Skip to contentSkip to conversation

arm_mpam: resctrl: Fixes for v7.1-rc1

PATCH RESEND In review James Morse 2 participants 5 messages Apr 20 MPAM driver

lore
Share & tools
lore

James Morse

author maintainer
#0 lore raw

cover arm_mpam: resctrl: Fixes for v7.1-rc1

Hi Catalin,

This series is the collected fixes for MPAM series in arm64/for-next/core.

These fix a unmount->remount problem with the CDP emulation, an uninitialised variable and a checker warning.

If its easier, there is a branch here:
https://git.kernel.org/pub/scm/linux/kernel/git/morse/linux.git mpam/glue/v7.0-rc3

Thanks,

James

Ben Horgan (2):
arm_mpam: resctrl: Fix the check for no monitor components found
arm_mpam: resctrl: Make resctrl_mon_ctx_waiters static

Zeng Heng (1):
arm_mpam: resctrl: Fix MBA CDP alloc_capable handling on unmount

 drivers/resctrl/mpam_resctrl.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)
Signature · 1 line
2.53.0
author maintainer
#1 lore raw

PATCH RESEND 1/3 arm_mpam: resctrl: Fix MBA CDP alloc_capable handling on unmount

From: Zeng Heng <[email protected]>

The code to set MBA's alloc_capable to true appears to be trying to restore alloc_capable on unmount. This can never work because resctrl_arch_set_cdp_enabled() is never invoked with RDT_RESOURCE_MBA as the rid parameter. Consequently, mpam_resctrl_controls[RDT_RESOURCE_MBA].cdp_enabled always remains false.

The alloc_capable setting in resctrl_arch_set_cdp_enabled() is to re-enable MBA if the caller opts in to separate control values using CDP for this resource. This doesn't happen today.

Add a comment to describe this.

However a bug remains where MBA allocation is permanently disabled after the mount with CDP option. Remounting without CDP cannot restore the MBA partition capability.

Add a check to re-enable MBA when CDP is disabled, which happens on unmount.

Fixes 6789fb99282c ("arm_mpam: resctrl: Add CDP emulation")

[ morse: Added comment for existing code, added hunk to fix this bug from
Ben H ]

---
 drivers/resctrl/mpam_resctrl.c | 8 ++++++++
 1 file changed, 8 insertions(+)
drivers/resctrl/mpam_resctrl.c +8 -0
index a9938006d0e6..4205fb2ee312 100644
@@ -220,10 +220,18 @@ int resctrl_arch_set_cdp_enabled(enum resctrl_res_level rid, bool enable)
 	if (cdp_enabled && !mpam_resctrl_controls[RDT_RESOURCE_MBA].cdp_enabled)
 		mpam_resctrl_controls[RDT_RESOURCE_MBA].resctrl_res.alloc_capable = false;
 
+	/*
+	 * If resctrl has attempted to enable CDP on MBA, re-enable MBA as two
+	 * configurations will be provided so there is no aliasing problem.
+	 */
 	if (mpam_resctrl_controls[RDT_RESOURCE_MBA].cdp_enabled &&
 	    mpam_resctrl_controls[RDT_RESOURCE_MBA].class)
 		mpam_resctrl_controls[RDT_RESOURCE_MBA].resctrl_res.alloc_capable = true;
 
+	/* On unmount when CDP is disabled, re-enable MBA */
+	if (!cdp_enabled && mpam_resctrl_controls[RDT_RESOURCE_MBA].class)
+		mpam_resctrl_controls[RDT_RESOURCE_MBA].resctrl_res.alloc_capable = true;
+
 	if (enable) {
 		if (mpam_partid_max < 1)
 			return -EINVAL;
Signature · 1 line
2.53.0
author maintainer
#2 lore raw

PATCH RESEND 2/3 arm_mpam: resctrl: Fix the check for no monitor components found

In reply to James Morse

From: Ben Horgan <[email protected]>

Dan Carpenter reports that, in mpam_resctrl_alloc_domain(), any_mon_comp is used in an 'if' condition when it may be uninitialized. Initialize it to NULL so that the check behaves correctly when no monitor components are found.

Fixes 264c285999fc ("arm_mpam: resctrl: Add monitor initialisation and domain boilerplate")
---
 drivers/resctrl/mpam_resctrl.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/resctrl/mpam_resctrl.c +1 -1
index 4205fb2ee312..1b0b37da12af 100644
@@ -1407,7 +1407,7 @@ mpam_resctrl_alloc_domain(unsigned int cpu, struct mpam_resctrl_res *res)
 	}
 
 	if (r->mon_capable) {
-		struct mpam_component *any_mon_comp;
+		struct mpam_component *any_mon_comp = NULL;
 		struct mpam_resctrl_mon *mon;
 		enum resctrl_event_id eventid;
 
Signature · 1 line
2.53.0
author maintainer
#3 lore raw

PATCH RESEND 3/3 arm_mpam: resctrl: Make resctrl_mon_ctx_waiters static

In reply to James Morse

From: Ben Horgan <[email protected]>

resctrl_mon_ctx_waiters is not used outside of this file, so make it static. This fixes the sparse warning:

drivers/resctrl/mpam_resctrl.c:25:1: warning: symbol 'resctrl_mon_ctx_waiters' was not declared. Should it be static?

Fixes 2a3c79c61539 ("arm_mpam: resctrl: Allow resctrl to allocate monitors")
---
 drivers/resctrl/mpam_resctrl.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/resctrl/mpam_resctrl.c +1 -1
index 1b0b37da12af..226ff6f532fa 100644
@@ -22,7 +22,7 @@
 
 #include "mpam_internal.h"
 
-DECLARE_WAIT_QUEUE_HEAD(resctrl_mon_ctx_waiters);
+static DECLARE_WAIT_QUEUE_HEAD(resctrl_mon_ctx_waiters);
 
 /*
  * The classes we've picked to map to resctrl resources, wrapped
Signature · 1 line
2.53.0
#4 lore raw
In reply to James Morse

Hi James,

Quoted 7 lines · James Morse

This series is the collected fixes for MPAM series in arm64/for-next/core.

These fix a unmount->remount problem with the CDP emulation, an uninitialised variable and a checker warning.

If its easier, there is a branch here:
https://git.kernel.org/pub/scm/linux/kernel/git/morse/linux.git mpam/glue/v7.0-rc3

I pulled the mpam/glue/v7.0-rc3+fixes into arm64 for-next/core.

Thanks!

Signature · 1 line
Catalin