKUN
-	tristate "Huawei Matebook E Go Embedded Controller driver"
-	depends on ARCH_QCOM || COMPILE_TEST
-	depends on I2C
-	depends on INPUT
-	depends on HWMON
-	select AUXILIARY_BUS
-
-	help
-	  Say Y here to enable the EC driver for the Huawei Matebook E Go
-	  which is a sc8280xp-based 2-in-1 tablet. The driver handles battery
-	  (information, charge control) and USB Type-C DP HPD events as well
-	  as some misc functions like the lid sensor and temperature sensors,
-	  etc.
-
-	  This driver provides battery and AC status support for the mentioned
-	  laptop where this information is not properly exposed via the
-	  standard ACPI devices.
-
-	  Say M or Y here to include this support.
-
-config EC_LENOVO_YOGA_C630
-	tristate "Lenovo Yoga C630 Embedded Controller driver"
-	depends on ARCH_QCOM || COMPILE_TEST
-	depends on I2C
-	select AUXILIARY_BUS
-	help
-	  Driver for the Embedded Controller in the Qualcomm Snapdragon-based
-	  Lenovo Yoga C630, which provides battery and power adapter
-	  information.
-
-	  This driver provides battery and AC status support for the mentioned
-	  laptop where this information is not properly exposed via the
-	  standard ACPI devices.
-
-	  Say M or Y here to include this support.
-
-endif # ARM64_PLATFORM_DEVICES
+source "drivers/platform/arm64/ec/Kconfig"
drivers/platform/arm64/Makefile +1 -8
index 46a99eba3264..ce840a8cf8cc 100644
@@ -1,10 +1,3 @@
  # SPDX-License-Identifier: GPL-2.0-only
-#
-# Makefile for linux/drivers/platform/arm64
-#
-# This dir should only include drivers for EC-like devices.
-#
 
-obj-$(CONFIG_EC_ACER_ASPIRE1)	+= acer-aspire1-ec.o
-obj-$(CONFIG_EC_HUAWEI_GAOKUN)	+= huawei-gaokun-ec.o
-obj-$(CONFIG_EC_LENOVO_YOGA_C630) += lenovo-yoga-c630.o
+obj-y             += ec/
drivers/platform/arm64/ec/Kconfig new file +7 -0
new file mode 100644
index 000000000000..06288aebc559
@@ -0,0 +1,73 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# EC-like Drivers for aarch64 based devices.
+#
+
+menuconfig ARM64_PLATFORM_DEVICES
+	bool "ARM64 Platform-Specific Device Drivers"

drivers/platform/arm64/ec/Kconfig line 17

Quoted 10 lines
drivers/platform/arm64/ec/Kconfig new file +10 -0
+	depends on ARM64 || COMPILE_TEST
+	default ARM64
+	help
+	  Say Y here to get to see options for platform-specific device drivers
+	  for arm64 based devices, primarily EC-like device drivers.
+	  This option alone does not add any kernel code.
+
+	  If you say N, all options in this submenu will be skipped and disabled.
+
+if ARM64_PLATFORM_DEVICES

Shouldn't this be kept in the directory above? By the description this would be expected to apply to all drivers in drivers/platfrom/arm64.> +

drivers/platform/arm64/ec/lenovo-yoga-c630.c

Quoted 83 lines
drivers/platform/arm64/ec/Kconfig new file +55 -0
+config EC_ACER_ASPIRE1
+	tristate "Acer Aspire 1 Embedded Controller driver"
+	depends on ARCH_QCOM || COMPILE_TEST
+	depends on I2C
+	depends on DRM
+	depends on POWER_SUPPLY
+	depends on INPUT
+	help
+	  Say Y here to enable the EC driver for the (Snapdragon-based)
+	  Acer Aspire 1 laptop. The EC handles battery and charging
+	  monitoring as well as some misc functions like the lid sensor
+	  and USB Type-C DP HPD events.
+
+	  This driver provides battery and AC status support for the mentioned
+	  laptop where this information is not properly exposed via the
+	  standard ACPI devices.
+
+config EC_HUAWEI_GAOKUN
+	tristate "Huawei Matebook E Go Embedded Controller driver"
+	depends on ARCH_QCOM || COMPILE_TEST
+	depends on I2C
+	depends on INPUT
+	depends on HWMON
+	select AUXILIARY_BUS
+
+	help
+	  Say Y here to enable the EC driver for the Huawei Matebook E Go
+	  which is a sc8280xp-based 2-in-1 tablet. The driver handles battery
+	  (information, charge control) and USB Type-C DP HPD events as well
+	  as some misc functions like the lid sensor and temperature sensors,
+	  etc.
+
+	  This driver provides battery and AC status support for the mentioned
+	  laptop where this information is not properly exposed via the
+	  standard ACPI devices.
+
+	  Say M or Y here to include this support.
+
+config EC_LENOVO_YOGA_C630
+	tristate "Lenovo Yoga C630 Embedded Controller driver"
+	depends on ARCH_QCOM || COMPILE_TEST
+	depends on I2C
+	select AUXILIARY_BUS
+	help
+	  Driver for the Embedded Controller in the Qualcomm Snapdragon-based
+	  Lenovo Yoga C630, which provides battery and power adapter
+	  information.
+
+	  This driver provides battery and AC status support for the mentioned
+	  laptop where this information is not properly exposed via the
+	  standard ACPI devices.
+
+	  Say M or Y here to include this support.
+
+endif # ARM64_PLATFORM_DEVICES
drivers/platform/arm64/ec/Makefile new file +10 -0
new file mode 100644
index 000000000000..b3a7c4096f08
@@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Makefile for linux/drivers/platform/arm64/ec
+#
+# This dir should only include drivers for EC-like devices.
+#
+
+obj-$(CONFIG_EC_ACER_ASPIRE1)	+= acer-aspire1-ec.o
+obj-$(CONFIG_EC_HUAWEI_GAOKUN)	+= huawei-gaokun-ec.o
+obj-$(CONFIG_EC_LENOVO_YOGA_C630) += lenovo-yoga-c630.o
drivers/platform/arm64/ec/acer-aspire1-ec.c renamed from drivers/platform/arm64/acer-aspire1-ec.c +0 -0
similarity index 100%
rename from drivers/platform/arm64/acer-aspire1-ec.c
rename to drivers/platform/arm64/ec/acer-aspire1-ec.c
drivers/platform/arm64/ec/huawei-gaokun-ec.c renamed from drivers/platform/arm64/huawei-gaokun-ec.c +0 -0
similarity index 100%
rename from drivers/platform/arm64/huawei-gaokun-ec.c
rename to drivers/platform/arm64/ec/huawei-gaokun-ec.c
drivers/platform/arm64/ec/lenovo-yoga-c630.c renamed from drivers/platform/arm64/lenovo-yoga-c630.c +0 -0
similarity index 100%
rename from drivers/platform/arm64/lenovo-yoga-c630.c
rename to drivers/platform/arm64/ec/lenovo-yoga-c630.c

Thanks,

Ben

author maintainer
#45 lore raw

Re: [RFC PATCH 12/36] platform: arm64: Move ec devices to an ec subdirectory

Hi Ben,

Quoted 36 lines · Ben Horgan
Quoted 35 lines · James Morse

commit 363c8aea257 "platform: Add ARM64 platform directory" added a subdirectory for arm64 platform devices, but claims that all such devices must be 'EC like'.

The arm64 MPAM driver manages an MMIO interface that appears in memory controllers, caches, IOMMU and connection points on the interconnect. It doesn't fit into any existing subsystem.

It would be convenient to use this subdirectory for drivers for other arm64 platform devices which aren't closely coupled to the architecture code and don't fit into any existing subsystem.

Move the existing code and maintainer entries to be under drivers/platform/arm64/ec. The MPAM driver will be added under drivers/platform/arm64/mpam.

drivers/platform/arm64/ec/Kconfig new file +12 -0
new file mode 100644
index 000000000000..06288aebc559
@@ -0,0 +1,73 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# EC-like Drivers for aarch64 based devices.
+#
+
+menuconfig ARM64_PLATFORM_DEVICES
+    bool "ARM64 Platform-Specific Device Drivers"
+    depends on ARM64 || COMPILE_TEST
+    default ARM64
+    help
+      Say Y here to get to see options for platform-specific device drivers
+      for arm64 based devices, primarily EC-like device drivers.
Quoted 8 lines

drivers/platform/arm64/ec/Kconfig line 17

Quoted 5 lines
drivers/platform/arm64/ec/Kconfig new file +5 -0
+      This option alone does not add any kernel code.
+
+      If you say N, all options in this submenu will be skipped and disabled.
+
+if ARM64_PLATFORM_DEVICES

Shouldn't this be kept in the directory above? By the description this would be expected to apply to all drivers in drivers/platfrom/arm64.

Doing that makes any MPAM options appear under 'Platform-Specific Device Drivers' too. I didn't to that as MPAM isn't specific to one platform, but doesn't fit under any of the bus or high level groups under drivers. (I briefly toyed with drivers/perf - as there isn't a drivers/qos).

Thanks,

James

author maintainer
#46 lore raw

RFC PATCH 13/36 arm_mpam: Add probe/remove for mpam msc driver and kbuild boiler plate

In reply to James Morse

Probing MPAM is convoluted. MSCs that are integrated with a CPU may only be accessible from those CPUs, and they may not be online. Touching the hardware early is pointless as MPAM can't be used until the system-wide common values for num_partid and num_pmg have been discovered.

Start with driver probe/remove and mapping the MSC.

---
 arch/arm64/Kconfig                          |   1 +
 drivers/platform/arm64/Kconfig              |   1 +
 drivers/platform/arm64/Makefile             |   1 +
 drivers/platform/arm64/mpam/Kconfig         |  10 +
 drivers/platform/arm64/mpam/Makefile        |   4 +
 drivers/platform/arm64/mpam/mpam_devices.c  | 336 ++++++++++++++++++++
 drivers/platform/arm64/mpam/mpam_internal.h |  62 ++++
 7 files changed, 415 insertions(+)
 create mode 100644 drivers/platform/arm64/mpam/Kconfig
 create mode 100644 drivers/platform/arm64/mpam/Makefile
 create mode 100644 drivers/platform/arm64/mpam/mpam_devices.c
 create mode 100644 drivers/platform/arm64/mpam/mpam_internal.h
arch/arm64/Kconfig +1 -0
index ad9a49a39e41..8abce7f4eb1e 100644
@@ -2060,6 +2060,7 @@ config ARM64_TLB_RANGE
 
 config ARM64_MPAM
 	bool "Enable support for MPAM"
+	select ARM64_MPAM_DRIVER
 	select ACPI_MPAM if ACPI
 	help
 	  Memory Partitioning and Monitoring is an optional extension
drivers/platform/arm64/Kconfig +1 -0
index 1eb8ab0855e5..16a927cf6ea2 100644
@@ -1,3 +1,4 @@
 # SPDX-License-Identifier: GPL-2.0-only
 
 source "drivers/platform/arm64/ec/Kconfig"
+source "drivers/platform/arm64/mpam/Kconfig"
drivers/platform/arm64/Makefile +1 -0
index ce840a8cf8cc..c6ec3bc6a100 100644
@@ -1,3 +1,4 @@
 # SPDX-License-Identifier: GPL-2.0-only
 
 obj-y             += ec/
+obj-y             += mpam/
drivers/platform/arm64/mpam/Kconfig new file +10 -0
new file mode 100644
index 000000000000..b63495d7da87
@@ -0,0 +1,10 @@
+# Confusingly, this is everything but the CPU bits of MPAM. CPU here means
+# CPU resources, not containers or cgroups etc.
+config ARM_CPU_RESCTRL
+	bool
+	depends on ARM64
+
+config ARM64_MPAM_DRIVER_DEBUG
+	bool "Enable debug messages from the MPAM driver."
+	help
+	  Say yes here to enable debug messages from the MPAM driver.
drivers/platform/arm64/mpam/Makefile new file +4 -0
new file mode 100644
index 000000000000..4255975c7724
@@ -0,0 +1,4 @@
+obj-$(CONFIG_ARM64_MPAM)			+= mpam.o
+mpam-y						+= mpam_devices.o
+
+cflags-$(CONFIG_ARM64_MPAM_DRIVER_DEBUG)	+= -DDEBUG
drivers/platform/arm64/mpam/mpam_devices.c new file +336 -0

336 lines not on this page · raw message

drivers/platform/arm64/mpam/mpam_internal.h new file +62 -0

62 lines not on this page · raw message

Signature · 1 line
2.39.5
maintainer
#47 lore raw

Re: [RFC PATCH 13/36] arm_mpam: Add probe/remove for mpam msc driver and kbuild boiler plate

Hi James,

Quoted 129 lines · James Morse

Probing MPAM is convoluted. MSCs that are integrated with a CPU may only be accessible from those CPUs, and they may not be online. Touching the hardware early is pointless as MPAM can't be used until the system-wide common values for num_partid and num_pmg have been discovered.

Start with driver probe/remove and mapping the MSC.

---
  arch/arm64/Kconfig                          |   1 +
  drivers/platform/arm64/Kconfig              |   1 +
  drivers/platform/arm64/Makefile             |   1 +
  drivers/platform/arm64/mpam/Kconfig         |  10 +
  drivers/platform/arm64/mpam/Makefile        |   4 +
  drivers/platform/arm64/mpam/mpam_devices.c  | 336 ++++++++++++++++++++
  drivers/platform/arm64/mpam/mpam_internal.h |  62 ++++
  7 files changed, 415 insertions(+)

create mode 100644 drivers/platform/arm64/mpam/Kconfig create mode 100644 drivers/platform/arm64/mpam/Makefile create mode 100644 drivers/platform/arm64/mpam/mpam_devices.c create mode 100644 drivers/platform/arm64/mpam/mpam_internal.h

arch/arm64/Kconfig +1 -0
index ad9a49a39e41..8abce7f4eb1e 100644
@@ -2060,6 +2060,7 @@ config ARM64_TLB_RANGE
 
  config ARM64_MPAM
  	bool "Enable support for MPAM"
+	select ARM64_MPAM_DRIVER
  	select ACPI_MPAM if ACPI
  	help
  	  Memory Partitioning and Monitoring is an optional extension
drivers/platform/arm64/Kconfig +1 -0
index 1eb8ab0855e5..16a927cf6ea2 100644
@@ -1,3 +1,4 @@
  # SPDX-License-Identifier: GPL-2.0-only
 
  source "drivers/platform/arm64/ec/Kconfig"
+source "drivers/platform/arm64/mpam/Kconfig"
drivers/platform/arm64/Makefile +1 -0
index ce840a8cf8cc..c6ec3bc6a100 100644
@@ -1,3 +1,4 @@
  # SPDX-License-Identifier: GPL-2.0-only
 
  obj-y             += ec/
+obj-y             += mpam/
drivers/platform/arm64/mpam/Kconfig new file +10 -0
new file mode 100644
index 000000000000..b63495d7da87
@@ -0,0 +1,10 @@
+# Confusingly, this is everything but the CPU bits of MPAM. CPU here means
+# CPU resources, not containers or cgroups etc.
+config ARM_CPU_RESCTRL
+	bool
+	depends on ARM64
+
+config ARM64_MPAM_DRIVER_DEBUG
+	bool "Enable debug messages from the MPAM driver."
+	help
+	  Say yes here to enable debug messages from the MPAM driver.
drivers/platform/arm64/mpam/Makefile new file +4 -0
new file mode 100644
index 000000000000..4255975c7724
@@ -0,0 +1,4 @@
+obj-$(CONFIG_ARM64_MPAM)			+= mpam.o
+mpam-y						+= mpam_devices.o
+
+cflags-$(CONFIG_ARM64_MPAM_DRIVER_DEBUG)	+= -DDEBUG
drivers/platform/arm64/mpam/mpam_devices.c new file +43 -0
new file mode 100644
index 000000000000..5b886ba54ba8
@@ -0,0 +1,336 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (C) 2025 Arm Ltd.
+
+#define pr_fmt(fmt) "%s:%s: " fmt, KBUILD_MODNAME, __func__
+
+#include <linux/acpi.h>
+#include <linux/arm_mpam.h>
+#include <linux/cacheinfo.h>
+#include <linux/cpu.h>
+#include <linux/cpumask.h>
+#include <linux/device.h>
+#include <linux/errno.h>
+#include <linux/gfp.h>
+#include <linux/list.h>
+#include <linux/lockdep.h>
+#include <linux/mutex.h>
+#include <linux/of.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/printk.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/srcu.h>
+#include <linux/types.h>
+
+#include <acpi/pcc.h>
+
+#include "mpam_internal.h"
+
+/*
+ * mpam_list_lock protects the SRCU lists when writing. Once the
+ * mpam_enabled key is enabled these lists are read-only,
+ * unless the error interrupt disables the driver.
+ */
+static DEFINE_MUTEX(mpam_list_lock);
+static LIST_HEAD(mpam_all_msc);
+
+static struct srcu_struct mpam_srcu;
+
+/* MPAM isn't available until all the MSC have been probed. */
+static u32 mpam_num_msc;
+
+static void mpam_discovery_complete(void)

drivers/platform/arm64/mpam/mpam_devices.c line 53

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c new file +10 -0
+{
+	pr_err("Discovered all MSC\n");
+}
+
+static int mpam_dt_count_msc(void)
+{
+	int count = 0;
+	struct device_node *np;
+
+	for_each_compatible_node(np, NULL, "arm,mpam-msc")

This will count even 'status = "disabled"' nodes. Add a check for that.

if (of_device_is_available(np))> + count++;

Quoted 340 lines
drivers/platform/arm64/mpam/mpam_devices.c new file +282 -0
+
+	return count;
+}
+
+static int mpam_dt_parse_resource(struct mpam_msc *msc, struct device_node *np,
+				  u32 ris_idx)
+{
+	int err = 0;
+	u32 level = 0;
+	unsigned long cache_id;
+	struct device_node *cache;
+
+	do {
+		if (of_device_is_compatible(np, "arm,mpam-cache")) {
+			cache = of_parse_phandle(np, "arm,mpam-device", 0);
+			if (!cache) {
+				pr_err("Failed to read phandle\n");
+				break;
+			}
+		} else if (of_device_is_compatible(np->parent, "cache")) {
+			cache = of_node_get(np->parent);
+		} else {
+			/* For now, only caches are supported */
+			cache = NULL;
+			break;
+		}
+
+		err = of_property_read_u32(cache, "cache-level", &level);
+		if (err) {
+			pr_err("Failed to read cache-level\n");
+			break;
+		}
+
+		cache_id = cache_of_calculate_id(cache);
+		if (cache_id == ~0UL) {
+			err = -ENOENT;
+			break;
+		}
+
+		err = mpam_ris_create(msc, ris_idx, MPAM_CLASS_CACHE, level,
+				      cache_id);
+	} while (0);
+	of_node_put(cache);
+
+	return err;
+}
+
+static int mpam_dt_parse_resources(struct mpam_msc *msc, void *ignored)
+{
+	int err, num_ris = 0;
+	const u32 *ris_idx_p;
+	struct device_node *iter, *np;
+
+	np = msc->pdev->dev.of_node;
+	for_each_child_of_node(np, iter) {
+		ris_idx_p = of_get_property(iter, "reg", NULL);
+		if (ris_idx_p) {
+			num_ris++;
+			err = mpam_dt_parse_resource(msc, iter, *ris_idx_p);
+			if (err) {
+				of_node_put(iter);
+				return err;
+			}
+		}
+	}
+
+	if (!num_ris)
+		mpam_dt_parse_resource(msc, np, 0);
+
+	return err;
+}
+
+/*
+ * An MSC can control traffic from a set of CPUs, but may only be accessible
+ * from a (hopefully wider) set of CPUs. The common reason for this is power
+ * management. If all the CPUs in a cluster are in PSCI:CPU_SUSPEND, the
+ * the corresponding cache may also be powered off. By making accesses from
+ * one of those CPUs, we ensure this isn't the case.
+ */
+static int update_msc_accessibility(struct mpam_msc *msc)
+{
+	struct device_node *parent;
+	u32 affinity_id;
+	int err;
+
+	if (!acpi_disabled) {
+		err = device_property_read_u32(&msc->pdev->dev, "cpu_affinity",
+					       &affinity_id);
+		if (err) {
+			cpumask_copy(&msc->accessibility, cpu_possible_mask);
+			err = 0;
+		} else {
+			err = acpi_pptt_get_cpus_from_container(affinity_id,
+								&msc->accessibility);
+		}
+
+		return err;
+	}
+
+	/* This depends on the path to of_node */
+	parent = of_get_parent(msc->pdev->dev.of_node);
+	if (parent == of_root) {
+		cpumask_copy(&msc->accessibility, cpu_possible_mask);
+		err = 0;
+	} else {
+		err = -EINVAL;
+		pr_err("Cannot determine accessibility of MSC: %s\n",
+		       dev_name(&msc->pdev->dev));
+	}
+	of_node_put(parent);
+
+	return err;
+}
+
+static int fw_num_msc;
+
+static void mpam_pcc_rx_callback(struct mbox_client *cl, void *msg)
+{
+	/* TODO: wake up tasks blocked on this MSC's PCC channel */
+}
+
+static void mpam_msc_drv_remove(struct platform_device *pdev)
+{
+	struct mpam_msc *msc = platform_get_drvdata(pdev);
+
+	if (!msc)
+		return;
+
+	mutex_lock(&mpam_list_lock);
+	mpam_num_msc--;
+	platform_set_drvdata(pdev, NULL);
+	list_del_rcu(&msc->glbl_list);
+	synchronize_srcu(&mpam_srcu);
+	devm_kfree(&pdev->dev, msc);
+	mutex_unlock(&mpam_list_lock);
+}
+
+static int mpam_msc_drv_probe(struct platform_device *pdev)
+{
+	int err;
+	struct mpam_msc *msc;
+	struct resource *msc_res;
+	void *plat_data = pdev->dev.platform_data;
+
+	mutex_lock(&mpam_list_lock);
+	do {
+		msc = devm_kzalloc(&pdev->dev, sizeof(*msc), GFP_KERNEL);
+		if (!msc) {
+			err = -ENOMEM;
+			break;
+		}
+
+		mutex_init(&msc->probe_lock);
+		mutex_init(&msc->part_sel_lock);
+		mutex_init(&msc->outer_mon_sel_lock);
+		raw_spin_lock_init(&msc->inner_mon_sel_lock);
+		msc->id = mpam_num_msc++;
+		msc->pdev = pdev;
+		INIT_LIST_HEAD_RCU(&msc->glbl_list);
+		INIT_LIST_HEAD_RCU(&msc->ris);
+
+		err = update_msc_accessibility(msc);
+		if (err)
+			break;
+		if (cpumask_empty(&msc->accessibility)) {
+			pr_err_once("msc:%u is not accessible from any CPU!",
+				    msc->id);
+			err = -EINVAL;
+			break;
+		}
+
+		if (device_property_read_u32(&pdev->dev, "pcc-channel",
+					     &msc->pcc_subspace_id))
+			msc->iface = MPAM_IFACE_MMIO;
+		else
+			msc->iface = MPAM_IFACE_PCC;
+
+		if (msc->iface == MPAM_IFACE_MMIO) {
+			void __iomem *io;
+
+			io = devm_platform_get_and_ioremap_resource(pdev, 0,
+								    &msc_res);
+			if (IS_ERR(io)) {
+				pr_err("Failed to map MSC base address\n");
+				err = PTR_ERR(io);
+				break;
+			}
+			msc->mapped_hwpage_sz = msc_res->end - msc_res->start;
+			msc->mapped_hwpage = io;
+		} else if (msc->iface == MPAM_IFACE_PCC) {
+			msc->pcc_cl.dev = &pdev->dev;
+			msc->pcc_cl.rx_callback = mpam_pcc_rx_callback;
+			msc->pcc_cl.tx_block = false;
+			msc->pcc_cl.tx_tout = 1000; /* 1s */
+			msc->pcc_cl.knows_txdone = false;
+
+			msc->pcc_chan = pcc_mbox_request_channel(&msc->pcc_cl,
+								 msc->pcc_subspace_id);
+			if (IS_ERR(msc->pcc_chan)) {
+				pr_err("Failed to request MSC PCC channel\n");
+				err = PTR_ERR(msc->pcc_chan);
+				break;
+			}
+		}
+
+		list_add_rcu(&msc->glbl_list, &mpam_all_msc);
+		platform_set_drvdata(pdev, msc);
+	} while (0);
+	mutex_unlock(&mpam_list_lock);
+
+	if (!err) {
+		/* Create RIS entries described by firmware */
+		if (!acpi_disabled)
+			err = acpi_mpam_parse_resources(msc, plat_data);
+		else
+			err = mpam_dt_parse_resources(msc, plat_data);
+	}
+
+	if (!err && fw_num_msc == mpam_num_msc)
+		mpam_discovery_complete();
+
+	if (err && msc)
+		mpam_msc_drv_remove(pdev);
+
+	return err;
+}
+
+static const struct of_device_id mpam_of_match[] = {
+	{ .compatible = "arm,mpam-msc", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, mpam_of_match);
+
+static struct platform_driver mpam_msc_driver = {
+	.driver = {
+		.name = "mpam_msc",
+		.of_match_table = of_match_ptr(mpam_of_match),
+	},
+	.probe = mpam_msc_drv_probe,
+	.remove = mpam_msc_drv_remove,
+};
+
+/*
+ * MSC that are hidden under caches are not created as platform devices
+ * as there is no cache driver. Caches are also special-cased in
+ * update_msc_accessibility().
+ */
+static void mpam_dt_create_foundling_msc(void)
+{
+	int err;
+	struct device_node *cache;
+
+	for_each_compatible_node(cache, NULL, "cache") {
+		err = of_platform_populate(cache, mpam_of_match, NULL, NULL);
+		if (err)
+			pr_err("Failed to create MSC devices under caches\n");
+	}
+}
+
+static int __init mpam_msc_driver_init(void)
+{
+	if (!system_supports_mpam())
+		return -EOPNOTSUPP;
+
+	init_srcu_struct(&mpam_srcu);
+
+	if (!acpi_disabled)
+		fw_num_msc = acpi_mpam_count_msc();
+	else
+		fw_num_msc = mpam_dt_count_msc();
+
+	if (fw_num_msc <= 0) {
+		pr_err("No MSC devices found in firmware\n");
+		return -EINVAL;
+	}
+
+	if (acpi_disabled)
+		mpam_dt_create_foundling_msc();
+
+	return platform_driver_register(&mpam_msc_driver);
+}
+subsys_initcall(mpam_msc_driver_init);
drivers/platform/arm64/mpam/mpam_internal.h new file +52 -0
new file mode 100644
index 000000000000..07e0f240eaca
@@ -0,0 +1,62 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+// Copyright (C) 2024 Arm Ltd.
+
+#ifndef MPAM_INTERNAL_H
+#define MPAM_INTERNAL_H
+
+#include <linux/arm_mpam.h>
+#include <linux/cpumask.h>
+#include <linux/io.h>
+#include <linux/mailbox_client.h>
+#include <linux/mutex.h>
+#include <linux/resctrl.h>
+#include <linux/sizes.h>
+
+struct mpam_msc {
+	/* member of mpam_all_msc */
+	struct list_head        glbl_list;
+
+	int			id;
+	struct platform_device *pdev;
+
+	/* Not modified after mpam_is_enabled() becomes true */
+	enum mpam_msc_iface	iface;
+	u32			pcc_subspace_id;
+	struct mbox_client	pcc_cl;
+	struct pcc_mbox_chan	*pcc_chan;
+	u32			nrdy_usec;
+	cpumask_t		accessibility;
+
+	/*
+	 * probe_lock is only take during discovery. After discovery these
+	 * properties become read-only and the lists are protected by SRCU.
+	 */
+	struct mutex		probe_lock;
+	unsigned long		ris_idxs[128 / BITS_PER_LONG];
+	u32			ris_max;
+
+	/* mpam_msc_ris of this component */
+	struct list_head	ris;
+
+	/*
+	 * part_sel_lock protects access to the MSC hardware registers that are
+	 * affected by MPAMCFG_PART_SEL. (including the ID registers that vary
+	 * by RIS).
+	 * If needed, take msc->lock first.
+	 */
+	struct mutex		part_sel_lock;
+
+	/*
+	 * mon_sel_lock protects access to the MSC hardware registers that are
+	 * affeted by MPAMCFG_MON_SEL.
+	 * If needed, take msc->lock first.

drivers/platform/arm64/mpam/mpam_internal.h line 62

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h new file +10 -0
+	 */
+	struct mutex		outer_mon_sel_lock;
+	raw_spinlock_t		inner_mon_sel_lock;
+	unsigned long		inner_mon_sel_flags;
+
+	void __iomem		*mapped_hwpage;
+	size_t			mapped_hwpage_sz;
+};
+
+#endif /* MPAM_INTERNAL_H */

Thanks,

Ben

author maintainer
#48 lore raw

Re: [RFC PATCH 13/36] arm_mpam: Add probe/remove for mpam msc driver and kbuild boiler plate

Hi Ben,

Quoted 26 lines · Ben Horgan
Quoted 15 lines · James Morse

Probing MPAM is convoluted. MSCs that are integrated with a CPU may only be accessible from those CPUs, and they may not be online. Touching the hardware early is pointless as MPAM can't be used until the system-wide common values for num_partid and num_pmg have been discovered.

Start with driver probe/remove and mapping the MSC.

drivers/platform/arm64/mpam/ renamed from drivers/platform/arm64/mpam/mpam_devices.c +0 -0

mpam_devices.c
new file mode 100644
index 000000000000..5b886ba54ba8

drivers/platform/arm64/mpam/mpam_devices.c line 53

Quoted 7 lines
drivers/platform/arm64/mpam/mpam_devices.c new file +6 -0
@@ -0,0 +1,336 @@
 
+static int mpam_dt_count_msc(void)
+{
+    int count = 0;
+    struct device_node *np;
+
+    for_each_compatible_node(np, NULL, "arm,mpam-msc")

This will count even 'status = "disabled"' nodes. Add a check for that.

Quoted 1 line

if (of_device_is_available(np))> +        count++;

Good spot, fixed - thanks.

Thanks,

James

Quoted 3 lines
Quoted 3 lines
drivers/platform/arm64/mpam/mpam_devices.c new file +3 -0
+
+    return count;
+}
maintainer
#49 lore raw

Re: [RFC PATCH 13/36] arm_mpam: Add probe/remove for mpam msc driver and kbuild boiler plate

In reply to James Morse
Quoted 10 lines · James Morse

Probing MPAM is convoluted. MSCs that are integrated with a CPU may only be accessible from those CPUs, and they may not be online. Touching the hardware early is pointless as MPAM can't be used until the system-wide common values for num_partid and num_pmg have been discovered.

Start with driver probe/remove and mapping the MSC.

Quoted 13 lines
---
 arch/arm64/Kconfig                          |   1 +
 drivers/platform/arm64/Kconfig              |   1 +
 drivers/platform/arm64/Makefile             |   1 +
 drivers/platform/arm64/mpam/Kconfig         |  10 +
 drivers/platform/arm64/mpam/Makefile        |   4 +
 drivers/platform/arm64/mpam/mpam_devices.c  | 336 ++++++++++++++++++++
 drivers/platform/arm64/mpam/mpam_internal.h |  62 ++++
 7 files changed, 415 insertions(+)
 create mode 100644 drivers/platform/arm64/mpam/Kconfig
 create mode 100644 drivers/platform/arm64/mpam/Makefile
 create mode 100644 drivers/platform/arm64/mpam/mpam_devices.c
 create mode 100644 drivers/platform/arm64/mpam/mpam_internal.h

Bikeshedding: why not drivers/resctrl to match fs/resctrl? We wouldn't need the previous patch either to move the arm64 platform drivers.

I'm not an expert on resctrl but the MPAM code looks more like a backend for the resctrl support, so it makes more sense to do as we did for other drivers like irqchip, iommu. You can create drivers/resctrl/arm64 if you want to keep them grouped.

Signature · 1 line
Catalin
author maintainer
#50 lore raw

Re: [RFC PATCH 13/36] arm_mpam: Add probe/remove for mpam msc driver and kbuild boiler plate

Hi Catalin,

Quoted 22 lines · Catalin Marinas
Quoted 8 lines · James Morse

Probing MPAM is convoluted. MSCs that are integrated with a CPU may only be accessible from those CPUs, and they may not be online. Touching the hardware early is pointless as MPAM can't be used until the system-wide common values for num_partid and num_pmg have been discovered.

Start with driver probe/remove and mapping the MSC.

Quoted 12 lines
 arch/arm64/Kconfig                          |   1 +
 drivers/platform/arm64/Kconfig              |   1 +
 drivers/platform/arm64/Makefile             |   1 +
 drivers/platform/arm64/mpam/Kconfig         |  10 +
 drivers/platform/arm64/mpam/Makefile        |   4 +
 drivers/platform/arm64/mpam/mpam_devices.c  | 336 ++++++++++++++++++++
 drivers/platform/arm64/mpam/mpam_internal.h |  62 ++++
 7 files changed, 415 insertions(+)
 create mode 100644 drivers/platform/arm64/mpam/Kconfig
 create mode 100644 drivers/platform/arm64/mpam/Makefile
 create mode 100644 drivers/platform/arm64/mpam/mpam_devices.c
 create mode 100644 drivers/platform/arm64/mpam/mpam_internal.h
Quoted 2 lines

Bikeshedding: why not drivers/resctrl to match fs/resctrl? We wouldn't need the previous patch either to move the arm64 platform drivers.

Initially because I don't see any other architecture having an MMIO interface to this stuff, and didn't want a 'top level' driver directory for a single driver. But, re-reading RISC-Vs CBQRI[0] it turns out that theirs is memory mapped...

Quoted 3 lines

I'm not an expert on resctrl but the MPAM code looks more like a backend for the resctrl support, so it makes more sense to do as we did for other drivers like irqchip, iommu.

Only because there are many irqchip or iommu. I'm not a fan of drivers/mpam, but drivers/resctrl would suit RISC-V too. (I'll check with Drew)

Thanks,

James

[0] https://patchew.org/linux/[email protected]/

#51 lore raw

Re: [RFC PATCH 13/36] arm_mpam: Add probe/remove for mpam msc driver and kbuild boiler plate

Quoted 25 lines · James Morse

Hi Catalin,

Quoted 21 lines · Catalin Marinas
Quoted 20 lines · James Morse

Probing MPAM is convoluted. MSCs that are integrated with a CPU may only be accessible from those CPUs, and they may not be online. Touching the hardware early is pointless as MPAM can't be used until the system-wide common values for num_partid and num_pmg have been discovered.

Start with driver probe/remove and mapping the MSC.

 arch/arm64/Kconfig                          |   1 +
 drivers/platform/arm64/Kconfig              |   1 +
 drivers/platform/arm64/Makefile             |   1 +
 drivers/platform/arm64/mpam/Kconfig         |  10 +
 drivers/platform/arm64/mpam/Makefile        |   4 +
 drivers/platform/arm64/mpam/mpam_devices.c  | 336 ++++++++++++++++++++
 drivers/platform/arm64/mpam/mpam_internal.h |  62 ++++
 7 files changed, 415 insertions(+)
 create mode 100644 drivers/platform/arm64/mpam/Kconfig
 create mode 100644 drivers/platform/arm64/mpam/Makefile
 create mode 100644 drivers/platform/arm64/mpam/mpam_devices.c
 create mode 100644 drivers/platform/arm64/mpam/mpam_internal.h
Quoted 6 lines
Quoted 2 lines

Bikeshedding: why not drivers/resctrl to match fs/resctrl? We wouldn't need the previous patch either to move the arm64 platform drivers.

Initially because I don't see any other architecture having an MMIO interface to this stuff, and didn't want a 'top level' driver directory for a single driver. But, re-reading RISC-Vs CBQRI[0] it turns out that theirs is memory mapped...

Yeah, all the cpus (e.g. harts) can access all the registers of the QoS controllers per the CBQRI spec [1].

The memory map for the example SoC in the proof-of-concept [2]:

  Base addr  Size
  0x4820000  4KB  Cluster 0 L2 cache controller
  0x4821000  4KB  Cluster 1 L2 cache controller
  0x4828000  4KB  Memory controller 0
  0x4829000  4KB  Memory controller 1
  0X482a000  4KB  Memory controller 2
  0X482b000  4KB  Shared LLC cache controller
Quoted 6 lines
Quoted 3 lines

I'm not an expert on resctrl but the MPAM code looks more like a backend for the resctrl support, so it makes more sense to do as we did for other drivers like irqchip, iommu.

Only because there are many irqchip or iommu. I'm not a fan of drivers/mpam, but drivers/resctrl would suit RISC-V too. (I'll check with Drew)

I think that is reasonable. In the proof-of-concept, I had the following structure, but I think there is a lot of room for improvement.

arch/riscv/kernel/qos/qos_resctrl.c
Implementation of the register interface described in the CBQRI spec along with the resctrl implementation. I should probably break this up into separate files for the CBQRI operations and the resctrl interface.

drivers/soc/foobar/foobar_cbqri_cache.c
DT-based driver for SoC cache controller that implements CBQRI

drivers/soc/foobar/foobar_cbqri_memory.c
DT-based driver for SoC memory controller that implements CBQRI

With all the great upstream progress, I've been meaning to rebase the RISC-V CBQRI support and post an RFC as its been a really long time. There is no public silicon yet that implements CBQRI but I think the possibility is getting closer. I've also been working on integrating ACPI support [3] using the new RQSC table, and I've been meaning to post an RFC for that too.

Thanks,
Drew

[1] https://github.com/riscv-non-isa/riscv-cbqri/releases/download/v1.0/riscv-cbqri.pdf
[2] https://lore.kernel.org/linux-riscv/[email protected]/
[3] https://lf-rise.atlassian.net/wiki/spaces/HOME/pages/433291272/ACPI+RQSC+Proof+of+Concept

author maintainer
#52 lore raw

RFC PATCH 14/36 arm_mpam: Add support for memory controller MSC on DT platforms

In reply to James Morse

From: Shanker Donthineni <[email protected]>

The device-tree binding has two examples for MSC associated with memory controllers. Add the support to discover the component_id from the device-tree and create 'memory' RIS.

[ morse: split out of a bigger patch, added affinity piece ]
---
 drivers/platform/arm64/mpam/mpam_devices.c | 67 +++++++++++++++-------
 1 file changed, 47 insertions(+), 20 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +47 -20
index 5b886ba54ba8..f5abd5f0d41a 100644
@@ -60,41 +60,63 @@ static int mpam_dt_parse_resource(struct mpam_msc *msc, struct device_node *np,
 				  u32 ris_idx)
 {
 	int err = 0;
-	u32 level = 0;
-	unsigned long cache_id;
-	struct device_node *cache;
+	u32 class_id = 0, component_id = 0;
+	struct device_node *cache = NULL, *memory = NULL;
+	enum mpam_class_types type = MPAM_CLASS_UNKNOWN;
 
 	do {
+		/* What kind of MSC is this? */
 		if (of_device_is_compatible(np, "arm,mpam-cache")) {
 			cache = of_parse_phandle(np, "arm,mpam-device", 0);
 			if (!cache) {
 				pr_err("Failed to read phandle\n");
 				break;
 			}
+			type = MPAM_CLASS_CACHE;
 		} else if (of_device_is_compatible(np->parent, "cache")) {
 			cache = of_node_get(np->parent);
+			type = MPAM_CLASS_CACHE;
+		} else if (of_device_is_compatible(np, "arm,mpam-memory")) {
+			memory = of_parse_phandle(np, "arm,mpam-device", 0);
+			if (!memory) {
+				pr_err("Failed to read phandle\n");
+				break;
+			}
+			type = MPAM_CLASS_MEMORY;
+		} else if (of_device_is_compatible(np, "arm,mpam-memory-controller-msc")) {
+			memory = of_node_get(np->parent);
+			type = MPAM_CLASS_MEMORY;
 		} else {
-			/* For now, only caches are supported */
-			cache = NULL;
+			/*
+			 * For now, only caches and memory controllers are
+			 * supported.
+			 */
 			break;
 		}
 
-		err = of_property_read_u32(cache, "cache-level", &level);
-		if (err) {
-			pr_err("Failed to read cache-level\n");
-			break;
-		}
-
-		cache_id = cache_of_calculate_id(cache);
-		if (cache_id == ~0UL) {
-			err = -ENOENT;
-			break;
+		/* Determine the class and component ids, based on type. */
+		if (type == MPAM_CLASS_CACHE) {
+			err = of_property_read_u32(cache, "cache-level", &class_id);
+			if (err) {
+				pr_err("Failed to read cache-level\n");
+				break;
+			}
+			component_id = cache_of_calculate_id(cache);
+			if (component_id == ~0UL) {
+				err = -ENOENT;
+				break;
+			}
+		} else if (type == MPAM_CLASS_MEMORY) {
+			err = of_node_to_nid(np);
+			component_id = (err == NUMA_NO_NODE) ? 0 : err;
+			class_id = 255;
 		}
 
-		err = mpam_ris_create(msc, ris_idx, MPAM_CLASS_CACHE, level,
-				      cache_id);
+		err = mpam_ris_create(msc, ris_idx, type, class_id,
+				      component_id);
 	} while (0);
 	of_node_put(cache);
+	of_node_put(memory);
 
 	return err;
 }
@@ -157,9 +179,14 @@ static int update_msc_accessibility(struct mpam_msc *msc)
 		cpumask_copy(&msc->accessibility, cpu_possible_mask);
 		err = 0;
 	} else {
-		err = -EINVAL;
-		pr_err("Cannot determine accessibility of MSC: %s\n",
-		       dev_name(&msc->pdev->dev));
+		if (of_device_is_compatible(parent, "memory")) {
+			cpumask_copy(&msc->accessibility, cpu_possible_mask);
+			err = 0;
+		} else {
+			err = -EINVAL;
+			pr_err("Cannot determine accessibility of MSC: %s\n",
+			       dev_name(&msc->pdev->dev));
+		}
 	}
 	of_node_put(parent);
 
Signature · 1 line
2.39.5
author maintainer
#53 lore raw

RFC PATCH 15/36 arm_mpam: Add the class and component structures for ris firmware described

In reply to James Morse

An MSC is a container of resources, each identified by their RIS index. Some RIS are described by firmware to provide their position in the system. Others are discovered when the driver probes the hardware.

To configure a resource it needs to be found by its class, e.g. 'L2'. There are two kinds of grouping, a class is a set of components, which are visible to user-space as there are likely to be multiple instances of the L2 cache. (e.g. one per cluster or package)

struct mpam_components are a set of struct mpam_vmsc. A vMSC groups the RIS in an MSC that control the same logical piece of hardware. (e.g. L2). This is to allow hardware implementations where two controls are presented as different RIS. Re-combining these RIS allows their feature bits to be or-ed. This structure is not visible outside mpam_devices.c

struct mpam_vmsc are then a set of struct mpam_msc_ris, which are not visible as each L2 cache may be composed of individual slices which need to be configured the same as the hardware is not able to distribute the configuration.

Add support for creating and destroying these structures.

A gfp is passed as the structures may need creating when a new RIS entry is discovered when probing the MSC.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 490 +++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |  91 ++++
 include/linux/arm_mpam.h                    |   8 +-
 3 files changed, 576 insertions(+), 13 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +483 -7
index f5abd5f0d41a..0d6d5180903b 100644
@@ -20,7 +20,6 @@
 #include <linux/printk.h>
 #include <linux/slab.h>
 #include <linux/spinlock.h>
-#include <linux/srcu.h>
 #include <linux/types.h>
 
 #include <acpi/pcc.h>
@@ -35,11 +34,485 @@
 static DEFINE_MUTEX(mpam_list_lock);
 static LIST_HEAD(mpam_all_msc);
 
-static struct srcu_struct mpam_srcu;
+struct srcu_struct mpam_srcu;
 
 /* MPAM isn't available until all the MSC have been probed. */
 static u32 mpam_num_msc;
 
+/*
+ * An MSC is a physical container for controls and monitors, each identified by
+ * their RIS index. These share a base-address, interrupts and some MMIO
+ * registers. A vMSC is a virtual container for RIS in an MSC that control or
+ * monitor the same thing. Members of a vMSC are all RIS in the same MSC, but
+ * not all RIS in an MSC share a vMSC.
+ * Components are a group of vMSC that control or monitor the same thing but
+ * are from different MSC, so have different base-address, interrupts etc.
+ * Classes are the set components of the same type.
+ *
+ * The features of a vMSC is the union of the RIS it contains.
+ * The features of a Class and Component are the common subset of the vMSC
+ * they contain.
+ *
+ * e.g. The system cache may have bandwidth controls on multiple interfaces,
+ * for regulating traffic from devices independently of traffic from CPUs.
+ * If these are two RIS in one MSC, they will be treated as controlling
+ * different things, and will not share a vMSC/component/class.
+ *
+ * e.g. The L2 may have one MSC and two RIS, one for cache-controls another
+ * for bandwidth. These two RIS are members of the same vMSC.
+ *
+ * e.g. The set of RIS that make up the L2 are grouped as a component. These
+ * are sometimes termed slices. They should be configured the same, as if there
+ * were only one.
+ *
+ * e.g. The SoC probably has more than one L2, each attached to a distinct set
+ * of CPUs. All the L2 components are grouped as a class.
+ *
+ * When creating an MSC, struct mpam_msc is added to the all mpam_all_msc list,
+ * then linked via struct mpam_ris to a vmsc, component and class.
+ * The same MSC may exist under different class->component->vmsc paths, but the
+ * RIS index will be unique.
+ */
+LIST_HEAD(mpam_classes);
+
+/* List of all objects that can be free()d after synchronise_srcu() */
+static LLIST_HEAD(mpam_garbage);
+
+#define init_garbage(x)	init_llist_node(&(x)->garbage.llist)
+
+static struct mpam_vmsc *
+mpam_vmsc_alloc(struct mpam_component *comp, struct mpam_msc *msc, gfp_t gfp)
+{
+	struct mpam_vmsc *vmsc;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	vmsc = kzalloc(sizeof(*vmsc), gfp);
+	if (!comp)
+		return ERR_PTR(-ENOMEM);
+	init_garbage(vmsc);
+
+	INIT_LIST_HEAD_RCU(&vmsc->ris);
+	INIT_LIST_HEAD_RCU(&vmsc->comp_list);
+	vmsc->comp = comp;
+	vmsc->msc = msc;
+
+	list_add_rcu(&vmsc->comp_list, &comp->vmsc);
+
+	return vmsc;
+}
+
+static struct mpam_vmsc *mpam_vmsc_get(struct mpam_component *comp,
+				       struct mpam_msc *msc, bool alloc,
+				       gfp_t gfp)
+{
+	struct mpam_vmsc *vmsc;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	list_for_each_entry(vmsc, &comp->vmsc, comp_list) {
+		if (vmsc->msc->id == msc->id)
+			return vmsc;
+	}
+
+	if (!alloc)
+		return ERR_PTR(-ENOENT);
+
+	return mpam_vmsc_alloc(comp, msc, gfp);
+}
+
+static struct mpam_component *
+mpam_component_alloc(struct mpam_class *class, int id, gfp_t gfp)
+{
+	struct mpam_component *comp;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	comp = kzalloc(sizeof(*comp), gfp);
+	if (!comp)
+		return ERR_PTR(-ENOMEM);
+	init_garbage(comp);
+
+	comp->comp_id = id;
+	INIT_LIST_HEAD_RCU(&comp->vmsc);
+	/* affinity is updated when ris are added */
+	INIT_LIST_HEAD_RCU(&comp->class_list);
+	comp->class = class;
+
+	list_add_rcu(&comp->class_list, &class->components);
+
+	return comp;
+}
+
+static struct mpam_component *
+mpam_component_get(struct mpam_class *class, int id, bool alloc, gfp_t gfp)
+{
+	struct mpam_component *comp;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	list_for_each_entry(comp, &class->components, class_list) {
+		if (comp->comp_id == id)
+			return comp;
+	}
+
+	if (!alloc)
+		return ERR_PTR(-ENOENT);
+
+	return mpam_component_alloc(class, id, gfp);
+}
+
+static struct mpam_class *
+mpam_class_alloc(u8 level_idx, enum mpam_class_types type, gfp_t gfp)
+{
+	struct mpam_class *class;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	class = kzalloc(sizeof(*class), gfp);
+	if (!class)
+		return ERR_PTR(-ENOMEM);
+	init_garbage(class);
+
+	INIT_LIST_HEAD_RCU(&class->components);
+	/* affinity is updated when ris are added */
+	class->level = level_idx;
+	class->type = type;
+	INIT_LIST_HEAD_RCU(&class->classes_list);
+
+	list_add_rcu(&class->classes_list, &mpam_classes);
+
+	return class;
+}
+
+static struct mpam_class *
+mpam_class_get(u8 level_idx, enum mpam_class_types type, bool alloc, gfp_t gfp)
+{
+	bool found = false;
+	struct mpam_class *class;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	list_for_each_entry(class, &mpam_classes, classes_list) {
+		if (class->type == type && class->level == level_idx) {
+			found = true;
+			break;
+		}
+	}
+
+	if (found)
+		return class;
+
+	if (!alloc)
+		return ERR_PTR(-ENOENT);
+
+	return mpam_class_alloc(level_idx, type, gfp);
+}
+
+#define add_to_garbage(x)				\
+do {							\
+	__typeof__(x) _x = x;				\
+	(_x)->garbage.to_free = (_x);			\
+	llist_add(&(_x)->garbage.llist, &mpam_garbage);	\
+} while (0)
+
+static void mpam_class_destroy(struct mpam_class *class)
+{
+	lockdep_assert_held(&mpam_list_lock);
+
+	list_del_rcu(&class->classes_list);
+	add_to_garbage(class);
+}
+
+static void mpam_comp_destroy(struct mpam_component *comp)
+{
+	struct mpam_class *class = comp->class;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	list_del_rcu(&comp->class_list);
+	add_to_garbage(comp);
+
+	if (list_empty(&class->components))
+		mpam_class_destroy(class);
+}
+
+static void mpam_vmsc_destroy(struct mpam_vmsc *vmsc)
+{
+	struct mpam_component *comp = vmsc->comp;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	list_del_rcu(&vmsc->comp_list);
+	add_to_garbage(vmsc);
+
+	if (list_empty(&comp->vmsc))
+		mpam_comp_destroy(comp);
+}
+
+static void mpam_ris_destroy(struct mpam_msc_ris *ris)
+{
+	struct mpam_vmsc *vmsc = ris->vmsc;
+	struct mpam_msc *msc = vmsc->msc;
+	struct platform_device *pdev = msc->pdev;
+	struct mpam_component *comp = vmsc->comp;
+	struct mpam_class *class = comp->class;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	cpumask_andnot(&comp->affinity, &comp->affinity, &ris->affinity);
+	cpumask_andnot(&class->affinity, &class->affinity, &ris->affinity);
+	clear_bit(ris->ris_idx, msc->ris_idxs);
+	list_del_rcu(&ris->vmsc_list);
+	list_del_rcu(&ris->msc_list);
+	add_to_garbage(ris);
+	ris->garbage.pdev = pdev;
+
+	if (list_empty(&vmsc->ris))
+		mpam_vmsc_destroy(vmsc);
+}
+
+/*
+ * There are two ways of reaching a struct mpam_msc_ris. Via the
+ * class->component->vmsc->ris, or via the msc.
+ * When destroying the msc, the other side needs unlinking and cleaning up too.
+ */
+static void mpam_msc_destroy(struct mpam_msc *msc)
+{
+	struct platform_device *pdev = msc->pdev;
+	struct mpam_msc_ris *ris, *tmp;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	list_del_rcu(&msc->glbl_list);
+	platform_set_drvdata(pdev, NULL);
+
+	list_for_each_entry_safe(ris, tmp, &msc->ris, msc_list)
+		mpam_ris_destroy(ris);
+
+	add_to_garbage(msc);
+	msc->garbage.pdev = pdev;
+}
+
+static void mpam_free_garbage(void)
+{
+	struct mpam_garbage *iter, *tmp;
+	struct llist_node *to_free = llist_del_all(&mpam_garbage);
+
+	if (!to_free)
+		return;
+
+	synchronize_srcu(&mpam_srcu);
+
+	llist_for_each_entry_safe(iter, tmp, to_free, llist) {
+		if (iter->pdev)
+			devm_kfree(&iter->pdev->dev, iter->to_free);
+		else
+			kfree(iter->to_free);
+	}
+}
+
+/* Called recursively to walk the list of caches from a particular CPU */
+static void __mpam_get_cpumask_from_cache_id(int cpu, struct device_node *cache_node,
+					     unsigned long cache_id,
+					     u32 cache_level,
+					     cpumask_t *affinity)
+{
+	int err;
+	u32 iter_level;
+	unsigned long iter_cache_id;
+	struct device_node *iter_node __free(device_node) = of_find_next_cache_node(cache_node);
+
+	if (!iter_node) {
+		pr_err("cpu %u next_cache_node returned NULL\n", cpu);
+		return;
+	}
+
+	err = of_property_read_u32(iter_node, "cache-level", &iter_level);
+	if (err)
+		return;
+
+	/*
+	 * get_cpu_cacheinfo_id() isn't ready until sometime
+	 * during device_initcall(). Use cache_of_calculate_id().
+	 */
+	iter_cache_id = cache_of_calculate_id(iter_node);
+	if (cache_id == ~0UL)
+		return;
+
+	if (iter_level == cache_level && iter_cache_id == cache_id)
+		cpumask_set_cpu(cpu, affinity);
+
+	__mpam_get_cpumask_from_cache_id(cpu, iter_node, cache_id, cache_level,
+					 affinity);
+}
+
+/*
+ * The cacheinfo structures are only populated when CPUs are online.
+ * This helper walks the device tree to include offline CPUs too.
+ */
+int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,
+				   cpumask_t *affinity)
+{
+	int cpu;
+
+	if (!acpi_disabled)
+		return acpi_pptt_get_cpumask_from_cache_id(cache_id, affinity);
+
+	for_each_possible_cpu(cpu) {
+		struct device_node *cpu_node __free(device_node) = of_get_cpu_node(cpu, NULL);
+		if (!cpu_node) {
+			pr_err("Failed to find cpu%d device node\n", cpu);
+			return -ENOENT;
+		}
+
+		__mpam_get_cpumask_from_cache_id(cpu, cpu_node, cache_id,
+						 cache_level, affinity);
+			continue;
+	}
+
+	return 0;
+}
+
+/*
+ * cpumask_of_node() only knows about online CPUs. This can't tell us whether
+ * a class is represented on all possible CPUs.
+ */
+static void get_cpumask_from_node_id(u32 node_id, cpumask_t *affinity)
+{
+	int cpu;
+
+	for_each_possible_cpu(cpu) {
+		if (node_id == cpu_to_node(cpu))
+			cpumask_set_cpu(cpu, affinity);
+	}
+}
+
+static int get_cpumask_from_cache(struct device_node *cache,
+				  cpumask_t *affinity)
+{
+	int err;
+	u32 cache_level;
+	unsigned long cache_id;
+
+	err = of_property_read_u32(cache, "cache-level", &cache_level);
+	if (err) {
+		pr_err("Failed to read cache-level from cache node\n");
+		return -ENOENT;
+	}
+
+	cache_id = cache_of_calculate_id(cache);
+	if (cache_id == ~0UL) {
+		pr_err("Failed to calculate cache-id from cache node\n");
+		return -ENOENT;
+	}
+
+	return mpam_get_cpumask_from_cache_id(cache_id, cache_level, affinity);
+}
+
+static int mpam_ris_get_affinity(struct mpam_msc *msc, cpumask_t *affinity,
+				 enum mpam_class_types type,
+				 struct mpam_class *class,
+				 struct mpam_component *comp)
+{
+	int err;
+
+	switch (type) {
+	case MPAM_CLASS_CACHE:
+		err = mpam_get_cpumask_from_cache_id(comp->comp_id, class->level,
+						     affinity);
+		if (err)
+			return err;
+
+		if (cpumask_empty(affinity))
+			pr_warn_once("%s no CPUs associated with cache node",
+				     dev_name(&msc->pdev->dev));
+
+		break;
+	case MPAM_CLASS_MEMORY:
+		get_cpumask_from_node_id(comp->comp_id, affinity);
+		/* affinity may be empty for CPU-less memory nodes */
+		break;
+	case MPAM_CLASS_UNKNOWN:
+		return 0;
+	}
+
+	cpumask_and(affinity, affinity, &msc->accessibility);
+
+	return 0;
+}
+
+static int mpam_ris_create_locked(struct mpam_msc *msc, u8 ris_idx,
+				  enum mpam_class_types type, u8 class_id,
+				  int component_id, gfp_t gfp)
+{
+	int err;
+	struct mpam_vmsc *vmsc;
+	struct mpam_msc_ris *ris;
+	struct mpam_class *class;
+	struct mpam_component *comp;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	if (test_and_set_bit(ris_idx, msc->ris_idxs))
+		return -EBUSY;
+
+	ris = devm_kzalloc(&msc->pdev->dev, sizeof(*ris), gfp);
+	if (!ris)
+		return -ENOMEM;
+	init_garbage(ris);
+
+	class = mpam_class_get(class_id, type, true, gfp);
+	if (IS_ERR(class))
+		return PTR_ERR(class);
+
+	comp = mpam_component_get(class, component_id, true, gfp);
+	if (IS_ERR(comp)) {
+		if (list_empty(&class->components))
+			mpam_class_destroy(class);
+		return PTR_ERR(comp);
+	}
+
+	vmsc = mpam_vmsc_get(comp, msc, true, gfp);
+	if (IS_ERR(vmsc)) {
+		if (list_empty(&comp->vmsc))
+			mpam_comp_destroy(comp);
+		return PTR_ERR(vmsc);
+	}
+
+	err = mpam_ris_get_affinity(msc, &ris->affinity, type, class, comp);
+	if (err) {
+		if (list_empty(&vmsc->ris))
+			mpam_vmsc_destroy(vmsc);
+		return err;
+	}
+
+	ris->ris_idx = ris_idx;
+	INIT_LIST_HEAD_RCU(&ris->vmsc_list);
+	ris->vmsc = vmsc;
+
+	cpumask_or(&comp->affinity, &comp->affinity, &ris->affinity);
+	cpumask_or(&class->affinity, &class->affinity, &ris->affinity);
+	list_add_rcu(&ris->vmsc_list, &vmsc->ris);
+
+	return 0;
+}
+
+int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
+		    enum mpam_class_types type, u8 class_id, int component_id)
+{
+	int err;
+
+	mutex_lock(&mpam_list_lock);
+	err = mpam_ris_create_locked(msc, ris_idx, type, class_id,
+				     component_id, GFP_KERNEL);
+	mutex_unlock(&mpam_list_lock);
+	if (err)
+		mpam_free_garbage();
+
+	return err;
+}
+
 static void mpam_discovery_complete(void)
 {
 	pr_err("Discovered all MSC\n");
@@ -179,7 +652,10 @@ static int update_msc_accessibility(struct mpam_msc *msc)
 		cpumask_copy(&msc->accessibility, cpu_possible_mask);
 		err = 0;
 	} else {
-		if (of_device_is_compatible(parent, "memory")) {
+		if (of_device_is_compatible(parent, "cache")) {
+			err = get_cpumask_from_cache(parent,
+						     &msc->accessibility);
+		} else if (of_device_is_compatible(parent, "memory")) {
 			cpumask_copy(&msc->accessibility, cpu_possible_mask);
 			err = 0;
 		} else {
@@ -209,11 +685,10 @@ static void mpam_msc_drv_remove(struct platform_device *pdev)
 
 	mutex_lock(&mpam_list_lock);
 	mpam_num_msc--;
-	platform_set_drvdata(pdev, NULL);
-	list_del_rcu(&msc->glbl_list);
-	synchronize_srcu(&mpam_srcu);
-	devm_kfree(&pdev->dev, msc);
+	mpam_msc_destroy(msc);
 	mutex_unlock(&mpam_list_lock);
+
+	mpam_free_garbage();
 }
 
 static int mpam_msc_drv_probe(struct platform_device *pdev)
@@ -230,6 +705,7 @@ static int mpam_msc_drv_probe(struct platform_device *pdev)
 			err = -ENOMEM;
 			break;
 		}
+		init_garbage(msc);
 
 		mutex_init(&msc->probe_lock);
 		mutex_init(&msc->part_sel_lock);
drivers/platform/arm64/mpam/mpam_internal.h +91 -0

107 lines not on this page · raw message

include/linux/arm_mpam.h +2 -6
index 0edefa6ba019..406a77be68cb 100644
@@ -36,11 +36,7 @@ static inline int acpi_mpam_parse_resources(struct mpam_msc *msc,
 static inline int acpi_mpam_count_msc(void) { return -EINVAL; }
 #endif
 
-static inline int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
-				  enum mpam_class_types type, u8 class_id,
-				  int component_id)
-{
-	return -EINVAL;
-}
+int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
+		    enum mpam_class_types type, u8 class_id, int component_id);
 
 #endif /* __LINUX_ARM_MPAM_H */
Signature · 1 line
2.39.5
author maintainer
#54 lore raw

RFC PATCH 16/36 arm_mpam: Add MPAM MSC register layout definitions

In reply to James Morse

Memory Partitioning and Monitoring (MPAM) has memory mapped devices (MSCs) with an identity/configuration page.

Add the definitions for these registers as offset within the page(s).

---
 drivers/platform/arm64/mpam/mpam_internal.h | 268 ++++++++++++++++++++
 1 file changed, 268 insertions(+)
drivers/platform/arm64/mpam/mpam_internal.h +268 -0
index d49bb884b433..9110c171d9d2 100644
@@ -150,4 +150,272 @@ extern struct list_head mpam_classes;
 int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,
 				   cpumask_t *affinity);
 
+/*
+ * MPAM MSCs have the following register layout. See:
+ * Arm Architecture Reference Manual Supplement - Memory System Resource
+ * Partitioning and Monitoring (MPAM), for Armv8-A. DDI 0598A.a
+ */
+#define MPAM_ARCHITECTURE_V1    0x10
+
+/* Memory mapped control pages: */
+/* ID Register offsets in the memory mapped page */
+#define MPAMF_IDR               0x0000  /* features id register */
+#define MPAMF_MSMON_IDR         0x0080  /* performance monitoring features */
+#define MPAMF_IMPL_IDR          0x0028  /* imp-def partitioning */
+#define MPAMF_CPOR_IDR          0x0030  /* cache-portion partitioning */
+#define MPAMF_CCAP_IDR          0x0038  /* cache-capacity partitioning */
+#define MPAMF_MBW_IDR           0x0040  /* mem-bw partitioning */
+#define MPAMF_PRI_IDR           0x0048  /* priority partitioning */
+#define MPAMF_CSUMON_IDR        0x0088  /* cache-usage monitor */
+#define MPAMF_MBWUMON_IDR       0x0090  /* mem-bw usage monitor */
+#define MPAMF_PARTID_NRW_IDR    0x0050  /* partid-narrowing */
+#define MPAMF_IIDR              0x0018  /* implementer id register */
+#define MPAMF_AIDR              0x0020  /* architectural id register */
+
+/* Configuration and Status Register offsets in the memory mapped page */
+#define MPAMCFG_PART_SEL        0x0100  /* partid to configure: */
+#define MPAMCFG_CPBM            0x1000  /* cache-portion config */
+#define MPAMCFG_CMAX            0x0108  /* cache-capacity config */
+#define MPAMCFG_CMIN            0x0110  /* cache-capacity config */
+#define MPAMCFG_MBW_MIN         0x0200  /* min mem-bw config */
+#define MPAMCFG_MBW_MAX         0x0208  /* max mem-bw config */
+#define MPAMCFG_MBW_WINWD       0x0220  /* mem-bw accounting window config */
+#define MPAMCFG_MBW_PBM         0x2000  /* mem-bw portion bitmap config */
+#define MPAMCFG_PRI             0x0400  /* priority partitioning config */
+#define MPAMCFG_MBW_PROP        0x0500  /* mem-bw stride config */
+#define MPAMCFG_INTPARTID       0x0600  /* partid-narrowing config */
+
+#define MSMON_CFG_MON_SEL       0x0800  /* monitor selector */
+#define MSMON_CFG_CSU_FLT       0x0810  /* cache-usage monitor filter */
+#define MSMON_CFG_CSU_CTL       0x0818  /* cache-usage monitor config */
+#define MSMON_CFG_MBWU_FLT      0x0820  /* mem-bw monitor filter */
+#define MSMON_CFG_MBWU_CTL      0x0828  /* mem-bw monitor config */
+#define MSMON_CSU               0x0840  /* current cache-usage */
+#define MSMON_CSU_CAPTURE       0x0848  /* last cache-usage value captured */
+#define MSMON_MBWU              0x0860  /* current mem-bw usage value */
+#define MSMON_MBWU_CAPTURE      0x0868  /* last mem-bw value captured */
+#define MSMON_CAPT_EVNT         0x0808  /* signal a capture event */
+#define MPAMF_ESR               0x00F8  /* error status register */
+#define MPAMF_ECR               0x00F0  /* error control register */
+
+/* MPAMF_IDR - MPAM features ID register */
+#define MPAMF_IDR_PARTID_MAX            GENMASK(15, 0)
+#define MPAMF_IDR_PMG_MAX               GENMASK(23, 16)
+#define MPAMF_IDR_HAS_CCAP_PART         BIT(24)
+#define MPAMF_IDR_HAS_CPOR_PART         BIT(25)
+#define MPAMF_IDR_HAS_MBW_PART          BIT(26)
+#define MPAMF_IDR_HAS_PRI_PART          BIT(27)
+#define MPAMF_IDR_HAS_EXT               BIT(28)
+#define MPAMF_IDR_HAS_IMPL_IDR          BIT(29)
+#define MPAMF_IDR_HAS_MSMON             BIT(30)
+#define MPAMF_IDR_HAS_PARTID_NRW        BIT(31)
+#define MPAMF_IDR_HAS_RIS               BIT(32)
+#define MPAMF_IDR_HAS_EXT_ESR           BIT(38)
+#define MPAMF_IDR_HAS_ESR               BIT(39)
+#define MPAMF_IDR_RIS_MAX               GENMASK(59, 56)
+
+/* MPAMF_MSMON_IDR - MPAM performance monitoring ID register */
+#define MPAMF_MSMON_IDR_MSMON_CSU               BIT(16)
+#define MPAMF_MSMON_IDR_MSMON_MBWU              BIT(17)
+#define MPAMF_MSMON_IDR_HAS_LOCAL_CAPT_EVNT     BIT(31)
+
+/* MPAMF_CPOR_IDR - MPAM features cache portion partitioning ID register */
+#define MPAMF_CPOR_IDR_CPBM_WD                  GENMASK(15, 0)
+
+/* MPAMF_CCAP_IDR - MPAM features cache capacity partitioning ID register */
+#define MPAMF_CCAP_IDR_HAS_CMAX_SOFTLIM         BIT(31)
+#define MPAMF_CCAP_IDR_NO_CMAX                  BIT(30)
+#define MPAMF_CCAP_IDR_HAS_CMIN                 BIT(29)
+#define MPAMF_CCAP_IDR_HAS_CASSOC               BIT(28)
+#define MPAMF_CCAP_IDR_CASSOC_WD                GENMASK(12, 8)
+#define MPAMF_CCAP_IDR_CMAX_WD                  GENMASK(5, 0)
+
+/* MPAMF_MBW_IDR - MPAM features memory bandwidth partitioning ID register */
+#define MPAMF_MBW_IDR_BWA_WD            GENMASK(5, 0)
+#define MPAMF_MBW_IDR_HAS_MIN           BIT(10)
+#define MPAMF_MBW_IDR_HAS_MAX           BIT(11)
+#define MPAMF_MBW_IDR_HAS_PBM           BIT(12)
+#define MPAMF_MBW_IDR_HAS_PROP          BIT(13)
+#define MPAMF_MBW_IDR_WINDWR            BIT(14)
+#define MPAMF_MBW_IDR_BWPBM_WD          GENMASK(28, 16)
+
+/* MPAMF_PRI_IDR - MPAM features priority partitioning ID register */
+#define MPAMF_PRI_IDR_HAS_INTPRI        BIT(0)
+#define MPAMF_PRI_IDR_INTPRI_0_IS_LOW   BIT(1)
+#define MPAMF_PRI_IDR_INTPRI_WD         GENMASK(9, 4)
+#define MPAMF_PRI_IDR_HAS_DSPRI         BIT(16)
+#define MPAMF_PRI_IDR_DSPRI_0_IS_LOW    BIT(17)
+#define MPAMF_PRI_IDR_DSPRI_WD          GENMASK(25, 20)
+
+/* MPAMF_CSUMON_IDR - MPAM cache storage usage monitor ID register */
+#define MPAMF_CSUMON_IDR_NUM_MON        GENMASK(15, 0)
+#define MPAMF_CSUMON_IDR_HAS_OFLOW_CAPT	BIT(24)
+#define MPAMF_CSUMON_IDR_HAS_CEVNT_OFLW	BIT(25)
+#define MPAMF_CSUMON_IDR_HAS_OFSR	BIT(26)
+#define MPAMF_CSUMON_IDR_HAS_OFLOW_LNKG	BIT(27)
+#define MPAMF_CSUMON_IDR_HAS_XCL	BIT(29)
+#define MPAMF_CSUMON_IDR_CSU_RO		BIT(30)
+#define MPAMF_CSUMON_IDR_HAS_CAPTURE    BIT(31)
+
+/* MPAMF_MBWUMON_IDR - MPAM memory bandwidth usage monitor ID register */
+#define MPAMF_MBWUMON_IDR_NUM_MON       GENMASK(15, 0)
+#define MPAMF_MBWUMON_IDR_HAS_RWBW      BIT(28)
+#define MPAMF_MBWUMON_IDR_LWD           BIT(29)
+#define MPAMF_MBWUMON_IDR_HAS_LONG      BIT(30)
+#define MPAMF_MBWUMON_IDR_HAS_CAPTURE   BIT(31)
+
+/* MPAMF_PARTID_NRW_IDR - MPAM PARTID narrowing ID register */
+#define MPAMF_PARTID_NRW_IDR_INTPARTID_MAX      GENMASK(15, 0)
+
+/* MPAMF_IIDR - MPAM implementation ID register */
+#define MPAMF_IIDR_PRODUCTID    GENMASK(31, 20)
+#define MPAMF_IIDR_PRODUCTID_SHIFT	20
+#define MPAMF_IIDR_VARIANT      GENMASK(19, 16)
+#define MPAMF_IIDR_VARIANT_SHIFT	16
+#define MPAMF_IIDR_REVISON      GENMASK(15, 12)
+#define MPAMF_IIDR_REVISON_SHIFT	12
+#define MPAMF_IIDR_IMPLEMENTER  GENMASK(11, 0)
+#define MPAMF_IIDR_IMPLEMENTER_SHIFT	0
+
+/* MPAMF_AIDR - MPAM architecture ID register */
+#define MPAMF_AIDR_ARCH_MAJOR_REV       GENMASK(7, 4)
+#define MPAMF_AIDR_ARCH_MINOR_REV       GENMASK(3, 0)
+
+/* MPAMCFG_PART_SEL - MPAM partition configuration selection register */
+#define MPAMCFG_PART_SEL_PARTID_SEL     GENMASK(15, 0)
+#define MPAMCFG_PART_SEL_INTERNAL       BIT(16)
+#define MPAMCFG_PART_SEL_RIS            GENMASK(27, 24)
+
+/* MPAMCFG_CMAX - MPAM cache capacity configuration register */
+#define MPAMCFG_CMAX_SOFTLIM            BIT(31)
+#define MPAMCFG_CMAX_CMAX               GENMASK(15, 0)
+
+/* MPAMCFG_CMIN - MPAM cache capacity configuration register */
+#define MPAMCFG_CMIN_CMIN               GENMASK(15, 0)
+
+/*
+ * MPAMCFG_MBW_MIN - MPAM memory minimum bandwidth partitioning configuration
+ *                   register
+ */
+#define MPAMCFG_MBW_MIN_MIN             GENMASK(15, 0)
+
+/*
+ * MPAMCFG_MBW_MAX - MPAM memory maximum bandwidth partitioning configuration
+ *                   register
+ */
+#define MPAMCFG_MBW_MAX_MAX             GENMASK(15, 0)
+#define MPAMCFG_MBW_MAX_HARDLIM         BIT(31)
+
+/*
+ * MPAMCFG_MBW_WINWD - MPAM memory bandwidth partitioning window width
+ *                     register
+ */
+#define MPAMCFG_MBW_WINWD_US_FRAC       GENMASK(7, 0)
+#define MPAMCFG_MBW_WINWD_US_INT        GENMASK(23, 8)
+
+/* MPAMCFG_PRI - MPAM priority partitioning configuration register */
+#define MPAMCFG_PRI_INTPRI              GENMASK(15, 0)
+#define MPAMCFG_PRI_DSPRI               GENMASK(31, 16)
+
+/*
+ * MPAMCFG_MBW_PROP - Memory bandwidth proportional stride partitioning
+ *                    configuration register
+ */
+#define MPAMCFG_MBW_PROP_STRIDEM1       GENMASK(15, 0)
+#define MPAMCFG_MBW_PROP_EN             BIT(31)
+
+/*
+ * MPAMCFG_INTPARTID - MPAM internal partition narrowing configuration register
+ */
+#define MPAMCFG_INTPARTID_INTPARTID     GENMASK(15, 0)
+#define MPAMCFG_INTPARTID_INTERNAL      BIT(16)
+
+/* MSMON_CFG_MON_SEL - Memory system performance monitor selection register */
+#define MSMON_CFG_MON_SEL_MON_SEL       GENMASK(15, 0)
+#define MSMON_CFG_MON_SEL_RIS           GENMASK(27, 24)
+
+/* MPAMF_ESR - MPAM Error Status Register */
+#define MPAMF_ESR_PARTID_OR_MON GENMASK(15, 0)
+#define MPAMF_ESR_PMG           GENMASK(23, 16)
+#define MPAMF_ESR_ERRCODE       GENMASK(27, 24)
+#define MPAMF_ESR_OVRWR         BIT(31)
+#define MPAMF_ESR_RIS           GENMASK(35, 32)
+
+/* MPAMF_ECR - MPAM Error Control Register */
+#define MPAMF_ECR_INTEN         BIT(0)
+
+/* Error conditions in accessing memory mapped registers */
+#define MPAM_ERRCODE_NONE                       0
+#define MPAM_ERRCODE_PARTID_SEL_RANGE           1
+#define MPAM_ERRCODE_REQ_PARTID_RANGE           2
+#define MPAM_ERRCODE_MSMONCFG_ID_RANGE          3
+#define MPAM_ERRCODE_REQ_PMG_RANGE              4
+#define MPAM_ERRCODE_MONITOR_RANGE              5
+#define MPAM_ERRCODE_INTPARTID_RANGE            6
+#define MPAM_ERRCODE_UNEXPECTED_INTERNAL        7
+
+/*
+ * MSMON_CFG_CSU_FLT - Memory system performance monitor configure cache storage
+ *                    usage monitor filter register
+ */
+#define MSMON_CFG_CSU_FLT_PARTID       GENMASK(15, 0)
+#define MSMON_CFG_CSU_FLT_PMG          GENMASK(23, 16)
+
+/*
+ * MSMON_CFG_CSU_CTL - Memory system performance monitor configure cache storage
+ *                    usage monitor control register
+ * MSMON_CFG_MBWU_CTL - Memory system performance monitor configure memory
+ *                     bandwidth usage monitor control register
+ */
+#define MSMON_CFG_x_CTL_TYPE           GENMASK(7, 0)
+#define MSMON_CFG_x_CTL_OFLOW_STATUS_L BIT(15)
+#define MSMON_CFG_x_CTL_MATCH_PARTID   BIT(16)
+#define MSMON_CFG_x_CTL_MATCH_PMG      BIT(17)
+#define MSMON_CFG_x_CTL_SCLEN          BIT(19)
+#define MSMON_CFG_x_CTL_SUBTYPE        GENMASK(23, 20)
+#define MSMON_CFG_x_CTL_OFLOW_FRZ      BIT(24)
+#define MSMON_CFG_x_CTL_OFLOW_INTR     BIT(25)
+#define MSMON_CFG_x_CTL_OFLOW_STATUS   BIT(26)
+#define MSMON_CFG_x_CTL_CAPT_RESET     BIT(27)
+#define MSMON_CFG_x_CTL_CAPT_EVNT      GENMASK(30, 28)
+#define MSMON_CFG_x_CTL_EN             BIT(31)
+
+#define MSMON_CFG_MBWU_CTL_TYPE_MBWU			0x42
+#define MSMON_CFG_MBWU_CTL_TYPE_CSU			0x43
+
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_NONE                 0
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_READ                 1
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_WRITE                2
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_BOTH                 3
+
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_MAX                  3
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_MASK                 0x3
+
+/*
+ * MSMON_CFG_MBWU_FLT - Memory system performance monitor configure memory
+ *                     bandwidth usage monitor filter register
+ */
+#define MSMON_CFG_MBWU_FLT_PARTID               GENMASK(15, 0)
+#define MSMON_CFG_MBWU_FLT_PMG                  GENMASK(23, 16)
+#define MSMON_CFG_MBWU_FLT_RWBW                 GENMASK(31, 30)
+
+/*
+ * MSMON_CSU - Memory system performance monitor cache storage usage monitor
+ *            register
+ * MSMON_CSU_CAPTURE -  Memory system performance monitor cache storage usage
+ *                     capture register
+ * MSMON_MBWU  - Memory system performance monitor memory bandwidth usage
+ *               monitor register
+ * MSMON_MBWU_CAPTURE - Memory system performance monitor memory bandwidth usage
+ *                     capture register
+ */
+#define MSMON___VALUE          GENMASK(30, 0)
+#define MSMON___NRDY           BIT(31)
+#define MSMON_MBWU_L_VALUE     GENMASK(62, 0)
+/*
+ * MSMON_CAPT_EVNT - Memory system performance monitoring capture event
+ *                  generation register
+ */
+#define MSMON_CAPT_EVNT_NOW    BIT(0)
+
 #endif /* MPAM_INTERNAL_H */
Signature · 1 line
2.39.5
#55 lore raw

RE: [RFC PATCH 16/36] arm_mpam: Add MPAM MSC register layout definitions

Hello James,

Quoted 285 lines

Memory Partitioning and Monitoring (MPAM) has memory mapped devices (MSCs) with an identity/configuration page.

Add the definitions for these registers as offset within the page(s).

---
 drivers/platform/arm64/mpam/mpam_internal.h | 268
++++++++++++++++++++
 1 file changed, 268 insertions(+)
drivers/platform/arm64/mpam/mpam_internal.h +0 -0
index d49bb884b433..9110c171d9d2 100644
@@ -150,4 +150,272 @@ extern struct list_head mpam_classes;  int

mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,

drivers/platform/arm64/mpam/mpam_internal.h +11 -0
 				   cpumask_t *affinity);
 
+/*
+ * MPAM MSCs have the following register layout. See:
+ * Arm Architecture Reference Manual Supplement - Memory System
+Resource
+ * Partitioning and Monitoring (MPAM), for Armv8-A. DDI 0598A.a  */
+#define MPAM_ARCHITECTURE_V1    0x10
+
+/* Memory mapped control pages: */
+/* ID Register offsets in the memory mapped page */
+#define MPAMF_IDR               0x0000  /* features id register */
+#define MPAMF_MSMON_IDR         0x0080  /* performance monitoring

features */

drivers/platform/arm64/mpam/mpam_internal.h +2 -0
+#define MPAMF_IMPL_IDR          0x0028  /* imp-def partitioning */
+#define MPAMF_CPOR_IDR          0x0030  /* cache-portion partitioning

*/
+#define MPAMF_CCAP_IDR 0x0038 /* cache-capacity
partitioning */

drivers/platform/arm64/mpam/mpam_internal.h +4 -0
+#define MPAMF_MBW_IDR           0x0040  /* mem-bw partitioning */
+#define MPAMF_PRI_IDR           0x0048  /* priority partitioning */
+#define MPAMF_CSUMON_IDR        0x0088  /* cache-usage monitor */
+#define MPAMF_MBWUMON_IDR       0x0090  /* mem-bw usage

monitor */

drivers/platform/arm64/mpam/mpam_internal.h +12 -0
+#define MPAMF_PARTID_NRW_IDR    0x0050  /* partid-narrowing */
+#define MPAMF_IIDR              0x0018  /* implementer id register */
+#define MPAMF_AIDR              0x0020  /* architectural id register */
+
+/* Configuration and Status Register offsets in the memory mapped page */
+#define MPAMCFG_PART_SEL        0x0100  /* partid to configure: */
+#define MPAMCFG_CPBM            0x1000  /* cache-portion config */
+#define MPAMCFG_CMAX            0x0108  /* cache-capacity config */
+#define MPAMCFG_CMIN            0x0110  /* cache-capacity config */
+#define MPAMCFG_MBW_MIN         0x0200  /* min mem-bw config */
+#define MPAMCFG_MBW_MAX         0x0208  /* max mem-bw config */
+#define MPAMCFG_MBW_WINWD       0x0220  /* mem-bw accounting

window config */
+#define MPAMCFG_MBW_PBM 0x2000 /* mem-bw portion
bitmap config */
+#define MPAMCFG_PRI 0x0400 /* priority partitioning config
*/
+#define MPAMCFG_MBW_PROP 0x0500 /* mem-bw stride config
*/
+#define MPAMCFG_INTPARTID 0x0600 /* partid-narrowing config
*/

drivers/platform/arm64/mpam/mpam_internal.h +3 -0
+
+#define MSMON_CFG_MON_SEL       0x0800  /* monitor selector */
+#define MSMON_CFG_CSU_FLT       0x0810  /* cache-usage monitor

filter */
+#define MSMON_CFG_CSU_CTL 0x0818 /* cache-usage monitor
config */
+#define MSMON_CFG_MBWU_FLT 0x0820 /* mem-bw monitor filter
*/
+#define MSMON_CFG_MBWU_CTL 0x0828 /* mem-bw monitor
config */

drivers/platform/arm64/mpam/mpam_internal.h +2 -0
+#define MSMON_CSU               0x0840  /* current cache-usage */
+#define MSMON_CSU_CAPTURE       0x0848  /* last cache-usage value

captured */
+#define MSMON_MBWU 0x0860 /* current mem-bw usage
value */
+#define MSMON_MBWU_CAPTURE 0x0868 /* last mem-bw value
captured */

drivers/platform/arm64/mpam/mpam_internal.h +26 -0
+#define MSMON_CAPT_EVNT         0x0808  /* signal a capture event */
+#define MPAMF_ESR               0x00F8  /* error status register */
+#define MPAMF_ECR               0x00F0  /* error control register */
+
+/* MPAMF_IDR - MPAM features ID register */
+#define MPAMF_IDR_PARTID_MAX            GENMASK(15, 0)
+#define MPAMF_IDR_PMG_MAX               GENMASK(23, 16)
+#define MPAMF_IDR_HAS_CCAP_PART         BIT(24)
+#define MPAMF_IDR_HAS_CPOR_PART         BIT(25)
+#define MPAMF_IDR_HAS_MBW_PART          BIT(26)
+#define MPAMF_IDR_HAS_PRI_PART          BIT(27)
+#define MPAMF_IDR_HAS_EXT               BIT(28)
+#define MPAMF_IDR_HAS_IMPL_IDR          BIT(29)
+#define MPAMF_IDR_HAS_MSMON             BIT(30)
+#define MPAMF_IDR_HAS_PARTID_NRW        BIT(31)
+#define MPAMF_IDR_HAS_RIS               BIT(32)
+#define MPAMF_IDR_HAS_EXT_ESR           BIT(38)
+#define MPAMF_IDR_HAS_ESR               BIT(39)
+#define MPAMF_IDR_RIS_MAX               GENMASK(59, 56)
+
+/* MPAMF_MSMON_IDR - MPAM performance monitoring ID register */
+#define MPAMF_MSMON_IDR_MSMON_CSU               BIT(16)
+#define MPAMF_MSMON_IDR_MSMON_MBWU              BIT(17)
+#define MPAMF_MSMON_IDR_HAS_LOCAL_CAPT_EVNT     BIT(31)
+
+/* MPAMF_CPOR_IDR - MPAM features cache portion partitioning ID

register */
+#define MPAMF_CPOR_IDR_CPBM_WD GENMASK(15,
0)

drivers/platform/arm64/mpam/mpam_internal.h +2 -0
+
+/* MPAMF_CCAP_IDR - MPAM features cache capacity partitioning ID

register */

drivers/platform/arm64/mpam/mpam_internal.h +5 -0
+#define MPAMF_CCAP_IDR_HAS_CMAX_SOFTLIM         BIT(31)
+#define MPAMF_CCAP_IDR_NO_CMAX                  BIT(30)
+#define MPAMF_CCAP_IDR_HAS_CMIN                 BIT(29)
+#define MPAMF_CCAP_IDR_HAS_CASSOC               BIT(28)
+#define MPAMF_CCAP_IDR_CASSOC_WD                GENMASK(12,

8)

drivers/platform/arm64/mpam/mpam_internal.h +3 -0
+#define MPAMF_CCAP_IDR_CMAX_WD                  GENMASK(5, 0)
+
+/* MPAMF_MBW_IDR - MPAM features memory bandwidth partitioning ID

register */

drivers/platform/arm64/mpam/mpam_internal.h +17 -0
+#define MPAMF_MBW_IDR_BWA_WD            GENMASK(5, 0)
+#define MPAMF_MBW_IDR_HAS_MIN           BIT(10)
+#define MPAMF_MBW_IDR_HAS_MAX           BIT(11)
+#define MPAMF_MBW_IDR_HAS_PBM           BIT(12)
+#define MPAMF_MBW_IDR_HAS_PROP          BIT(13)
+#define MPAMF_MBW_IDR_WINDWR            BIT(14)
+#define MPAMF_MBW_IDR_BWPBM_WD          GENMASK(28, 16)
+
+/* MPAMF_PRI_IDR - MPAM features priority partitioning ID register */
+#define MPAMF_PRI_IDR_HAS_INTPRI        BIT(0)
+#define MPAMF_PRI_IDR_INTPRI_0_IS_LOW   BIT(1)
+#define MPAMF_PRI_IDR_INTPRI_WD         GENMASK(9, 4)
+#define MPAMF_PRI_IDR_HAS_DSPRI         BIT(16)
+#define MPAMF_PRI_IDR_DSPRI_0_IS_LOW    BIT(17)
+#define MPAMF_PRI_IDR_DSPRI_WD          GENMASK(25, 20)
+
+/* MPAMF_CSUMON_IDR - MPAM cache storage usage monitor ID register

*/

drivers/platform/arm64/mpam/mpam_internal.h +10 -0
+#define MPAMF_CSUMON_IDR_NUM_MON        GENMASK(15, 0)
+#define MPAMF_CSUMON_IDR_HAS_OFLOW_CAPT	BIT(24)
+#define MPAMF_CSUMON_IDR_HAS_CEVNT_OFLW	BIT(25)
+#define MPAMF_CSUMON_IDR_HAS_OFSR	BIT(26)
+#define MPAMF_CSUMON_IDR_HAS_OFLOW_LNKG	BIT(27)
+#define MPAMF_CSUMON_IDR_HAS_XCL	BIT(29)
+#define MPAMF_CSUMON_IDR_CSU_RO		BIT(30)
+#define MPAMF_CSUMON_IDR_HAS_CAPTURE    BIT(31)
+
+/* MPAMF_MBWUMON_IDR - MPAM memory bandwidth usage monitor ID

register */

drivers/platform/arm64/mpam/mpam_internal.h +8 -0
+#define MPAMF_MBWUMON_IDR_NUM_MON       GENMASK(15, 0)
+#define MPAMF_MBWUMON_IDR_HAS_RWBW      BIT(28)
+#define MPAMF_MBWUMON_IDR_LWD           BIT(29)
+#define MPAMF_MBWUMON_IDR_HAS_LONG      BIT(30)
+#define MPAMF_MBWUMON_IDR_HAS_CAPTURE   BIT(31)
+
+/* MPAMF_PARTID_NRW_IDR - MPAM PARTID narrowing ID register */
+#define MPAMF_PARTID_NRW_IDR_INTPARTID_MAX      GENMASK(15,

0)

drivers/platform/arm64/mpam/mpam_internal.h +16 -0
+
+/* MPAMF_IIDR - MPAM implementation ID register */
+#define MPAMF_IIDR_PRODUCTID    GENMASK(31, 20)
+#define MPAMF_IIDR_PRODUCTID_SHIFT	20
+#define MPAMF_IIDR_VARIANT      GENMASK(19, 16)
+#define MPAMF_IIDR_VARIANT_SHIFT	16
+#define MPAMF_IIDR_REVISON      GENMASK(15, 12)
+#define MPAMF_IIDR_REVISON_SHIFT	12
+#define MPAMF_IIDR_IMPLEMENTER  GENMASK(11, 0)
+#define MPAMF_IIDR_IMPLEMENTER_SHIFT	0
+
+/* MPAMF_AIDR - MPAM architecture ID register */
+#define MPAMF_AIDR_ARCH_MAJOR_REV       GENMASK(7, 4)
+#define MPAMF_AIDR_ARCH_MINOR_REV       GENMASK(3, 0)
+
+/* MPAMCFG_PART_SEL - MPAM partition configuration selection register

*/

drivers/platform/arm64/mpam/mpam_internal.h +13 -0
+#define MPAMCFG_PART_SEL_PARTID_SEL     GENMASK(15, 0)
+#define MPAMCFG_PART_SEL_INTERNAL       BIT(16)
+#define MPAMCFG_PART_SEL_RIS            GENMASK(27, 24)
+
+/* MPAMCFG_CMAX - MPAM cache capacity configuration register */
+#define MPAMCFG_CMAX_SOFTLIM            BIT(31)
+#define MPAMCFG_CMAX_CMAX               GENMASK(15, 0)
+
+/* MPAMCFG_CMIN - MPAM cache capacity configuration register */
+#define MPAMCFG_CMIN_CMIN               GENMASK(15, 0)
+
+/*
+ * MPAMCFG_MBW_MIN - MPAM memory minimum bandwidth partitioning

configuration
+ * register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +4 -0
+#define MPAMCFG_MBW_MIN_MIN             GENMASK(15, 0)
+
+/*
+ * MPAMCFG_MBW_MAX - MPAM memory maximum bandwidth

partitioning configuration
+ * register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +5 -0
+#define MPAMCFG_MBW_MAX_MAX             GENMASK(15, 0)
+#define MPAMCFG_MBW_MAX_HARDLIM         BIT(31)
+
+/*
+ * MPAMCFG_MBW_WINWD - MPAM memory bandwidth partitioning

window width
+ * register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +9 -0
+#define MPAMCFG_MBW_WINWD_US_FRAC       GENMASK(7, 0)
+#define MPAMCFG_MBW_WINWD_US_INT        GENMASK(23, 8)
+
+/* MPAMCFG_PRI - MPAM priority partitioning configuration register */
+#define MPAMCFG_PRI_INTPRI              GENMASK(15, 0)
+#define MPAMCFG_PRI_DSPRI               GENMASK(31, 16)
+
+/*
+ * MPAMCFG_MBW_PROP - Memory bandwidth proportional stride

partitioning
+ * configuration register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +10 -0
+#define MPAMCFG_MBW_PROP_STRIDEM1       GENMASK(15, 0)
+#define MPAMCFG_MBW_PROP_EN             BIT(31)
+
+/*
+ * MPAMCFG_INTPARTID - MPAM internal partition narrowing configuration
+register  */
+#define MPAMCFG_INTPARTID_INTPARTID     GENMASK(15, 0)
+#define MPAMCFG_INTPARTID_INTERNAL      BIT(16)
+
+/* MSMON_CFG_MON_SEL - Memory system performance monitor

selection register */

drivers/platform/arm64/mpam/mpam_internal.h +25 -0
+#define MSMON_CFG_MON_SEL_MON_SEL       GENMASK(15, 0)
+#define MSMON_CFG_MON_SEL_RIS           GENMASK(27, 24)
+
+/* MPAMF_ESR - MPAM Error Status Register */ #define
+MPAMF_ESR_PARTID_OR_MON GENMASK(15, 0)
+#define MPAMF_ESR_PMG           GENMASK(23, 16)
+#define MPAMF_ESR_ERRCODE       GENMASK(27, 24)
+#define MPAMF_ESR_OVRWR         BIT(31)
+#define MPAMF_ESR_RIS           GENMASK(35, 32)
+
+/* MPAMF_ECR - MPAM Error Control Register */
+#define MPAMF_ECR_INTEN         BIT(0)
+
+/* Error conditions in accessing memory mapped registers */
+#define MPAM_ERRCODE_NONE                       0
+#define MPAM_ERRCODE_PARTID_SEL_RANGE           1
+#define MPAM_ERRCODE_REQ_PARTID_RANGE           2
+#define MPAM_ERRCODE_MSMONCFG_ID_RANGE          3
+#define MPAM_ERRCODE_REQ_PMG_RANGE              4
+#define MPAM_ERRCODE_MONITOR_RANGE              5
+#define MPAM_ERRCODE_INTPARTID_RANGE            6
+#define MPAM_ERRCODE_UNEXPECTED_INTERNAL        7
+
+/*
+ * MSMON_CFG_CSU_FLT - Memory system performance monitor configure

cache storage
+ * usage monitor filter register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +5 -0
+#define MSMON_CFG_CSU_FLT_PARTID       GENMASK(15, 0)
+#define MSMON_CFG_CSU_FLT_PMG          GENMASK(23, 16)
+
+/*
+ * MSMON_CFG_CSU_CTL - Memory system performance monitor configure

cache storage
+ * usage monitor control register
+ * MSMON_CFG_MBWU_CTL - Memory system performance monitor configure memory
+ * bandwidth usage monitor control register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +14 -0
+#define MSMON_CFG_x_CTL_TYPE           GENMASK(7, 0)
+#define MSMON_CFG_x_CTL_OFLOW_STATUS_L BIT(15)
+#define MSMON_CFG_x_CTL_MATCH_PARTID   BIT(16)
+#define MSMON_CFG_x_CTL_MATCH_PMG      BIT(17)
+#define MSMON_CFG_x_CTL_SCLEN          BIT(19)
+#define MSMON_CFG_x_CTL_SUBTYPE        GENMASK(23, 20)
+#define MSMON_CFG_x_CTL_OFLOW_FRZ      BIT(24)
+#define MSMON_CFG_x_CTL_OFLOW_INTR     BIT(25)
+#define MSMON_CFG_x_CTL_OFLOW_STATUS   BIT(26)
+#define MSMON_CFG_x_CTL_CAPT_RESET     BIT(27)
+#define MSMON_CFG_x_CTL_CAPT_EVNT      GENMASK(30, 28)
+#define MSMON_CFG_x_CTL_EN             BIT(31)
+
+#define MSMON_CFG_MBWU_CTL_TYPE_MBWU
Quoted 3 lines
	0x42
+#define MSMON_CFG_MBWU_CTL_TYPE_CSU
	0x43

MSMON_CFG_CSU_CTL_TYPE_CSU?

Best regards,
Shaopeng TAN

Quoted 48 lines
drivers/platform/arm64/mpam/mpam_internal.h +11 -0
+
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_NONE                 0
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_READ                 1
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_WRITE                2
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_BOTH                 3
+
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_MAX                  3
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_MASK                 0x3
+
+/*
+ * MSMON_CFG_MBWU_FLT - Memory system performance monitor

configure memory
+ * bandwidth usage monitor filter register
+ */
+#define MSMON_CFG_MBWU_FLT_PARTID GENMASK(15,
0)
+#define MSMON_CFG_MBWU_FLT_PMG GENMASK(23,
16)
+#define MSMON_CFG_MBWU_FLT_RWBW GENMASK(31,
30)

drivers/platform/arm64/mpam/mpam_internal.h +3 -0
+
+/*
+ * MSMON_CSU - Memory system performance monitor cache storage usage

monitor
+ * register
+ * MSMON_CSU_CAPTURE - Memory system performance monitor cache storage usage
+ * capture register
+ * MSMON_MBWU - Memory system performance monitor memory bandwidth usage
+ * monitor register
+ * MSMON_MBWU_CAPTURE - Memory system performance monitor memory bandwidth usage
+ * capture register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +5 -0
+#define MSMON___VALUE          GENMASK(30, 0)
+#define MSMON___NRDY           BIT(31)
+#define MSMON_MBWU_L_VALUE     GENMASK(62, 0)
+/*
+ * MSMON_CAPT_EVNT - Memory system performance monitoring capture

event
+ * generation register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +2 -1
+#define MSMON_CAPT_EVNT_NOW    BIT(0)
+
 #endif /* MPAM_INTERNAL_H */
--

2.39.5

author maintainer
#56 lore raw

Re: [RFC PATCH 16/36] arm_mpam: Add MPAM MSC register layout definitions

Hi Shaopeng,

Quoted 31 lines · Shaopeng Tan (Fujitsu)
Quoted 31 lines

Memory Partitioning and Monitoring (MPAM) has memory mapped devices (MSCs) with an identity/configuration page.

Add the definitions for these registers as offset within the page(s).

drivers/platform/arm64/mpam/mpam_internal.h +2 -0
index d49bb884b433..9110c171d9d2 100644
@@ -150,4 +150,272 @@ extern struct list_head mpam_classes;  int
 
+/*
+ * MSMON_CFG_CSU_CTL - Memory system performance monitor configure

cache storage
+ * usage monitor control register
+ * MSMON_CFG_MBWU_CTL - Memory system performance monitor configure memory
+ * bandwidth usage monitor control register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +11 -0
+#define MSMON_CFG_x_CTL_TYPE           GENMASK(7, 0)
+#define MSMON_CFG_x_CTL_OFLOW_STATUS_L BIT(15)
+#define MSMON_CFG_x_CTL_MATCH_PARTID   BIT(16)
+#define MSMON_CFG_x_CTL_MATCH_PMG      BIT(17)
+#define MSMON_CFG_x_CTL_SCLEN          BIT(19)
+#define MSMON_CFG_x_CTL_SUBTYPE        GENMASK(23, 20)
+#define MSMON_CFG_x_CTL_OFLOW_FRZ      BIT(24)
+#define MSMON_CFG_x_CTL_OFLOW_INTR     BIT(25)
+#define MSMON_CFG_x_CTL_OFLOW_STATUS   BIT(26)
+#define MSMON_CFG_x_CTL_CAPT_RESET     BIT(27)
+#define MSMON_CFG_x_CTL_CAPT_EVNT      GENMASK(30, 28)
Quoted 8 lines
Quoted 6 lines
drivers/platform/arm64/mpam/mpam_internal.h +3 -0
+#define MSMON_CFG_x_CTL_EN             BIT(31)
+
+#define MSMON_CFG_MBWU_CTL_TYPE_MBWU
	0x42
+#define MSMON_CFG_MBWU_CTL_TYPE_CSU
	0x43

MSMON_CFG_CSU_CTL_TYPE_CSU?

Yup, copy-and-paste error. Thanks!

James

maintainer
#57 lore raw

Re: [RFC PATCH 16/36] arm_mpam: Add MPAM MSC register layout definitions

In reply to James Morse

Hi James,

Nit: The file uses a mixture of tabs and spaces.

Quoted 15 lines · James Morse

Memory Partitioning and Monitoring (MPAM) has memory mapped devices (MSCs) with an identity/configuration page.

Add the definitions for these registers as offset within the page(s).

---
  drivers/platform/arm64/mpam/mpam_internal.h | 268 ++++++++++++++++++++
  1 file changed, 268 insertions(+)
drivers/platform/arm64/mpam/mpam_internal.h +0 -0
index d49bb884b433..9110c171d9d2 100644

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 156

Quoted 7 lines
drivers/platform/arm64/mpam/mpam_internal.h +4 -0
@@ -150,4 +150,272 @@ extern struct list_head mpam_classes;
  int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,
  				   cpumask_t *affinity);
 
+/*
+ * MPAM MSCs have the following register layout. See:
+ * Arm Architecture Reference Manual Supplement - Memory System Resource
+ * Partitioning and Monitoring (MPAM), for Armv8-A. DDI 0598A.a

I've been checking this against https://developer.arm.com/documentation/ihi0099/latest/ as that looks to be the current document although hopefully the contents are non-contradictory.> + */

Quoted 41 lines
drivers/platform/arm64/mpam/mpam_internal.h +41 -0
+#define MPAM_ARCHITECTURE_V1    0x10
+
+/* Memory mapped control pages: */
+/* ID Register offsets in the memory mapped page */
+#define MPAMF_IDR               0x0000  /* features id register */
+#define MPAMF_MSMON_IDR         0x0080  /* performance monitoring features */
+#define MPAMF_IMPL_IDR          0x0028  /* imp-def partitioning */
+#define MPAMF_CPOR_IDR          0x0030  /* cache-portion partitioning */
+#define MPAMF_CCAP_IDR          0x0038  /* cache-capacity partitioning */
+#define MPAMF_MBW_IDR           0x0040  /* mem-bw partitioning */
+#define MPAMF_PRI_IDR           0x0048  /* priority partitioning */
+#define MPAMF_CSUMON_IDR        0x0088  /* cache-usage monitor */
+#define MPAMF_MBWUMON_IDR       0x0090  /* mem-bw usage monitor */
+#define MPAMF_PARTID_NRW_IDR    0x0050  /* partid-narrowing */
+#define MPAMF_IIDR              0x0018  /* implementer id register */
+#define MPAMF_AIDR              0x0020  /* architectural id register */
+
+/* Configuration and Status Register offsets in the memory mapped page */
+#define MPAMCFG_PART_SEL        0x0100  /* partid to configure: */
+#define MPAMCFG_CPBM            0x1000  /* cache-portion config */
+#define MPAMCFG_CMAX            0x0108  /* cache-capacity config */
+#define MPAMCFG_CMIN            0x0110  /* cache-capacity config */
+#define MPAMCFG_MBW_MIN         0x0200  /* min mem-bw config */
+#define MPAMCFG_MBW_MAX         0x0208  /* max mem-bw config */
+#define MPAMCFG_MBW_WINWD       0x0220  /* mem-bw accounting window config */
+#define MPAMCFG_MBW_PBM         0x2000  /* mem-bw portion bitmap config */
+#define MPAMCFG_PRI             0x0400  /* priority partitioning config */
+#define MPAMCFG_MBW_PROP        0x0500  /* mem-bw stride config */
+#define MPAMCFG_INTPARTID       0x0600  /* partid-narrowing config */
+
+#define MSMON_CFG_MON_SEL       0x0800  /* monitor selector */
+#define MSMON_CFG_CSU_FLT       0x0810  /* cache-usage monitor filter */
+#define MSMON_CFG_CSU_CTL       0x0818  /* cache-usage monitor config */
+#define MSMON_CFG_MBWU_FLT      0x0820  /* mem-bw monitor filter */
+#define MSMON_CFG_MBWU_CTL      0x0828  /* mem-bw monitor config */
+#define MSMON_CSU               0x0840  /* current cache-usage */
+#define MSMON_CSU_CAPTURE       0x0848  /* last cache-usage value captured */
+#define MSMON_MBWU              0x0860  /* current mem-bw usage value */
+#define MSMON_MBWU_CAPTURE      0x0868  /* last mem-bw value captured */
+#define MSMON_CAPT_EVNT         0x0808  /* signal a capture event */
+#define MPAMF_ESR               0x00F8  /* error status register */

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 208

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h +10 -0
+#define MPAMF_ECR               0x00F0  /* error control register */
+
+/* MPAMF_IDR - MPAM features ID register */
+#define MPAMF_IDR_PARTID_MAX            GENMASK(15, 0)
+#define MPAMF_IDR_PMG_MAX               GENMASK(23, 16)
+#define MPAMF_IDR_HAS_CCAP_PART         BIT(24)
+#define MPAMF_IDR_HAS_CPOR_PART         BIT(25)
+#define MPAMF_IDR_HAS_MBW_PART          BIT(26)
+#define MPAMF_IDR_HAS_PRI_PART          BIT(27)
+#define MPAMF_IDR_HAS_EXT               BIT(28)

MPAMF_IDR_EXT. The field name is ext rather than has_ext. > +#define MPAMF_IDR_HAS_IMPL_IDR BIT(29)

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 213

Quoted 4 lines
drivers/platform/arm64/mpam/mpam_internal.h +4 -0
+#define MPAMF_IDR_HAS_MSMON             BIT(30)
+#define MPAMF_IDR_HAS_PARTID_NRW        BIT(31)
+#define MPAMF_IDR_HAS_RIS               BIT(32)
+#define MPAMF_IDR_HAS_EXT_ESR           BIT(38)

MPAMF_IDR_HAS_EXTD_ESR. Missing D.> +#define MPAMF_IDR_HAS_ESR

    BIT(39)
Quoted 7 lines
drivers/platform/arm64/mpam/mpam_internal.h +7 -0
+#define MPAMF_IDR_RIS_MAX               GENMASK(59, 56)
+
+/* MPAMF_MSMON_IDR - MPAM performance monitoring ID register */
+#define MPAMF_MSMON_IDR_MSMON_CSU               BIT(16)
+#define MPAMF_MSMON_IDR_MSMON_MBWU              BIT(17)
+#define MPAMF_MSMON_IDR_HAS_LOCAL_CAPT_EVNT     BIT(31)
+

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 231

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h +10 -0
+/* MPAMF_CPOR_IDR - MPAM features cache portion partitioning ID register */
+#define MPAMF_CPOR_IDR_CPBM_WD                  GENMASK(15, 0)
+
+/* MPAMF_CCAP_IDR - MPAM features cache capacity partitioning ID register */
+#define MPAMF_CCAP_IDR_HAS_CMAX_SOFTLIM         BIT(31)
+#define MPAMF_CCAP_IDR_NO_CMAX                  BIT(30)
+#define MPAMF_CCAP_IDR_HAS_CMIN                 BIT(29)
+#define MPAMF_CCAP_IDR_HAS_CASSOC               BIT(28)
+#define MPAMF_CCAP_IDR_CASSOC_WD                GENMASK(12, 8)
+#define MPAMF_CCAP_IDR_CMAX_WD                  GENMASK(5, 0)

nit: Field ordering differs from the other registers.> +

Quoted 96 lines
drivers/platform/arm64/mpam/mpam_internal.h +96 -0
+/* MPAMF_MBW_IDR - MPAM features memory bandwidth partitioning ID register */
+#define MPAMF_MBW_IDR_BWA_WD            GENMASK(5, 0)
+#define MPAMF_MBW_IDR_HAS_MIN           BIT(10)
+#define MPAMF_MBW_IDR_HAS_MAX           BIT(11)
+#define MPAMF_MBW_IDR_HAS_PBM           BIT(12)
+#define MPAMF_MBW_IDR_HAS_PROP          BIT(13)
+#define MPAMF_MBW_IDR_WINDWR            BIT(14)
+#define MPAMF_MBW_IDR_BWPBM_WD          GENMASK(28, 16)
+
+/* MPAMF_PRI_IDR - MPAM features priority partitioning ID register */
+#define MPAMF_PRI_IDR_HAS_INTPRI        BIT(0)
+#define MPAMF_PRI_IDR_INTPRI_0_IS_LOW   BIT(1)
+#define MPAMF_PRI_IDR_INTPRI_WD         GENMASK(9, 4)
+#define MPAMF_PRI_IDR_HAS_DSPRI         BIT(16)
+#define MPAMF_PRI_IDR_DSPRI_0_IS_LOW    BIT(17)
+#define MPAMF_PRI_IDR_DSPRI_WD          GENMASK(25, 20)
+
+/* MPAMF_CSUMON_IDR - MPAM cache storage usage monitor ID register */
+#define MPAMF_CSUMON_IDR_NUM_MON        GENMASK(15, 0)
+#define MPAMF_CSUMON_IDR_HAS_OFLOW_CAPT	BIT(24)
+#define MPAMF_CSUMON_IDR_HAS_CEVNT_OFLW	BIT(25)
+#define MPAMF_CSUMON_IDR_HAS_OFSR	BIT(26)
+#define MPAMF_CSUMON_IDR_HAS_OFLOW_LNKG	BIT(27)
+#define MPAMF_CSUMON_IDR_HAS_XCL	BIT(29)
+#define MPAMF_CSUMON_IDR_CSU_RO		BIT(30)
+#define MPAMF_CSUMON_IDR_HAS_CAPTURE    BIT(31)
+
+/* MPAMF_MBWUMON_IDR - MPAM memory bandwidth usage monitor ID register */
+#define MPAMF_MBWUMON_IDR_NUM_MON       GENMASK(15, 0)
+#define MPAMF_MBWUMON_IDR_HAS_RWBW      BIT(28)
+#define MPAMF_MBWUMON_IDR_LWD           BIT(29)
+#define MPAMF_MBWUMON_IDR_HAS_LONG      BIT(30)
+#define MPAMF_MBWUMON_IDR_HAS_CAPTURE   BIT(31)
+
+/* MPAMF_PARTID_NRW_IDR - MPAM PARTID narrowing ID register */
+#define MPAMF_PARTID_NRW_IDR_INTPARTID_MAX      GENMASK(15, 0)
+
+/* MPAMF_IIDR - MPAM implementation ID register */
+#define MPAMF_IIDR_PRODUCTID    GENMASK(31, 20)
+#define MPAMF_IIDR_PRODUCTID_SHIFT	20
+#define MPAMF_IIDR_VARIANT      GENMASK(19, 16)
+#define MPAMF_IIDR_VARIANT_SHIFT	16
+#define MPAMF_IIDR_REVISON      GENMASK(15, 12)
+#define MPAMF_IIDR_REVISON_SHIFT	12
+#define MPAMF_IIDR_IMPLEMENTER  GENMASK(11, 0)
+#define MPAMF_IIDR_IMPLEMENTER_SHIFT	0
+
+/* MPAMF_AIDR - MPAM architecture ID register */
+#define MPAMF_AIDR_ARCH_MAJOR_REV       GENMASK(7, 4)
+#define MPAMF_AIDR_ARCH_MINOR_REV       GENMASK(3, 0)
+
+/* MPAMCFG_PART_SEL - MPAM partition configuration selection register */
+#define MPAMCFG_PART_SEL_PARTID_SEL     GENMASK(15, 0)
+#define MPAMCFG_PART_SEL_INTERNAL       BIT(16)
+#define MPAMCFG_PART_SEL_RIS            GENMASK(27, 24)
+
+/* MPAMCFG_CMAX - MPAM cache capacity configuration register */
+#define MPAMCFG_CMAX_SOFTLIM            BIT(31)
+#define MPAMCFG_CMAX_CMAX               GENMASK(15, 0)
+
+/* MPAMCFG_CMIN - MPAM cache capacity configuration register */
+#define MPAMCFG_CMIN_CMIN               GENMASK(15, 0)
+
+/*
+ * MPAMCFG_MBW_MIN - MPAM memory minimum bandwidth partitioning configuration
+ *                   register
+ */
+#define MPAMCFG_MBW_MIN_MIN             GENMASK(15, 0)
+
+/*
+ * MPAMCFG_MBW_MAX - MPAM memory maximum bandwidth partitioning configuration
+ *                   register
+ */
+#define MPAMCFG_MBW_MAX_MAX             GENMASK(15, 0)
+#define MPAMCFG_MBW_MAX_HARDLIM         BIT(31)
+
+/*
+ * MPAMCFG_MBW_WINWD - MPAM memory bandwidth partitioning window width
+ *                     register
+ */
+#define MPAMCFG_MBW_WINWD_US_FRAC       GENMASK(7, 0)
+#define MPAMCFG_MBW_WINWD_US_INT        GENMASK(23, 8)
+
+/* MPAMCFG_PRI - MPAM priority partitioning configuration register */
+#define MPAMCFG_PRI_INTPRI              GENMASK(15, 0)
+#define MPAMCFG_PRI_DSPRI               GENMASK(31, 16)
+
+/*
+ * MPAMCFG_MBW_PROP - Memory bandwidth proportional stride partitioning
+ *                    configuration register
+ */
+#define MPAMCFG_MBW_PROP_STRIDEM1       GENMASK(15, 0)
+#define MPAMCFG_MBW_PROP_EN             BIT(31)
+
+/*
+ * MPAMCFG_INTPARTID - MPAM internal partition narrowing configuration register

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes;

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h +10 -0
+ */
+#define MPAMCFG_INTPARTID_INTPARTID     GENMASK(15, 0)
+#define MPAMCFG_INTPARTID_INTERNAL      BIT(16)
+
+/* MSMON_CFG_MON_SEL - Memory system performance monitor selection register */
+#define MSMON_CFG_MON_SEL_MON_SEL       GENMASK(15, 0)
+#define MSMON_CFG_MON_SEL_RIS           GENMASK(27, 24)
+
+/* MPAMF_ESR - MPAM Error Status Register */
+#define MPAMF_ESR_PARTID_OR_MON GENMASK(15, 0)Probably a better name but PARTID_MON is in the specification.> +#define 

MPAMF_ESR_PMG GENMASK(23, 16)

Quoted 22 lines
drivers/platform/arm64/mpam/mpam_internal.h +22 -0
+#define MPAMF_ESR_ERRCODE       GENMASK(27, 24)
+#define MPAMF_ESR_OVRWR         BIT(31)
+#define MPAMF_ESR_RIS           GENMASK(35, 32)
+
+/* MPAMF_ECR - MPAM Error Control Register */
+#define MPAMF_ECR_INTEN         BIT(0)
+
+/* Error conditions in accessing memory mapped registers */
+#define MPAM_ERRCODE_NONE                       0
+#define MPAM_ERRCODE_PARTID_SEL_RANGE           1
+#define MPAM_ERRCODE_REQ_PARTID_RANGE           2
+#define MPAM_ERRCODE_MSMONCFG_ID_RANGE          3
+#define MPAM_ERRCODE_REQ_PMG_RANGE              4
+#define MPAM_ERRCODE_MONITOR_RANGE              5
+#define MPAM_ERRCODE_INTPARTID_RANGE            6
+#define MPAM_ERRCODE_UNEXPECTED_INTERNAL        7
+
+/*
+ * MSMON_CFG_CSU_FLT - Memory system performance monitor configure cache storage
+ *                    usage monitor filter register
+ */
+#define MSMON_CFG_CSU_FLT_PARTID       GENMASK(15, 0)

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 371

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h +10 -0
+#define MSMON_CFG_CSU_FLT_PMG          GENMASK(23, 16)
+
+/*
+ * MSMON_CFG_CSU_CTL - Memory system performance monitor configure cache storage
+ *                    usage monitor control register
+ * MSMON_CFG_MBWU_CTL - Memory system performance monitor configure memory
+ *                     bandwidth usage monitor control register
+ */
+#define MSMON_CFG_x_CTL_TYPE           GENMASK(7, 0)
+#define MSMON_CFG_x_CTL_OFLOW_STATUS_L BIT(15)

No OFLOW_STATUS_L for csu.> +#define MSMON_CFG_x_CTL_MATCH_PARTID BIT(16)

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 375

Quoted 3 lines
drivers/platform/arm64/mpam/mpam_internal.h +3 -0
+#define MSMON_CFG_x_CTL_MATCH_PMG      BIT(17)
+#define MSMON_CFG_x_CTL_SCLEN          BIT(19)
+#define MSMON_CFG_x_CTL_SUBTYPE        GENMASK(23, 20)

GENMASK(22,20)> +#define MSMON_CFG_x_CTL_OFLOW_FRZ BIT(24)

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 389

Quoted 13 lines
drivers/platform/arm64/mpam/mpam_internal.h +13 -0
+#define MSMON_CFG_x_CTL_OFLOW_INTR     BIT(25)
+#define MSMON_CFG_x_CTL_OFLOW_STATUS   BIT(26)
+#define MSMON_CFG_x_CTL_CAPT_RESET     BIT(27)
+#define MSMON_CFG_x_CTL_CAPT_EVNT      GENMASK(30, 28)
+#define MSMON_CFG_x_CTL_EN             BIT(31)
+
+#define MSMON_CFG_MBWU_CTL_TYPE_MBWU			0x42
+#define MSMON_CFG_MBWU_CTL_TYPE_CSU			0x43
+
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_NONE                 0
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_READ                 1
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_WRITE                2
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_BOTH                 3

I'm not sure where these come from? SUBTYPE is marked unused in the spec. Remove?> +

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 392

Quoted 2 lines
drivers/platform/arm64/mpam/mpam_internal.h +2 -0
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_MAX                  3
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_MASK                 0x3

Remove for same reason.> +

Quoted 11 lines
drivers/platform/arm64/mpam/mpam_internal.h +11 -0
+/*
+ * MSMON_CFG_MBWU_FLT - Memory system performance monitor configure memory
+ *                     bandwidth usage monitor filter register
+ */
+#define MSMON_CFG_MBWU_FLT_PARTID               GENMASK(15, 0)
+#define MSMON_CFG_MBWU_FLT_PMG                  GENMASK(23, 16)
+#define MSMON_CFG_MBWU_FLT_RWBW                 GENMASK(31, 30)
+
+/*
+ * MSMON_CSU - Memory system performance monitor cache storage usage monitor
+ *            register

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 414

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h +10 -0
+ * MSMON_CSU_CAPTURE -  Memory system performance monitor cache storage usage
+ *                     capture register
+ * MSMON_MBWU  - Memory system performance monitor memory bandwidth usage
+ *               monitor register
+ * MSMON_MBWU_CAPTURE - Memory system performance monitor memory bandwidth usage
+ *                     capture register
+ */
+#define MSMON___VALUE          GENMASK(30, 0)
+#define MSMON___NRDY           BIT(31)
+#define MSMON_MBWU_L_VALUE     GENMASK(62, 0)

This gets renamed in the series. I think all registers layout definitions can be added in this commit.> +/*

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 421

Quoted 6 lines
drivers/platform/arm64/mpam/mpam_internal.h +5 -0
+ * MSMON_CAPT_EVNT - Memory system performance monitoring capture event
+ *                  generation register
+ */
+#define MSMON_CAPT_EVNT_NOW    BIT(0)
+
  #endif /* MPAM_INTERNAL_H */

Thanks,

Ben

author maintainer
#58 lore raw

Re: [RFC PATCH 16/36] arm_mpam: Add MPAM MSC register layout definitions

Hi Ben,

Quoted 6 lines · Ben Horgan
Quoted 4 lines · James Morse

Memory Partitioning and Monitoring (MPAM) has memory mapped devices (MSCs) with an identity/configuration page.

Add the definitions for these registers as offset within the page(s).

Quoted 16 lines
Quoted 8 lines
drivers/platform/arm64/mpam/ renamed from drivers/platform/arm64/mpam/mpam_internal.h +0 -0

mpam_internal.h
index d49bb884b433..9110c171d9d2 100644

drivers/platform/arm64/mpam/mpam_internal.h +0 -0
@@ -150,4 +150,272 @@ extern struct list_head mpam_classes;
   int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,
                      cpumask_t *affinity);
Quoted 4 lines

  +/*
+ * MPAM MSCs have the following register layout. See:
+ * Arm Architecture Reference Manual Supplement - Memory System Resource
+ * Partitioning and Monitoring (MPAM), for Armv8-A. DDI 0598A.a

I've been checking this against https://developer.arm.com/documentation/ihi0099/latest/ as that looks to be the current document although hopefully the contents are non- contradictory.

Yeah, A.a was the first release, they then split that document up and the CPU half got consumed by the arm-arm. (pacman, but in pdf form).
I've updated this comment to point to the more modern 'system component specification', aka "all the mmio stuff".

Quoted 10 lines

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 208

Quoted 8 lines
drivers/platform/arm64/mpam/mpam_internal.h +8 -0
+/* MPAMF_IDR - MPAM features ID register */
+#define MPAMF_IDR_PARTID_MAX            GENMASK(15, 0)
+#define MPAMF_IDR_PMG_MAX               GENMASK(23, 16)
+#define MPAMF_IDR_HAS_CCAP_PART         BIT(24)
+#define MPAMF_IDR_HAS_CPOR_PART         BIT(25)
+#define MPAMF_IDR_HAS_MBW_PART          BIT(26)
+#define MPAMF_IDR_HAS_PRI_PART          BIT(27)
+#define MPAMF_IDR_HAS_EXT               BIT(28)

MPAMF_IDR_EXT. The field name is ext rather than has_ext.

Making it the odd one out!
I'll change this in the hope that one day this sort of thing can be generated.

Quoted 7 lines

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 213

Quoted 5 lines
drivers/platform/arm64/mpam/mpam_internal.h +5 -0
+#define MPAMF_IDR_HAS_IMPL_IDR          BIT(29)
+#define MPAMF_IDR_HAS_MSMON             BIT(30)
+#define MPAMF_IDR_HAS_PARTID_NRW        BIT(31)
+#define MPAMF_IDR_HAS_RIS               BIT(32)
+#define MPAMF_IDR_HAS_EXT_ESR           BIT(38)

MPAMF_IDR_HAS_EXTD_ESR. Missing D.

Fixed.

Quoted 12 lines
Quoted 12 lines
drivers/platform/arm64/mpam/mpam_internal.h +12 -0
+#define MPAMF_IDR_HAS_ESR     BIT(39)
+#define MPAMF_IDR_RIS_MAX               GENMASK(59, 56)
+
+/* MPAMF_MSMON_IDR - MPAM performance monitoring ID register */
+#define MPAMF_MSMON_IDR_MSMON_CSU               BIT(16)
+#define MPAMF_MSMON_IDR_MSMON_MBWU              BIT(17)
+#define MPAMF_MSMON_IDR_HAS_LOCAL_CAPT_EVNT     BIT(31)
+
+/* MPAMF_CPOR_IDR - MPAM features cache portion partitioning ID register */
+#define MPAMF_CPOR_IDR_CPBM_WD                  GENMASK(15, 0)
+
+/* MPAMF_CCAP_IDR - MPAM features cache capacity partitioning ID register */
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 231

Quoted 6 lines
drivers/platform/arm64/mpam/mpam_internal.h +6 -0
+#define MPAMF_CCAP_IDR_HAS_CMAX_SOFTLIM         BIT(31)
+#define MPAMF_CCAP_IDR_NO_CMAX                  BIT(30)
+#define MPAMF_CCAP_IDR_HAS_CMIN                 BIT(29)
+#define MPAMF_CCAP_IDR_HAS_CASSOC               BIT(28)
+#define MPAMF_CCAP_IDR_CASSOC_WD                GENMASK(12, 8)
+#define MPAMF_CCAP_IDR_CMAX_WD                  GENMASK(5, 0)

nit: Field ordering differs from the other registers.

Fixed,

[..]

Quoted 3 lines

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 338

Quoted 3 lines
drivers/platform/arm64/mpam/mpam_internal.h +2 -0
+/* MPAMF_ESR - MPAM Error Status Register */
 
+#define MPAMF_ESR_PARTID_OR_MON GENMASK(15, 0)

Probably a better name but PARTID_MON is in the specification.

Fixed,

[..]

Quoted 10 lines

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 371

Quoted 8 lines
drivers/platform/arm64/mpam/mpam_internal.h +8 -0
+/*
+ * MSMON_CFG_CSU_CTL - Memory system performance monitor configure cache storage
+ *                    usage monitor control register
+ * MSMON_CFG_MBWU_CTL - Memory system performance monitor configure memory
+ *                     bandwidth usage monitor control register
+ */
+#define MSMON_CFG_x_CTL_TYPE           GENMASK(7, 0)
+#define MSMON_CFG_x_CTL_OFLOW_STATUS_L BIT(15)

No OFLOW_STATUS_L for csu.

You're suggesting there shouldn't be an 'x' in the middle? Sure.

Overflow is nonsense for the CSU 'counters' as they don't count up, so can't overflow. (and yet they have an overflow status bit!)

Quoted 7 lines

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 375

Quoted 5 lines
drivers/platform/arm64/mpam/mpam_internal.h +4 -0
+#define MSMON_CFG_x_CTL_MATCH_PARTID   BIT(16)
+#define MSMON_CFG_x_CTL_MATCH_PMG      BIT(17)
+#define MSMON_CFG_x_CTL_SCLEN          BIT(19)
 
+#define MSMON_CFG_x_CTL_SUBTYPE        GENMASK(23, 20)

GENMASK(22,20)> +#define MSMON_CFG_x_CTL_OFLOW_FRZ      BIT(24)

Quoted 1 line

+#define MSMON_CFG_x_CTL_OFLOW_INTR     BIT(25)

(Are you using Outlook?)

Quoted 6 lines
Quoted 6 lines
drivers/platform/arm64/mpam/mpam_internal.h +6 -0
+#define MSMON_CFG_x_CTL_OFLOW_STATUS   BIT(26)
+#define MSMON_CFG_x_CTL_CAPT_RESET     BIT(27)
+#define MSMON_CFG_x_CTL_CAPT_EVNT      GENMASK(30, 28)
+#define MSMON_CFG_x_CTL_EN             BIT(31)
+
+#define MSMON_CFG_MBWU_CTL_TYPE_MBWU            0x42
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 389

Quoted 6 lines
drivers/platform/arm64/mpam/mpam_internal.h +6 -0
+#define MSMON_CFG_MBWU_CTL_TYPE_CSU            0x43
+
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_NONE                 0
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_READ                 1
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_WRITE                2
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_BOTH                 3

I'm not sure where these come from? SUBTYPE is marked unused in the spec. Remove?> +

Looks like an earlier version of RWBW feature. I'll rip these out.

Quoted 23 lines

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 392

Quoted 2 lines
drivers/platform/arm64/mpam/mpam_internal.h +2 -0
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_MAX                  3
+#define MSMON_CFG_MBWU_CTL_SUBTYPE_MASK                 0x3

Remove for same reason.

Quoted 17 lines
drivers/platform/arm64/mpam/mpam_internal.h +17 -0
+
+/*
+ * MSMON_CFG_MBWU_FLT - Memory system performance monitor configure memory
+ *                     bandwidth usage monitor filter register
+ */
+#define MSMON_CFG_MBWU_FLT_PARTID               GENMASK(15, 0)
+#define MSMON_CFG_MBWU_FLT_PMG                  GENMASK(23, 16)
+#define MSMON_CFG_MBWU_FLT_RWBW                 GENMASK(31, 30)
+
+/*
+ * MSMON_CSU - Memory system performance monitor cache storage usage monitor
+ *            register
+ * MSMON_CSU_CAPTURE -  Memory system performance monitor cache storage usage
+ *                     capture register
+ * MSMON_MBWU  - Memory system performance monitor memory bandwidth usage
+ *               monitor register
+ * MSMON_MBWU_CAPTURE - Memory system performance monitor memory bandwidth usage
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_internal.h extern struct list_head mpam_classes; · line 414

Quoted 5 lines

+ *                     capture register
+ */

drivers/platform/arm64/mpam/mpam_internal.h +3 -0
+#define MSMON___VALUE          GENMASK(30, 0)
+#define MSMON___NRDY           BIT(31)
+#define MSMON_MBWU_L_VALUE     GENMASK(62, 0)

This gets renamed in the series. I think all registers layout definitions can be added in this commit.

Yup, I've pulled that hunk in. (and spotted there is another) Adding them all here was the plan. (and there will be a few that don't get used)

Thanks for going through all those!

James

author maintainer
#59 lore raw

RFC PATCH 17/36 arm_mpam: Add cpuhp callbacks to probe MSC hardware

In reply to James Morse

Because an MSC can only by accessed from the CPUs in its cpu-affinity set we need to be running on one of those CPUs to probe the MSC hardware.

Do this work in the cpuhp callback. Probing the hardware will only happen before MPAM is enabled, walk all the MSCs and probe those we can reach that haven't already been probed.

Later once MPAM is enabled, this cpuhp callback will be replaced by one that avoids the global list.

Enabling a static key will also take the cpuhp lock, so can't be done from the cpuhp callback. Whenever a new MSC has been probed schedule work to test if all the MSCs have now been probed.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 149 +++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |   8 +-
 2 files changed, 152 insertions(+), 5 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +146 -3
index 0d6d5180903b..89434ae3efa6 100644
@@ -4,6 +4,7 @@
 #define pr_fmt(fmt) "%s:%s: " fmt, KBUILD_MODNAME, __func__
 
 #include <linux/acpi.h>
+#include <linux/atomic.h>
 #include <linux/arm_mpam.h>
 #include <linux/cacheinfo.h>
 #include <linux/cpu.h>
@@ -21,6 +22,7 @@
 #include <linux/slab.h>
 #include <linux/spinlock.h>
 #include <linux/types.h>
+#include <linux/workqueue.h>
 
 #include <acpi/pcc.h>
 
@@ -39,6 +41,16 @@ struct srcu_struct mpam_srcu;
 /* MPAM isn't available until all the MSC have been probed. */
 static u32 mpam_num_msc;
 
+static int mpam_cpuhp_state;
+static DEFINE_MUTEX(mpam_cpuhp_state_lock);
+
+/*
+ * mpam is enabled once all devices have been probed from CPU online callbacks,
+ * scheduled via this work_struct. If access to an MSC depends on a CPU that
+ * was not brought online at boot, this can happen surprisingly late.
+ */
+static DECLARE_WORK(mpam_enable_work, &mpam_enable);
+
 /*
  * An MSC is a physical container for controls and monitors, each identified by
  * their RIS index. These share a base-address, interrupts and some MMIO
@@ -78,6 +90,22 @@ LIST_HEAD(mpam_classes);
 /* List of all objects that can be free()d after synchronise_srcu() */
 static LLIST_HEAD(mpam_garbage);
 
+static u32 __mpam_read_reg(struct mpam_msc *msc, u16 reg)
+{
+	WARN_ON_ONCE(reg > msc->mapped_hwpage_sz);
+	WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(), &msc->accessibility));
+
+	return readl_relaxed(msc->mapped_hwpage + reg);
+}
+
+static inline u32 _mpam_read_partsel_reg(struct mpam_msc *msc, u16 reg)
+{
+	lockdep_assert_held_once(&msc->part_sel_lock);
+	return __mpam_read_reg(msc, reg);
+}
+
+#define mpam_read_partsel_reg(msc, reg)        _mpam_read_partsel_reg(msc, MPAMF_##reg)
+
 #define init_garbage(x)	init_llist_node(&(x)->garbage.llist)
 
 static struct mpam_vmsc *
@@ -513,9 +541,84 @@ int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
 	return err;
 }
 
-static void mpam_discovery_complete(void)
+static int mpam_msc_hw_probe(struct mpam_msc *msc)
 {
-	pr_err("Discovered all MSC\n");
+	u64 idr;
+	int err;
+
+	lockdep_assert_held(&msc->probe_lock);
+
+	mutex_lock(&msc->part_sel_lock);
+	idr = mpam_read_partsel_reg(msc, AIDR);
+	if ((idr & MPAMF_AIDR_ARCH_MAJOR_REV) != MPAM_ARCHITECTURE_V1) {
+		pr_err_once("%s does not match MPAM architecture v1.0\n",
+			    dev_name(&msc->pdev->dev));
+		err = -EIO;
+	} else {
+		msc->probed = true;
+		err = 0;
+	}
+	mutex_unlock(&msc->part_sel_lock);
+
+	return err;
+}
+
+static int mpam_cpu_online(unsigned int cpu)
+{
+	return 0;
+}
+
+/* Before mpam is enabled, try to probe new MSC */
+static int mpam_discovery_cpu_online(unsigned int cpu)
+{
+	int err = 0;
+	struct mpam_msc *msc;
+	bool new_device_probed = false;
+
+	mutex_lock(&mpam_list_lock);
+	list_for_each_entry(msc, &mpam_all_msc, glbl_list) {
+		if (!cpumask_test_cpu(cpu, &msc->accessibility))
+			continue;
+
+		mutex_lock(&msc->probe_lock);
+		if (!msc->probed)
+			err = mpam_msc_hw_probe(msc);
+		mutex_unlock(&msc->probe_lock);
+
+		if (!err)
+			new_device_probed = true;
+		else
+			break; // mpam_broken
+	}
+	mutex_unlock(&mpam_list_lock);
+
+	if (new_device_probed && !err)
+		schedule_work(&mpam_enable_work);
+
+	return err;
+}
+
+static int mpam_cpu_offline(unsigned int cpu)
+{
+	return 0;
+}
+
+static void mpam_register_cpuhp_callbacks(int (*online)(unsigned int online),
+					  int (*offline)(unsigned int offline))
+{
+	mutex_lock(&mpam_cpuhp_state_lock);
+	if (mpam_cpuhp_state) {
+		cpuhp_remove_state(mpam_cpuhp_state);
+		mpam_cpuhp_state = 0;
+	}
+
+	mpam_cpuhp_state = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "mpam:online",
+					     online, offline);
+	if (mpam_cpuhp_state <= 0) {
+		pr_err("Failed to register cpuhp callbacks");
+		mpam_cpuhp_state = 0;
+	}
+	mutex_unlock(&mpam_cpuhp_state_lock);
 }
 
 static int mpam_dt_count_msc(void)
@@ -774,7 +877,7 @@ static int mpam_msc_drv_probe(struct platform_device *pdev)
 	}
 
 	if (!err && fw_num_msc == mpam_num_msc)
-		mpam_discovery_complete();
+		mpam_register_cpuhp_callbacks(&mpam_discovery_cpu_online, NULL);
 
 	if (err && msc)
 		mpam_msc_drv_remove(pdev);
@@ -797,6 +900,46 @@ static struct platform_driver mpam_msc_driver = {
 	.remove = mpam_msc_drv_remove,
 };
 
+static void mpam_enable_once(void)
+{
+	mutex_lock(&mpam_cpuhp_state_lock);
+	cpuhp_remove_state(mpam_cpuhp_state);
+	mpam_cpuhp_state = 0;
+	mutex_unlock(&mpam_cpuhp_state_lock);
+
+	mpam_register_cpuhp_callbacks(mpam_cpu_online, mpam_cpu_offline);
+
+	pr_info("MPAM enabled\n");
+}
+
+/*
+ * Enable mpam once all devices have been probed.
+ * Scheduled by mpam_discovery_cpu_online() once all devices have been created.
+ * Also scheduled when new devices are probed when new CPUs come online.
+ */
+void mpam_enable(struct work_struct *work)
+{
+	static atomic_t once;
+	struct mpam_msc *msc;
+	bool all_devices_probed = true;
+
+	/* Have we probed all the hw devices? */
+	mutex_lock(&mpam_list_lock);
+	list_for_each_entry(msc, &mpam_all_msc, glbl_list) {
+		mutex_lock(&msc->probe_lock);
+		if (!msc->probed)
+			all_devices_probed = false;
+		mutex_unlock(&msc->probe_lock);
+
+		if (!all_devices_probed)
+			break;
+	}
+	mutex_unlock(&mpam_list_lock);
+
+	if (all_devices_probed && !atomic_fetch_inc(&once))
+		mpam_enable_once();
+}
+
 /*
  * MSC that are hidden under caches are not created as platform devices
  * as there is no cache driver. Caches are also special-cased in
drivers/platform/arm64/mpam/mpam_internal.h +6 -2
index 9110c171d9d2..f56e69ff8397 100644
@@ -49,6 +49,7 @@ struct mpam_msc {
 	 * properties become read-only and the lists are protected by SRCU.
 	 */
 	struct mutex		probe_lock;
+	bool			probed;
 	unsigned long		ris_idxs[128 / BITS_PER_LONG];
 	u32			ris_max;
 
@@ -59,14 +60,14 @@ struct mpam_msc {
 	 * part_sel_lock protects access to the MSC hardware registers that are
 	 * affected by MPAMCFG_PART_SEL. (including the ID registers that vary
 	 * by RIS).
-	 * If needed, take msc->lock first.
+	 * If needed, take msc->probe_lock first.
 	 */
 	struct mutex		part_sel_lock;
 
 	/*
 	 * mon_sel_lock protects access to the MSC hardware registers that are
 	 * affeted by MPAMCFG_MON_SEL.
-	 * If needed, take msc->lock first.
+	 * If needed, take msc->probe_lock first.
 	 */
 	struct mutex		outer_mon_sel_lock;
 	raw_spinlock_t		inner_mon_sel_lock;
@@ -147,6 +148,9 @@ struct mpam_msc_ris {
 extern struct srcu_struct mpam_srcu;
 extern struct list_head mpam_classes;
 
+/* Scheduled work callback to enable mpam once all MSC have been probed */
+void mpam_enable(struct work_struct *work);
+
 int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,
 				   cpumask_t *affinity);
 
Signature · 1 line
2.39.5
maintainer
#60 lore raw

Re: [RFC PATCH 17/36] arm_mpam: Add cpuhp callbacks to probe MSC hardware

Hi James,

Quoted 90 lines · James Morse

Because an MSC can only by accessed from the CPUs in its cpu-affinity set we need to be running on one of those CPUs to probe the MSC hardware.

Do this work in the cpuhp callback. Probing the hardware will only happen before MPAM is enabled, walk all the MSCs and probe those we can reach that haven't already been probed.

Later once MPAM is enabled, this cpuhp callback will be replaced by one that avoids the global list.

Enabling a static key will also take the cpuhp lock, so can't be done from the cpuhp callback. Whenever a new MSC has been probed schedule work to test if all the MSCs have now been probed.

---
  drivers/platform/arm64/mpam/mpam_devices.c  | 149 +++++++++++++++++++-
  drivers/platform/arm64/mpam/mpam_internal.h |   8 +-
  2 files changed, 152 insertions(+), 5 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +29 -2
index 0d6d5180903b..89434ae3efa6 100644
@@ -4,6 +4,7 @@
  #define pr_fmt(fmt) "%s:%s: " fmt, KBUILD_MODNAME, __func__
 
  #include <linux/acpi.h>
+#include <linux/atomic.h>
  #include <linux/arm_mpam.h>
  #include <linux/cacheinfo.h>
  #include <linux/cpu.h>
@@ -21,6 +22,7 @@
  #include <linux/slab.h>
  #include <linux/spinlock.h>
  #include <linux/types.h>
+#include <linux/workqueue.h>
 
  #include <acpi/pcc.h>
 
@@ -39,6 +41,16 @@ struct srcu_struct mpam_srcu;
  /* MPAM isn't available until all the MSC have been probed. */
  static u32 mpam_num_msc;
 
+static int mpam_cpuhp_state;
+static DEFINE_MUTEX(mpam_cpuhp_state_lock);
+
+/*
+ * mpam is enabled once all devices have been probed from CPU online callbacks,
+ * scheduled via this work_struct. If access to an MSC depends on a CPU that
+ * was not brought online at boot, this can happen surprisingly late.
+ */
+static DECLARE_WORK(mpam_enable_work, &mpam_enable);
+
  /*
   * An MSC is a physical container for controls and monitors, each identified by
   * their RIS index. These share a base-address, interrupts and some MMIO
@@ -78,6 +90,22 @@ LIST_HEAD(mpam_classes);
  /* List of all objects that can be free()d after synchronise_srcu() */
  static LLIST_HEAD(mpam_garbage);
 
+static u32 __mpam_read_reg(struct mpam_msc *msc, u16 reg)
+{
+	WARN_ON_ONCE(reg > msc->mapped_hwpage_sz);
+	WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(), &msc->accessibility));
+
+	return readl_relaxed(msc->mapped_hwpage + reg);
+}
+
+static inline u32 _mpam_read_partsel_reg(struct mpam_msc *msc, u16 reg)
+{
+	lockdep_assert_held_once(&msc->part_sel_lock);
+	return __mpam_read_reg(msc, reg);
+}
+
+#define mpam_read_partsel_reg(msc, reg)        _mpam_read_partsel_reg(msc, MPAMF_##reg)
+
  #define init_garbage(x)	init_llist_node(&(x)->garbage.llist)
 
  static struct mpam_vmsc *
@@ -513,9 +541,84 @@ int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
  	return err;
  }
 
-static void mpam_discovery_complete(void)
+static int mpam_msc_hw_probe(struct mpam_msc *msc)
  {
-	pr_err("Discovered all MSC\n");

drivers/platform/arm64/mpam/mpam_devices.c int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx, · line 555

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+	u64 idr;
+	int err;
+
+	lockdep_assert_held(&msc->probe_lock);
+
+	mutex_lock(&msc->part_sel_lock);
+	idr = mpam_read_partsel_reg(msc, AIDR);
+	if ((idr & MPAMF_AIDR_ARCH_MAJOR_REV) != MPAM_ARCHITECTURE_V1) {
+		pr_err_once("%s does not match MPAM architecture v1.0\n",
+			    dev_name(&msc->pdev->dev));

The error message need only mention the major revision. You've added support for v1.1 and v1.0.> + err = -EIO;

drivers/platform/arm64/mpam/mpam_devices.c int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx, · line 564

Quoted 8 lines
drivers/platform/arm64/mpam/mpam_devices.c +8 -0
+	} else {
+		msc->probed = true;
+		err = 0;
+	}
+	mutex_unlock(&msc->part_sel_lock);
+
+	return err;
+}

[snip]
Thanks,

Ben

author maintainer
#61 lore raw

Re: [RFC PATCH 17/36] arm_mpam: Add cpuhp callbacks to probe MSC hardware

Hi Ben,

Quoted 33 lines · Ben Horgan
Quoted 32 lines · James Morse

Because an MSC can only by accessed from the CPUs in its cpu-affinity set we need to be running on one of those CPUs to probe the MSC hardware.

Do this work in the cpuhp callback. Probing the hardware will only happen before MPAM is enabled, walk all the MSCs and probe those we can reach that haven't already been probed.

Later once MPAM is enabled, this cpuhp callback will be replaced by one that avoids the global list.

Enabling a static key will also take the cpuhp lock, so can't be done from the cpuhp callback. Whenever a new MSC has been probed schedule work to test if all the MSCs have now been probed.

drivers/platform/arm64/mpam/ renamed from drivers/platform/arm64/mpam/mpam_devices.c +0 -0

mpam_devices.c
index 0d6d5180903b..89434ae3efa6 100644

drivers/platform/arm64/mpam/mpam_devices.c +6 -1
@@ -513,9 +541,84 @@ int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
       return err;
   }
   -static void mpam_discovery_complete(void)
+static int mpam_msc_hw_probe(struct mpam_msc *msc)
   {
-    pr_err("Discovered all MSC\n");
+    u64 idr;
+    int err;
+
+    lockdep_assert_held(&msc->probe_lock);
+
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_devices.c int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx, · line 555

Quoted 5 lines
drivers/platform/arm64/mpam/mpam_devices.c +5 -0
+    mutex_lock(&msc->part_sel_lock);
+    idr = mpam_read_partsel_reg(msc, AIDR);
+    if ((idr & MPAMF_AIDR_ARCH_MAJOR_REV) != MPAM_ARCHITECTURE_V1) {
+        pr_err_once("%s does not match MPAM architecture v1.0\n",
+                dev_name(&msc->pdev->dev));

The error message need only mention the major revision. You've added support for v1.1 and v1.0.

Makes sense,

Thanks,

James

#62 lore raw

Re: [RFC PATCH 17/36] arm_mpam: Add cpuhp callbacks to probe MSC hardware

In reply to James Morse

Hi James,

Quoted 179 lines

Because an MSC can only by accessed from the CPUs in its cpu-affinity set we need to be running on one of those CPUs to probe the MSC hardware.

Do this work in the cpuhp callback. Probing the hardware will only happen before MPAM is enabled, walk all the MSCs and probe those we can reach that haven't already been probed.

Later once MPAM is enabled, this cpuhp callback will be replaced by one that avoids the global list.

Enabling a static key will also take the cpuhp lock, so can't be done from the cpuhp callback. Whenever a new MSC has been probed schedule work to test if all the MSCs have now been probed.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 149 +++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |   8 +-
 2 files changed, 152 insertions(+), 5 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +106 -3
index 0d6d5180903b..89434ae3efa6 100644
@@ -4,6 +4,7 @@
 #define pr_fmt(fmt) "%s:%s: " fmt, KBUILD_MODNAME, __func__
 
 #include <linux/acpi.h>
+#include <linux/atomic.h>
 #include <linux/arm_mpam.h>
 #include <linux/cacheinfo.h>
 #include <linux/cpu.h>
@@ -21,6 +22,7 @@
 #include <linux/slab.h>
 #include <linux/spinlock.h>
 #include <linux/types.h>
+#include <linux/workqueue.h>
 
 #include <acpi/pcc.h>
 
@@ -39,6 +41,16 @@ struct srcu_struct mpam_srcu;
 /* MPAM isn't available until all the MSC have been probed. */
 static u32 mpam_num_msc;
 
+static int mpam_cpuhp_state;
+static DEFINE_MUTEX(mpam_cpuhp_state_lock);
+
+/*
+ * mpam is enabled once all devices have been probed from CPU online callbacks,
+ * scheduled via this work_struct. If access to an MSC depends on a CPU that
+ * was not brought online at boot, this can happen surprisingly late.
+ */
+static DECLARE_WORK(mpam_enable_work, &mpam_enable);
+
 /*
  * An MSC is a physical container for controls and monitors, each identified by
  * their RIS index. These share a base-address, interrupts and some MMIO
@@ -78,6 +90,22 @@ LIST_HEAD(mpam_classes);
 /* List of all objects that can be free()d after synchronise_srcu() */
 static LLIST_HEAD(mpam_garbage);
 
+static u32 __mpam_read_reg(struct mpam_msc *msc, u16 reg)
+{
+	WARN_ON_ONCE(reg > msc->mapped_hwpage_sz);
+	WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(), &msc->accessibility));
+
+	return readl_relaxed(msc->mapped_hwpage + reg);
+}
+
+static inline u32 _mpam_read_partsel_reg(struct mpam_msc *msc, u16 reg)
+{
+	lockdep_assert_held_once(&msc->part_sel_lock);
+	return __mpam_read_reg(msc, reg);
+}
+
+#define mpam_read_partsel_reg(msc, reg)        _mpam_read_partsel_reg(msc, MPAMF_##reg)
+
 #define init_garbage(x)	init_llist_node(&(x)->garbage.llist)
 
 static struct mpam_vmsc *
@@ -513,9 +541,84 @@ int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
 	return err;
 }
 
-static void mpam_discovery_complete(void)
+static int mpam_msc_hw_probe(struct mpam_msc *msc)
 {
-	pr_err("Discovered all MSC\n");
+	u64 idr;
+	int err;
+
+	lockdep_assert_held(&msc->probe_lock);
+
+	mutex_lock(&msc->part_sel_lock);
+	idr = mpam_read_partsel_reg(msc, AIDR);
+	if ((idr & MPAMF_AIDR_ARCH_MAJOR_REV) != MPAM_ARCHITECTURE_V1) {
+		pr_err_once("%s does not match MPAM architecture v1.0\n",
+			    dev_name(&msc->pdev->dev));
+		err = -EIO;
+	} else {
+		msc->probed = true;
+		err = 0;
+	}
+	mutex_unlock(&msc->part_sel_lock);
+
+	return err;
+}
+
+static int mpam_cpu_online(unsigned int cpu)
+{
+	return 0;
+}
+
+/* Before mpam is enabled, try to probe new MSC */
+static int mpam_discovery_cpu_online(unsigned int cpu)
+{
+	int err = 0;
+	struct mpam_msc *msc;
+	bool new_device_probed = false;
+
+	mutex_lock(&mpam_list_lock);
+	list_for_each_entry(msc, &mpam_all_msc, glbl_list) {
+		if (!cpumask_test_cpu(cpu, &msc->accessibility))
+			continue;
+
+		mutex_lock(&msc->probe_lock);
+		if (!msc->probed)
+			err = mpam_msc_hw_probe(msc);
+		mutex_unlock(&msc->probe_lock);
+
+		if (!err)
+			new_device_probed = true;
+		else
+			break; // mpam_broken
+	}
+	mutex_unlock(&mpam_list_lock);
+
+	if (new_device_probed && !err)
+		schedule_work(&mpam_enable_work);
+
+	return err;
+}
+
+static int mpam_cpu_offline(unsigned int cpu)
+{
+	return 0;
+}
+
+static void mpam_register_cpuhp_callbacks(int (*online)(unsigned int online),
+					  int (*offline)(unsigned int offline))
+{
+	mutex_lock(&mpam_cpuhp_state_lock);
+	if (mpam_cpuhp_state) {
+		cpuhp_remove_state(mpam_cpuhp_state);
+		mpam_cpuhp_state = 0;
+	}
+
+	mpam_cpuhp_state = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "mpam:online",
+					     online, offline);
+	if (mpam_cpuhp_state <= 0) {
+		pr_err("Failed to register cpuhp callbacks");
+		mpam_cpuhp_state = 0;
+	}
+	mutex_unlock(&mpam_cpuhp_state_lock);
 }
 
 static int mpam_dt_count_msc(void)
@@ -774,7 +877,7 @@ static int mpam_msc_drv_probe(struct platform_device *pdev)
 	}
 
 	if (!err && fw_num_msc == mpam_num_msc)
-		mpam_discovery_complete();
+		mpam_register_cpuhp_callbacks(&mpam_discovery_cpu_online, NULL);
 
 	if (err && msc)
 		mpam_msc_drv_remove(pdev);

drivers/platform/arm64/mpam/mpam_devices.c static struct platform_driver mpam_msc_driver = { · line 908

Quoted 9 lines
drivers/platform/arm64/mpam/mpam_devices.c +6 -0
@@ -797,6 +900,46 @@ static struct platform_driver mpam_msc_driver = {
 	.remove = mpam_msc_drv_remove,
 };
 
+static void mpam_enable_once(void)
+{
+	mutex_lock(&mpam_cpuhp_state_lock);
+	cpuhp_remove_state(mpam_cpuhp_state);
+	mpam_cpuhp_state = 0;
+	mutex_unlock(&mpam_cpuhp_state_lock);

Deleting the above 4 lines?
The mpam_cpuhp_state will be removed firstly in mpam_register_cpuhp_callbacks if the mpam_cpuhp_state isn't 0.

drivers/platform/arm64/mpam/mpam_devices.c static struct platform_driver mpam_msc_driver = { · line 914

Quoted 6 lines
drivers/platform/arm64/mpam/mpam_devices.c +6 -0
+
+	mpam_register_cpuhp_callbacks(mpam_cpu_online, mpam_cpu_offline);
+
+	pr_info("MPAM enabled\n");
+}
+

[snip]

Regards,
Baisheng

author maintainer
#63 lore raw

Re: [RFC PATCH 17/36] arm_mpam: Add cpuhp callbacks to probe MSC hardware

Hi Baisheng,

Quoted 41 lines · Baisheng Gao
Quoted 40 lines

Because an MSC can only by accessed from the CPUs in its cpu-affinity set we need to be running on one of those CPUs to probe the MSC hardware.

Do this work in the cpuhp callback. Probing the hardware will only happen before MPAM is enabled, walk all the MSCs and probe those we can reach that haven't already been probed.

Later once MPAM is enabled, this cpuhp callback will be replaced by one that avoids the global list.

Enabling a static key will also take the cpuhp lock, so can't be done from the cpuhp callback. Whenever a new MSC has been probed schedule work to test if all the MSCs have now been probed.

drivers/platform/arm64/mpam/mpam_devices.c +16 -0
index 0d6d5180903b..89434ae3efa6 100644
@@ -513,9 +541,84 @@ int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
+static void mpam_register_cpuhp_callbacks(int (*online)(unsigned int online),
+					  int (*offline)(unsigned int offline))
+{
+	mutex_lock(&mpam_cpuhp_state_lock);
+	if (mpam_cpuhp_state) {
+		cpuhp_remove_state(mpam_cpuhp_state);
+		mpam_cpuhp_state = 0;
+	}
+
+	mpam_cpuhp_state = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "mpam:online",
+					     online, offline);
+	if (mpam_cpuhp_state <= 0) {
+		pr_err("Failed to register cpuhp callbacks");
+		mpam_cpuhp_state = 0;
+	}
+	mutex_unlock(&mpam_cpuhp_state_lock);
 }
 
 static int mpam_dt_count_msc(void)
Quoted 11 lines

drivers/platform/arm64/mpam/mpam_devices.c static struct platform_driver mpam_msc_driver = { · line 905

Quoted 7 lines
drivers/platform/arm64/mpam/mpam_devices.c +6 -0
@@ -797,6 +900,46 @@ static struct platform_driver mpam_msc_driver = {
+static void mpam_enable_once(void)
+{
+	mutex_lock(&mpam_cpuhp_state_lock);
+	cpuhp_remove_state(mpam_cpuhp_state);
+	mpam_cpuhp_state = 0;
+	mutex_unlock(&mpam_cpuhp_state_lock);

Deleting the above 4 lines?
The mpam_cpuhp_state will be removed firstly in mpam_register_cpuhp_callbacks if the mpam_cpuhp_state isn't 0.

Yup - this is a pointless appendage because of the way the code evolved! Thanks for catching it

James

Quoted 5 lines
Quoted 5 lines
drivers/platform/arm64/mpam/mpam_devices.c +5 -0
+
+	mpam_register_cpuhp_callbacks(mpam_cpu_online, mpam_cpu_offline);
+
+	pr_info("MPAM enabled\n");
+}
#64 lore raw

Re: [RFC PATCH 17/36] arm_mpam: Add cpuhp callbacks to probe MSC hardware

In reply to James Morse
Quoted 11 lines · On Fri, 11 Jul 2025 18:36:29 +0000 James Morse <[email protected]> wrote:

Because an MSC can only by accessed from the CPUs in its cpu-affinity set we need to be running on one of those CPUs to probe the MSC hardware.

Do this work in the cpuhp callback. Probing the hardware will only happen before MPAM is enabled, walk all the MSCs and probe those we can reach that haven't already been probed.

Later once MPAM is enabled, this cpuhp callback will be replaced by one that avoids the global list.

Quoted 6 lines

Enabling a static key will also take the cpuhp lock, so can't be done from the cpuhp callback. Whenever a new MSC has been probed schedule work to test if all the MSCs have now been probed.

Hi James,

One trivial thing noticed whilst testing..

Quoted 7 lines
drivers/platform/arm64/mpam/mpam_devices.c +1 -1
@@ -513,9 +541,84 @@ int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
 	return err;
 }
 
-static void mpam_discovery_complete(void)
+static int mpam_msc_hw_probe(struct mpam_msc *msc)
 {

drivers/platform/arm64/mpam/mpam_devices.c int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx, · line 554

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +9 -1
-	pr_err("Discovered all MSC\n");
+	u64 idr;
+	int err;
+
+	lockdep_assert_held(&msc->probe_lock);
+
+	mutex_lock(&msc->part_sel_lock);
+	idr = mpam_read_partsel_reg(msc, AIDR);
+	if ((idr & MPAMF_AIDR_ARCH_MAJOR_REV) != MPAM_ARCHITECTURE_V1) {
+		pr_err_once("%s does not match MPAM architecture v1.0\n",

You are only checking MAJOR REV which is probably the right thing to do but in that case maybe change the message to be v1.x ?

Quoted 11 lines
drivers/platform/arm64/mpam/mpam_devices.c +11 -0
+			    dev_name(&msc->pdev->dev));
+		err = -EIO;
+	} else {
+		msc->probed = true;
+		err = 0;
+	}
+	mutex_unlock(&msc->part_sel_lock);
+
+	return err;
+}
+
author maintainer
#65 lore raw

RFC PATCH 18/36 arm_mpam: Probe MSCs to find the supported partid/pmg values

In reply to James Morse

CPUs can generate traffic with a range of PARTID and PMG values, but each MSC may have its own maximum size for these fields. Before MPAM can be used, the driver needs to probe each RIS on each MSC, to find the system-wide smallest value that can be used.

While doing this, RIS entries that firmware didn't describe are create under MPAM_CLASS_UNKNOWN.

While we're here, implement the mpam_register_requestor() call for the arch code. Future callers of this will tell us about the SMMU and ITS.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 172 +++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |   6 +
 include/linux/arm_mpam.h                    |  14 ++
 3 files changed, 185 insertions(+), 7 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +165 -7
index 89434ae3efa6..8646fb85ad09 100644
@@ -6,6 +6,7 @@
 #include <linux/acpi.h>
 #include <linux/atomic.h>
 #include <linux/arm_mpam.h>
+#include <linux/bitfield.h>
 #include <linux/cacheinfo.h>
 #include <linux/cpu.h>
 #include <linux/cpumask.h>
@@ -44,6 +45,15 @@ static u32 mpam_num_msc;
 static int mpam_cpuhp_state;
 static DEFINE_MUTEX(mpam_cpuhp_state_lock);
 
+/*
+ * The smallest common values for any CPU or MSC in the system.
+ * Generating traffic outside this range will result in screaming interrupts.
+ */
+u16 mpam_partid_max;
+u8 mpam_pmg_max;
+static bool partid_max_init, partid_max_published;
+static DEFINE_SPINLOCK(partid_max_lock);
+
 /*
  * mpam is enabled once all devices have been probed from CPU online callbacks,
  * scheduled via this work_struct. If access to an MSC depends on a CPU that
@@ -106,6 +116,74 @@ static inline u32 _mpam_read_partsel_reg(struct mpam_msc *msc, u16 reg)
 
 #define mpam_read_partsel_reg(msc, reg)        _mpam_read_partsel_reg(msc, MPAMF_##reg)
 
+static void __mpam_write_reg(struct mpam_msc *msc, u16 reg, u32 val)
+{
+	WARN_ON_ONCE(reg + sizeof(u32) > msc->mapped_hwpage_sz);
+	WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(), &msc->accessibility));
+
+	writel_relaxed(val, msc->mapped_hwpage + reg);
+}
+
+static inline void _mpam_write_partsel_reg(struct mpam_msc *msc, u16 reg, u32 val)
+{
+	lockdep_assert_held_once(&msc->part_sel_lock);
+	__mpam_write_reg(msc, reg, val);
+}
+#define mpam_write_partsel_reg(msc, reg, val)  _mpam_write_partsel_reg(msc, MPAMCFG_##reg, val)
+
+static u64 mpam_msc_read_idr(struct mpam_msc *msc)
+{
+	u64 idr_high = 0, idr_low;
+
+	lockdep_assert_held(&msc->part_sel_lock);
+
+	idr_low = mpam_read_partsel_reg(msc, IDR);
+	if (FIELD_GET(MPAMF_IDR_HAS_EXT, idr_low))
+		idr_high = mpam_read_partsel_reg(msc, IDR + 4);
+
+	return (idr_high << 32) | idr_low;
+}
+
+static void __mpam_part_sel_raw(u32 partsel, struct mpam_msc *msc)
+{
+	lockdep_assert_held(&msc->part_sel_lock);
+
+	mpam_write_partsel_reg(msc, PART_SEL, partsel);
+}
+
+static void __mpam_part_sel(u8 ris_idx, u16 partid, struct mpam_msc *msc)
+{
+	u32 partsel = FIELD_PREP(MPAMCFG_PART_SEL_RIS, ris_idx) |
+		      FIELD_PREP(MPAMCFG_PART_SEL_PARTID_SEL, partid);
+
+	__mpam_part_sel_raw(partsel, msc);
+}
+
+int mpam_register_requestor(u16 partid_max, u8 pmg_max)
+{
+	int err = 0;
+
+	lockdep_assert_irqs_enabled();
+
+	spin_lock(&partid_max_lock);
+	if (!partid_max_init) {
+		mpam_partid_max = partid_max;
+		mpam_pmg_max = pmg_max;
+		partid_max_init = true;
+	} else if (!partid_max_published) {
+		mpam_partid_max = min(mpam_partid_max, partid_max);
+		mpam_pmg_max = min(mpam_pmg_max, pmg_max);
+	} else {
+		/* New requestors can't lower the values */
+		if (partid_max < mpam_partid_max || pmg_max < mpam_pmg_max)
+			err = -EBUSY;
+	}
+	spin_unlock(&partid_max_lock);
+
+	return err;
+}
+EXPORT_SYMBOL(mpam_register_requestor);
+
 #define init_garbage(x)	init_llist_node(&(x)->garbage.llist)
 
 static struct mpam_vmsc *
@@ -522,6 +600,7 @@ static int mpam_ris_create_locked(struct mpam_msc *msc, u8 ris_idx,
 	cpumask_or(&comp->affinity, &comp->affinity, &ris->affinity);
 	cpumask_or(&class->affinity, &class->affinity, &ris->affinity);
 	list_add_rcu(&ris->vmsc_list, &vmsc->ris);
+	list_add_rcu(&ris->msc_list, &msc->ris);
 
 	return 0;
 }
@@ -541,10 +620,37 @@ int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
 	return err;
 }
 
+static struct mpam_msc_ris *mpam_get_or_create_ris(struct mpam_msc *msc,
+						   u8 ris_idx)
+{
+	int err;
+	struct mpam_msc_ris *ris, *found = ERR_PTR(-ENOENT);
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	if (!test_bit(ris_idx, msc->ris_idxs)) {
+		err = mpam_ris_create_locked(msc, ris_idx, MPAM_CLASS_UNKNOWN,
+					     0, 0, GFP_ATOMIC);
+		if (err)
+			return ERR_PTR(err);
+	}
+
+	list_for_each_entry(ris, &msc->ris, msc_list) {
+		if (ris->ris_idx == ris_idx) {
+			found = ris;
+			break;
+		}
+	}
+
+	return found;
+}
+
 static int mpam_msc_hw_probe(struct mpam_msc *msc)
 {
 	u64 idr;
-	int err;
+	u16 partid_max;
+	u8 ris_idx, pmg_max;
+	struct mpam_msc_ris *ris;
 
 	lockdep_assert_held(&msc->probe_lock);
 
@@ -553,14 +659,42 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
 	if ((idr & MPAMF_AIDR_ARCH_MAJOR_REV) != MPAM_ARCHITECTURE_V1) {
 		pr_err_once("%s does not match MPAM architecture v1.0\n",
 			    dev_name(&msc->pdev->dev));
-		err = -EIO;
-	} else {
-		msc->probed = true;
-		err = 0;
+		mutex_unlock(&msc->part_sel_lock);
+		return -EIO;
 	}
+
+	idr = mpam_msc_read_idr(msc);
 	mutex_unlock(&msc->part_sel_lock);
+	msc->ris_max = FIELD_GET(MPAMF_IDR_RIS_MAX, idr);
 
-	return err;
+	/* Use these values so partid/pmg always starts with a valid value */
+	msc->partid_max = FIELD_GET(MPAMF_IDR_PARTID_MAX, idr);
+	msc->pmg_max = FIELD_GET(MPAMF_IDR_PMG_MAX, idr);
+
+	for (ris_idx = 0; ris_idx <= msc->ris_max; ris_idx++) {
+		mutex_lock(&msc->part_sel_lock);
+		__mpam_part_sel(ris_idx, 0, msc);
+		idr = mpam_msc_read_idr(msc);
+		mutex_unlock(&msc->part_sel_lock);
+
+		partid_max = FIELD_GET(MPAMF_IDR_PARTID_MAX, idr);
+		pmg_max = FIELD_GET(MPAMF_IDR_PMG_MAX, idr);
+		msc->partid_max = min(msc->partid_max, partid_max);
+		msc->pmg_max = min(msc->pmg_max, pmg_max);
+
+		ris = mpam_get_or_create_ris(msc, ris_idx);
+		if (IS_ERR(ris))
+			return PTR_ERR(ris);
+	}
+
+	spin_lock(&partid_max_lock);
+	mpam_partid_max = min(mpam_partid_max, msc->partid_max);
+	mpam_pmg_max = min(mpam_pmg_max, msc->pmg_max);
+	spin_unlock(&partid_max_lock);
+
+	msc->probed = true;
+
+	return 0;
 }
 
 static int mpam_cpu_online(unsigned int cpu)
@@ -907,9 +1041,18 @@ static void mpam_enable_once(void)
 	mpam_cpuhp_state = 0;
 	mutex_unlock(&mpam_cpuhp_state_lock);
 
+	/*
+	 * Once the cpuhp callbacks have been changed, mpam_partid_max can no
+	 * longer change.
+	 */
+	spin_lock(&partid_max_lock);
+	partid_max_published = true;
+	spin_unlock(&partid_max_lock);
+
 	mpam_register_cpuhp_callbacks(mpam_cpu_online, mpam_cpu_offline);
 
-	pr_info("MPAM enabled\n");
+	printk(KERN_INFO "MPAM enabled with %u partid and %u pmg\n",
+	       mpam_partid_max + 1, mpam_pmg_max + 1);
 }
 
 /*
@@ -959,11 +1102,25 @@ static void mpam_dt_create_foundling_msc(void)
 
 static int __init mpam_msc_driver_init(void)
 {
+	bool mpam_not_available = false;
+
 	if (!system_supports_mpam())
 		return -EOPNOTSUPP;
 
 	init_srcu_struct(&mpam_srcu);
 
+	/*
+	 * If the MPAM CPU interface is not implemented, or reserved by
+	 * firmware, there is no point touching the rest of the hardware.
+	 */
+	spin_lock(&partid_max_lock);
+	if (!partid_max_init || (!mpam_partid_max && !mpam_pmg_max))
+		mpam_not_available = true;
+	spin_unlock(&partid_max_lock);
+
+	if (mpam_not_available)
+		return 0;
+
 	if (!acpi_disabled)
 		fw_num_msc = acpi_mpam_count_msc();
 	else
@@ -979,4 +1136,5 @@ static int __init mpam_msc_driver_init(void)
 
 	return platform_driver_register(&mpam_msc_driver);
 }
+/* Must occur after arm64_mpam_register_cpus() from arch_initcall() */
 subsys_initcall(mpam_msc_driver_init);
drivers/platform/arm64/mpam/mpam_internal.h +6 -0
index f56e69ff8397..eb5cc6775d54 100644
@@ -50,6 +50,8 @@ struct mpam_msc {
 	 */
 	struct mutex		probe_lock;
 	bool			probed;
+	u16			partid_max;
+	u8			pmg_max;
 	unsigned long		ris_idxs[128 / BITS_PER_LONG];
 	u32			ris_max;
 
@@ -148,6 +150,10 @@ struct mpam_msc_ris {
 extern struct srcu_struct mpam_srcu;
 extern struct list_head mpam_classes;
 
+/* System wide partid/pmg values */
+extern u16 mpam_partid_max;
+extern u8 mpam_pmg_max;
+
 /* Scheduled work callback to enable mpam once all MSC have been probed */
 void mpam_enable(struct work_struct *work);
 
include/linux/arm_mpam.h +14 -0
index 406a77be68cb..8af93794c7a2 100644
@@ -39,4 +39,18 @@ static inline int acpi_mpam_count_msc(void) { return -EINVAL; }
 int mpam_ris_create(struct mpam_msc *msc, u8 ris_idx,
 		    enum mpam_class_types type, u8 class_id, int component_id);
 
+/**
+ * mpam_register_requestor() - Register a requestor with the MPAM driver
+ * @partid_max:		The maximum PARTID value the requestor can generate.
+ * @pmg_max:		The maximum PMG value the requestor can generate.
+ *
+ * Registers a requestor with the MPAM driver to ensure the chosen system-wide
+ * minimum PARTID and PMG values will allow the requestors features to be used.
+ *
+ * Returns an error if the registration is too late, and a larger PARTID/PMG
+ * value has been advertised to user-space. In this case the requestor should
+ * not use its MPAM features. Returns 0 on success.
+ */
+int mpam_register_requestor(u16 partid_max, u8 pmg_max);
+
 #endif /* __LINUX_ARM_MPAM_H */
Signature · 1 line
2.39.5
author maintainer
#66 lore raw

RFC PATCH 19/36 arm_mpam: Add helpers for managing the locking around the mon_sel registers

In reply to James Morse

The MSC MON_SEL register needs to be accessed from hardirq context by the PMU drivers, making an irqsave spinlock the obvious lock to protect these registers. On systems with SCMI mailboxes it must be able to sleep, meaning a mutex must be used.

Clearly these two can't exist at the same time.

Add helpers for the MON_SEL locking. The outer lock must be taken in a pre-emptible context before the inner lock can be taken. On systems with SCMI mailboxes where the MON_SEL accesses must sleep - the inner lock will fail to be 'taken' if the caller is unable to sleep. This will allow the PMU driver to fail without having to check the interface type of each MSC.

---
 drivers/platform/arm64/mpam/mpam_internal.h | 57 ++++++++++++++++++++-
 1 file changed, 56 insertions(+), 1 deletion(-)
drivers/platform/arm64/mpam/mpam_internal.h +56 -1
index eb5cc6775d54..42a454d5f914 100644
@@ -68,10 +68,19 @@ struct mpam_msc {
 
 	/*
 	 * mon_sel_lock protects access to the MSC hardware registers that are
-	 * affeted by MPAMCFG_MON_SEL.
+	 * affected by MPAMCFG_MON_SEL, and the mbwu_state.
+	 * Both the 'inner' and 'outer' must be taken.
+	 * For real MMIO MSC, the outer lock is unnecessary - but keeps the
+	 * code common with:
+	 * Firmware backed MSC need to sleep when accessing the MSC, which
+	 * means some code-paths will always fail. For these MSC the outer
+	 * lock is providing the protection, and the inner lock fails to
+	 * be taken if the task is unable to sleep.
+	 *
 	 * If needed, take msc->probe_lock first.
 	 */
 	struct mutex		outer_mon_sel_lock;
+	bool			outer_lock_held;
 	raw_spinlock_t		inner_mon_sel_lock;
 	unsigned long		inner_mon_sel_flags;
 
@@ -81,6 +90,52 @@ struct mpam_msc {
 	struct mpam_garbage	garbage;
 };
 
+static inline bool __must_check mpam_mon_sel_inner_lock(struct mpam_msc *msc)
+{
+	/*
+	 * The outer lock may be taken by a CPU that then issues an IPI to run
+	 * a helper that takes the inner lock. lockdep can't help us here.
+	 */
+	WARN_ON_ONCE(!msc->outer_lock_held);
+
+	if (msc->iface == MPAM_IFACE_MMIO) {
+		raw_spin_lock_irqsave(&msc->inner_mon_sel_lock, msc->inner_mon_sel_flags);
+		return true;
+	}
+
+	/* Accesses must fail if we are not pre-emptible */
+	return !!preemptible();
+}
+
+static inline void mpam_mon_sel_inner_unlock(struct mpam_msc *msc)
+{
+	WARN_ON_ONCE(!msc->outer_lock_held);
+
+	if (msc->iface == MPAM_IFACE_MMIO)
+		raw_spin_unlock_irqrestore(&msc->inner_mon_sel_lock, msc->inner_mon_sel_flags);
+}
+
+static inline void mpam_mon_sel_outer_lock(struct mpam_msc *msc)
+{
+	mutex_lock(&msc->outer_mon_sel_lock);
+	msc->outer_lock_held = true;
+}
+
+static inline void mpam_mon_sel_outer_unlock(struct mpam_msc *msc)
+{
+	msc->outer_lock_held = false;
+	mutex_unlock(&msc->outer_mon_sel_lock);
+}
+
+static inline void mpam_mon_sel_lock_held(struct mpam_msc *msc)
+{
+	WARN_ON_ONCE(!msc->outer_lock_held);
+	if (msc->iface == MPAM_IFACE_MMIO)
+		lockdep_assert_held_once(&msc->inner_mon_sel_lock);
+	else
+		lockdep_assert_preemption_enabled();
+}
+
 struct mpam_class {
 	/* mpam_components in this class */
 	struct list_head	components;
Signature · 1 line
2.39.5
author maintainer
#67 lore raw

RFC PATCH 20/36 arm_mpam: Probe the hardware features resctrl supports

In reply to James Morse

Expand the probing support with the control and monitor types we can use with resctrl.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 154 +++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |  53 +++++++
 2 files changed, 206 insertions(+), 1 deletion(-)
drivers/platform/arm64/mpam/mpam_devices.c +153 -1
index 8646fb85ad09..61911831ab39 100644
@@ -102,7 +102,7 @@ static LLIST_HEAD(mpam_garbage);
 
 static u32 __mpam_read_reg(struct mpam_msc *msc, u16 reg)
 {
-	WARN_ON_ONCE(reg > msc->mapped_hwpage_sz);
+	WARN_ON_ONCE(reg + sizeof(u32) > msc->mapped_hwpage_sz);
 	WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(), &msc->accessibility));
 
 	return readl_relaxed(msc->mapped_hwpage + reg);
@@ -131,6 +131,20 @@ static inline void _mpam_write_partsel_reg(struct mpam_msc *msc, u16 reg, u32 va
 }
 #define mpam_write_partsel_reg(msc, reg, val)  _mpam_write_partsel_reg(msc, MPAMCFG_##reg, val)
 
+static inline u32 _mpam_read_monsel_reg(struct mpam_msc *msc, u16 reg)
+{
+	mpam_mon_sel_lock_held(msc);
+	return __mpam_read_reg(msc, reg);
+}
+#define mpam_read_monsel_reg(msc, reg) _mpam_read_monsel_reg(msc, MSMON_##reg)
+
+static inline void _mpam_write_monsel_reg(struct mpam_msc *msc, u16 reg, u32 val)
+{
+	mpam_mon_sel_lock_held(msc);
+	__mpam_write_reg(msc, reg, val);
+}
+#define mpam_write_monsel_reg(msc, reg, val)   _mpam_write_monsel_reg(msc, MSMON_##reg, val)
+
 static u64 mpam_msc_read_idr(struct mpam_msc *msc)
 {
 	u64 idr_high = 0, idr_low;
@@ -645,6 +659,137 @@ static struct mpam_msc_ris *mpam_get_or_create_ris(struct mpam_msc *msc,
 	return found;
 }
 
+/*
+ * IHI009A.a has this nugget: "If a monitor does not support automatic behaviour
+ * of NRDY, software can use this bit for any purpose" - so hardware might not
+ * implement this - but it isn't RES0.
+ *
+ * Try and see what values stick in this bit. If we can write either value,
+ * its probably not implemented by hardware.
+ */
+#define mpam_ris_hw_probe_hw_nrdy(_ris, _mon_reg, _result)			\
+do {										\
+	u32 now;								\
+	u64 mon_sel;								\
+	bool can_set, can_clear;						\
+	struct mpam_msc *_msc = _ris->vmsc->msc;				\
+										\
+	if (WARN_ON_ONCE(!mpam_mon_sel_inner_lock(_msc))) {			\
+		_result = false;						\
+		break;								\
+	}									\
+	mon_sel = FIELD_PREP(MSMON_CFG_MON_SEL_MON_SEL, 0) |			\
+		  FIELD_PREP(MSMON_CFG_MON_SEL_RIS, _ris->ris_idx);		\
+	mpam_write_monsel_reg(_msc, CFG_MON_SEL, mon_sel);			\
+										\
+	mpam_write_monsel_reg(_msc, _mon_reg, MSMON___NRDY);			\
+	now = mpam_read_monsel_reg(_msc, _mon_reg);				\
+	can_set = now & MSMON___NRDY;						\
+										\
+	mpam_write_monsel_reg(_msc, _mon_reg, 0);				\
+	now = mpam_read_monsel_reg(_msc, _mon_reg);				\
+	can_clear = !(now & MSMON___NRDY);					\
+	mpam_mon_sel_inner_unlock(_msc);					\
+										\
+	_result = (!can_set || !can_clear);					\
+} while (0)
+
+static void mpam_ris_hw_probe(struct mpam_msc_ris *ris)
+{
+	int err;
+	struct mpam_msc *msc = ris->vmsc->msc;
+	struct mpam_props *props = &ris->props;
+
+	lockdep_assert_held(&msc->probe_lock);
+	lockdep_assert_held(&msc->part_sel_lock);
+
+	/* Cache Portion partitioning */
+	if (FIELD_GET(MPAMF_IDR_HAS_CPOR_PART, ris->idr)) {
+		u32 cpor_features = mpam_read_partsel_reg(msc, CPOR_IDR);
+
+		props->cpbm_wd = FIELD_GET(MPAMF_CPOR_IDR_CPBM_WD, cpor_features);
+		if (props->cpbm_wd)
+			mpam_set_feature(mpam_feat_cpor_part, props);
+	}
+
+	/* Memory bandwidth partitioning */
+	if (FIELD_GET(MPAMF_IDR_HAS_MBW_PART, ris->idr)) {
+		u32 mbw_features = mpam_read_partsel_reg(msc, MBW_IDR);
+
+		/* portion bitmap resolution */
+		props->mbw_pbm_bits = FIELD_GET(MPAMF_MBW_IDR_BWPBM_WD, mbw_features);
+		if (props->mbw_pbm_bits &&
+		    FIELD_GET(MPAMF_MBW_IDR_HAS_PBM, mbw_features))
+			mpam_set_feature(mpam_feat_mbw_part, props);
+
+		props->bwa_wd = FIELD_GET(MPAMF_MBW_IDR_BWA_WD, mbw_features);
+		if (props->bwa_wd && FIELD_GET(MPAMF_MBW_IDR_HAS_MAX, mbw_features))
+			mpam_set_feature(mpam_feat_mbw_max, props);
+	}
+
+	/* Performance Monitoring */
+	if (FIELD_GET(MPAMF_IDR_HAS_MSMON, ris->idr)) {
+		u32 msmon_features = mpam_read_partsel_reg(msc, MSMON_IDR);
+
+		/*
+		 * If the firmware max-nrdy-us property is missing, the
+		 * CSU counters can't be used. Should we wait forever?
+		 */
+		err = device_property_read_u32(&msc->pdev->dev,
+					       "arm,not-ready-us",
+					       &msc->nrdy_usec);
+
+		if (FIELD_GET(MPAMF_MSMON_IDR_MSMON_CSU, msmon_features)) {
+			u32 csumonidr;
+
+			csumonidr = mpam_read_partsel_reg(msc, CSUMON_IDR);
+			props->num_csu_mon = FIELD_GET(MPAMF_CSUMON_IDR_NUM_MON, csumonidr);
+			if (props->num_csu_mon) {
+				bool hw_managed;
+
+				mpam_set_feature(mpam_feat_msmon_csu, props);
+
+				/* Is NRDY hardware managed? */
+				mpam_mon_sel_outer_lock(msc);
+				mpam_ris_hw_probe_hw_nrdy(ris, CSU, hw_managed);
+				mpam_mon_sel_outer_unlock(msc);
+				if (hw_managed)
+					mpam_set_feature(mpam_feat_msmon_csu_hw_nrdy, props);
+			}
+
+			/*
+			 * Accept the missing firmware property if NRDY appears
+			 * un-implemented.
+			 */
+			if (err && mpam_has_feature(mpam_feat_msmon_csu_hw_nrdy, props))
+				pr_err_once("Counters are not usable because not-ready timeout was not provided by firmware.");
+		}
+		if (FIELD_GET(MPAMF_MSMON_IDR_MSMON_MBWU, msmon_features)) {
+			bool hw_managed;
+			u32 mbwumonidr = mpam_read_partsel_reg(msc, MBWUMON_IDR);
+
+			props->num_mbwu_mon = FIELD_GET(MPAMF_MBWUMON_IDR_NUM_MON, mbwumonidr);
+			if (props->num_mbwu_mon)
+				mpam_set_feature(mpam_feat_msmon_mbwu, props);
+
+			if (FIELD_GET(MPAMF_MBWUMON_IDR_HAS_RWBW, mbwumonidr))
+				mpam_set_feature(mpam_feat_msmon_mbwu_rwbw, props);
+
+			/* Is NRDY hardware managed? */
+			mpam_mon_sel_outer_lock(msc);
+			mpam_ris_hw_probe_hw_nrdy(ris, MBWU, hw_managed);
+			mpam_mon_sel_outer_unlock(msc);
+			if (hw_managed)
+				mpam_set_feature(mpam_feat_msmon_mbwu_hw_nrdy, props);
+
+			/*
+			 * Don't warn about any missing firmware property for
+			 * MBWU NRDY - it doesn't make any sense!
+			 */
+		}
+	}
+}
+
 static int mpam_msc_hw_probe(struct mpam_msc *msc)
 {
 	u64 idr;
@@ -665,6 +810,7 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
 
 	idr = mpam_msc_read_idr(msc);
 	mutex_unlock(&msc->part_sel_lock);
+
 	msc->ris_max = FIELD_GET(MPAMF_IDR_RIS_MAX, idr);
 
 	/* Use these values so partid/pmg always starts with a valid value */
@@ -685,6 +831,12 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
 		ris = mpam_get_or_create_ris(msc, ris_idx);
 		if (IS_ERR(ris))
 			return PTR_ERR(ris);
+		ris->idr = idr;
+
+		mutex_lock(&msc->part_sel_lock);
+		__mpam_part_sel(ris_idx, 0, msc);
+		mpam_ris_hw_probe(ris);
+		mutex_unlock(&msc->part_sel_lock);
 	}
 
 	spin_lock(&partid_max_lock);
drivers/platform/arm64/mpam/mpam_internal.h +53 -0

71 lines not on this page · raw message

Signature · 1 line
2.39.5
maintainer
#68 lore raw

Re: [RFC PATCH 20/36] arm_mpam: Probe the hardware features resctrl supports

Hi James,

Quoted 72 lines · James Morse

Expand the probing support with the control and monitor types we can use with resctrl.

---
  drivers/platform/arm64/mpam/mpam_devices.c  | 154 +++++++++++++++++++-
  drivers/platform/arm64/mpam/mpam_internal.h |  53 +++++++
  2 files changed, 206 insertions(+), 1 deletion(-)
drivers/platform/arm64/mpam/mpam_devices.c +39 -1
index 8646fb85ad09..61911831ab39 100644
@@ -102,7 +102,7 @@ static LLIST_HEAD(mpam_garbage);
 
  static u32 __mpam_read_reg(struct mpam_msc *msc, u16 reg)
  {
-	WARN_ON_ONCE(reg > msc->mapped_hwpage_sz);
+	WARN_ON_ONCE(reg + sizeof(u32) > msc->mapped_hwpage_sz);
  	WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(), &msc->accessibility));
 
  	return readl_relaxed(msc->mapped_hwpage + reg);
@@ -131,6 +131,20 @@ static inline void _mpam_write_partsel_reg(struct mpam_msc *msc, u16 reg, u32 va
  }
  #define mpam_write_partsel_reg(msc, reg, val)  _mpam_write_partsel_reg(msc, MPAMCFG_##reg, val)
 
+static inline u32 _mpam_read_monsel_reg(struct mpam_msc *msc, u16 reg)
+{
+	mpam_mon_sel_lock_held(msc);
+	return __mpam_read_reg(msc, reg);
+}
+#define mpam_read_monsel_reg(msc, reg) _mpam_read_monsel_reg(msc, MSMON_##reg)
+
+static inline void _mpam_write_monsel_reg(struct mpam_msc *msc, u16 reg, u32 val)
+{
+	mpam_mon_sel_lock_held(msc);
+	__mpam_write_reg(msc, reg, val);
+}
+#define mpam_write_monsel_reg(msc, reg, val)   _mpam_write_monsel_reg(msc, MSMON_##reg, val)
+
  static u64 mpam_msc_read_idr(struct mpam_msc *msc)
  {
  	u64 idr_high = 0, idr_low;
@@ -645,6 +659,137 @@ static struct mpam_msc_ris *mpam_get_or_create_ris(struct mpam_msc *msc,
  	return found;
  }
 
+/*
+ * IHI009A.a has this nugget: "If a monitor does not support automatic behaviour
+ * of NRDY, software can use this bit for any purpose" - so hardware might not
+ * implement this - but it isn't RES0.
+ *
+ * Try and see what values stick in this bit. If we can write either value,
+ * its probably not implemented by hardware.
+ */
+#define mpam_ris_hw_probe_hw_nrdy(_ris, _mon_reg, _result)			\
+do {										\
+	u32 now;								\
+	u64 mon_sel;								\
+	bool can_set, can_clear;						\
+	struct mpam_msc *_msc = _ris->vmsc->msc;				\
+										\
+	if (WARN_ON_ONCE(!mpam_mon_sel_inner_lock(_msc))) {			\
+		_result = false;						\
+		break;								\
+	}									\
+	mon_sel = FIELD_PREP(MSMON_CFG_MON_SEL_MON_SEL, 0) |			\
+		  FIELD_PREP(MSMON_CFG_MON_SEL_RIS, _ris->ris_idx);		\
+	mpam_write_monsel_reg(_msc, CFG_MON_SEL, mon_sel);			\
+										\
+	mpam_write_monsel_reg(_msc, _mon_reg, MSMON___NRDY);			\

drivers/platform/arm64/mpam/mpam_devices.c static struct mpam_msc_ris *mpam_get_or_create_ris(struct mpam_msc *msc, · line 695

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+	now = mpam_read_monsel_reg(_msc, _mon_reg);				\
+	can_set = now & MSMON___NRDY;						\
+										\
+	mpam_write_monsel_reg(_msc, _mon_reg, 0);				\
+	now = mpam_read_monsel_reg(_msc, _mon_reg);				\
+	can_clear = !(now & MSMON___NRDY);					\
+	mpam_mon_sel_inner_unlock(_msc);					\
+										\
+	_result = (!can_set || !can_clear);					\
+} while (0)

It is a bit surprising that something that looks like a function modifies a boolean passed by value. Consider continuing the pattern you have above:
#define mpam_ris_hw_probe_hw_nrdy(_ris, _mon_reg, _result) _mpam_ris_hw_probe_hw_nrdy(_ris, MSMON##_mon_reg, _result)

with signature:
void _mpam_ris_hw_probe_hw_nrdy(struct mpam_msc *msc, u16 reg, bool *hw_managed);

and using the _mpam functions from the new _mpam_ris_hw_probe_hw_nrdy().

Quoted 199 lines
drivers/platform/arm64/mpam/mpam_devices.c +104 -0
+
+static void mpam_ris_hw_probe(struct mpam_msc_ris *ris)
+{
+	int err;
+	struct mpam_msc *msc = ris->vmsc->msc;
+	struct mpam_props *props = &ris->props;
+
+	lockdep_assert_held(&msc->probe_lock);
+	lockdep_assert_held(&msc->part_sel_lock);
+
+	/* Cache Portion partitioning */
+	if (FIELD_GET(MPAMF_IDR_HAS_CPOR_PART, ris->idr)) {
+		u32 cpor_features = mpam_read_partsel_reg(msc, CPOR_IDR);
+
+		props->cpbm_wd = FIELD_GET(MPAMF_CPOR_IDR_CPBM_WD, cpor_features);
+		if (props->cpbm_wd)
+			mpam_set_feature(mpam_feat_cpor_part, props);
+	}
+
+	/* Memory bandwidth partitioning */
+	if (FIELD_GET(MPAMF_IDR_HAS_MBW_PART, ris->idr)) {
+		u32 mbw_features = mpam_read_partsel_reg(msc, MBW_IDR);
+
+		/* portion bitmap resolution */
+		props->mbw_pbm_bits = FIELD_GET(MPAMF_MBW_IDR_BWPBM_WD, mbw_features);
+		if (props->mbw_pbm_bits &&
+		    FIELD_GET(MPAMF_MBW_IDR_HAS_PBM, mbw_features))
+			mpam_set_feature(mpam_feat_mbw_part, props);
+
+		props->bwa_wd = FIELD_GET(MPAMF_MBW_IDR_BWA_WD, mbw_features);
+		if (props->bwa_wd && FIELD_GET(MPAMF_MBW_IDR_HAS_MAX, mbw_features))
+			mpam_set_feature(mpam_feat_mbw_max, props);
+	}
+
+	/* Performance Monitoring */
+	if (FIELD_GET(MPAMF_IDR_HAS_MSMON, ris->idr)) {
+		u32 msmon_features = mpam_read_partsel_reg(msc, MSMON_IDR);
+
+		/*
+		 * If the firmware max-nrdy-us property is missing, the
+		 * CSU counters can't be used. Should we wait forever?
+		 */
+		err = device_property_read_u32(&msc->pdev->dev,
+					       "arm,not-ready-us",
+					       &msc->nrdy_usec);
+
+		if (FIELD_GET(MPAMF_MSMON_IDR_MSMON_CSU, msmon_features)) {
+			u32 csumonidr;
+
+			csumonidr = mpam_read_partsel_reg(msc, CSUMON_IDR);
+			props->num_csu_mon = FIELD_GET(MPAMF_CSUMON_IDR_NUM_MON, csumonidr);
+			if (props->num_csu_mon) {
+				bool hw_managed;
+
+				mpam_set_feature(mpam_feat_msmon_csu, props);
+
+				/* Is NRDY hardware managed? */
+				mpam_mon_sel_outer_lock(msc);
+				mpam_ris_hw_probe_hw_nrdy(ris, CSU, hw_managed);
+				mpam_mon_sel_outer_unlock(msc);
+				if (hw_managed)
+					mpam_set_feature(mpam_feat_msmon_csu_hw_nrdy, props);
+			}
+
+			/*
+			 * Accept the missing firmware property if NRDY appears
+			 * un-implemented.
+			 */
+			if (err && mpam_has_feature(mpam_feat_msmon_csu_hw_nrdy, props))
+				pr_err_once("Counters are not usable because not-ready timeout was not provided by firmware.");
+		}
+		if (FIELD_GET(MPAMF_MSMON_IDR_MSMON_MBWU, msmon_features)) {
+			bool hw_managed;
+			u32 mbwumonidr = mpam_read_partsel_reg(msc, MBWUMON_IDR);
+
+			props->num_mbwu_mon = FIELD_GET(MPAMF_MBWUMON_IDR_NUM_MON, mbwumonidr);
+			if (props->num_mbwu_mon)
+				mpam_set_feature(mpam_feat_msmon_mbwu, props);
+
+			if (FIELD_GET(MPAMF_MBWUMON_IDR_HAS_RWBW, mbwumonidr))
+				mpam_set_feature(mpam_feat_msmon_mbwu_rwbw, props);
+
+			/* Is NRDY hardware managed? */
+			mpam_mon_sel_outer_lock(msc);
+			mpam_ris_hw_probe_hw_nrdy(ris, MBWU, hw_managed);
+			mpam_mon_sel_outer_unlock(msc);
+			if (hw_managed)
+				mpam_set_feature(mpam_feat_msmon_mbwu_hw_nrdy, props);
+
+			/*
+			 * Don't warn about any missing firmware property for
+			 * MBWU NRDY - it doesn't make any sense!
+			 */
+		}
+	}
+}
+
  static int mpam_msc_hw_probe(struct mpam_msc *msc)
  {
  	u64 idr;
@@ -665,6 +810,7 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
 
  	idr = mpam_msc_read_idr(msc);
  	mutex_unlock(&msc->part_sel_lock);
+
  	msc->ris_max = FIELD_GET(MPAMF_IDR_RIS_MAX, idr);
 
  	/* Use these values so partid/pmg always starts with a valid value */
@@ -685,6 +831,12 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
  		ris = mpam_get_or_create_ris(msc, ris_idx);
  		if (IS_ERR(ris))
  			return PTR_ERR(ris);
+		ris->idr = idr;
+
+		mutex_lock(&msc->part_sel_lock);
+		__mpam_part_sel(ris_idx, 0, msc);
+		mpam_ris_hw_probe(ris);
+		mutex_unlock(&msc->part_sel_lock);
  	}
 
  	spin_lock(&partid_max_lock);
drivers/platform/arm64/mpam/mpam_internal.h +53 -0
index 42a454d5f914..ae6fd1f62cc4 100644
@@ -136,6 +136,55 @@ static inline void mpam_mon_sel_lock_held(struct mpam_msc *msc)
  		lockdep_assert_preemption_enabled();
  }
 
+/*
+ * When we compact the supported features, we don't care what they are.
+ * Storing them as a bitmap makes life easy.
+ */
+typedef u16 mpam_features_t;
+
+/* Bits for mpam_features_t */
+enum mpam_device_features {
+	mpam_feat_ccap_part = 0,
+	mpam_feat_cpor_part,
+	mpam_feat_mbw_part,
+	mpam_feat_mbw_min,
+	mpam_feat_mbw_max,
+	mpam_feat_mbw_prop,
+	mpam_feat_msmon,
+	mpam_feat_msmon_csu,
+	mpam_feat_msmon_csu_capture,
+	mpam_feat_msmon_csu_hw_nrdy,
+	mpam_feat_msmon_mbwu,
+	mpam_feat_msmon_mbwu_capture,
+	mpam_feat_msmon_mbwu_rwbw,
+	mpam_feat_msmon_mbwu_hw_nrdy,
+	mpam_feat_msmon_capt,
+	MPAM_FEATURE_LAST,
+};
+#define MPAM_ALL_FEATURES      ((1 << MPAM_FEATURE_LAST) - 1)
+
+struct mpam_props {
+	mpam_features_t		features;
+
+	u16			cpbm_wd;
+	u16			mbw_pbm_bits;
+	u16			bwa_wd;
+	u16			num_csu_mon;
+	u16			num_mbwu_mon;
+};
+
+static inline bool mpam_has_feature(enum mpam_device_features feat,
+				    struct mpam_props *props)
+{
+	return (1 << feat) & props->features;
+}
+
+static inline void mpam_set_feature(enum mpam_device_features feat,
+				    struct mpam_props *props)
+{
+	props->features |= (1 << feat);
+}
+
  struct mpam_class {
  	/* mpam_components in this class */
  	struct list_head	components;
@@ -175,6 +224,8 @@ struct mpam_vmsc {
  	/* mpam_msc_ris in this vmsc */
  	struct list_head	ris;
 
+	struct mpam_props	props;
+
  	/* All RIS in this vMSC are members of this MSC */
  	struct mpam_msc		*msc;
 
@@ -186,6 +237,8 @@ struct mpam_vmsc {
 
  struct mpam_msc_ris {
  	u8			ris_idx;
+	u64			idr;
+	struct mpam_props	props;
 
  	cpumask_t		affinity;
 
Signature · 3 lines
Thanks,

Ben
#69 lore raw

Re: [RFC PATCH 20/36] arm_mpam: Probe the hardware features resctrl supports

}

Quoted 41 lines
Quoted 25 lines
drivers/platform/arm64/mpam/mpam_devices.c +24 -0
+/*
+ * IHI009A.a has this nugget: "If a monitor does not support automatic behaviour
+ * of NRDY, software can use this bit for any purpose" - so hardware might not
+ * implement this - but it isn't RES0.
+ *
+ * Try and see what values stick in this bit. If we can write either value,
+ * its probably not implemented by hardware.
+ */
+#define mpam_ris_hw_probe_hw_nrdy(_ris, _mon_reg, _result)			\
+do {										\
+	u32 now;								\
+	u64 mon_sel;								\
+	bool can_set, can_clear;						\
+	struct mpam_msc *_msc = _ris->vmsc->msc;				\
+										\
+	if (WARN_ON_ONCE(!mpam_mon_sel_inner_lock(_msc))) {			\
+		_result = false;						\
+		break;								\
+	}									\
+	mon_sel = FIELD_PREP(MSMON_CFG_MON_SEL_MON_SEL, 0) |			\
+		  FIELD_PREP(MSMON_CFG_MON_SEL_RIS, _ris->ris_idx);		\
+	mpam_write_monsel_reg(_msc, CFG_MON_SEL, mon_sel);			\
+										\
+	mpam_write_monsel_reg(_msc, _mon_reg, MSMON___NRDY);			\

drivers/platform/arm64/mpam/mpam_devices.c static struct mpam_msc_ris *mpam_get_or_create_ris(struct mpam_msc *msc, · line 695

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+	now = mpam_read_monsel_reg(_msc, _mon_reg);				\
+	can_set = now & MSMON___NRDY;						\
+										\
+	mpam_write_monsel_reg(_msc, _mon_reg, 0);				\
+	now = mpam_read_monsel_reg(_msc, _mon_reg);				\
+	can_clear = !(now & MSMON___NRDY);					\
+	mpam_mon_sel_inner_unlock(_msc);					\
+										\
+	_result = (!can_set || !can_clear);					\
+} while (0)  

It is a bit surprising that something that looks like a function modifies a boolean passed by value. Consider continuing the pattern you have above:
#define mpam_ris_hw_probe_hw_nrdy(_ris, _mon_reg, _result) _mpam_ris_hw_probe_hw_nrdy(_ris, MSMON##_mon_reg, _result)

Quoted 6 lines

with signature:
void _mpam_ris_hw_probe_hw_nrdy(struct mpam_msc *msc, u16 reg, bool *hw_managed);

and using the _mpam functions from the new _mpam_ris_hw_probe_hw_nrdy().

Agreed that this is ugly. Only a tiny bit of macro stuff is actually going on here. I'd make it function.

If you really want to construct MSMON_CSU etc then wrap that helper with a macro that builds reg from the name.

I might have missed something though in converting this.
The version I have has some other changes though so not trivial to post here :(

author maintainer
#70 lore raw

Re: [RFC PATCH 20/36] arm_mpam: Probe the hardware features resctrl supports

In reply to Ben Horgan

Hi Ben,

Quoted 52 lines · Ben Horgan
Quoted 37 lines · James Morse

Expand the probing support with the control and monitor types we can use with resctrl.

drivers/platform/arm64/mpam/ renamed from drivers/platform/arm64/mpam/mpam_devices.c +0 -0

mpam_devices.c
index 8646fb85ad09..61911831ab39 100644

drivers/platform/arm64/mpam/mpam_devices.c +0 -0
@@ -645,6 +659,137 @@ static struct mpam_msc_ris *mpam_get_or_create_ris(struct mpam_msc
 *msc,
       return found;
   }

  +/*
+ * IHI009A.a has this nugget: "If a monitor does not support automatic behaviour
+ * of NRDY, software can use this bit for any purpose" - so hardware might not
+ * implement this - but it isn't RES0.
+ *
+ * Try and see what values stick in this bit. If we can write either value,
+ * its probably not implemented by hardware.
+ */

drivers/platform/arm64/mpam/mpam_devices.c +16 -0
+#define mpam_ris_hw_probe_hw_nrdy(_ris, _mon_reg, _result)            \
+do {                                        \
+    u32 now;                                \
+    u64 mon_sel;                                \
+    bool can_set, can_clear;                        \
+    struct mpam_msc *_msc = _ris->vmsc->msc;                \
+                                        \
+    if (WARN_ON_ONCE(!mpam_mon_sel_inner_lock(_msc))) {            \
+        _result = false;                        \
+        break;                                \
+    }                                    \
+    mon_sel = FIELD_PREP(MSMON_CFG_MON_SEL_MON_SEL, 0) |            \
+          FIELD_PREP(MSMON_CFG_MON_SEL_RIS, _ris->ris_idx);        \
+    mpam_write_monsel_reg(_msc, CFG_MON_SEL, mon_sel);            \
+                                        \
+    mpam_write_monsel_reg(_msc, _mon_reg, MSMON___NRDY);            \

drivers/platform/arm64/mpam/mpam_devices.c static struct mpam_msc_ris *mpam_get_or_create_ris(struct mpam_msc *msc, · line 695

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+    now = mpam_read_monsel_reg(_msc, _mon_reg);                \
+    can_set = now & MSMON___NRDY;                        \
+                                        \
+    mpam_write_monsel_reg(_msc, _mon_reg, 0);                \
+    now = mpam_read_monsel_reg(_msc, _mon_reg);                \
+    can_clear = !(now & MSMON___NRDY);                    \
+    mpam_mon_sel_inner_unlock(_msc);                    \
+                                        \
+    _result = (!can_set || !can_clear);                    \
+} while (0)

It is a bit surprising that something that looks like a function modifies a boolean passed by value.

Quoted 8 lines

Consider continuing the pattern you have above:
#define mpam_ris_hw_probe_hw_nrdy(_ris, _mon_reg, _result) _mpam_ris_hw_probe_hw_nrdy(_ris, MSMON##_mon_reg, _result)

with signature:
void _mpam_ris_hw_probe_hw_nrdy(struct mpam_msc *msc, u16 reg, bool *hw_managed);

and using the _mpam functions from the new _mpam_ris_hw_probe_hw_nrdy().

With that, it may as well return the result.
Done.

Thanks,

James

maintainer
#71 lore raw

Re: [RFC PATCH 20/36] arm_mpam: Probe the hardware features resctrl supports

In reply to James Morse

Hi James,

Quoted 227 lines · James Morse

Expand the probing support with the control and monitor types we can use with resctrl.

---
  drivers/platform/arm64/mpam/mpam_devices.c  | 154 +++++++++++++++++++-
  drivers/platform/arm64/mpam/mpam_internal.h |  53 +++++++
  2 files changed, 206 insertions(+), 1 deletion(-)
drivers/platform/arm64/mpam/mpam_devices.c +153 -1
index 8646fb85ad09..61911831ab39 100644
@@ -102,7 +102,7 @@ static LLIST_HEAD(mpam_garbage);
 
  static u32 __mpam_read_reg(struct mpam_msc *msc, u16 reg)
  {
-	WARN_ON_ONCE(reg > msc->mapped_hwpage_sz);
+	WARN_ON_ONCE(reg + sizeof(u32) > msc->mapped_hwpage_sz);
  	WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(), &msc->accessibility));
 
  	return readl_relaxed(msc->mapped_hwpage + reg);
@@ -131,6 +131,20 @@ static inline void _mpam_write_partsel_reg(struct mpam_msc *msc, u16 reg, u32 va
  }
  #define mpam_write_partsel_reg(msc, reg, val)  _mpam_write_partsel_reg(msc, MPAMCFG_##reg, val)
 
+static inline u32 _mpam_read_monsel_reg(struct mpam_msc *msc, u16 reg)
+{
+	mpam_mon_sel_lock_held(msc);
+	return __mpam_read_reg(msc, reg);
+}
+#define mpam_read_monsel_reg(msc, reg) _mpam_read_monsel_reg(msc, MSMON_##reg)
+
+static inline void _mpam_write_monsel_reg(struct mpam_msc *msc, u16 reg, u32 val)
+{
+	mpam_mon_sel_lock_held(msc);
+	__mpam_write_reg(msc, reg, val);
+}
+#define mpam_write_monsel_reg(msc, reg, val)   _mpam_write_monsel_reg(msc, MSMON_##reg, val)
+
  static u64 mpam_msc_read_idr(struct mpam_msc *msc)
  {
  	u64 idr_high = 0, idr_low;
@@ -645,6 +659,137 @@ static struct mpam_msc_ris *mpam_get_or_create_ris(struct mpam_msc *msc,
  	return found;
  }
 
+/*
+ * IHI009A.a has this nugget: "If a monitor does not support automatic behaviour
+ * of NRDY, software can use this bit for any purpose" - so hardware might not
+ * implement this - but it isn't RES0.
+ *
+ * Try and see what values stick in this bit. If we can write either value,
+ * its probably not implemented by hardware.
+ */
+#define mpam_ris_hw_probe_hw_nrdy(_ris, _mon_reg, _result)			\
+do {										\
+	u32 now;								\
+	u64 mon_sel;								\
+	bool can_set, can_clear;						\
+	struct mpam_msc *_msc = _ris->vmsc->msc;				\
+										\
+	if (WARN_ON_ONCE(!mpam_mon_sel_inner_lock(_msc))) {			\
+		_result = false;						\
+		break;								\
+	}									\
+	mon_sel = FIELD_PREP(MSMON_CFG_MON_SEL_MON_SEL, 0) |			\
+		  FIELD_PREP(MSMON_CFG_MON_SEL_RIS, _ris->ris_idx);		\
+	mpam_write_monsel_reg(_msc, CFG_MON_SEL, mon_sel);			\
+										\
+	mpam_write_monsel_reg(_msc, _mon_reg, MSMON___NRDY);			\
+	now = mpam_read_monsel_reg(_msc, _mon_reg);				\
+	can_set = now & MSMON___NRDY;						\
+										\
+	mpam_write_monsel_reg(_msc, _mon_reg, 0);				\
+	now = mpam_read_monsel_reg(_msc, _mon_reg);				\
+	can_clear = !(now & MSMON___NRDY);					\
+	mpam_mon_sel_inner_unlock(_msc);					\
+										\
+	_result = (!can_set || !can_clear);					\
+} while (0)
+
+static void mpam_ris_hw_probe(struct mpam_msc_ris *ris)
+{
+	int err;
+	struct mpam_msc *msc = ris->vmsc->msc;
+	struct mpam_props *props = &ris->props;
+
+	lockdep_assert_held(&msc->probe_lock);
+	lockdep_assert_held(&msc->part_sel_lock);
+
+	/* Cache Portion partitioning */
+	if (FIELD_GET(MPAMF_IDR_HAS_CPOR_PART, ris->idr)) {
+		u32 cpor_features = mpam_read_partsel_reg(msc, CPOR_IDR);
+
+		props->cpbm_wd = FIELD_GET(MPAMF_CPOR_IDR_CPBM_WD, cpor_features);
+		if (props->cpbm_wd)
+			mpam_set_feature(mpam_feat_cpor_part, props);
+	}
+
+	/* Memory bandwidth partitioning */
+	if (FIELD_GET(MPAMF_IDR_HAS_MBW_PART, ris->idr)) {
+		u32 mbw_features = mpam_read_partsel_reg(msc, MBW_IDR);
+
+		/* portion bitmap resolution */
+		props->mbw_pbm_bits = FIELD_GET(MPAMF_MBW_IDR_BWPBM_WD, mbw_features);
+		if (props->mbw_pbm_bits &&
+		    FIELD_GET(MPAMF_MBW_IDR_HAS_PBM, mbw_features))
+			mpam_set_feature(mpam_feat_mbw_part, props);
+
+		props->bwa_wd = FIELD_GET(MPAMF_MBW_IDR_BWA_WD, mbw_features);
+		if (props->bwa_wd && FIELD_GET(MPAMF_MBW_IDR_HAS_MAX, mbw_features))
+			mpam_set_feature(mpam_feat_mbw_max, props);
+	}
+
+	/* Performance Monitoring */
+	if (FIELD_GET(MPAMF_IDR_HAS_MSMON, ris->idr)) {
+		u32 msmon_features = mpam_read_partsel_reg(msc, MSMON_IDR);
+
+		/*
+		 * If the firmware max-nrdy-us property is missing, the
+		 * CSU counters can't be used. Should we wait forever?
+		 */
+		err = device_property_read_u32(&msc->pdev->dev,
+					       "arm,not-ready-us",
+					       &msc->nrdy_usec);
+
+		if (FIELD_GET(MPAMF_MSMON_IDR_MSMON_CSU, msmon_features)) {
+			u32 csumonidr;
+
+			csumonidr = mpam_read_partsel_reg(msc, CSUMON_IDR);
+			props->num_csu_mon = FIELD_GET(MPAMF_CSUMON_IDR_NUM_MON, csumonidr);
+			if (props->num_csu_mon) {
+				bool hw_managed;
+
+				mpam_set_feature(mpam_feat_msmon_csu, props);
+
+				/* Is NRDY hardware managed? */
+				mpam_mon_sel_outer_lock(msc);
+				mpam_ris_hw_probe_hw_nrdy(ris, CSU, hw_managed);
+				mpam_mon_sel_outer_unlock(msc);
+				if (hw_managed)
+					mpam_set_feature(mpam_feat_msmon_csu_hw_nrdy, props);
+			}
+
+			/*
+			 * Accept the missing firmware property if NRDY appears
+			 * un-implemented.
+			 */
+			if (err && mpam_has_feature(mpam_feat_msmon_csu_hw_nrdy, props))
+				pr_err_once("Counters are not usable because not-ready timeout was not provided by firmware.");
+		}
+		if (FIELD_GET(MPAMF_MSMON_IDR_MSMON_MBWU, msmon_features)) {
+			bool hw_managed;
+			u32 mbwumonidr = mpam_read_partsel_reg(msc, MBWUMON_IDR);
+
+			props->num_mbwu_mon = FIELD_GET(MPAMF_MBWUMON_IDR_NUM_MON, mbwumonidr);
+			if (props->num_mbwu_mon)
+				mpam_set_feature(mpam_feat_msmon_mbwu, props);
+
+			if (FIELD_GET(MPAMF_MBWUMON_IDR_HAS_RWBW, mbwumonidr))
+				mpam_set_feature(mpam_feat_msmon_mbwu_rwbw, props);
+
+			/* Is NRDY hardware managed? */
+			mpam_mon_sel_outer_lock(msc);
+			mpam_ris_hw_probe_hw_nrdy(ris, MBWU, hw_managed);
+			mpam_mon_sel_outer_unlock(msc);
+			if (hw_managed)
+				mpam_set_feature(mpam_feat_msmon_mbwu_hw_nrdy, props);
+
+			/*
+			 * Don't warn about any missing firmware property for
+			 * MBWU NRDY - it doesn't make any sense!
+			 */
+		}
+	}
+}
+
  static int mpam_msc_hw_probe(struct mpam_msc *msc)
  {
  	u64 idr;
@@ -665,6 +810,7 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
 
  	idr = mpam_msc_read_idr(msc);
  	mutex_unlock(&msc->part_sel_lock);
+
  	msc->ris_max = FIELD_GET(MPAMF_IDR_RIS_MAX, idr);
 
  	/* Use these values so partid/pmg always starts with a valid value */
@@ -685,6 +831,12 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
  		ris = mpam_get_or_create_ris(msc, ris_idx);
  		if (IS_ERR(ris))
  			return PTR_ERR(ris);
+		ris->idr = idr;
+
+		mutex_lock(&msc->part_sel_lock);
+		__mpam_part_sel(ris_idx, 0, msc);
+		mpam_ris_hw_probe(ris);
+		mutex_unlock(&msc->part_sel_lock);
  	}
 
  	spin_lock(&partid_max_lock);
drivers/platform/arm64/mpam/mpam_internal.h +16 -0
index 42a454d5f914..ae6fd1f62cc4 100644
@@ -136,6 +136,55 @@ static inline void mpam_mon_sel_lock_held(struct mpam_msc *msc)
  		lockdep_assert_preemption_enabled();
  }
 
+/*
+ * When we compact the supported features, we don't care what they are.
+ * Storing them as a bitmap makes life easy.
+ */
+typedef u16 mpam_features_t;
+
+/* Bits for mpam_features_t */
+enum mpam_device_features {
+	mpam_feat_ccap_part = 0,
+	mpam_feat_cpor_part,
+	mpam_feat_mbw_part,
+	mpam_feat_mbw_min,
+	mpam_feat_mbw_max,
+	mpam_feat_mbw_prop,
+	mpam_feat_msmon,
+	mpam_feat_msmon_csu,

drivers/platform/arm64/mpam/mpam_internal.h static inline void mpam_mon_sel_lock_held(struct mpam_msc *msc) · line 164

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h +10 -0
+	mpam_feat_msmon_csu_capture,
+	mpam_feat_msmon_csu_hw_nrdy,
+	mpam_feat_msmon_mbwu,
+	mpam_feat_msmon_mbwu_capture,
+	mpam_feat_msmon_mbwu_rwbw,
+	mpam_feat_msmon_mbwu_hw_nrdy,
+	mpam_feat_msmon_capt,
+	MPAM_FEATURE_LAST,
+};
+#define MPAM_ALL_FEATURES      ((1 << MPAM_FEATURE_LAST) - 1)

Consider a static assert to check the type is big enough.

static_assert(BITS_PER_TYPE(mpam_features_t) >= MPAM_FEATURE_LAST);

Quoted 36 lines
drivers/platform/arm64/mpam/mpam_internal.h +25 -0
+
+struct mpam_props {
+	mpam_features_t		features;
+
+	u16			cpbm_wd;
+	u16			mbw_pbm_bits;
+	u16			bwa_wd;
+	u16			num_csu_mon;
+	u16			num_mbwu_mon;
+};
+
+static inline bool mpam_has_feature(enum mpam_device_features feat,
+				    struct mpam_props *props)
+{
+	return (1 << feat) & props->features;
+}
+
+static inline void mpam_set_feature(enum mpam_device_features feat,
+				    struct mpam_props *props)
+{
+	props->features |= (1 << feat);
+}
+
  struct mpam_class {
  	/* mpam_components in this class */
  	struct list_head	components;
@@ -175,6 +224,8 @@ struct mpam_vmsc {
  	/* mpam_msc_ris in this vmsc */
  	struct list_head	ris;
 
+	struct mpam_props	props;
+
  	/* All RIS in this vMSC are members of this MSC */
  	struct mpam_msc		*msc;
 

drivers/platform/arm64/mpam/mpam_internal.h struct mpam_vmsc { · line 244

Quoted 8 lines
drivers/platform/arm64/mpam/mpam_internal.h +2 -0
@@ -186,6 +237,8 @@ struct mpam_vmsc {
 
  struct mpam_msc_ris {
  	u8			ris_idx;
+	u64			idr;
+	struct mpam_props	props;
 
  	cpumask_t		affinity;
 

Thanks,

Ben

author maintainer
#72 lore raw

Re: [RFC PATCH 20/36] arm_mpam: Probe the hardware features resctrl supports

Hi Ben,

Quoted 41 lines · Ben Horgan
Quoted 27 lines · James Morse

Expand the probing support with the control and monitor types we can use with resctrl.

drivers/platform/arm64/mpam/ renamed from drivers/platform/arm64/mpam/mpam_internal.h +0 -0

mpam_internal.h
index 42a454d5f914..ae6fd1f62cc4 100644

drivers/platform/arm64/mpam/mpam_internal.h +15 -0
@@ -136,6 +136,55 @@ static inline void mpam_mon_sel_lock_held(struct mpam_msc *msc)
          lockdep_assert_preemption_enabled();
  }
  +/*
+ * When we compact the supported features, we don't care what they are.
+ * Storing them as a bitmap makes life easy.
+ */
+typedef u16 mpam_features_t;
+
+/* Bits for mpam_features_t */
+enum mpam_device_features {
+    mpam_feat_ccap_part = 0,
+    mpam_feat_cpor_part,
+    mpam_feat_mbw_part,
+    mpam_feat_mbw_min,
+    mpam_feat_mbw_max,
+    mpam_feat_mbw_prop,
+    mpam_feat_msmon,
+    mpam_feat_msmon_csu,

drivers/platform/arm64/mpam/mpam_internal.h static inline void mpam_mon_sel_lock_held(struct mpam_msc *msc) · line 163

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h +10 -0
+    mpam_feat_msmon_csu_capture,
+    mpam_feat_msmon_csu_hw_nrdy,
+    mpam_feat_msmon_mbwu,
+    mpam_feat_msmon_mbwu_capture,
+    mpam_feat_msmon_mbwu_rwbw,
+    mpam_feat_msmon_mbwu_hw_nrdy,
+    mpam_feat_msmon_capt,
+    MPAM_FEATURE_LAST,
+};
+#define MPAM_ALL_FEATURES      ((1 << MPAM_FEATURE_LAST) - 1)

Consider a static assert to check the type is big enough.

Quoted 1 line

static_assert(BITS_PER_TYPE(mpam_features_t) >= MPAM_FEATURE_LAST);

Fancy.
There used to be an uglier one - not sure what happened to it!

Thanks,

James

author maintainer
#73 lore raw

RFC PATCH 21/36 arm_mpam: Merge supported features during mpam_enable() into mpam_class

In reply to James Morse

To make a decision about whether to expose an mpam class as a resctrl resource we need to know its overall supported features and properties.

Once we've probed all the resources, we can walk the tree and produced overall values by merging the bitmaps. This eliminates features that are only supported by some MSC that make up a component or class.

If bitmap properties are mismatched within a component we cannot support the mismatched feature.

Care has to be taken as vMSC may hold mismatched RIS.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 215 ++++++++++++++++++++
 drivers/platform/arm64/mpam/mpam_internal.h |   8 +
 2 files changed, 223 insertions(+)
drivers/platform/arm64/mpam/mpam_devices.c +215 -0
index 61911831ab39..7b042a35405a 100644
@@ -1186,8 +1186,223 @@ static struct platform_driver mpam_msc_driver = {
 	.remove = mpam_msc_drv_remove,
 };
 
+/* Any of these features mean the BWA_WD field is valid. */
+static bool mpam_has_bwa_wd_feature(struct mpam_props *props)
+{
+	if (mpam_has_feature(mpam_feat_mbw_min, props))
+		return true;
+	if (mpam_has_feature(mpam_feat_mbw_max, props))
+		return true;
+	if (mpam_has_feature(mpam_feat_mbw_prop, props))
+		return true;
+	return false;
+}
+
+#define MISMATCHED_HELPER(parent, child, helper, field, alias)		\
+	helper(parent) &&						\
+	((helper(child) && (parent)->field != (child)->field) ||	\
+	 (!helper(child) && !(alias)))
+
+#define MISMATCHED_FEAT(parent, child, feat, field, alias)		     \
+	mpam_has_feature((feat), (parent)) &&				     \
+	((mpam_has_feature((feat), (child)) && (parent)->field != (child)->field) || \
+	 (!mpam_has_feature((feat), (child)) && !(alias)))
+
+#define CAN_MERGE_FEAT(parent, child, feat, alias)			\
+	(alias) && !mpam_has_feature((feat), (parent)) &&		\
+	mpam_has_feature((feat), (child))
+
+/*
+ * Combime two props fields.
+ * If this is for controls that alias the same resource, it is safe to just
+ * copy the values over. If two aliasing controls implement the same scheme
+ * a safe value must be picked.
+ * For non-aliasing controls, these control different resources, and the
+ * resulting safe value must be compatible with both. When merging values in
+ * the tree, all the aliasing resources must be handled first.
+ * On mismatch, parent is modified.
+ */
+static void __props_mismatch(struct mpam_props *parent,
+			     struct mpam_props *child, bool alias)
+{
+	if (CAN_MERGE_FEAT(parent, child, mpam_feat_cpor_part, alias)) {
+		parent->cpbm_wd = child->cpbm_wd;
+	} else if (MISMATCHED_FEAT(parent, child, mpam_feat_cpor_part,
+				   cpbm_wd, alias)) {
+		pr_debug("%s cleared cpor_part\n", __func__);
+		mpam_clear_feature(mpam_feat_cpor_part, &parent->features);
+		parent->cpbm_wd = 0;
+	}
+
+	if (CAN_MERGE_FEAT(parent, child, mpam_feat_mbw_part, alias)) {
+		parent->mbw_pbm_bits = child->mbw_pbm_bits;
+	} else if (MISMATCHED_FEAT(parent, child, mpam_feat_mbw_part,
+				   mbw_pbm_bits, alias)) {
+		pr_debug("%s cleared mbw_part\n", __func__);
+		mpam_clear_feature(mpam_feat_mbw_part, &parent->features);
+		parent->mbw_pbm_bits = 0;
+	}
+
+	/* bwa_wd is a count of bits, fewer bits means less precision */
+	if (alias && !mpam_has_bwa_wd_feature(parent) && mpam_has_bwa_wd_feature(child)) {
+		parent->bwa_wd = child->bwa_wd;
+	} else if (MISMATCHED_HELPER(parent, child, mpam_has_bwa_wd_feature,
+				     bwa_wd, alias)) {
+		pr_debug("%s took the min bwa_wd\n", __func__);
+		parent->bwa_wd = min(parent->bwa_wd, child->bwa_wd);
+	}
+
+	/* For num properties, take the minimum */
+	if (CAN_MERGE_FEAT(parent, child, mpam_feat_msmon_csu, alias)) {
+		parent->num_csu_mon = child->num_csu_mon;
+	} else if (MISMATCHED_FEAT(parent, child, mpam_feat_msmon_csu,
+				   num_csu_mon, alias)) {
+		pr_debug("%s took the min num_csu_mon\n", __func__);
+		parent->num_csu_mon = min(parent->num_csu_mon, child->num_csu_mon);
+	}
+
+	if (CAN_MERGE_FEAT(parent, child, mpam_feat_msmon_mbwu, alias)) {
+		parent->num_mbwu_mon = child->num_mbwu_mon;
+	} else if (MISMATCHED_FEAT(parent, child, mpam_feat_msmon_mbwu,
+				   num_mbwu_mon, alias)) {
+		pr_debug("%s took the min num_mbwu_mon\n", __func__);
+		parent->num_mbwu_mon = min(parent->num_mbwu_mon, child->num_mbwu_mon);
+	}
+
+	if (alias) {
+		/* Merge features for aliased resources */
+		parent->features |= child->features;
+	} else {
+		/* Clear missing features for non aliasing */
+		parent->features &= child->features;
+	}
+}
+
+/*
+ * If a vmsc doesn't match class feature/configuration, do the right thing(tm).
+ * For 'num' properties we can just take the minimum.
+ * For properties where the mismatched unused bits would make a difference, we
+ * nobble the class feature, as we can't configure all the resources.
+ * e.g. The L3 cache is composed of two resources with 13 and 17 portion
+ * bitmaps respectively.
+ */
+static void
+__class_props_mismatch(struct mpam_class *class, struct mpam_vmsc *vmsc)
+{
+	struct mpam_props *cprops = &class->props;
+	struct mpam_props *vprops = &vmsc->props;
+
+	lockdep_assert_held(&mpam_list_lock); /* we modify class */
+
+	pr_debug("%s: Merging features for class:0x%lx &= vmsc:0x%lx\n",
+		 dev_name(&vmsc->msc->pdev->dev),
+		 (long)cprops->features, (long)vprops->features);
+
+	/* Take the safe value for any common features */
+	__props_mismatch(cprops, vprops, false);
+}
+
+static void
+__vmsc_props_mismatch(struct mpam_vmsc *vmsc, struct mpam_msc_ris *ris)
+{
+	struct mpam_props *rprops = &ris->props;
+	struct mpam_props *vprops = &vmsc->props;
+
+	lockdep_assert_held(&mpam_list_lock); /* we modify vmsc */
+
+	pr_debug("%s: Merging features for vmsc:0x%lx |= ris:0x%lx\n",
+		 dev_name(&vmsc->msc->pdev->dev),
+		 (long)vprops->features, (long)rprops->features);
+
+	/*
+	 * Merge mismatched features - Copy any features that aren't common,
+	 * but take the safe value for any common features.
+	 */
+	__props_mismatch(vprops, rprops, true);
+}
+
+/*
+ * Copy the first component's first vMSC's properties and features to the
+ * class. __class_props_mismatch() will remove conflicts.
+ * It is not possible to have a class with no components, or a component with
+ * no resources. The vMSC properties have already been built.
+ */
+static void mpam_enable_init_class_features(struct mpam_class *class)
+{
+	struct mpam_vmsc *vmsc;
+	struct mpam_component *comp;
+
+	comp = list_first_entry_or_null(&class->components,
+					struct mpam_component, class_list);
+	if (WARN_ON(!comp))
+		return;
+
+	vmsc = list_first_entry_or_null(&comp->vmsc,
+					struct mpam_vmsc, comp_list);
+	if (WARN_ON(!vmsc))
+		return;
+
+	class->props = vmsc->props;
+}
+
+static void mpam_enable_merge_vmsc_features(struct mpam_component *comp)
+{
+	struct mpam_vmsc *vmsc;
+	struct mpam_msc_ris *ris;
+	struct mpam_class *class = comp->class;
+
+	list_for_each_entry(vmsc, &comp->vmsc, comp_list) {
+		list_for_each_entry(ris, &vmsc->ris, vmsc_list) {
+			__vmsc_props_mismatch(vmsc, ris);
+			class->nrdy_usec = max(class->nrdy_usec,
+					       vmsc->msc->nrdy_usec);
+		}
+	}
+}
+
+static void mpam_enable_merge_class_features(struct mpam_component *comp)
+{
+	struct mpam_vmsc *vmsc;
+	struct mpam_class *class = comp->class;
+
+	list_for_each_entry(vmsc, &comp->vmsc, comp_list)
+		__class_props_mismatch(class, vmsc);
+}
+
+/*
+ * Merge all the common resource features into class.
+ * vmsc features are bitwise-or'd together, this must be done first.
+ * Next the class features are the bitwise-and of all the vmsc features.
+ * Other features are the min/max as appropriate.
+ *
+ * To avoid walking the whole tree twice, the class->nrdy_usec property is
+ * updated when working with the vmsc as it is a max(), and doesn't need
+ * initialising first.
+ */
+static void mpam_enable_merge_features(struct list_head *all_classes_list)
+{
+	struct mpam_class *class;
+	struct mpam_component *comp;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	list_for_each_entry(class, all_classes_list, classes_list) {
+		list_for_each_entry(comp, &class->components, class_list)
+			mpam_enable_merge_vmsc_features(comp);
+
+		mpam_enable_init_class_features(class);
+
+		list_for_each_entry(comp, &class->components, class_list)
+			mpam_enable_merge_class_features(comp);
+	}
+}
+
 static void mpam_enable_once(void)
 {
+	mutex_lock(&mpam_list_lock);
+	mpam_enable_merge_features(&mpam_classes);
+	mutex_unlock(&mpam_list_lock);
+
 	mutex_lock(&mpam_cpuhp_state_lock);
 	cpuhp_remove_state(mpam_cpuhp_state);
 	mpam_cpuhp_state = 0;
drivers/platform/arm64/mpam/mpam_internal.h +8 -0
index ae6fd1f62cc4..be56234b84b4 100644
@@ -185,12 +185,20 @@ static inline void mpam_set_feature(enum mpam_device_features feat,
 	props->features |= (1 << feat);
 }
 
+static inline void mpam_clear_feature(enum mpam_device_features feat,
+				      mpam_features_t *supported)
+{
+	*supported &= ~(1 << feat);
+}
+
 struct mpam_class {
 	/* mpam_components in this class */
 	struct list_head	components;
 
 	cpumask_t		affinity;
 
+	struct mpam_props	props;
+	u32			nrdy_usec;
 	u8			level;
 	enum mpam_class_types	type;
 
Signature · 1 line
2.39.5
maintainer
#74 lore raw

Re: [RFC PATCH 21/36] arm_mpam: Merge supported features during mpam_enable() into mpam_class

Hi James,

Quoted 6 lines · James Morse

To make a decision about whether to expose an mpam class as a resctrl resource we need to know its overall supported features and properties.

Once we've probed all the resources, we can walk the tree and produced overall values by merging the bitmaps. This

nit: s/produced/produce/

Quoted 40 lines

eliminates features that are only supported by some MSC that make up a component or class.

If bitmap properties are mismatched within a component we cannot support the mismatched feature.

Care has to be taken as vMSC may hold mismatched RIS.

---
  drivers/platform/arm64/mpam/mpam_devices.c  | 215 ++++++++++++++++++++
  drivers/platform/arm64/mpam/mpam_internal.h |   8 +
  2 files changed, 223 insertions(+)
drivers/platform/arm64/mpam/mpam_devices.c +18 -0
index 61911831ab39..7b042a35405a 100644
@@ -1186,8 +1186,223 @@ static struct platform_driver mpam_msc_driver = {
  	.remove = mpam_msc_drv_remove,
  };
 
+/* Any of these features mean the BWA_WD field is valid. */
+static bool mpam_has_bwa_wd_feature(struct mpam_props *props)
+{
+	if (mpam_has_feature(mpam_feat_mbw_min, props))
+		return true;
+	if (mpam_has_feature(mpam_feat_mbw_max, props))
+		return true;
+	if (mpam_has_feature(mpam_feat_mbw_prop, props))
+		return true;
+	return false;
+}
+
+#define MISMATCHED_HELPER(parent, child, helper, field, alias)		\
+	helper(parent) &&						\
+	((helper(child) && (parent)->field != (child)->field) ||	\
+	 (!helper(child) && !(alias)))
+
+#define MISMATCHED_FEAT(parent, child, feat, field, alias)		     \

drivers/platform/arm64/mpam/mpam_devices.c static struct platform_driver mpam_msc_driver = { · line 1216

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+	mpam_has_feature((feat), (parent)) &&				     \
+	((mpam_has_feature((feat), (child)) && (parent)->field != (child)->field) || \
+	 (!mpam_has_feature((feat), (child)) && !(alias)))
+
+#define CAN_MERGE_FEAT(parent, child, feat, alias)			\
+	(alias) && !mpam_has_feature((feat), (parent)) &&		\
+	mpam_has_feature((feat), (child))
+
+/*
+ * Combime two props fields.

nit: s/combime/combine/

Quoted 207 lines
drivers/platform/arm64/mpam/mpam_devices.c +187 -0
+ * If this is for controls that alias the same resource, it is safe to just
+ * copy the values over. If two aliasing controls implement the same scheme
+ * a safe value must be picked.
+ * For non-aliasing controls, these control different resources, and the
+ * resulting safe value must be compatible with both. When merging values in
+ * the tree, all the aliasing resources must be handled first.
+ * On mismatch, parent is modified.
+ */
+static void __props_mismatch(struct mpam_props *parent,
+			     struct mpam_props *child, bool alias)
+{
+	if (CAN_MERGE_FEAT(parent, child, mpam_feat_cpor_part, alias)) {
+		parent->cpbm_wd = child->cpbm_wd;
+	} else if (MISMATCHED_FEAT(parent, child, mpam_feat_cpor_part,
+				   cpbm_wd, alias)) {
+		pr_debug("%s cleared cpor_part\n", __func__);
+		mpam_clear_feature(mpam_feat_cpor_part, &parent->features);
+		parent->cpbm_wd = 0;
+	}
+
+	if (CAN_MERGE_FEAT(parent, child, mpam_feat_mbw_part, alias)) {
+		parent->mbw_pbm_bits = child->mbw_pbm_bits;
+	} else if (MISMATCHED_FEAT(parent, child, mpam_feat_mbw_part,
+				   mbw_pbm_bits, alias)) {
+		pr_debug("%s cleared mbw_part\n", __func__);
+		mpam_clear_feature(mpam_feat_mbw_part, &parent->features);
+		parent->mbw_pbm_bits = 0;
+	}
+
+	/* bwa_wd is a count of bits, fewer bits means less precision */
+	if (alias && !mpam_has_bwa_wd_feature(parent) && mpam_has_bwa_wd_feature(child)) {
+		parent->bwa_wd = child->bwa_wd;
+	} else if (MISMATCHED_HELPER(parent, child, mpam_has_bwa_wd_feature,
+				     bwa_wd, alias)) {
+		pr_debug("%s took the min bwa_wd\n", __func__);
+		parent->bwa_wd = min(parent->bwa_wd, child->bwa_wd);
+	}
+
+	/* For num properties, take the minimum */
+	if (CAN_MERGE_FEAT(parent, child, mpam_feat_msmon_csu, alias)) {
+		parent->num_csu_mon = child->num_csu_mon;
+	} else if (MISMATCHED_FEAT(parent, child, mpam_feat_msmon_csu,
+				   num_csu_mon, alias)) {
+		pr_debug("%s took the min num_csu_mon\n", __func__);
+		parent->num_csu_mon = min(parent->num_csu_mon, child->num_csu_mon);
+	}
+
+	if (CAN_MERGE_FEAT(parent, child, mpam_feat_msmon_mbwu, alias)) {
+		parent->num_mbwu_mon = child->num_mbwu_mon;
+	} else if (MISMATCHED_FEAT(parent, child, mpam_feat_msmon_mbwu,
+				   num_mbwu_mon, alias)) {
+		pr_debug("%s took the min num_mbwu_mon\n", __func__);
+		parent->num_mbwu_mon = min(parent->num_mbwu_mon, child->num_mbwu_mon);
+	}
+
+	if (alias) {
+		/* Merge features for aliased resources */
+		parent->features |= child->features;
+	} else {
+		/* Clear missing features for non aliasing */
+		parent->features &= child->features;
+	}
+}
+
+/*
+ * If a vmsc doesn't match class feature/configuration, do the right thing(tm).
+ * For 'num' properties we can just take the minimum.
+ * For properties where the mismatched unused bits would make a difference, we
+ * nobble the class feature, as we can't configure all the resources.
+ * e.g. The L3 cache is composed of two resources with 13 and 17 portion
+ * bitmaps respectively.
+ */
+static void
+__class_props_mismatch(struct mpam_class *class, struct mpam_vmsc *vmsc)
+{
+	struct mpam_props *cprops = &class->props;
+	struct mpam_props *vprops = &vmsc->props;
+
+	lockdep_assert_held(&mpam_list_lock); /* we modify class */
+
+	pr_debug("%s: Merging features for class:0x%lx &= vmsc:0x%lx\n",
+		 dev_name(&vmsc->msc->pdev->dev),
+		 (long)cprops->features, (long)vprops->features);
+
+	/* Take the safe value for any common features */
+	__props_mismatch(cprops, vprops, false);
+}
+
+static void
+__vmsc_props_mismatch(struct mpam_vmsc *vmsc, struct mpam_msc_ris *ris)
+{
+	struct mpam_props *rprops = &ris->props;
+	struct mpam_props *vprops = &vmsc->props;
+
+	lockdep_assert_held(&mpam_list_lock); /* we modify vmsc */
+
+	pr_debug("%s: Merging features for vmsc:0x%lx |= ris:0x%lx\n",
+		 dev_name(&vmsc->msc->pdev->dev),
+		 (long)vprops->features, (long)rprops->features);
+
+	/*
+	 * Merge mismatched features - Copy any features that aren't common,
+	 * but take the safe value for any common features.
+	 */
+	__props_mismatch(vprops, rprops, true);
+}
+
+/*
+ * Copy the first component's first vMSC's properties and features to the
+ * class. __class_props_mismatch() will remove conflicts.
+ * It is not possible to have a class with no components, or a component with
+ * no resources. The vMSC properties have already been built.
+ */
+static void mpam_enable_init_class_features(struct mpam_class *class)
+{
+	struct mpam_vmsc *vmsc;
+	struct mpam_component *comp;
+
+	comp = list_first_entry_or_null(&class->components,
+					struct mpam_component, class_list);
+	if (WARN_ON(!comp))
+		return;
+
+	vmsc = list_first_entry_or_null(&comp->vmsc,
+					struct mpam_vmsc, comp_list);
+	if (WARN_ON(!vmsc))
+		return;
+
+	class->props = vmsc->props;
+}
+
+static void mpam_enable_merge_vmsc_features(struct mpam_component *comp)
+{
+	struct mpam_vmsc *vmsc;
+	struct mpam_msc_ris *ris;
+	struct mpam_class *class = comp->class;
+
+	list_for_each_entry(vmsc, &comp->vmsc, comp_list) {
+		list_for_each_entry(ris, &vmsc->ris, vmsc_list) {
+			__vmsc_props_mismatch(vmsc, ris);
+			class->nrdy_usec = max(class->nrdy_usec,
+					       vmsc->msc->nrdy_usec);
+		}
+	}
+}
+
+static void mpam_enable_merge_class_features(struct mpam_component *comp)
+{
+	struct mpam_vmsc *vmsc;
+	struct mpam_class *class = comp->class;
+
+	list_for_each_entry(vmsc, &comp->vmsc, comp_list)
+		__class_props_mismatch(class, vmsc);
+}
+
+/*
+ * Merge all the common resource features into class.
+ * vmsc features are bitwise-or'd together, this must be done first.
+ * Next the class features are the bitwise-and of all the vmsc features.
+ * Other features are the min/max as appropriate.
+ *
+ * To avoid walking the whole tree twice, the class->nrdy_usec property is
+ * updated when working with the vmsc as it is a max(), and doesn't need
+ * initialising first.
+ */
+static void mpam_enable_merge_features(struct list_head *all_classes_list)
+{
+	struct mpam_class *class;
+	struct mpam_component *comp;
+
+	lockdep_assert_held(&mpam_list_lock);
+
+	list_for_each_entry(class, all_classes_list, classes_list) {
+		list_for_each_entry(comp, &class->components, class_list)
+			mpam_enable_merge_vmsc_features(comp);
+
+		mpam_enable_init_class_features(class);
+
+		list_for_each_entry(comp, &class->components, class_list)
+			mpam_enable_merge_class_features(comp);
+	}
+}
+
  static void mpam_enable_once(void)
  {
+	mutex_lock(&mpam_list_lock);
+	mpam_enable_merge_features(&mpam_classes);
+	mutex_unlock(&mpam_list_lock);
+
  	mutex_lock(&mpam_cpuhp_state_lock);
  	cpuhp_remove_state(mpam_cpuhp_state);
  	mpam_cpuhp_state = 0;
drivers/platform/arm64/mpam/mpam_internal.h +6 -0
index ae6fd1f62cc4..be56234b84b4 100644
@@ -185,12 +185,20 @@ static inline void mpam_set_feature(enum mpam_device_features feat,
  	props->features |= (1 << feat);
  }
 
+static inline void mpam_clear_feature(enum mpam_device_features feat,
+				      mpam_features_t *supported)
+{
+	*supported &= ~(1 << feat);
+}
+
  struct mpam_class {

drivers/platform/arm64/mpam/mpam_internal.h static inline void mpam_set_feature(enum mpam_device_features feat, · line 204

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h +2 -0
  	/* mpam_components in this class */
  	struct list_head	components;
 
  	cpumask_t		affinity;
 
+	struct mpam_props	props;
+	u32			nrdy_usec;
  	u8			level;
  	enum mpam_class_types	type;
 

Thanks,

Ben

author maintainer
#75 lore raw

RFC PATCH 22/36 arm_mpam: Reset MSC controls from cpu hp callbacks

In reply to James Morse

When a CPU comes online, it may bring a newly accessible MSC with it. Only the default partid has its value reset by hardware, and even then the MSC might not have been reset since its config was previously dirtyied. e.g. Kexec.

Any in-use partid must have its configuration restored, or reset. In-use partids may be held in caches and evicted later.

MSC are also reset when CPUs are taken offline to cover cases where firmware doesn't reset the MSC over reboot using UEFI, or kexec where there is no firmware involvement.

If the configuration for a RIS has not been touched since it was brought online, it does not need resetting again.

To reset, write the maximum values for all discovered controls.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 124 +++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |   8 ++
 2 files changed, 131 insertions(+), 1 deletion(-)
drivers/platform/arm64/mpam/mpam_devices.c +123 -1
index 7b042a35405a..d014dbe0ab96 100644
@@ -7,6 +7,7 @@
 #include <linux/atomic.h>
 #include <linux/arm_mpam.h>
 #include <linux/bitfield.h>
+#include <linux/bitmap.h>
 #include <linux/cacheinfo.h>
 #include <linux/cpu.h>
 #include <linux/cpumask.h>
@@ -849,8 +850,116 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
 	return 0;
 }
 
+static void mpam_reset_msc_bitmap(struct mpam_msc *msc, u16 reg, u16 wd)
+{
+	u32 num_words, msb;
+	u32 bm = ~0;
+	int i;
+
+	lockdep_assert_held(&msc->part_sel_lock);
+
+	if (wd == 0)
+		return;
+
+	/*
+	 * Write all ~0 to all but the last 32bit-word, which may
+	 * have fewer bits...
+	 */
+	num_words = DIV_ROUND_UP(wd, 32);
+	for (i = 0; i < num_words - 1; i++, reg += sizeof(bm))
+		__mpam_write_reg(msc, reg, bm);
+
+	/*
+	 * ....and then the last (maybe) partial 32bit word. When wd is a
+	 * multiple of 32, msb should be 31 to write a full 32bit word.
+	 */
+	msb = (wd - 1) % 32;
+	bm = GENMASK(msb, 0);
+	if (bm)
+		__mpam_write_reg(msc, reg, bm);
+}
+
+static void mpam_reset_ris_partid(struct mpam_msc_ris *ris, u16 partid)
+{
+	u16 bwa_fract = MPAMCFG_MBW_MAX_MAX;
+	struct mpam_msc *msc = ris->vmsc->msc;
+	struct mpam_props *rprops = &ris->props;
+
+	mpam_assert_srcu_read_lock_held();
+
+	mutex_lock(&msc->part_sel_lock);
+	__mpam_part_sel(ris->ris_idx, partid, msc);
+
+	if (mpam_has_feature(mpam_feat_cpor_part, rprops))
+		mpam_reset_msc_bitmap(msc, MPAMCFG_CPBM, rprops->cpbm_wd);
+
+	if (mpam_has_feature(mpam_feat_mbw_part, rprops))
+		mpam_reset_msc_bitmap(msc, MPAMCFG_MBW_PBM, rprops->mbw_pbm_bits);
+
+	if (mpam_has_feature(mpam_feat_mbw_min, rprops))
+		mpam_write_partsel_reg(msc, MBW_MIN, 0);
+
+	if (mpam_has_feature(mpam_feat_mbw_max, rprops))
+		mpam_write_partsel_reg(msc, MBW_MAX, bwa_fract);
+
+	if (mpam_has_feature(mpam_feat_mbw_prop, rprops))
+		mpam_write_partsel_reg(msc, MBW_PROP, bwa_fract);
+	mutex_unlock(&msc->part_sel_lock);
+}
+
+static void mpam_reset_ris(struct mpam_msc_ris *ris)
+{
+	u16 partid, partid_max;
+
+	mpam_assert_srcu_read_lock_held();
+
+	if (ris->in_reset_state)
+		return;
+
+	spin_lock(&partid_max_lock);
+	partid_max = mpam_partid_max;
+	spin_unlock(&partid_max_lock);
+	for (partid = 0; partid < partid_max; partid++)
+		mpam_reset_ris_partid(ris, partid);
+}
+
+static void mpam_reset_msc(struct mpam_msc *msc, bool online)
+{
+	int idx;
+	struct mpam_msc_ris *ris;
+
+	mpam_assert_srcu_read_lock_held();
+
+	mpam_mon_sel_outer_lock(msc);
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(ris, &msc->ris, msc_list, srcu_read_lock_held(&mpam_srcu)) {
+		mpam_reset_ris(ris);
+
+		/*
+		 * Set in_reset_state when coming online. The reset state
+		 * for non-zero partid may be lost while the CPUs are offline.
+		 */
+		ris->in_reset_state = online;
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+	mpam_mon_sel_outer_unlock(msc);
+}
+
 static int mpam_cpu_online(unsigned int cpu)
 {
+	int idx;
+	struct mpam_msc *msc;
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		if (!cpumask_test_cpu(cpu, &msc->accessibility))
+			continue;
+
+		if (atomic_fetch_inc(&msc->online_refs) == 0)
+			mpam_reset_msc(msc, true);
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
 	return 0;
 }
 
@@ -886,6 +995,19 @@ static int mpam_discovery_cpu_online(unsigned int cpu)
 
 static int mpam_cpu_offline(unsigned int cpu)
 {
+	int idx;
+	struct mpam_msc *msc;
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		if (!cpumask_test_cpu(cpu, &msc->accessibility))
+			continue;
+
+		if (atomic_dec_and_test(&msc->online_refs))
+			mpam_reset_msc(msc, false);
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
 	return 0;
 }
 
@@ -1419,7 +1541,7 @@ static void mpam_enable_once(void)
 	mpam_register_cpuhp_callbacks(mpam_cpu_online, mpam_cpu_offline);
 
 	printk(KERN_INFO "MPAM enabled with %u partid and %u pmg\n",
-	       mpam_partid_max + 1, mpam_pmg_max + 1);
+		READ_ONCE(mpam_partid_max) + 1, mpam_pmg_max + 1);
 }
 
 /*
drivers/platform/arm64/mpam/mpam_internal.h +8 -0
index be56234b84b4..f3cc88136524 100644
@@ -5,6 +5,7 @@
 #define MPAM_INTERNAL_H
 
 #include <linux/arm_mpam.h>
+#include <linux/atomic.h>
 #include <linux/cpumask.h>
 #include <linux/io.h>
 #include <linux/llist.h>
@@ -43,6 +44,7 @@ struct mpam_msc {
 	struct pcc_mbox_chan	*pcc_chan;
 	u32			nrdy_usec;
 	cpumask_t		accessibility;
+	atomic_t		online_refs;
 
 	/*
 	 * probe_lock is only take during discovery. After discovery these
@@ -247,6 +249,7 @@ struct mpam_msc_ris {
 	u8			ris_idx;
 	u64			idr;
 	struct mpam_props	props;
+	bool			in_reset_state;
 
 	cpumask_t		affinity;
 
@@ -266,6 +269,11 @@ struct mpam_msc_ris {
 extern struct srcu_struct mpam_srcu;
 extern struct list_head mpam_classes;
 
+static inline void mpam_assert_srcu_read_lock_held(void)
+{
+	WARN_ON_ONCE(!srcu_read_lock_held((&mpam_srcu)));
+}
+
 /* System wide partid/pmg values */
 extern u16 mpam_partid_max;
 extern u8 mpam_pmg_max;
Signature · 1 line
2.39.5
maintainer
#76 lore raw

Re: [RFC PATCH 22/36] arm_mpam: Reset MSC controls from cpu hp callbacks

Hi James,

Quoted 57 lines · James Morse

When a CPU comes online, it may bring a newly accessible MSC with it. Only the default partid has its value reset by hardware, and even then the MSC might not have been reset since its config was previously dirtyied. e.g. Kexec.

Any in-use partid must have its configuration restored, or reset. In-use partids may be held in caches and evicted later.

MSC are also reset when CPUs are taken offline to cover cases where firmware doesn't reset the MSC over reboot using UEFI, or kexec where there is no firmware involvement.

If the configuration for a RIS has not been touched since it was brought online, it does not need resetting again.

To reset, write the maximum values for all discovered controls.

---
  drivers/platform/arm64/mpam/mpam_devices.c  | 124 +++++++++++++++++++-
  drivers/platform/arm64/mpam/mpam_internal.h |   8 ++
  2 files changed, 131 insertions(+), 1 deletion(-)
drivers/platform/arm64/mpam/mpam_devices.c +18 -0
index 7b042a35405a..d014dbe0ab96 100644
@@ -7,6 +7,7 @@
  #include <linux/atomic.h>
  #include <linux/arm_mpam.h>
  #include <linux/bitfield.h>
+#include <linux/bitmap.h>
  #include <linux/cacheinfo.h>
  #include <linux/cpu.h>
  #include <linux/cpumask.h>
@@ -849,8 +850,116 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
  	return 0;
  }
 
+static void mpam_reset_msc_bitmap(struct mpam_msc *msc, u16 reg, u16 wd)
+{
+	u32 num_words, msb;
+	u32 bm = ~0;
+	int i;
+
+	lockdep_assert_held(&msc->part_sel_lock);
+
+	if (wd == 0)
+		return;
+
+	/*
+	 * Write all ~0 to all but the last 32bit-word, which may
+	 * have fewer bits...
+	 */
+	num_words = DIV_ROUND_UP(wd, 32);
+	for (i = 0; i < num_words - 1; i++, reg += sizeof(bm))

drivers/platform/arm64/mpam/mpam_devices.c static int mpam_msc_hw_probe(struct mpam_msc *msc) · line 879

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+		__mpam_write_reg(msc, reg, bm);
+
+	/*
+	 * ....and then the last (maybe) partial 32bit word. When wd is a
+	 * multiple of 32, msb should be 31 to write a full 32bit word.
+	 */
+	msb = (wd - 1) % 32;
+	bm = GENMASK(msb, 0);
+	if (bm)
+		__mpam_write_reg(msc, reg, bm);

Drop the 'if' as the 0 bit will always be part of the mask.

Quoted 107 lines
drivers/platform/arm64/mpam/mpam_devices.c +94 -0
+}
+
+static void mpam_reset_ris_partid(struct mpam_msc_ris *ris, u16 partid)
+{
+	u16 bwa_fract = MPAMCFG_MBW_MAX_MAX;
+	struct mpam_msc *msc = ris->vmsc->msc;
+	struct mpam_props *rprops = &ris->props;
+
+	mpam_assert_srcu_read_lock_held();
+
+	mutex_lock(&msc->part_sel_lock);
+	__mpam_part_sel(ris->ris_idx, partid, msc);
+
+	if (mpam_has_feature(mpam_feat_cpor_part, rprops))
+		mpam_reset_msc_bitmap(msc, MPAMCFG_CPBM, rprops->cpbm_wd);
+
+	if (mpam_has_feature(mpam_feat_mbw_part, rprops))
+		mpam_reset_msc_bitmap(msc, MPAMCFG_MBW_PBM, rprops->mbw_pbm_bits);
+
+	if (mpam_has_feature(mpam_feat_mbw_min, rprops))
+		mpam_write_partsel_reg(msc, MBW_MIN, 0);
+
+	if (mpam_has_feature(mpam_feat_mbw_max, rprops))
+		mpam_write_partsel_reg(msc, MBW_MAX, bwa_fract);
+
+	if (mpam_has_feature(mpam_feat_mbw_prop, rprops))
+		mpam_write_partsel_reg(msc, MBW_PROP, bwa_fract);
+	mutex_unlock(&msc->part_sel_lock);
+}
+
+static void mpam_reset_ris(struct mpam_msc_ris *ris)
+{
+	u16 partid, partid_max;
+
+	mpam_assert_srcu_read_lock_held();
+
+	if (ris->in_reset_state)
+		return;
+
+	spin_lock(&partid_max_lock);
+	partid_max = mpam_partid_max;
+	spin_unlock(&partid_max_lock);
+	for (partid = 0; partid < partid_max; partid++)
+		mpam_reset_ris_partid(ris, partid);
+}
+
+static void mpam_reset_msc(struct mpam_msc *msc, bool online)
+{
+	int idx;
+	struct mpam_msc_ris *ris;
+
+	mpam_assert_srcu_read_lock_held();
+
+	mpam_mon_sel_outer_lock(msc);
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(ris, &msc->ris, msc_list, srcu_read_lock_held(&mpam_srcu)) {
+		mpam_reset_ris(ris);
+
+		/*
+		 * Set in_reset_state when coming online. The reset state
+		 * for non-zero partid may be lost while the CPUs are offline.
+		 */
+		ris->in_reset_state = online;
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+	mpam_mon_sel_outer_unlock(msc);
+}
+
  static int mpam_cpu_online(unsigned int cpu)
  {
+	int idx;
+	struct mpam_msc *msc;
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		if (!cpumask_test_cpu(cpu, &msc->accessibility))
+			continue;
+
+		if (atomic_fetch_inc(&msc->online_refs) == 0)
+			mpam_reset_msc(msc, true);
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
  	return 0;
  }
 
@@ -886,6 +995,19 @@ static int mpam_discovery_cpu_online(unsigned int cpu)
 
  static int mpam_cpu_offline(unsigned int cpu)
  {
+	int idx;
+	struct mpam_msc *msc;
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		if (!cpumask_test_cpu(cpu, &msc->accessibility))
+			continue;
+
+		if (atomic_dec_and_test(&msc->online_refs))
+			mpam_reset_msc(msc, false);
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
  	return 0;
  }
 

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_once(void) · line 1544

Quoted 5 lines
drivers/platform/arm64/mpam/mpam_devices.c +1 -1
@@ -1419,7 +1541,7 @@ static void mpam_enable_once(void)
  	mpam_register_cpuhp_callbacks(mpam_cpu_online, mpam_cpu_offline);
 
  	printk(KERN_INFO "MPAM enabled with %u partid and %u pmg\n",
-	       mpam_partid_max + 1, mpam_pmg_max + 1);
+		READ_ONCE(mpam_partid_max) + 1, mpam_pmg_max + 1);

Belongs in 'arm_mpam: Probe MSCs to find the supported partid/pmg values'.

Quoted 33 lines
drivers/platform/arm64/mpam/mpam_devices.c +0 -0
  }
 
  /*
drivers/platform/arm64/mpam/mpam_internal.h +3 -0
index be56234b84b4..f3cc88136524 100644
@@ -5,6 +5,7 @@
  #define MPAM_INTERNAL_H
 
  #include <linux/arm_mpam.h>
+#include <linux/atomic.h>
  #include <linux/cpumask.h>
  #include <linux/io.h>
  #include <linux/llist.h>
@@ -43,6 +44,7 @@ struct mpam_msc {
  	struct pcc_mbox_chan	*pcc_chan;
  	u32			nrdy_usec;
  	cpumask_t		accessibility;
+	atomic_t		online_refs;
 
  	/*
  	 * probe_lock is only take during discovery. After discovery these
@@ -247,6 +249,7 @@ struct mpam_msc_ris {
  	u8			ris_idx;
  	u64			idr;
  	struct mpam_props	props;
+	bool			in_reset_state;
 
  	cpumask_t		affinity;
 
@@ -266,6 +269,11 @@ struct mpam_msc_ris {
  extern struct srcu_struct mpam_srcu;

drivers/platform/arm64/mpam/mpam_internal.h struct mpam_msc_ris { · line 279

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_internal.h +5 -0
  extern struct list_head mpam_classes;
 
+static inline void mpam_assert_srcu_read_lock_held(void)
+{
+	WARN_ON_ONCE(!srcu_read_lock_held((&mpam_srcu)));
+}
+
  /* System wide partid/pmg values */
  extern u16 mpam_partid_max;
  extern u8 mpam_pmg_max;

Thanks,

Ben

author maintainer
#77 lore raw

Re: [RFC PATCH 22/36] arm_mpam: Reset MSC controls from cpu hp callbacks

Hi Ben,

Quoted 46 lines · Ben Horgan
Quoted 45 lines · James Morse

When a CPU comes online, it may bring a newly accessible MSC with it. Only the default partid has its value reset by hardware, and even then the MSC might not have been reset since its config was previously dirtyied. e.g. Kexec.

Any in-use partid must have its configuration restored, or reset. In-use partids may be held in caches and evicted later.

MSC are also reset when CPUs are taken offline to cover cases where firmware doesn't reset the MSC over reboot using UEFI, or kexec where there is no firmware involvement.

If the configuration for a RIS has not been touched since it was brought online, it does not need resetting again.

To reset, write the maximum values for all discovered controls.

drivers/platform/arm64/mpam/ renamed from drivers/platform/arm64/mpam/mpam_devices.c +0 -0

mpam_devices.c
index 7b042a35405a..d014dbe0ab96 100644

drivers/platform/arm64/mpam/mpam_devices.c +19 -0
@@ -849,8 +850,116 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
      return 0;
  }
  +static void mpam_reset_msc_bitmap(struct mpam_msc *msc, u16 reg, u16 wd)
+{
+    u32 num_words, msb;
+    u32 bm = ~0;
+    int i;
+
+    lockdep_assert_held(&msc->part_sel_lock);
+
+    if (wd == 0)
+        return;
+
+    /*
+     * Write all ~0 to all but the last 32bit-word, which may
+     * have fewer bits...
+     */
+    num_words = DIV_ROUND_UP(wd, 32);
+    for (i = 0; i < num_words - 1; i++, reg += sizeof(bm))
+        __mpam_write_reg(msc, reg, bm);
+
+    /*
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_devices.c static int mpam_msc_hw_probe(struct mpam_msc *msc) · line 879

Quoted 7 lines
drivers/platform/arm64/mpam/mpam_devices.c +7 -0
+     * ....and then the last (maybe) partial 32bit word. When wd is a
+     * multiple of 32, msb should be 31 to write a full 32bit word.
+     */
+    msb = (wd - 1) % 32;
+    bm = GENMASK(msb, 0);
+    if (bm)
+        __mpam_write_reg(msc, reg, bm);

Drop the 'if' as the 0 bit will always be part of the mask.

Yup, this was a harmless leftover of the versions that tried to optionally write the left over bits.

Quoted 7 lines

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_once(void) · line 1544

Quoted 5 lines
drivers/platform/arm64/mpam/mpam_devices.c +1 -1
  @@ -1419,7 +1541,7 @@ static void mpam_enable_once(void)
      mpam_register_cpuhp_callbacks(mpam_cpu_online, mpam_cpu_offline);
        printk(KERN_INFO "MPAM enabled with %u partid and %u pmg\n",
-           mpam_partid_max + 1, mpam_pmg_max + 1);
+        READ_ONCE(mpam_partid_max) + 1, mpam_pmg_max + 1);

Belongs in 'arm_mpam: Probe MSCs to find the supported partid/pmg values'.

That value is now protected by a lock, and can't change at this point, so it no longer needs the READ_ONCE() anyway.

Thanks,

James

author maintainer
#78 lore raw

RFC PATCH 23/36 arm_mpam: Add a helper to touch an MSC from any CPU

In reply to James Morse

Resetting RIS entries from the cpuhp callback is easy as the callback occurs on the correct CPU. This won't be true for any other caller that wants to reset or configure an MSC.

Add a helper that schedules the provided function if necessary. Prevent the cpuhp callbacks from changing the MSC state by taking the cpuhp lock.

---
 drivers/platform/arm64/mpam/mpam_devices.c | 37 ++++++++++++++++++++--
 1 file changed, 34 insertions(+), 3 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +34 -3
index d014dbe0ab96..2e32e54cc081 100644
@@ -907,20 +907,51 @@ static void mpam_reset_ris_partid(struct mpam_msc_ris *ris, u16 partid)
 	mutex_unlock(&msc->part_sel_lock);
 }
 
-static void mpam_reset_ris(struct mpam_msc_ris *ris)
+/*
+ * Called via smp_call_on_cpu() to prevent migration, while still being
+ * pre-emptible.
+ */
+static int mpam_reset_ris(void *arg)
 {
 	u16 partid, partid_max;
+	struct mpam_msc_ris *ris = arg;
 
 	mpam_assert_srcu_read_lock_held();
 
 	if (ris->in_reset_state)
-		return;
+		return 0;
 
 	spin_lock(&partid_max_lock);
 	partid_max = mpam_partid_max;
 	spin_unlock(&partid_max_lock);
 	for (partid = 0; partid < partid_max; partid++)
 		mpam_reset_ris_partid(ris, partid);
+
+	return 0;
+}
+
+/*
+ * Get the preferred CPU for this MSC. If it is accessible from this CPU,
+ * this CPU is preferred. This can be preempted/migrated, it will only result
+ * in more work.
+ */
+static int mpam_get_msc_preferred_cpu(struct mpam_msc *msc)
+{
+	int cpu = raw_smp_processor_id();
+
+	if (cpumask_test_cpu(cpu, &msc->accessibility))
+		return cpu;
+
+	return cpumask_first_and(&msc->accessibility, cpu_online_mask);
+}
+
+static int mpam_touch_msc(struct mpam_msc *msc, int (*fn)(void *a), void *arg)
+{
+	lockdep_assert_irqs_enabled();
+	lockdep_assert_cpus_held();
+	mpam_assert_srcu_read_lock_held();
+
+	return smp_call_on_cpu(mpam_get_msc_preferred_cpu(msc), fn, arg, true);
 }
 
 static void mpam_reset_msc(struct mpam_msc *msc, bool online)
@@ -933,7 +964,7 @@ static void mpam_reset_msc(struct mpam_msc *msc, bool online)
 	mpam_mon_sel_outer_lock(msc);
 	idx = srcu_read_lock(&mpam_srcu);
 	list_for_each_entry_srcu(ris, &msc->ris, msc_list, srcu_read_lock_held(&mpam_srcu)) {
-		mpam_reset_ris(ris);
+		mpam_touch_msc(msc, &mpam_reset_ris, ris);
 
 		/*
 		 * Set in_reset_state when coming online. The reset state
Signature · 1 line
2.39.5
author maintainer
#79 lore raw

RFC PATCH 24/36 arm_mpam: Extend reset logic to allow devices to be reset any time

In reply to James Morse

cpuhp callbacks aren't the only time the MSC configuration may need to be reset. Resctrl has an API call to reset a class.
If an MPAM error interrupt arrives it indicates the driver has misprogrammed an MSC. The safest thing to do is reset all the MSCs and disable MPAM.

Add a helper to reset RIS via their class. Call this from mpam_disable(), which can be scheduled from the error interrupt handler.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 62 ++++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |  1 +
 2 files changed, 61 insertions(+), 2 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +60 -2
index 2e32e54cc081..145535cd4732 100644
@@ -916,8 +916,6 @@ static int mpam_reset_ris(void *arg)
 	u16 partid, partid_max;
 	struct mpam_msc_ris *ris = arg;
 
-	mpam_assert_srcu_read_lock_held();
-
 	if (ris->in_reset_state)
 		return 0;
 
@@ -1575,6 +1573,66 @@ static void mpam_enable_once(void)
 		READ_ONCE(mpam_partid_max) + 1, mpam_pmg_max + 1);
 }
 
+static void mpam_reset_component_locked(struct mpam_component *comp)
+{
+	int idx;
+	struct mpam_msc *msc;
+	struct mpam_vmsc *vmsc;
+	struct mpam_msc_ris *ris;
+
+	might_sleep();
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_rcu(vmsc, &comp->vmsc, comp_list) {
+		msc = vmsc->msc;
+
+		list_for_each_entry_rcu(ris, &vmsc->ris, vmsc_list) {
+			if (!ris->in_reset_state)
+				mpam_touch_msc(msc, mpam_reset_ris, ris);
+			ris->in_reset_state = true;
+		}
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+}
+
+static void mpam_reset_class_locked(struct mpam_class *class)
+{
+	int idx;
+	struct mpam_component *comp;
+
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_rcu(comp, &class->components, class_list)
+		mpam_reset_component_locked(comp);
+	srcu_read_unlock(&mpam_srcu, idx);
+}
+
+static void mpam_reset_class(struct mpam_class *class)
+{
+	cpus_read_lock();
+	mpam_reset_class_locked(class);
+	cpus_read_unlock();
+}
+
+/*
+ * Called in response to an error IRQ.
+ * All of MPAMs errors indicate a software bug, restore any modified
+ * controls to their reset values.
+ */
+void mpam_disable(void)
+{
+	int idx;
+	struct mpam_class *class;
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(class, &mpam_classes, classes_list,
+				 srcu_read_lock_held(&mpam_srcu))
+		mpam_reset_class(class);
+	srcu_read_unlock(&mpam_srcu, idx);
+}
+
 /*
  * Enable mpam once all devices have been probed.
  * Scheduled by mpam_discovery_cpu_online() once all devices have been created.
drivers/platform/arm64/mpam/mpam_internal.h +1 -0
index f3cc88136524..de05eece0a31 100644
@@ -280,6 +280,7 @@ extern u8 mpam_pmg_max;
 
 /* Scheduled work callback to enable mpam once all MSC have been probed */
 void mpam_enable(struct work_struct *work);
+void mpam_disable(void);
 
 int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,
 				   cpumask_t *affinity);
Signature · 1 line
2.39.5
maintainer
#80 lore raw

Re: [RFC PATCH 24/36] arm_mpam: Extend reset logic to allow devices to be reset any time

Hi James,

Quoted 81 lines · James Morse

cpuhp callbacks aren't the only time the MSC configuration may need to be reset. Resctrl has an API call to reset a class.
If an MPAM error interrupt arrives it indicates the driver has misprogrammed an MSC. The safest thing to do is reset all the MSCs and disable MPAM.

Add a helper to reset RIS via their class. Call this from mpam_disable(), which can be scheduled from the error interrupt handler.

---
  drivers/platform/arm64/mpam/mpam_devices.c  | 62 ++++++++++++++++++++-
  drivers/platform/arm64/mpam/mpam_internal.h |  1 +
  2 files changed, 61 insertions(+), 2 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +49 -2
index 2e32e54cc081..145535cd4732 100644
@@ -916,8 +916,6 @@ static int mpam_reset_ris(void *arg)
  	u16 partid, partid_max;
  	struct mpam_msc_ris *ris = arg;
 
-	mpam_assert_srcu_read_lock_held();
-
  	if (ris->in_reset_state)
  		return 0;
 
@@ -1575,6 +1573,66 @@ static void mpam_enable_once(void)
  		READ_ONCE(mpam_partid_max) + 1, mpam_pmg_max + 1);
  }
 
+static void mpam_reset_component_locked(struct mpam_component *comp)
+{
+	int idx;
+	struct mpam_msc *msc;
+	struct mpam_vmsc *vmsc;
+	struct mpam_msc_ris *ris;
+
+	might_sleep();
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_rcu(vmsc, &comp->vmsc, comp_list) {
+		msc = vmsc->msc;
+
+		list_for_each_entry_rcu(ris, &vmsc->ris, vmsc_list) {
+			if (!ris->in_reset_state)
+				mpam_touch_msc(msc, mpam_reset_ris, ris);
+			ris->in_reset_state = true;
+		}
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+}
+
+static void mpam_reset_class_locked(struct mpam_class *class)
+{
+	int idx;
+	struct mpam_component *comp;
+
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_rcu(comp, &class->components, class_list)
+		mpam_reset_component_locked(comp);
+	srcu_read_unlock(&mpam_srcu, idx);
+}
+
+static void mpam_reset_class(struct mpam_class *class)
+{
+	cpus_read_lock();
+	mpam_reset_class_locked(class);
+	cpus_read_unlock();
+}
+
+/*
+ * Called in response to an error IRQ.
+ * All of MPAMs errors indicate a software bug, restore any modified
+ * controls to their reset values.
+ */
+void mpam_disable(void)

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_once(void) · line 1634

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+{
+	int idx;
+	struct mpam_class *class;
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(class, &mpam_classes, classes_list,
+				 srcu_read_lock_held(&mpam_srcu))
+		mpam_reset_class(class);
+	srcu_read_unlock(&mpam_srcu, idx);
+}

Consider moving to the next patch where you introduce interrupt support.

Quoted 9 lines
drivers/platform/arm64/mpam/mpam_devices.c +1 -0
+
  /*
   * Enable mpam once all devices have been probed.
   * Scheduled by mpam_discovery_cpu_online() once all devices have been created.
drivers/platform/arm64/mpam/mpam_internal.h +0 -0
index f3cc88136524..de05eece0a31 100644

drivers/platform/arm64/mpam/mpam_internal.h extern u8 mpam_pmg_max; · line 286

Quoted 7 lines
drivers/platform/arm64/mpam/mpam_internal.h +1 -0
@@ -280,6 +280,7 @@ extern u8 mpam_pmg_max;
 
  /* Scheduled work callback to enable mpam once all MSC have been probed */
  void mpam_enable(struct work_struct *work);
+void mpam_disable(void);
 
  int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,
  				   cpumask_t *affinity);

Thanks,

Ben

author maintainer
#81 lore raw

Re: [RFC PATCH 24/36] arm_mpam: Extend reset logic to allow devices to be reset any time

Hi Ben,

Quoted 24 lines · Ben Horgan
Quoted 23 lines · James Morse

cpuhp callbacks aren't the only time the MSC configuration may need to be reset. Resctrl has an API call to reset a class.
If an MPAM error interrupt arrives it indicates the driver has misprogrammed an MSC. The safest thing to do is reset all the MSCs and disable MPAM.

Add a helper to reset RIS via their class. Call this from mpam_disable(), which can be scheduled from the error interrupt handler.

drivers/platform/arm64/mpam/ renamed from drivers/platform/arm64/mpam/mpam_devices.c +0 -0

mpam_devices.c
index 2e32e54cc081..145535cd4732 100644

drivers/platform/arm64/mpam/mpam_devices.c +9 -0
 
+/*
+ * Called in response to an error IRQ.
+ * All of MPAMs errors indicate a software bug, restore any modified
+ * controls to their reset values.
+ */
+void mpam_disable(void)
+{
+    int idx;
+    struct mpam_class *class;
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_once(void) · line 1634

Quoted 7 lines
drivers/platform/arm64/mpam/mpam_devices.c +7 -0
+
+    idx = srcu_read_lock(&mpam_srcu);
+    list_for_each_entry_srcu(class, &mpam_classes, classes_list,
+                 srcu_read_lock_held(&mpam_srcu))
+        mpam_reset_class(class);
+    srcu_read_unlock(&mpam_srcu, idx);
+}

Consider moving to the next patch where you introduce interrupt support.

I pulled these changes out of that patch to try and make it simpler! Doing that would leave a bunch of static functions unused.

Thanks,

James

author maintainer
#82 lore raw

RFC PATCH 25/36 arm_mpam: Register and enable IRQs

In reply to James Morse

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 304 +++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |   9 +-
 2 files changed, 307 insertions(+), 6 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +299 -5
index 145535cd4732..af19cc25d16e 100644
@@ -14,6 +14,9 @@
 #include <linux/device.h>
 #include <linux/errno.h>
 #include <linux/gfp.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/irqdesc.h>
 #include <linux/list.h>
 #include <linux/lockdep.h>
 #include <linux/mutex.h>
@@ -62,6 +65,12 @@ static DEFINE_SPINLOCK(partid_max_lock);
  */
 static DECLARE_WORK(mpam_enable_work, &mpam_enable);
 
+/*
+ * All mpam error interrupts indicate a software bug. On receipt, disable the
+ * driver.
+ */
+static DECLARE_WORK(mpam_broken_work, &mpam_disable);
+
 /*
  * An MSC is a physical container for controls and monitors, each identified by
  * their RIS index. These share a base-address, interrupts and some MMIO
@@ -159,6 +168,24 @@ static u64 mpam_msc_read_idr(struct mpam_msc *msc)
 	return (idr_high << 32) | idr_low;
 }
 
+static void mpam_msc_zero_esr(struct mpam_msc *msc)
+{
+	__mpam_write_reg(msc, MPAMF_ESR, 0);
+	if (msc->has_extd_esr)
+		__mpam_write_reg(msc, MPAMF_ESR + 4, 0);
+}
+
+static u64 mpam_msc_read_esr(struct mpam_msc *msc)
+{
+	u64 esr_high = 0, esr_low;
+
+	esr_low = __mpam_read_reg(msc, MPAMF_ESR);
+	if (msc->has_extd_esr)
+		esr_high = __mpam_read_reg(msc, MPAMF_ESR + 4);
+
+	return (esr_high << 32) | esr_low;
+}
+
 static void __mpam_part_sel_raw(u32 partsel, struct mpam_msc *msc)
 {
 	lockdep_assert_held(&msc->part_sel_lock);
@@ -405,12 +432,12 @@ static void mpam_msc_destroy(struct mpam_msc *msc)
 
 	lockdep_assert_held(&mpam_list_lock);
 
-	list_del_rcu(&msc->glbl_list);
-	platform_set_drvdata(pdev, NULL);
-
 	list_for_each_entry_safe(ris, tmp, &msc->ris, msc_list)
 		mpam_ris_destroy(ris);
 
+	list_del_rcu(&msc->glbl_list);
+	platform_set_drvdata(pdev, NULL);
+
 	add_to_garbage(msc);
 	msc->garbage.pdev = pdev;
 }
@@ -828,6 +855,7 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
 		pmg_max = FIELD_GET(MPAMF_IDR_PMG_MAX, idr);
 		msc->partid_max = min(msc->partid_max, partid_max);
 		msc->pmg_max = min(msc->pmg_max, pmg_max);
+		msc->has_extd_esr = FIELD_GET(MPAMF_IDR_HAS_EXT_ESR, idr);
 
 		ris = mpam_get_or_create_ris(msc, ris_idx);
 		if (IS_ERR(ris))
@@ -974,6 +1002,13 @@ static void mpam_reset_msc(struct mpam_msc *msc, bool online)
 	mpam_mon_sel_outer_unlock(msc);
 }
 
+static void _enable_percpu_irq(void *_irq)
+{
+	int *irq = _irq;
+
+	enable_percpu_irq(*irq, IRQ_TYPE_NONE);
+}
+
 static int mpam_cpu_online(unsigned int cpu)
 {
 	int idx;
@@ -984,6 +1019,9 @@ static int mpam_cpu_online(unsigned int cpu)
 		if (!cpumask_test_cpu(cpu, &msc->accessibility))
 			continue;
 
+		if (msc->reenable_error_ppi)
+			_enable_percpu_irq(&msc->reenable_error_ppi);
+
 		if (atomic_fetch_inc(&msc->online_refs) == 0)
 			mpam_reset_msc(msc, true);
 	}
@@ -1032,6 +1070,9 @@ static int mpam_cpu_offline(unsigned int cpu)
 		if (!cpumask_test_cpu(cpu, &msc->accessibility))
 			continue;
 
+		if (msc->reenable_error_ppi)
+			disable_percpu_irq(msc->reenable_error_ppi);
+
 		if (atomic_dec_and_test(&msc->online_refs))
 			mpam_reset_msc(msc, false);
 	}
@@ -1058,6 +1099,51 @@ static void mpam_register_cpuhp_callbacks(int (*online)(unsigned int online),
 	mutex_unlock(&mpam_cpuhp_state_lock);
 }
 
+static int __setup_ppi(struct mpam_msc *msc)
+{
+	int cpu;
+
+	msc->error_dev_id = alloc_percpu_gfp(struct mpam_msc *, GFP_KERNEL);
+	if (!msc->error_dev_id)
+		return -ENOMEM;
+
+	for_each_cpu(cpu, &msc->accessibility) {
+		struct mpam_msc *empty = *per_cpu_ptr(msc->error_dev_id, cpu);
+
+		if (empty) {
+			pr_err_once("%s shares PPI with %s!\n",
+				    dev_name(&msc->pdev->dev),
+				    dev_name(&empty->pdev->dev));
+			return -EBUSY;
+		}
+		*per_cpu_ptr(msc->error_dev_id, cpu) = msc;
+	}
+
+	return 0;
+}
+
+static int mpam_msc_setup_error_irq(struct mpam_msc *msc)
+{
+	int irq;
+
+	irq = platform_get_irq_byname_optional(msc->pdev, "error");
+	if (irq <= 0)
+		return 0;
+
+	/* Allocate and initialise the percpu device pointer for PPI */
+	if (irq_is_percpu(irq))
+		return __setup_ppi(msc);
+
+	/* sanity check: shared interrupts can be routed anywhere? */
+	if (!cpumask_equal(&msc->accessibility, cpu_possible_mask)) {
+		pr_err_once("msc:%u is a private resource with a shared error interrupt",
+			    msc->id);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 static int mpam_dt_count_msc(void)
 {
 	int count = 0;
@@ -1266,6 +1352,10 @@ static int mpam_msc_drv_probe(struct platform_device *pdev)
 			break;
 		}
 
+		err = mpam_msc_setup_error_irq(msc);
+		if (err)
+			break;
+
 		if (device_property_read_u32(&pdev->dev, "pcc-channel",
 					     &msc->pcc_subspace_id))
 			msc->iface = MPAM_IFACE_MMIO;
@@ -1548,11 +1638,193 @@ static void mpam_enable_merge_features(struct list_head *all_classes_list)
 	}
 }
 
+static char *mpam_errcode_names[16] = {
+	[0] = "No error",
+	[1] = "PARTID_SEL_Range",
+	[2] = "Req_PARTID_Range",
+	[3] = "MSMONCFG_ID_RANGE",
+	[4] = "Req_PMG_Range",
+	[5] = "Monitor_Range",
+	[6] = "intPARTID_Range",
+	[7] = "Unexpected_INTERNAL",
+	[8] = "Undefined_RIS_PART_SEL",
+	[9] = "RIS_No_Control",
+	[10] = "Undefined_RIS_MON_SEL",
+	[11] = "RIS_No_Monitor",
+	[12 ... 15] = "Reserved"
+};
+
+static int mpam_enable_msc_ecr(void *_msc)
+{
+	struct mpam_msc *msc = _msc;
+
+	__mpam_write_reg(msc, MPAMF_ECR, 1);
+
+	return 0;
+}
+
+static int mpam_disable_msc_ecr(void *_msc)
+{
+	struct mpam_msc *msc = _msc;
+
+	__mpam_write_reg(msc, MPAMF_ECR, 0);
+
+	return 0;
+}
+
+static irqreturn_t __mpam_irq_handler(int irq, struct mpam_msc *msc)
+{
+	u64 reg;
+	u16 partid;
+	u8 errcode, pmg, ris;
+
+	if (WARN_ON_ONCE(!msc) ||
+	    WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(),
+					   &msc->accessibility)))
+		return IRQ_NONE;
+
+	reg = mpam_msc_read_esr(msc);
+
+	errcode = FIELD_GET(MPAMF_ESR_ERRCODE, reg);
+	if (!errcode)
+		return IRQ_NONE;
+
+	/* Clear level triggered irq */
+	mpam_msc_zero_esr(msc);
+
+	partid = FIELD_GET(MPAMF_ESR_PARTID_OR_MON, reg);
+	pmg = FIELD_GET(MPAMF_ESR_PMG, reg);
+	ris = FIELD_GET(MPAMF_ESR_PMG, reg);
+
+	pr_err("error irq from msc:%u '%s', partid:%u, pmg: %u, ris: %u\n",
+	       msc->id, mpam_errcode_names[errcode], partid, pmg, ris);
+
+	if (irq_is_percpu(irq)) {
+		mpam_disable_msc_ecr(msc);
+		schedule_work(&mpam_broken_work);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_WAKE_THREAD;
+}
+
+static irqreturn_t mpam_ppi_handler(int irq, void *dev_id)
+{
+	struct mpam_msc *msc = *(struct mpam_msc **)dev_id;
+
+	return __mpam_irq_handler(irq, msc);
+}
+
+static irqreturn_t mpam_spi_handler(int irq, void *dev_id)
+{
+	struct mpam_msc *msc = dev_id;
+
+	return __mpam_irq_handler(irq, msc);
+}
+
+static irqreturn_t mpam_disable_thread(int irq, void *dev_id);
+
+static int mpam_register_irqs(void)
+{
+	int err, irq, idx;
+	struct mpam_msc *msc;
+
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)
+			continue;
+
+		/* The MPAM spec says the interrupt can be SPI, PPI or LPI */
+		/* We anticipate sharing the interrupt with other MSCs */
+		if (irq_is_percpu(irq)) {
+			err = request_percpu_irq(irq, &mpam_ppi_handler,
+						 "mpam:msc:error",
+						 msc->error_dev_id);
+			if (err)
+				return err;
+
+			msc->reenable_error_ppi = irq;
+			smp_call_function_many(&msc->accessibility,
+					       &_enable_percpu_irq, &irq,
+					       true);
+		} else {
+			err = devm_request_threaded_irq(&msc->pdev->dev, irq,
+							&mpam_spi_handler,
+							&mpam_disable_thread,
+							IRQF_SHARED,
+							"mpam:msc:error", msc);
+			if (err)
+				return err;
+		}
+
+		msc->error_irq_requested = true;
+		mpam_touch_msc(msc, mpam_enable_msc_ecr, msc);
+		msc->error_irq_hw_enabled = true;
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
+	return 0;
+}
+
+static void mpam_unregister_irqs(void)
+{
+	int irq, idx;
+	struct mpam_msc *msc;
+
+	cpus_read_lock();
+	/* take the lock as free_irq() can sleep */
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)
+			continue;
+
+		if (msc->error_irq_hw_enabled) {
+			mpam_touch_msc(msc, mpam_disable_msc_ecr, msc);
+			msc->error_irq_hw_enabled = false;
+		}
+
+		if (msc->error_irq_requested) {
+			if (irq_is_percpu(irq)) {
+				msc->reenable_error_ppi = 0;
+				free_percpu_irq(irq, msc->error_dev_id);
+			} else {
+				devm_free_irq(&msc->pdev->dev, irq, msc);
+			}
+			msc->error_irq_requested = false;
+		}
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+	cpus_read_unlock();
+}
+
 static void mpam_enable_once(void)
 {
+	int err;
+
+	/*
+	 * If all the MSC have been probed, enabling the IRQs happens next.
+	 * That involves cross-calling to a CPU that can reach the MSC, and
+	 * the locks must be taken in this order:
+	 */
+	cpus_read_lock();
 	mutex_lock(&mpam_list_lock);
 	mpam_enable_merge_features(&mpam_classes);
+
+	err = mpam_register_irqs();
+	if (err)
+		pr_warn("Failed to register irqs: %d\n", err);
+
 	mutex_unlock(&mpam_list_lock);
+	cpus_read_unlock();
+
+	if (err) {
+		schedule_work(&mpam_broken_work);
+		return;
+	}
 
 	mutex_lock(&mpam_cpuhp_state_lock);
 	cpuhp_remove_state(mpam_cpuhp_state);
@@ -1621,16 +1893,39 @@ static void mpam_reset_class(struct mpam_class *class)
  * All of MPAMs errors indicate a software bug, restore any modified
  * controls to their reset values.
  */
-void mpam_disable(void)
+static irqreturn_t mpam_disable_thread(int irq, void *dev_id)
 {
 	int idx;
 	struct mpam_class *class;
+	struct mpam_msc *msc, *tmp;
+
+	mutex_lock(&mpam_cpuhp_state_lock);
+	if (mpam_cpuhp_state) {
+		cpuhp_remove_state(mpam_cpuhp_state);
+		mpam_cpuhp_state = 0;
+	}
+	mutex_unlock(&mpam_cpuhp_state_lock);
+
+	mpam_unregister_irqs();
 
 	idx = srcu_read_lock(&mpam_srcu);
 	list_for_each_entry_srcu(class, &mpam_classes, classes_list,
 				 srcu_read_lock_held(&mpam_srcu))
 		mpam_reset_class(class);
 	srcu_read_unlock(&mpam_srcu, idx);
+
+	mutex_lock(&mpam_list_lock);
+	list_for_each_entry_safe(msc, tmp, &mpam_all_msc, glbl_list)
+		mpam_msc_destroy(msc);
+	mutex_unlock(&mpam_list_lock);
+	mpam_free_garbage();
+
+	return IRQ_HANDLED;
+}
+
+void mpam_disable(struct work_struct *ignored)
+{
+	mpam_disable_thread(0, NULL);
 }
 
 /*
@@ -1644,7 +1939,6 @@ void mpam_enable(struct work_struct *work)
 	struct mpam_msc *msc;
 	bool all_devices_probed = true;
 
-	/* Have we probed all the hw devices? */
 	mutex_lock(&mpam_list_lock);
 	list_for_each_entry(msc, &mpam_all_msc, glbl_list) {
 		mutex_lock(&msc->probe_lock);
drivers/platform/arm64/mpam/mpam_internal.h +8 -1
index de05eece0a31..e1c6a2676b54 100644
@@ -44,6 +44,11 @@ struct mpam_msc {
 	struct pcc_mbox_chan	*pcc_chan;
 	u32			nrdy_usec;
 	cpumask_t		accessibility;
+	bool			has_extd_esr;
+
+	int				reenable_error_ppi;
+	struct mpam_msc * __percpu	*error_dev_id;
+
 	atomic_t		online_refs;
 
 	/*
@@ -52,6 +57,8 @@ struct mpam_msc {
 	 */
 	struct mutex		probe_lock;
 	bool			probed;
+	bool			error_irq_requested;
+	bool			error_irq_hw_enabled;
 	u16			partid_max;
 	u8			pmg_max;
 	unsigned long		ris_idxs[128 / BITS_PER_LONG];
@@ -280,7 +287,7 @@ extern u8 mpam_pmg_max;
 
 /* Scheduled work callback to enable mpam once all MSC have been probed */
 void mpam_enable(struct work_struct *work);
-void mpam_disable(void);
+void mpam_disable(struct work_struct *work);
 
 int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,
 				   cpumask_t *affinity);
Signature · 1 line
2.39.5
#83 lore raw

RE: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

Hello James,

Quoted 248 lines

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 304
+++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |   9 +-
 2 files changed, 307 insertions(+), 6 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +28 -3
index 145535cd4732..af19cc25d16e 100644
@@ -14,6 +14,9 @@
 #include <linux/device.h>
 #include <linux/errno.h>
 #include <linux/gfp.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/irqdesc.h>
 #include <linux/list.h>
 #include <linux/lockdep.h>
 #include <linux/mutex.h>
@@ -62,6 +65,12 @@ static DEFINE_SPINLOCK(partid_max_lock);
  */
 static DECLARE_WORK(mpam_enable_work, &mpam_enable);
 
+/*
+ * All mpam error interrupts indicate a software bug. On receipt,
+disable the
+ * driver.
+ */
+static DECLARE_WORK(mpam_broken_work, &mpam_disable);
+
 /*
  * An MSC is a physical container for controls and monitors, each identified by
  * their RIS index. These share a base-address, interrupts and some MMIO
@@ -159,6 +168,24 @@ static u64 mpam_msc_read_idr(struct mpam_msc
 *msc)
 	return (idr_high << 32) | idr_low;
 }
 
+static void mpam_msc_zero_esr(struct mpam_msc *msc) {
+	__mpam_write_reg(msc, MPAMF_ESR, 0);
+	if (msc->has_extd_esr)
+		__mpam_write_reg(msc, MPAMF_ESR + 4, 0); }
+
+static u64 mpam_msc_read_esr(struct mpam_msc *msc) {
+	u64 esr_high = 0, esr_low;
+
+	esr_low = __mpam_read_reg(msc, MPAMF_ESR);
+	if (msc->has_extd_esr)
+		esr_high = __mpam_read_reg(msc, MPAMF_ESR + 4);
+
+	return (esr_high << 32) | esr_low;
+}
+
 static void __mpam_part_sel_raw(u32 partsel, struct mpam_msc *msc)  {
 	lockdep_assert_held(&msc->part_sel_lock);
@@ -405,12 +432,12 @@ static void mpam_msc_destroy(struct mpam_msc
 *msc)
 
 	lockdep_assert_held(&mpam_list_lock);
 
-	list_del_rcu(&msc->glbl_list);
-	platform_set_drvdata(pdev, NULL);
-
 	list_for_each_entry_safe(ris, tmp, &msc->ris, msc_list)
 		mpam_ris_destroy(ris);
 
+	list_del_rcu(&msc->glbl_list);
+	platform_set_drvdata(pdev, NULL);
+
 	add_to_garbage(msc);
 	msc->garbage.pdev = pdev;

}

drivers/platform/arm64/mpam/mpam_devices.c +1 -0
@@ -828,6 +855,7 @@ static int mpam_msc_hw_probe(struct mpam_msc
 *msc)
 		pmg_max = FIELD_GET(MPAMF_IDR_PMG_MAX, idr);
 		msc->partid_max = min(msc->partid_max, partid_max);
 		msc->pmg_max = min(msc->pmg_max, pmg_max);
+		msc->has_extd_esr =
 FIELD_GET(MPAMF_IDR_HAS_EXT_ESR, idr);
 
drivers/platform/arm64/mpam/mpam_devices.c +47 -0
 		ris = mpam_get_or_create_ris(msc, ris_idx);
 		if (IS_ERR(ris))
@@ -974,6 +1002,13 @@ static void mpam_reset_msc(struct mpam_msc
 *msc, bool online)
 	mpam_mon_sel_outer_unlock(msc);
 }
 
+static void _enable_percpu_irq(void *_irq) {
+	int *irq = _irq;
+
+	enable_percpu_irq(*irq, IRQ_TYPE_NONE); }
+
 static int mpam_cpu_online(unsigned int cpu)  {
 	int idx;
@@ -984,6 +1019,9 @@ static int mpam_cpu_online(unsigned int cpu)
 		if (!cpumask_test_cpu(cpu, &msc->accessibility))
 			continue;
 
+		if (msc->reenable_error_ppi)
+			_enable_percpu_irq(&msc->reenable_error_ppi);
+
 		if (atomic_fetch_inc(&msc->online_refs) == 0)
 			mpam_reset_msc(msc, true);
 	}
@@ -1032,6 +1070,9 @@ static int mpam_cpu_offline(unsigned int cpu)
 		if (!cpumask_test_cpu(cpu, &msc->accessibility))
 			continue;
 
+		if (msc->reenable_error_ppi)
+			disable_percpu_irq(msc->reenable_error_ppi);
+
 		if (atomic_dec_and_test(&msc->online_refs))
 			mpam_reset_msc(msc, false);
 	}
@@ -1058,6 +1099,51 @@ static void mpam_register_cpuhp_callbacks(int
 (*online)(unsigned int online),
 	mutex_unlock(&mpam_cpuhp_state_lock);
 }
 
+static int __setup_ppi(struct mpam_msc *msc) {
+	int cpu;
+
+	msc->error_dev_id = alloc_percpu_gfp(struct mpam_msc *,
 GFP_KERNEL);
+	if (!msc->error_dev_id)
+		return -ENOMEM;
+
+	for_each_cpu(cpu, &msc->accessibility) {
+		struct mpam_msc *empty = *per_cpu_ptr(msc->error_dev_id,
 cpu);
+
+		if (empty) {
+			pr_err_once("%s shares PPI with %s!\n",
+				    dev_name(&msc->pdev->dev),
+				    dev_name(&empty->pdev->dev));
+			return -EBUSY;
+		}
+		*per_cpu_ptr(msc->error_dev_id, cpu) = msc;
+	}
+
+	return 0;
+}
+
+static int mpam_msc_setup_error_irq(struct mpam_msc *msc) {
+	int irq;
+
+	irq = platform_get_irq_byname_optional(msc->pdev, "error");
+	if (irq <= 0)
+		return 0;
+
+	/* Allocate and initialise the percpu device pointer for PPI */
+	if (irq_is_percpu(irq))
+		return __setup_ppi(msc);
+
+	/* sanity check: shared interrupts can be routed anywhere? */
+	if (!cpumask_equal(&msc->accessibility, cpu_possible_mask)) {
+		pr_err_once("msc:%u is a private resource with a shared error

interrupt",

drivers/platform/arm64/mpam/mpam_devices.c +11 -0
+			    msc->id);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 static int mpam_dt_count_msc(void)
 {
 	int count = 0;
@@ -1266,6 +1352,10 @@ static int mpam_msc_drv_probe(struct
 platform_device *pdev)
 			break;
 		}
 
+		err = mpam_msc_setup_error_irq(msc);
+		if (err)
+			break;
+
 		if (device_property_read_u32(&pdev->dev, "pcc-channel",
 					     &msc->pcc_subspace_id))
drivers/platform/arm64/mpam/mpam_devices.c +44 -0
 			msc->iface = MPAM_IFACE_MMIO;
@@ -1548,11 +1638,193 @@ static void mpam_enable_merge_features(struct
 list_head *all_classes_list)
 	}
 }
 
+static char *mpam_errcode_names[16] = {
+	[0] = "No error",
+	[1] = "PARTID_SEL_Range",
+	[2] = "Req_PARTID_Range",
+	[3] = "MSMONCFG_ID_RANGE",
+	[4] = "Req_PMG_Range",
+	[5] = "Monitor_Range",
+	[6] = "intPARTID_Range",
+	[7] = "Unexpected_INTERNAL",
+	[8] = "Undefined_RIS_PART_SEL",
+	[9] = "RIS_No_Control",
+	[10] = "Undefined_RIS_MON_SEL",
+	[11] = "RIS_No_Monitor",
+	[12 ... 15] = "Reserved"
+};
+
+static int mpam_enable_msc_ecr(void *_msc) {
+	struct mpam_msc *msc = _msc;
+
+	__mpam_write_reg(msc, MPAMF_ECR, 1);
+
+	return 0;
+}
+
+static int mpam_disable_msc_ecr(void *_msc) {
+	struct mpam_msc *msc = _msc;
+
+	__mpam_write_reg(msc, MPAMF_ECR, 0);
+
+	return 0;
+}
+
+static irqreturn_t __mpam_irq_handler(int irq, struct mpam_msc *msc) {
+	u64 reg;
+	u16 partid;
+	u8 errcode, pmg, ris;
+
+	if (WARN_ON_ONCE(!msc) ||
+	    WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(),
+					   &msc->accessibility)))
+		return IRQ_NONE;
+
+	reg = mpam_msc_read_esr(msc);
+

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct · line 1695

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+	errcode = FIELD_GET(MPAMF_ESR_ERRCODE, reg);
+	if (!errcode)
+		return IRQ_NONE;
+
+	/* Clear level triggered irq */
+	mpam_msc_zero_esr(msc);
+
+	partid = FIELD_GET(MPAMF_ESR_PARTID_OR_MON, reg);
+	pmg = FIELD_GET(MPAMF_ESR_PMG, reg);
+	ris = FIELD_GET(MPAMF_ESR_PMG, reg);

MPAMF_ESR_RIS?

Best regards,
Shaopeng TAN

Quoted 223 lines
drivers/platform/arm64/mpam/mpam_devices.c +145 -2
+	pr_err("error irq from msc:%u '%s', partid:%u, pmg: %u, ris: %u\n",
+	       msc->id, mpam_errcode_names[errcode], partid, pmg, ris);
+
+	if (irq_is_percpu(irq)) {
+		mpam_disable_msc_ecr(msc);
+		schedule_work(&mpam_broken_work);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_WAKE_THREAD;
+}
+
+static irqreturn_t mpam_ppi_handler(int irq, void *dev_id) {
+	struct mpam_msc *msc = *(struct mpam_msc **)dev_id;
+
+	return __mpam_irq_handler(irq, msc);
+}
+
+static irqreturn_t mpam_spi_handler(int irq, void *dev_id) {
+	struct mpam_msc *msc = dev_id;
+
+	return __mpam_irq_handler(irq, msc);
+}
+
+static irqreturn_t mpam_disable_thread(int irq, void *dev_id);
+
+static int mpam_register_irqs(void)
+{
+	int err, irq, idx;
+	struct mpam_msc *msc;
+
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list,
 srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)
+			continue;
+
+		/* The MPAM spec says the interrupt can be SPI, PPI or LPI */
+		/* We anticipate sharing the interrupt with other MSCs */
+		if (irq_is_percpu(irq)) {
+			err = request_percpu_irq(irq, &mpam_ppi_handler,
+						 "mpam:msc:error",
+						 msc->error_dev_id);
+			if (err)
+				return err;
+
+			msc->reenable_error_ppi = irq;
+			smp_call_function_many(&msc->accessibility,
+					       &_enable_percpu_irq, &irq,
+					       true);
+		} else {
+			err =
 devm_request_threaded_irq(&msc->pdev->dev, irq,
+
 	&mpam_spi_handler,
+
 	&mpam_disable_thread,
+							IRQF_SHARED,
+							"mpam:msc:error",
 msc);
+			if (err)
+				return err;
+		}
+
+		msc->error_irq_requested = true;
+		mpam_touch_msc(msc, mpam_enable_msc_ecr, msc);
+		msc->error_irq_hw_enabled = true;
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
+	return 0;
+}
+
+static void mpam_unregister_irqs(void)
+{
+	int irq, idx;
+	struct mpam_msc *msc;
+
+	cpus_read_lock();
+	/* take the lock as free_irq() can sleep */
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list,
 srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)
+			continue;
+
+		if (msc->error_irq_hw_enabled) {
+			mpam_touch_msc(msc, mpam_disable_msc_ecr,
 msc);
+			msc->error_irq_hw_enabled = false;
+		}
+
+		if (msc->error_irq_requested) {
+			if (irq_is_percpu(irq)) {
+				msc->reenable_error_ppi = 0;
+				free_percpu_irq(irq, msc->error_dev_id);
+			} else {
+				devm_free_irq(&msc->pdev->dev, irq, msc);
+			}
+			msc->error_irq_requested = false;
+		}
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+	cpus_read_unlock();
+}
+
 static void mpam_enable_once(void)
 {
+	int err;
+
+	/*
+	 * If all the MSC have been probed, enabling the IRQs happens next.
+	 * That involves cross-calling to a CPU that can reach the MSC, and
+	 * the locks must be taken in this order:
+	 */
+	cpus_read_lock();
 	mutex_lock(&mpam_list_lock);
 	mpam_enable_merge_features(&mpam_classes);
+
+	err = mpam_register_irqs();
+	if (err)
+		pr_warn("Failed to register irqs: %d\n", err);
+
 	mutex_unlock(&mpam_list_lock);
+	cpus_read_unlock();
+
+	if (err) {
+		schedule_work(&mpam_broken_work);
+		return;
+	}
 
 	mutex_lock(&mpam_cpuhp_state_lock);
 	cpuhp_remove_state(mpam_cpuhp_state);
@@ -1621,16 +1893,39 @@ static void mpam_reset_class(struct mpam_class
 *class)
  * All of MPAMs errors indicate a software bug, restore any modified
  * controls to their reset values.
  */
-void mpam_disable(void)
+static irqreturn_t mpam_disable_thread(int irq, void *dev_id)
 {
 	int idx;
 	struct mpam_class *class;
+	struct mpam_msc *msc, *tmp;
+
+	mutex_lock(&mpam_cpuhp_state_lock);
+	if (mpam_cpuhp_state) {
+		cpuhp_remove_state(mpam_cpuhp_state);
+		mpam_cpuhp_state = 0;
+	}
+	mutex_unlock(&mpam_cpuhp_state_lock);
+
+	mpam_unregister_irqs();
 
 	idx = srcu_read_lock(&mpam_srcu);
 	list_for_each_entry_srcu(class, &mpam_classes, classes_list,
 				 srcu_read_lock_held(&mpam_srcu))
 		mpam_reset_class(class);
 	srcu_read_unlock(&mpam_srcu, idx);
+
+	mutex_lock(&mpam_list_lock);
+	list_for_each_entry_safe(msc, tmp, &mpam_all_msc, glbl_list)
+		mpam_msc_destroy(msc);
+	mutex_unlock(&mpam_list_lock);
+	mpam_free_garbage();
+
+	return IRQ_HANDLED;
+}
+
+void mpam_disable(struct work_struct *ignored) {
+	mpam_disable_thread(0, NULL);
 }
 
 /*
@@ -1644,7 +1939,6 @@ void mpam_enable(struct work_struct *work)
 	struct mpam_msc *msc;
 	bool all_devices_probed = true;
 
-	/* Have we probed all the hw devices? */
 	mutex_lock(&mpam_list_lock);
 	list_for_each_entry(msc, &mpam_all_msc, glbl_list) {
 		mutex_lock(&msc->probe_lock);
drivers/platform/arm64/mpam/mpam_internal.h +8 -1
index de05eece0a31..e1c6a2676b54 100644
@@ -44,6 +44,11 @@ struct mpam_msc {
 	struct pcc_mbox_chan	*pcc_chan;
 	u32			nrdy_usec;
 	cpumask_t		accessibility;
+	bool			has_extd_esr;
+
+	int				reenable_error_ppi;
+	struct mpam_msc * __percpu	*error_dev_id;
+
 	atomic_t		online_refs;
 
 	/*
@@ -52,6 +57,8 @@ struct mpam_msc {
 	 */
 	struct mutex		probe_lock;
 	bool			probed;
+	bool			error_irq_requested;
+	bool			error_irq_hw_enabled;
 	u16			partid_max;
 	u8			pmg_max;
 	unsigned long		ris_idxs[128 / BITS_PER_LONG];
@@ -280,7 +287,7 @@ extern u8 mpam_pmg_max;
 
 /* Scheduled work callback to enable mpam once all MSC have been probed
 */  void mpam_enable(struct work_struct *work); -void mpam_disable(void);
+void mpam_disable(struct work_struct *work);
 
 int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32
 cache_level,
 				   cpumask_t *affinity);
--

2.39.5

author maintainer
#84 lore raw

Re: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

Hi Shaopeng,

Quoted 35 lines · Shaopeng Tan (Fujitsu)
Quoted 35 lines

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

drivers/platform/arm64/mpam/mpam_devices.c +15 -0
index 145535cd4732..af19cc25d16e 100644
 
+static irqreturn_t __mpam_irq_handler(int irq, struct mpam_msc *msc) {
+	u64 reg;
+	u16 partid;
+	u8 errcode, pmg, ris;
+
+	if (WARN_ON_ONCE(!msc) ||
+	    WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(),
+					   &msc->accessibility)))
+		return IRQ_NONE;
+
+	reg = mpam_msc_read_esr(msc);
+
+	errcode = FIELD_GET(MPAMF_ESR_ERRCODE, reg);
+	if (!errcode)
+		return IRQ_NONE;
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1697

Quoted 7 lines
drivers/platform/arm64/mpam/mpam_devices.c +7 -0
+
+	/* Clear level triggered irq */
+	mpam_msc_zero_esr(msc);
+
+	partid = FIELD_GET(MPAMF_ESR_PARTID_OR_MON, reg);
+	pmg = FIELD_GET(MPAMF_ESR_PMG, reg);
+	ris = FIELD_GET(MPAMF_ESR_PMG, reg);

MPAMF_ESR_RIS?

Yup, thanks for catching that!

James

#85 lore raw

RE: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

In reply to James Morse

Hello James,

Quoted 242 lines

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 304
+++++++++++++++++++-
 drivers/platform/arm64/mpam/mpam_internal.h |   9 +-
 2 files changed, 307 insertions(+), 6 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +28 -3
index 145535cd4732..af19cc25d16e 100644
@@ -14,6 +14,9 @@
 #include <linux/device.h>
 #include <linux/errno.h>
 #include <linux/gfp.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/irqdesc.h>
 #include <linux/list.h>
 #include <linux/lockdep.h>
 #include <linux/mutex.h>
@@ -62,6 +65,12 @@ static DEFINE_SPINLOCK(partid_max_lock);
  */
 static DECLARE_WORK(mpam_enable_work, &mpam_enable);
 
+/*
+ * All mpam error interrupts indicate a software bug. On receipt,
+disable the
+ * driver.
+ */
+static DECLARE_WORK(mpam_broken_work, &mpam_disable);
+
 /*
  * An MSC is a physical container for controls and monitors, each identified by
  * their RIS index. These share a base-address, interrupts and some MMIO
@@ -159,6 +168,24 @@ static u64 mpam_msc_read_idr(struct mpam_msc
 *msc)
 	return (idr_high << 32) | idr_low;
 }
 
+static void mpam_msc_zero_esr(struct mpam_msc *msc) {
+	__mpam_write_reg(msc, MPAMF_ESR, 0);
+	if (msc->has_extd_esr)
+		__mpam_write_reg(msc, MPAMF_ESR + 4, 0); }
+
+static u64 mpam_msc_read_esr(struct mpam_msc *msc) {
+	u64 esr_high = 0, esr_low;
+
+	esr_low = __mpam_read_reg(msc, MPAMF_ESR);
+	if (msc->has_extd_esr)
+		esr_high = __mpam_read_reg(msc, MPAMF_ESR + 4);
+
+	return (esr_high << 32) | esr_low;
+}
+
 static void __mpam_part_sel_raw(u32 partsel, struct mpam_msc *msc)  {
 	lockdep_assert_held(&msc->part_sel_lock);
@@ -405,12 +432,12 @@ static void mpam_msc_destroy(struct mpam_msc
 *msc)
 
 	lockdep_assert_held(&mpam_list_lock);
 
-	list_del_rcu(&msc->glbl_list);
-	platform_set_drvdata(pdev, NULL);
-
 	list_for_each_entry_safe(ris, tmp, &msc->ris, msc_list)
 		mpam_ris_destroy(ris);
 
+	list_del_rcu(&msc->glbl_list);
+	platform_set_drvdata(pdev, NULL);
+
 	add_to_garbage(msc);
 	msc->garbage.pdev = pdev;

}

drivers/platform/arm64/mpam/mpam_devices.c +1 -0
@@ -828,6 +855,7 @@ static int mpam_msc_hw_probe(struct mpam_msc
 *msc)
 		pmg_max = FIELD_GET(MPAMF_IDR_PMG_MAX, idr);
 		msc->partid_max = min(msc->partid_max, partid_max);
 		msc->pmg_max = min(msc->pmg_max, pmg_max);
+		msc->has_extd_esr =
 FIELD_GET(MPAMF_IDR_HAS_EXT_ESR, idr);
 
drivers/platform/arm64/mpam/mpam_devices.c +47 -0
 		ris = mpam_get_or_create_ris(msc, ris_idx);
 		if (IS_ERR(ris))
@@ -974,6 +1002,13 @@ static void mpam_reset_msc(struct mpam_msc
 *msc, bool online)
 	mpam_mon_sel_outer_unlock(msc);
 }
 
+static void _enable_percpu_irq(void *_irq) {
+	int *irq = _irq;
+
+	enable_percpu_irq(*irq, IRQ_TYPE_NONE); }
+
 static int mpam_cpu_online(unsigned int cpu)  {
 	int idx;
@@ -984,6 +1019,9 @@ static int mpam_cpu_online(unsigned int cpu)
 		if (!cpumask_test_cpu(cpu, &msc->accessibility))
 			continue;
 
+		if (msc->reenable_error_ppi)
+			_enable_percpu_irq(&msc->reenable_error_ppi);
+
 		if (atomic_fetch_inc(&msc->online_refs) == 0)
 			mpam_reset_msc(msc, true);
 	}
@@ -1032,6 +1070,9 @@ static int mpam_cpu_offline(unsigned int cpu)
 		if (!cpumask_test_cpu(cpu, &msc->accessibility))
 			continue;
 
+		if (msc->reenable_error_ppi)
+			disable_percpu_irq(msc->reenable_error_ppi);
+
 		if (atomic_dec_and_test(&msc->online_refs))
 			mpam_reset_msc(msc, false);
 	}
@@ -1058,6 +1099,51 @@ static void mpam_register_cpuhp_callbacks(int
 (*online)(unsigned int online),
 	mutex_unlock(&mpam_cpuhp_state_lock);
 }
 
+static int __setup_ppi(struct mpam_msc *msc) {
+	int cpu;
+
+	msc->error_dev_id = alloc_percpu_gfp(struct mpam_msc *,
 GFP_KERNEL);
+	if (!msc->error_dev_id)
+		return -ENOMEM;
+
+	for_each_cpu(cpu, &msc->accessibility) {
+		struct mpam_msc *empty = *per_cpu_ptr(msc->error_dev_id,
 cpu);
+
+		if (empty) {
+			pr_err_once("%s shares PPI with %s!\n",
+				    dev_name(&msc->pdev->dev),
+				    dev_name(&empty->pdev->dev));
+			return -EBUSY;
+		}
+		*per_cpu_ptr(msc->error_dev_id, cpu) = msc;
+	}
+
+	return 0;
+}
+
+static int mpam_msc_setup_error_irq(struct mpam_msc *msc) {
+	int irq;
+
+	irq = platform_get_irq_byname_optional(msc->pdev, "error");
+	if (irq <= 0)
+		return 0;
+
+	/* Allocate and initialise the percpu device pointer for PPI */
+	if (irq_is_percpu(irq))
+		return __setup_ppi(msc);
+
+	/* sanity check: shared interrupts can be routed anywhere? */
+	if (!cpumask_equal(&msc->accessibility, cpu_possible_mask)) {
+		pr_err_once("msc:%u is a private resource with a shared error

interrupt",

drivers/platform/arm64/mpam/mpam_devices.c +11 -0
+			    msc->id);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 static int mpam_dt_count_msc(void)
 {
 	int count = 0;
@@ -1266,6 +1352,10 @@ static int mpam_msc_drv_probe(struct
 platform_device *pdev)
 			break;
 		}
 
+		err = mpam_msc_setup_error_irq(msc);
+		if (err)
+			break;
+
 		if (device_property_read_u32(&pdev->dev, "pcc-channel",
 					     &msc->pcc_subspace_id))
drivers/platform/arm64/mpam/mpam_devices.c +37 -0
 			msc->iface = MPAM_IFACE_MMIO;
@@ -1548,11 +1638,193 @@ static void mpam_enable_merge_features(struct
 list_head *all_classes_list)
 	}
 }
 
+static char *mpam_errcode_names[16] = {
+	[0] = "No error",
+	[1] = "PARTID_SEL_Range",
+	[2] = "Req_PARTID_Range",
+	[3] = "MSMONCFG_ID_RANGE",
+	[4] = "Req_PMG_Range",
+	[5] = "Monitor_Range",
+	[6] = "intPARTID_Range",
+	[7] = "Unexpected_INTERNAL",
+	[8] = "Undefined_RIS_PART_SEL",
+	[9] = "RIS_No_Control",
+	[10] = "Undefined_RIS_MON_SEL",
+	[11] = "RIS_No_Monitor",
+	[12 ... 15] = "Reserved"
+};
+
+static int mpam_enable_msc_ecr(void *_msc) {
+	struct mpam_msc *msc = _msc;
+
+	__mpam_write_reg(msc, MPAMF_ECR, 1);
+
+	return 0;
+}
+
+static int mpam_disable_msc_ecr(void *_msc) {
+	struct mpam_msc *msc = _msc;
+
+	__mpam_write_reg(msc, MPAMF_ECR, 0);
+
+	return 0;
+}
+
+static irqreturn_t __mpam_irq_handler(int irq, struct mpam_msc *msc) {
+	u64 reg;
+	u16 partid;
+	u8 errcode, pmg, ris;
+

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct · line 1688

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+	if (WARN_ON_ONCE(!msc) ||
+	    WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(),
+					   &msc->accessibility)))
+		return IRQ_NONE;
+
+	reg = mpam_msc_read_esr(msc);
+
+	errcode = FIELD_GET(MPAMF_ESR_ERRCODE, reg);
+	if (!errcode)
+		return IRQ_NONE;

In general, I think there is no problem.
However, the initial value of MPAMF_ESR_ERRCODE may not be 0 on some chips. It is better to initialize when loading the MPAM driver.

Best regards,
Shaopeng TAN

Quoted 230 lines
drivers/platform/arm64/mpam/mpam_devices.c +152 -2
+	/* Clear level triggered irq */
+	mpam_msc_zero_esr(msc);
+
+	partid = FIELD_GET(MPAMF_ESR_PARTID_OR_MON, reg);
+	pmg = FIELD_GET(MPAMF_ESR_PMG, reg);
+	ris = FIELD_GET(MPAMF_ESR_PMG, reg);
+
+	pr_err("error irq from msc:%u '%s', partid:%u, pmg: %u, ris: %u\n",
+	       msc->id, mpam_errcode_names[errcode], partid, pmg, ris);
+
+	if (irq_is_percpu(irq)) {
+		mpam_disable_msc_ecr(msc);
+		schedule_work(&mpam_broken_work);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_WAKE_THREAD;
+}
+
+static irqreturn_t mpam_ppi_handler(int irq, void *dev_id) {
+	struct mpam_msc *msc = *(struct mpam_msc **)dev_id;
+
+	return __mpam_irq_handler(irq, msc);
+}
+
+static irqreturn_t mpam_spi_handler(int irq, void *dev_id) {
+	struct mpam_msc *msc = dev_id;
+
+	return __mpam_irq_handler(irq, msc);
+}
+
+static irqreturn_t mpam_disable_thread(int irq, void *dev_id);
+
+static int mpam_register_irqs(void)
+{
+	int err, irq, idx;
+	struct mpam_msc *msc;
+
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list,
 srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)
+			continue;
+
+		/* The MPAM spec says the interrupt can be SPI, PPI or LPI */
+		/* We anticipate sharing the interrupt with other MSCs */
+		if (irq_is_percpu(irq)) {
+			err = request_percpu_irq(irq, &mpam_ppi_handler,
+						 "mpam:msc:error",
+						 msc->error_dev_id);
+			if (err)
+				return err;
+
+			msc->reenable_error_ppi = irq;
+			smp_call_function_many(&msc->accessibility,
+					       &_enable_percpu_irq, &irq,
+					       true);
+		} else {
+			err =
 devm_request_threaded_irq(&msc->pdev->dev, irq,
+
 	&mpam_spi_handler,
+
 	&mpam_disable_thread,
+							IRQF_SHARED,
+							"mpam:msc:error",
 msc);
+			if (err)
+				return err;
+		}
+
+		msc->error_irq_requested = true;
+		mpam_touch_msc(msc, mpam_enable_msc_ecr, msc);
+		msc->error_irq_hw_enabled = true;
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
+	return 0;
+}
+
+static void mpam_unregister_irqs(void)
+{
+	int irq, idx;
+	struct mpam_msc *msc;
+
+	cpus_read_lock();
+	/* take the lock as free_irq() can sleep */
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list,
 srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)
+			continue;
+
+		if (msc->error_irq_hw_enabled) {
+			mpam_touch_msc(msc, mpam_disable_msc_ecr,
 msc);
+			msc->error_irq_hw_enabled = false;
+		}
+
+		if (msc->error_irq_requested) {
+			if (irq_is_percpu(irq)) {
+				msc->reenable_error_ppi = 0;
+				free_percpu_irq(irq, msc->error_dev_id);
+			} else {
+				devm_free_irq(&msc->pdev->dev, irq, msc);
+			}
+			msc->error_irq_requested = false;
+		}
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+	cpus_read_unlock();
+}
+
 static void mpam_enable_once(void)
 {
+	int err;
+
+	/*
+	 * If all the MSC have been probed, enabling the IRQs happens next.
+	 * That involves cross-calling to a CPU that can reach the MSC, and
+	 * the locks must be taken in this order:
+	 */
+	cpus_read_lock();
 	mutex_lock(&mpam_list_lock);
 	mpam_enable_merge_features(&mpam_classes);
+
+	err = mpam_register_irqs();
+	if (err)
+		pr_warn("Failed to register irqs: %d\n", err);
+
 	mutex_unlock(&mpam_list_lock);
+	cpus_read_unlock();
+
+	if (err) {
+		schedule_work(&mpam_broken_work);
+		return;
+	}
 
 	mutex_lock(&mpam_cpuhp_state_lock);
 	cpuhp_remove_state(mpam_cpuhp_state);
@@ -1621,16 +1893,39 @@ static void mpam_reset_class(struct mpam_class
 *class)
  * All of MPAMs errors indicate a software bug, restore any modified
  * controls to their reset values.
  */
-void mpam_disable(void)
+static irqreturn_t mpam_disable_thread(int irq, void *dev_id)
 {
 	int idx;
 	struct mpam_class *class;
+	struct mpam_msc *msc, *tmp;
+
+	mutex_lock(&mpam_cpuhp_state_lock);
+	if (mpam_cpuhp_state) {
+		cpuhp_remove_state(mpam_cpuhp_state);
+		mpam_cpuhp_state = 0;
+	}
+	mutex_unlock(&mpam_cpuhp_state_lock);
+
+	mpam_unregister_irqs();
 
 	idx = srcu_read_lock(&mpam_srcu);
 	list_for_each_entry_srcu(class, &mpam_classes, classes_list,
 				 srcu_read_lock_held(&mpam_srcu))
 		mpam_reset_class(class);
 	srcu_read_unlock(&mpam_srcu, idx);
+
+	mutex_lock(&mpam_list_lock);
+	list_for_each_entry_safe(msc, tmp, &mpam_all_msc, glbl_list)
+		mpam_msc_destroy(msc);
+	mutex_unlock(&mpam_list_lock);
+	mpam_free_garbage();
+
+	return IRQ_HANDLED;
+}
+
+void mpam_disable(struct work_struct *ignored) {
+	mpam_disable_thread(0, NULL);
 }
 
 /*
@@ -1644,7 +1939,6 @@ void mpam_enable(struct work_struct *work)
 	struct mpam_msc *msc;
 	bool all_devices_probed = true;
 
-	/* Have we probed all the hw devices? */
 	mutex_lock(&mpam_list_lock);
 	list_for_each_entry(msc, &mpam_all_msc, glbl_list) {
 		mutex_lock(&msc->probe_lock);
drivers/platform/arm64/mpam/mpam_internal.h +8 -1
index de05eece0a31..e1c6a2676b54 100644
@@ -44,6 +44,11 @@ struct mpam_msc {
 	struct pcc_mbox_chan	*pcc_chan;
 	u32			nrdy_usec;
 	cpumask_t		accessibility;
+	bool			has_extd_esr;
+
+	int				reenable_error_ppi;
+	struct mpam_msc * __percpu	*error_dev_id;
+
 	atomic_t		online_refs;
 
 	/*
@@ -52,6 +57,8 @@ struct mpam_msc {
 	 */
 	struct mutex		probe_lock;
 	bool			probed;
+	bool			error_irq_requested;
+	bool			error_irq_hw_enabled;
 	u16			partid_max;
 	u8			pmg_max;
 	unsigned long		ris_idxs[128 / BITS_PER_LONG];
@@ -280,7 +287,7 @@ extern u8 mpam_pmg_max;
 
 /* Scheduled work callback to enable mpam once all MSC have been probed
 */  void mpam_enable(struct work_struct *work); -void mpam_disable(void);
+void mpam_disable(struct work_struct *work);
 
 int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32
 cache_level,
 				   cpumask_t *affinity);
--

2.39.5

author maintainer
#86 lore raw

Re: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

Hi Shaoepeng,

Quoted 29 lines · Shaopeng Tan (Fujitsu)
Quoted 29 lines

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

drivers/platform/arm64/mpam/mpam_devices.c +9 -0
index 145535cd4732..af19cc25d16e 100644
 
+static irqreturn_t __mpam_irq_handler(int irq, struct mpam_msc *msc) {
+	u64 reg;
+	u16 partid;
+	u8 errcode, pmg, ris;
+
+	if (WARN_ON_ONCE(!msc) ||
+	    WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(),
+					   &msc->accessibility)))
+		return IRQ_NONE;
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1690

Quoted 6 lines
drivers/platform/arm64/mpam/mpam_devices.c +6 -0
+
+	reg = mpam_msc_read_esr(msc);
+
+	errcode = FIELD_GET(MPAMF_ESR_ERRCODE, reg);
+	if (!errcode)
+		return IRQ_NONE;

In general, I think there is no problem. However, the initial value of MPAMF_ESR_ERRCODE may not be 0 on some chips. It is better to initialize when loading the MPAM driver

Hmm, the architecture doesn't say - but if this weren't true, you get spurious fatal errors. More interesting is if the bootloader (unlikey) or pervious kexec kernel (feasible) were using MPAM and triggered an error. We'd ned up disabling the driver when there was no bug...

I'll add something to the hw_probe() path.

Thanks,

James

#87 lore raw

Re: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

In reply to James Morse
Quoted 11 lines · On Fri, 11 Jul 2025 18:36:37 +0000 James Morse <[email protected]> wrote:

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Quoted 6 lines

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

Sparse gives an imbalance warning in mpam_register_irqs()

Quoted 11 lines
drivers/platform/arm64/mpam/mpam_devices.c +11 -0
+static int mpam_register_irqs(void)
+{
+	int err, irq, idx;
+	struct mpam_msc *msc;
+
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1747

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+			continue;
+
+		/* The MPAM spec says the interrupt can be SPI, PPI or LPI */
+		/* We anticipate sharing the interrupt with other MSCs */
+		if (irq_is_percpu(irq)) {
+			err = request_percpu_irq(irq, &mpam_ppi_handler,
+						 "mpam:msc:error",
+						 msc->error_dev_id);
+			if (err)
+				return err;

Looks like the srcu_read_lock is still held.

There is a DEFINE_LOCK_GUARD_1() in srcu.h so you can do

	guard(srcu)(&mpam_srcu, idx);

I think and not worry about releasing it in errors or the good path.

Quoted 23 lines
drivers/platform/arm64/mpam/mpam_devices.c +23 -0
+
+			msc->reenable_error_ppi = irq;
+			smp_call_function_many(&msc->accessibility,
+					       &_enable_percpu_irq, &irq,
+					       true);
+		} else {
+			err = devm_request_threaded_irq(&msc->pdev->dev, irq,
+							&mpam_spi_handler,
+							&mpam_disable_thread,
+							IRQF_SHARED,
+							"mpam:msc:error", msc);
+			if (err)
+				return err;
+		}
+
+		msc->error_irq_requested = true;
+		mpam_touch_msc(msc, mpam_enable_msc_ecr, msc);
+		msc->error_irq_hw_enabled = true;
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
+	return 0;
+}
author maintainer
#88 lore raw

Re: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

Hi Jonathan,

Quoted 33 lines · Jonathan Cameron
Quoted 8 lines · On Fri, 11 Jul 2025 18:36:37 +0000 James Morse <[email protected]> wrote:

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

Quoted 5 lines

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

Sparse gives an imbalance warning in mpam_register_irqs()

Quoted 15 lines
drivers/platform/arm64/mpam/mpam_devices.c +15 -0
+static int mpam_register_irqs(void)
+{
+	int err, irq, idx;
+	struct mpam_msc *msc;
+
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)
+			continue;
+
+		/* The MPAM spec says the interrupt can be SPI, PPI or LPI */
+		/* We anticipate sharing the interrupt with other MSCs */
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1747

Quoted 6 lines
drivers/platform/arm64/mpam/mpam_devices.c +6 -0
+		if (irq_is_percpu(irq)) {
+			err = request_percpu_irq(irq, &mpam_ppi_handler,
+						 "mpam:msc:error",
+						 msc->error_dev_id);
+			if (err)
+				return err;

Looks like the srcu_read_lock is still held.

Oops,

Quoted 5 lines

There is a DEFINE_LOCK_GUARD_1() in srcu.h so you can do

	guard(srcu)(&mpam_srcu, idx);

I think and not worry about releasing it in errors or the good path.

Sure ... but having the compiler chose when to release locks makes me nervous!

Thanks,

James

maintainer
#89 lore raw

Re: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

In reply to James Morse

Hi James,

Quoted 210 lines · James Morse

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

---
  drivers/platform/arm64/mpam/mpam_devices.c  | 304 +++++++++++++++++++-
  drivers/platform/arm64/mpam/mpam_internal.h |   9 +-
  2 files changed, 307 insertions(+), 6 deletions(-)
drivers/platform/arm64/mpam/mpam_devices.c +104 -3
index 145535cd4732..af19cc25d16e 100644
@@ -14,6 +14,9 @@
  #include <linux/device.h>
  #include <linux/errno.h>
  #include <linux/gfp.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/irqdesc.h>
  #include <linux/list.h>
  #include <linux/lockdep.h>
  #include <linux/mutex.h>
@@ -62,6 +65,12 @@ static DEFINE_SPINLOCK(partid_max_lock);
   */
  static DECLARE_WORK(mpam_enable_work, &mpam_enable);
 
+/*
+ * All mpam error interrupts indicate a software bug. On receipt, disable the
+ * driver.
+ */
+static DECLARE_WORK(mpam_broken_work, &mpam_disable);
+
  /*
   * An MSC is a physical container for controls and monitors, each identified by
   * their RIS index. These share a base-address, interrupts and some MMIO
@@ -159,6 +168,24 @@ static u64 mpam_msc_read_idr(struct mpam_msc *msc)
  	return (idr_high << 32) | idr_low;
  }
 
+static void mpam_msc_zero_esr(struct mpam_msc *msc)
+{
+	__mpam_write_reg(msc, MPAMF_ESR, 0);
+	if (msc->has_extd_esr)
+		__mpam_write_reg(msc, MPAMF_ESR + 4, 0);
+}
+
+static u64 mpam_msc_read_esr(struct mpam_msc *msc)
+{
+	u64 esr_high = 0, esr_low;
+
+	esr_low = __mpam_read_reg(msc, MPAMF_ESR);
+	if (msc->has_extd_esr)
+		esr_high = __mpam_read_reg(msc, MPAMF_ESR + 4);
+
+	return (esr_high << 32) | esr_low;
+}
+
  static void __mpam_part_sel_raw(u32 partsel, struct mpam_msc *msc)
  {
  	lockdep_assert_held(&msc->part_sel_lock);
@@ -405,12 +432,12 @@ static void mpam_msc_destroy(struct mpam_msc *msc)
 
  	lockdep_assert_held(&mpam_list_lock);
 
-	list_del_rcu(&msc->glbl_list);
-	platform_set_drvdata(pdev, NULL);
-
  	list_for_each_entry_safe(ris, tmp, &msc->ris, msc_list)
  		mpam_ris_destroy(ris);
 
+	list_del_rcu(&msc->glbl_list);
+	platform_set_drvdata(pdev, NULL);
+
  	add_to_garbage(msc);
  	msc->garbage.pdev = pdev;
  }
@@ -828,6 +855,7 @@ static int mpam_msc_hw_probe(struct mpam_msc *msc)
  		pmg_max = FIELD_GET(MPAMF_IDR_PMG_MAX, idr);
  		msc->partid_max = min(msc->partid_max, partid_max);
  		msc->pmg_max = min(msc->pmg_max, pmg_max);
+		msc->has_extd_esr = FIELD_GET(MPAMF_IDR_HAS_EXT_ESR, idr);
 
  		ris = mpam_get_or_create_ris(msc, ris_idx);
  		if (IS_ERR(ris))
@@ -974,6 +1002,13 @@ static void mpam_reset_msc(struct mpam_msc *msc, bool online)
  	mpam_mon_sel_outer_unlock(msc);
  }
 
+static void _enable_percpu_irq(void *_irq)
+{
+	int *irq = _irq;
+
+	enable_percpu_irq(*irq, IRQ_TYPE_NONE);
+}
+
  static int mpam_cpu_online(unsigned int cpu)
  {
  	int idx;
@@ -984,6 +1019,9 @@ static int mpam_cpu_online(unsigned int cpu)
  		if (!cpumask_test_cpu(cpu, &msc->accessibility))
  			continue;
 
+		if (msc->reenable_error_ppi)
+			_enable_percpu_irq(&msc->reenable_error_ppi);
+
  		if (atomic_fetch_inc(&msc->online_refs) == 0)
  			mpam_reset_msc(msc, true);
  	}
@@ -1032,6 +1070,9 @@ static int mpam_cpu_offline(unsigned int cpu)
  		if (!cpumask_test_cpu(cpu, &msc->accessibility))
  			continue;
 
+		if (msc->reenable_error_ppi)
+			disable_percpu_irq(msc->reenable_error_ppi);
+
  		if (atomic_dec_and_test(&msc->online_refs))
  			mpam_reset_msc(msc, false);
  	}
@@ -1058,6 +1099,51 @@ static void mpam_register_cpuhp_callbacks(int (*online)(unsigned int online),
  	mutex_unlock(&mpam_cpuhp_state_lock);
  }
 
+static int __setup_ppi(struct mpam_msc *msc)
+{
+	int cpu;
+
+	msc->error_dev_id = alloc_percpu_gfp(struct mpam_msc *, GFP_KERNEL);
+	if (!msc->error_dev_id)
+		return -ENOMEM;
+
+	for_each_cpu(cpu, &msc->accessibility) {
+		struct mpam_msc *empty = *per_cpu_ptr(msc->error_dev_id, cpu);
+
+		if (empty) {
+			pr_err_once("%s shares PPI with %s!\n",
+				    dev_name(&msc->pdev->dev),
+				    dev_name(&empty->pdev->dev));
+			return -EBUSY;
+		}
+		*per_cpu_ptr(msc->error_dev_id, cpu) = msc;
+	}
+
+	return 0;
+}
+
+static int mpam_msc_setup_error_irq(struct mpam_msc *msc)
+{
+	int irq;
+
+	irq = platform_get_irq_byname_optional(msc->pdev, "error");
+	if (irq <= 0)
+		return 0;
+
+	/* Allocate and initialise the percpu device pointer for PPI */
+	if (irq_is_percpu(irq))
+		return __setup_ppi(msc);
+
+	/* sanity check: shared interrupts can be routed anywhere? */
+	if (!cpumask_equal(&msc->accessibility, cpu_possible_mask)) {
+		pr_err_once("msc:%u is a private resource with a shared error interrupt",
+			    msc->id);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
  static int mpam_dt_count_msc(void)
  {
  	int count = 0;
@@ -1266,6 +1352,10 @@ static int mpam_msc_drv_probe(struct platform_device *pdev)
  			break;
  		}
 
+		err = mpam_msc_setup_error_irq(msc);
+		if (err)
+			break;
+
  		if (device_property_read_u32(&pdev->dev, "pcc-channel",
  					     &msc->pcc_subspace_id))
  			msc->iface = MPAM_IFACE_MMIO;
@@ -1548,11 +1638,193 @@ static void mpam_enable_merge_features(struct list_head *all_classes_list)
  	}
  }
 
+static char *mpam_errcode_names[16] = {
+	[0] = "No error",
+	[1] = "PARTID_SEL_Range",
+	[2] = "Req_PARTID_Range",
+	[3] = "MSMONCFG_ID_RANGE",
+	[4] = "Req_PMG_Range",
+	[5] = "Monitor_Range",
+	[6] = "intPARTID_Range",
+	[7] = "Unexpected_INTERNAL",
+	[8] = "Undefined_RIS_PART_SEL",
+	[9] = "RIS_No_Control",

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1661

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+	[10] = "Undefined_RIS_MON_SEL",
+	[11] = "RIS_No_Monitor",
+	[12 ... 15] = "Reserved"
+};
+
+static int mpam_enable_msc_ecr(void *_msc)
+{
+	struct mpam_msc *msc = _msc;
+
+	__mpam_write_reg(msc, MPAMF_ECR, 1);

You can use MPAMF_ECR_INTEN.

Quoted 211 lines
drivers/platform/arm64/mpam/mpam_devices.c +185 -1
+
+	return 0;
+}
+
+static int mpam_disable_msc_ecr(void *_msc)
+{
+	struct mpam_msc *msc = _msc;
+
+	__mpam_write_reg(msc, MPAMF_ECR, 0);
+
+	return 0;
+}
+
+static irqreturn_t __mpam_irq_handler(int irq, struct mpam_msc *msc)
+{
+	u64 reg;
+	u16 partid;
+	u8 errcode, pmg, ris;
+
+	if (WARN_ON_ONCE(!msc) ||
+	    WARN_ON_ONCE(!cpumask_test_cpu(smp_processor_id(),
+					   &msc->accessibility)))
+		return IRQ_NONE;
+
+	reg = mpam_msc_read_esr(msc);
+
+	errcode = FIELD_GET(MPAMF_ESR_ERRCODE, reg);
+	if (!errcode)
+		return IRQ_NONE;
+
+	/* Clear level triggered irq */
+	mpam_msc_zero_esr(msc);
+
+	partid = FIELD_GET(MPAMF_ESR_PARTID_OR_MON, reg);
+	pmg = FIELD_GET(MPAMF_ESR_PMG, reg);
+	ris = FIELD_GET(MPAMF_ESR_PMG, reg);
+
+	pr_err("error irq from msc:%u '%s', partid:%u, pmg: %u, ris: %u\n",
+	       msc->id, mpam_errcode_names[errcode], partid, pmg, ris);
+
+	if (irq_is_percpu(irq)) {
+		mpam_disable_msc_ecr(msc);
+		schedule_work(&mpam_broken_work);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_WAKE_THREAD;
+}
+
+static irqreturn_t mpam_ppi_handler(int irq, void *dev_id)
+{
+	struct mpam_msc *msc = *(struct mpam_msc **)dev_id;
+
+	return __mpam_irq_handler(irq, msc);
+}
+
+static irqreturn_t mpam_spi_handler(int irq, void *dev_id)
+{
+	struct mpam_msc *msc = dev_id;
+
+	return __mpam_irq_handler(irq, msc);
+}
+
+static irqreturn_t mpam_disable_thread(int irq, void *dev_id);
+
+static int mpam_register_irqs(void)
+{
+	int err, irq, idx;
+	struct mpam_msc *msc;
+
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)
+			continue;
+
+		/* The MPAM spec says the interrupt can be SPI, PPI or LPI */
+		/* We anticipate sharing the interrupt with other MSCs */
+		if (irq_is_percpu(irq)) {
+			err = request_percpu_irq(irq, &mpam_ppi_handler,
+						 "mpam:msc:error",
+						 msc->error_dev_id);
+			if (err)
+				return err;
+
+			msc->reenable_error_ppi = irq;
+			smp_call_function_many(&msc->accessibility,
+					       &_enable_percpu_irq, &irq,
+					       true);
+		} else {
+			err = devm_request_threaded_irq(&msc->pdev->dev, irq,
+							&mpam_spi_handler,
+							&mpam_disable_thread,
+							IRQF_SHARED,
+							"mpam:msc:error", msc);
+			if (err)
+				return err;
+		}
+
+		msc->error_irq_requested = true;
+		mpam_touch_msc(msc, mpam_enable_msc_ecr, msc);
+		msc->error_irq_hw_enabled = true;
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
+	return 0;
+}
+
+static void mpam_unregister_irqs(void)
+{
+	int irq, idx;
+	struct mpam_msc *msc;
+
+	cpus_read_lock();
+	/* take the lock as free_irq() can sleep */
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)
+			continue;
+
+		if (msc->error_irq_hw_enabled) {
+			mpam_touch_msc(msc, mpam_disable_msc_ecr, msc);
+			msc->error_irq_hw_enabled = false;
+		}
+
+		if (msc->error_irq_requested) {
+			if (irq_is_percpu(irq)) {
+				msc->reenable_error_ppi = 0;
+				free_percpu_irq(irq, msc->error_dev_id);
+			} else {
+				devm_free_irq(&msc->pdev->dev, irq, msc);
+			}
+			msc->error_irq_requested = false;
+		}
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+	cpus_read_unlock();
+}
+
  static void mpam_enable_once(void)
  {
+	int err;
+
+	/*
+	 * If all the MSC have been probed, enabling the IRQs happens next.
+	 * That involves cross-calling to a CPU that can reach the MSC, and
+	 * the locks must be taken in this order:
+	 */
+	cpus_read_lock();
  	mutex_lock(&mpam_list_lock);
  	mpam_enable_merge_features(&mpam_classes);
+
+	err = mpam_register_irqs();
+	if (err)
+		pr_warn("Failed to register irqs: %d\n", err);
+
  	mutex_unlock(&mpam_list_lock);
+	cpus_read_unlock();
+
+	if (err) {
+		schedule_work(&mpam_broken_work);
+		return;
+	}
 
  	mutex_lock(&mpam_cpuhp_state_lock);
  	cpuhp_remove_state(mpam_cpuhp_state);
@@ -1621,16 +1893,39 @@ static void mpam_reset_class(struct mpam_class *class)
   * All of MPAMs errors indicate a software bug, restore any modified
   * controls to their reset values.
   */
-void mpam_disable(void)
+static irqreturn_t mpam_disable_thread(int irq, void *dev_id)
  {
  	int idx;
  	struct mpam_class *class;
+	struct mpam_msc *msc, *tmp;
+
+	mutex_lock(&mpam_cpuhp_state_lock);
+	if (mpam_cpuhp_state) {
+		cpuhp_remove_state(mpam_cpuhp_state);
+		mpam_cpuhp_state = 0;
+	}
+	mutex_unlock(&mpam_cpuhp_state_lock);
+
+	mpam_unregister_irqs();
 
  	idx = srcu_read_lock(&mpam_srcu);
  	list_for_each_entry_srcu(class, &mpam_classes, classes_list,
  				 srcu_read_lock_held(&mpam_srcu))
  		mpam_reset_class(class);
  	srcu_read_unlock(&mpam_srcu, idx);
+
+	mutex_lock(&mpam_list_lock);
+	list_for_each_entry_safe(msc, tmp, &mpam_all_msc, glbl_list)
+		mpam_msc_destroy(msc);
+	mutex_unlock(&mpam_list_lock);
+	mpam_free_garbage();
+
+	return IRQ_HANDLED;
+}
+
+void mpam_disable(struct work_struct *ignored)
+{
+	mpam_disable_thread(0, NULL);
  }
 
  /*

drivers/platform/arm64/mpam/mpam_devices.c void mpam_enable(struct work_struct *work) · line 1647

Quoted 4 lines
drivers/platform/arm64/mpam/mpam_devices.c +0 -1
@@ -1644,7 +1939,6 @@ void mpam_enable(struct work_struct *work)
  	struct mpam_msc *msc;
  	bool all_devices_probed = true;
 
-	/* Have we probed all the hw devices? */

Stray change. Keep the comment or remove it in the patch that introduced it.

Quoted 37 lines
drivers/platform/arm64/mpam/mpam_devices.c +0 -0
  	mutex_lock(&mpam_list_lock);
  	list_for_each_entry(msc, &mpam_all_msc, glbl_list) {
  		mutex_lock(&msc->probe_lock);
drivers/platform/arm64/mpam/mpam_internal.h +8 -1
index de05eece0a31..e1c6a2676b54 100644
@@ -44,6 +44,11 @@ struct mpam_msc {
  	struct pcc_mbox_chan	*pcc_chan;
  	u32			nrdy_usec;
  	cpumask_t		accessibility;
+	bool			has_extd_esr;
+
+	int				reenable_error_ppi;
+	struct mpam_msc * __percpu	*error_dev_id;
+
  	atomic_t		online_refs;
 
  	/*
@@ -52,6 +57,8 @@ struct mpam_msc {
  	 */
  	struct mutex		probe_lock;
  	bool			probed;
+	bool			error_irq_requested;
+	bool			error_irq_hw_enabled;
  	u16			partid_max;
  	u8			pmg_max;
  	unsigned long		ris_idxs[128 / BITS_PER_LONG];
@@ -280,7 +287,7 @@ extern u8 mpam_pmg_max;
 
  /* Scheduled work callback to enable mpam once all MSC have been probed */
  void mpam_enable(struct work_struct *work);
-void mpam_disable(void);
+void mpam_disable(struct work_struct *work);
 
  int mpam_get_cpumask_from_cache_id(unsigned long cache_id, u32 cache_level,
  				   cpumask_t *affinity);
Signature · 3 lines
Thanks,

Ben
author maintainer
#90 lore raw

Re: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

Hi Ben,

Quoted 21 lines · Ben Horgan
Quoted 19 lines · James Morse

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

drivers/platform/arm64/mpam/ renamed from drivers/platform/arm64/mpam/mpam_devices.c +0 -0

mpam_devices.c
index 145535cd4732..af19cc25d16e 100644

drivers/platform/arm64/mpam/mpam_devices.c +0 -0
@@ -1548,11 +1638,193 @@ static void mpam_enable_merge_features(struct list_head
Quoted 6 lines

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1661

Quoted 5 lines
drivers/platform/arm64/mpam/mpam_devices.c +5 -0
+static int mpam_enable_msc_ecr(void *_msc)
+{
+    struct mpam_msc *msc = _msc;
+
+    __mpam_write_reg(msc, MPAMF_ECR, 1);

You can use MPAMF_ECR_INTEN.

Sure,

Quoted 9 lines

drivers/platform/arm64/mpam/mpam_devices.c void mpam_enable(struct work_struct *work) · line 1941

Quoted 8 lines
drivers/platform/arm64/mpam/mpam_devices.c +3 -0
+
+    return 0;
+}
 
@@ -1644,7 +1939,6 @@ void mpam_enable(struct work_struct *work)
      struct mpam_msc *msc;
      bool all_devices_probed = true;
  -    /* Have we probed all the hw devices? */

Stray change. Keep the comment or remove it in the patch that introduced it.

Fixed.

Thanks,

James

reviewer
#91 lore raw

Re: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

In reply to James Morse

Hi, James,

Quoted 17 lines · James Morse

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

[SNIP]

Quoted 11 lines
drivers/platform/arm64/mpam/mpam_devices.c +11 -0
+static int mpam_register_irqs(void)
+{
+	int err, irq, idx;
+	struct mpam_msc *msc;
+
+	lockdep_assert_cpus_held();
+
+	idx = srcu_read_lock(&mpam_srcu);
+	list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list, srcu_read_lock_held(&mpam_srcu)) {
+		irq = platform_get_irq_byname_optional(msc->pdev, "error");
+		if (irq <= 0)

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1747

Quoted 10 lines
drivers/platform/arm64/mpam/mpam_devices.c +10 -0
+			continue;
+
+		/* The MPAM spec says the interrupt can be SPI, PPI or LPI */
+		/* We anticipate sharing the interrupt with other MSCs */
+		if (irq_is_percpu(irq)) {
+			err = request_percpu_irq(irq, &mpam_ppi_handler,
+						 "mpam:msc:error",
+						 msc->error_dev_id);
+			if (err)
+				return err;

But right now mpam_srcu is still being locked. Need to unlock it before return.

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1760

Quoted 13 lines
drivers/platform/arm64/mpam/mpam_devices.c +13 -0
+
+			msc->reenable_error_ppi = irq;
+			smp_call_function_many(&msc->accessibility,
+					       &_enable_percpu_irq, &irq,
+					       true);
+		} else {
+			err = devm_request_threaded_irq(&msc->pdev->dev, irq,
+							&mpam_spi_handler,
+							&mpam_disable_thread,
+							IRQF_SHARED,
+							"mpam:msc:error", msc);
+			if (err)
+				return err;

Ditto.

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1771

Quoted 11 lines
drivers/platform/arm64/mpam/mpam_devices.c +11 -0
+		}
+
+		msc->error_irq_requested = true;
+		mpam_touch_msc(msc, mpam_enable_msc_ecr, msc);
+		msc->error_irq_hw_enabled = true;
+	}
+	srcu_read_unlock(&mpam_srcu, idx);
+
+	return 0;
+}
+

[SNIP]

Thanks.

-Fenghua

author maintainer
#92 lore raw

Re: [RFC PATCH 25/36] arm_mpam: Register and enable IRQs

Hi Fenghua,

Quoted 30 lines · Fenghua Yu
Quoted 29 lines · James Morse

Register and enable error IRQs. All the MPAM error interrupts indicate a software bug, e.g. out of range partid. If the error interrupt is ever signalled, attempt to disable MPAM.

Only the irq handler accesses the ESR register, so no locking is needed. The work to disable MPAM after an error needs to happen at process context, use a threaded interrupt.

There is no support for percpu threaded interrupts, for now schedule the work to be done from the irq handler.

Enabling the IRQs in the MSC may involve cross calling to a CPU that can access the MSC.

drivers/platform/arm64/mpam/mpam_devices.c +9 -0
+static int mpam_register_irqs(void)
+{
+    int err, irq, idx;
+    struct mpam_msc *msc;
+
+    lockdep_assert_cpus_held();
+
+    idx = srcu_read_lock(&mpam_srcu);
+    list_for_each_entry_srcu(msc, &mpam_all_msc, glbl_list,

srcu_read_lock_held(&mpam_srcu)) {

drivers/platform/arm64/mpam/mpam_devices.c +5 -0
+        irq = platform_get_irq_byname_optional(msc->pdev, "error");
+        if (irq <= 0)
+            continue;
+
+        /* The MPAM spec says the interrupt can be SPI, PPI or LPI */
Quoted 8 lines

drivers/platform/arm64/mpam/mpam_devices.c static void mpam_enable_merge_features(struct list_head *all_classes_list) · line 1747

Quoted 7 lines
drivers/platform/arm64/mpam/mpam_devices.c +7 -0
+        /* We anticipate sharing the interrupt with other MSCs */
+        if (irq_is_percpu(irq)) {
+            err = request_percpu_irq(irq, &mpam_ppi_handler,
+                         "mpam:msc:error",
+                         msc->error_dev_id);
+            if (err)
+                return err;

But right now mpam_srcu is still being locked. Need to unlock it before return.

Yup, Jonathan's srcu guard runes solve that in a future proof way.

Thanks,

James

author maintainer
#93 lore raw

RFC PATCH 26/36 arm_mpam: Use a static key to indicate when mpam is enabled

In reply to James Morse

Once all the MSC have been probed, the system wide usable number of PARTID is known and the configuration arrays can be allocated.

After this point, checking all the MSC have been probed is pointless, and the cpuhp callbacks should restore the configuration, instead of just resetting the MSC.

Add a static key to enable this behaviour. This will also allow MPAM to be disabled in repsonse to an error, and the architecture code to enable/disable the context switch of the MPAM system registers.

---
 drivers/platform/arm64/mpam/mpam_devices.c  | 8 ++++++++
 drivers/platform/arm64/mpam/mpam_internal.h | 8 ++++++++
 2 files changed, 16 insertions(+)
drivers/platform/arm64/mpam/mpam_devices.c +8 -0
index af19cc25d16e..bb3695eb84e9 100644
@@ -33,6 +33,8 @@
 
 #include "mpam_internal.h"
 
+DEFINE_STATIC_KEY_FALSE(mpam_enabled); /* TODO: move to arch code */
+
 /*
  * mpam_list_lock protects the SRCU lists when writing. Once the
  * mpam_enabled key is enabled these lists are read-only,
@@ -1037,6 +1039,9 @@ static int mpam_discovery_cpu_online(unsigned int cpu)
 	struct mpam_msc *msc;
 	bool new_device_probed = false;
 
+	if (mpam_is_enabled())
+		return 0;
+
 	mutex_lock(&mpam_list_lock);
 	list_for_each_entry(msc, &mpam_all_msc, glbl_list) {
 		if (!cpumask_test_cpu(cpu, &msc->accessibility))
@@ -1839,6 +1844,7 @@ static void mpam_enable_once(void)
 	partid_max_published = true;
 	spin_unlock(&partid_max_lock);
 
+	static_branch_enable(&mpam_enabled);
 	mpam_register_cpuhp_callbacks(mpam_cpu_online, mpam_cpu_offline);
 
 	printk(KERN_INFO "MPAM enabled with %u partid and %u pmg\n",
@@ -1906,6 +1912,8 @@ static irqreturn_t mpam_disable_thread(int irq, void *dev_id)
 	}
 	mutex_unlock(&mpam_cpuhp_state_lock);
 
+	static_branch_disable(&mpam_enabled);
+
 	mpam_unregister_irqs();
 
 	idx = srcu_read_lock(&mpam_srcu);
drivers/platform/arm64/mpam/mpam_internal.h +8 -0
index e1c6a2676b54..1a24424b48df 100644
@@ -8,6 +8,7 @@
 #include <linux/atomic.h>
 #include <linux/cpumask.h>
 #include <linux/io.h>
+#include <linux/jump_label.h>
 #include <linux/llist.h>
 #include <linux/mailbox_client.h>
 #include <linux/mutex.h>
@@ -15,6 +16,13 @@
 #include <linux/sizes.h>
 #include <linux/srcu.h>
 
+DECLARE_STATIC_KEY_FALSE(mpam_enabled);
+
+static inline bool mpam_is_enabled(void)
+{
+	return static_branch_likely(&mpam_enabled);
+}
+
 /*
  * Structures protected by SRCU may not be freed for a surprising amount of
  * time (especially if perf is running). To ensure the MPAM error interrupt can
Signature · 1 line
2.39.5