Skip to contentSkip to conversation

ACPI: processor driver vs. core

PATCH Superseded Alex Chiang 3 participants 17 messages Jan 25, 2010 – Feb 3, 2010

Subsystems (4)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

cover ACPI: processor driver vs. core

This series cleans up some of the mess I made when introducing early _PDC.

The major change is renaming processor_core.c to processor_driver.c, and then renaming processor_pdc.c to processor_core.c.

The idea is that the code in processor_core.c will always be built statically into the kernel (as long as ACPI is configured), while allowing the ACPI processor driver to remain modular (if so desired).

We do this because part of the cleanups involves teaching the early _PDC evaluation code how to determine if a processor is physically present or not -- aka enumeration -- and that sort of code doesn't really belong in a file named processor_pdc.

There are quite a few checkpatch errors in the first rename patch, and I'll look at cleaning those up in a later series.

/ac

---

Alex Chiang (12):
ACPI: processor: mv processor_core.c processor_driver.c
ACPI: processor: mv processor_pdc.c processor_core.c
ACPI: processor: export acpi_get_cpuid()
ACPI: processor: move acpi_get_cpuid into processor_core.c
ACPI: processor: add internal processor_physically_present()
ACPI: processor: remove early _PDC optin quirks
ACPI: processor: driver doesn't need to evaluate _PDC
ACPI: processor: refactor internal map_lapic_id()
ACPI: processor: refactor internal map_x2apic_id()
ACPI: processor: refactor internal map_lsapic_id()
ACPI: processor: push file static MADT pointer into internal map_madt_entry()
ACPI: processor core: style and sparse cleanups

 Documentation/kernel-parameters.txt |    4 
 arch/ia64/kernel/acpi.c             |    3 
 arch/x86/kernel/acpi/boot.c         |    3 
 drivers/acpi/Makefile               |    4 
 drivers/acpi/processor_core.c       | 1149 +++++------------------------------
 drivers/acpi/processor_driver.c     |  976 ++++++++++++++++++++++++++++++
 drivers/acpi/processor_pdc.c        |  209 ------
 include/acpi/processor.h            |   10 
 8 files changed, 1172 insertions(+), 1186 deletions(-)
 create mode 100644 drivers/acpi/processor_driver.c
 delete mode 100644 drivers/acpi/processor_pdc.c
#1 lore raw

PATCH 1/12 ACPI: processor: mv processor_core.c processor_driver.c

The ACPI processor driver can be built as a module. But it has pieces of code that should always be built statically into the kernel.

The plan is for processor_core.c to contain the static bits while processor_driver.c contains the module-like bits.

Since the bulk of the code in the current processor_core.c is module-like, first step is to rename the file to processor_driver.c

Next step will re-create processor_core.c and cherry-pick out the static bits.

Cc Venkatesh Pallipadi <[email protected]>

---

 drivers/acpi/Makefile           |    2 
 drivers/acpi/processor_core.c   | 1140 ---------------------------------------
 drivers/acpi/processor_driver.c | 1140 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 1141 insertions(+), 1141 deletions(-)
 delete mode 100644 drivers/acpi/processor_core.c
 create mode 100644 drivers/acpi/processor_driver.c
drivers/acpi/Makefile +1 -1
index 66cc3f3..6b363a5 100644
@@ -61,7 +61,7 @@ obj-$(CONFIG_ACPI_SBS)		+= sbs.o
 obj-$(CONFIG_ACPI_POWER_METER)	+= power_meter.o
 
 # processor has its own "processor." module_param namespace
-processor-y			:= processor_core.o processor_throttling.o
+processor-y			:= processor_driver.o processor_throttling.o
 processor-y			+= processor_idle.o processor_thermal.o
 processor-$(CONFIG_CPU_FREQ)	+= processor_perflib.o
 
drivers/acpi/processor_core.c deleted +0 -1140

1,140 lines not on this page · raw message

drivers/acpi/processor_driver.c new file +1140 -0

1,140 lines not on this page · raw message

#2 lore raw

PATCH 2/12 ACPI: processor: mv processor_pdc.c processor_core.c

In reply to Alex Chiang

We've renamed the old processor_core.c to processor_driver.c, to convey the idea that it can be built modular and has driver-like bits.

Now let's re-create a processor_core.c for the bits needed statically by the rest of the kernel. The contents of processor_pdc.c are a good starting spot, so let's just rename that file and complete our three card monte.

Cc Venkatesh Pallipadi <[email protected]>

---

 drivers/acpi/Makefile         |    2 
 drivers/acpi/processor_core.c |  209 +++++++++++++++++++++++++++++++++++++++++
 drivers/acpi/processor_pdc.c  |  209 -----------------------------------------
 include/acpi/processor.h      |    2 
 4 files changed, 211 insertions(+), 211 deletions(-)
 create mode 100644 drivers/acpi/processor_core.c
 delete mode 100644 drivers/acpi/processor_pdc.c
drivers/acpi/Makefile +1 -1
index 6b363a5..a8d8998 100644
@@ -32,7 +32,7 @@ acpi-$(CONFIG_ACPI_SLEEP)	+= proc.o
 #
 acpi-y				+= bus.o glue.o
 acpi-y				+= scan.o
-acpi-y				+= processor_pdc.o
+acpi-y				+= processor_core.o
 acpi-y				+= ec.o
 acpi-$(CONFIG_ACPI_DOCK)	+= dock.o
 acpi-y				+= pci_root.o pci_link.o pci_irq.o pci_bind.o
drivers/acpi/processor_core.c new file +209 -0

209 lines not on this page · raw message

drivers/acpi/processor_pdc.c deleted +0 -209

209 lines not on this page · raw message

include/acpi/processor.h +1 -1
index 0ea5ef4..d767006 100644
@@ -320,7 +320,7 @@ static inline int acpi_processor_get_bios_limit(int cpu, unsigned int *limit)
 
 #endif				/* CONFIG_CPU_FREQ */
 
-/* in processor_pdc.c */
+/* in processor_core.c */
 void acpi_processor_set_pdc(acpi_handle handle);
 
 /* in processor_throttling.c */
#3 lore raw

PATCH 3/12 ACPI: processor: export acpi_get_cpuid()

In reply to Alex Chiang

Rename static get_cpu_id() to acpi_get_cpuid() and export it.

This change also gives us an opportunity to remove the #ifndef CONFIG_SMP from processor_driver.c and into a header file where it properly belongs.

Cc Venkatesh Pallipadi <[email protected]>

---

 drivers/acpi/processor_driver.c |   10 ++++------
 include/acpi/processor.h        |    8 ++++++++
 2 files changed, 12 insertions(+), 6 deletions(-)
drivers/acpi/processor_driver.c +4 -6
index d2f285b..404fab4 100644
@@ -359,10 +359,7 @@ static inline int acpi_processor_remove_fs(struct acpi_device *device)
 
 /* Use the acpiid in MADT to map cpus in case of SMP */
 
-#ifndef CONFIG_SMP
-static int get_cpu_id(acpi_handle handle, int type, u32 acpi_id) { return -1; }
-#else
-
+#ifdef CONFIG_SMP
 static struct acpi_table_madt *madt;
 
 static int map_lapic_id(struct acpi_subtable_header *entry,
@@ -494,7 +491,7 @@ exit:
 	return apic_id;
 }
 
-static int get_cpu_id(acpi_handle handle, int type, u32 acpi_id)
+int acpi_get_cpuid(acpi_handle handle, int type, u32 acpi_id)
 {
 	int i;
 	int apic_id = -1;
@@ -511,6 +508,7 @@ static int get_cpu_id(acpi_handle handle, int type, u32 acpi_id)
 	}
 	return -1;
 }
+EXPORT_SYMBOL_GPL(acpi_get_cpuid);
 #endif
 
 /* --------------------------------------------------------------------------
@@ -577,7 +575,7 @@ static int acpi_processor_get_info(struct acpi_device *device)
 		device_declaration = 1;
 		pr->acpi_id = value;
 	}
-	cpu_index = get_cpu_id(pr->handle, device_declaration, pr->acpi_id);
+	cpu_index = acpi_get_cpuid(pr->handle, device_declaration, pr->acpi_id);
 
 	/* Handle UP system running SMP kernel, with no LAPIC in MADT */
 	if (!cpu0_initialized && (cpu_index == -1) &&
include/acpi/processor.h +8 -0
index d767006..3b66978 100644
@@ -322,6 +322,14 @@ static inline int acpi_processor_get_bios_limit(int cpu, unsigned int *limit)
 
 /* in processor_core.c */
 void acpi_processor_set_pdc(acpi_handle handle);
+#ifdef CONFIG_SMP
+int acpi_get_cpuid(acpi_handle, int type, u32 acpi_id);
+#else
+static inline int acpi_get_cpuid(acpi_handle handle, int type, u32 acpi_id)
+{
+	return -1;
+}
+#endif
 
 /* in processor_throttling.c */
 int acpi_processor_tstate_has_changed(struct acpi_processor *pr);
#4 lore raw

PATCH 4/12 ACPI: processor: move acpi_get_cpuid into processor_core.c

In reply to Alex Chiang

Enumerating processors (via MADT/_MAT) belongs in the processor core, which is always built-in, rather than living in the processor driver which may not be built.

Cc Venkatesh Pallipadi <[email protected]>

---

 drivers/acpi/processor_core.c   |  160 +++++++++++++++++++++++++++++++++++++++
 drivers/acpi/processor_driver.c |  159 ---------------------------------------
 2 files changed, 160 insertions(+), 159 deletions(-)
drivers/acpi/processor_core.c +160 -0
index 598bb7f..ca12d70 100644
@@ -42,6 +42,159 @@ static struct dmi_system_id __cpuinitdata processor_idle_dmi_table[] = {
 	{},
 };
 
+#ifdef CONFIG_SMP
+static struct acpi_table_madt *madt;
+
+static int map_lapic_id(struct acpi_subtable_header *entry,
+		 u32 acpi_id, int *apic_id)
+{
+	struct acpi_madt_local_apic *lapic =
+		(struct acpi_madt_local_apic *)entry;
+	if ((lapic->lapic_flags & ACPI_MADT_ENABLED) &&
+	    lapic->processor_id == acpi_id) {
+		*apic_id = lapic->id;
+		return 1;
+	}
+	return 0;
+}
+
+static int map_x2apic_id(struct acpi_subtable_header *entry,
+			 int device_declaration, u32 acpi_id, int *apic_id)
+{
+	struct acpi_madt_local_x2apic *apic =
+		(struct acpi_madt_local_x2apic *)entry;
+	u32 tmp = apic->local_apic_id;
+
+	/* Only check enabled APICs*/
+	if (!(apic->lapic_flags & ACPI_MADT_ENABLED))
+		return 0;
+
+	/* Device statement declaration type */
+	if (device_declaration) {
+		if (apic->uid == acpi_id)
+			goto found;
+	}
+
+	return 0;
+found:
+	*apic_id = tmp;
+	return 1;
+}
+
+static int map_lsapic_id(struct acpi_subtable_header *entry,
+		int device_declaration, u32 acpi_id, int *apic_id)
+{
+	struct acpi_madt_local_sapic *lsapic =
+		(struct acpi_madt_local_sapic *)entry;
+	u32 tmp = (lsapic->id << 8) | lsapic->eid;
+
+	/* Only check enabled APICs*/
+	if (!(lsapic->lapic_flags & ACPI_MADT_ENABLED))
+		return 0;
+
+	/* Device statement declaration type */
+	if (device_declaration) {
+		if (entry->length < 16)
+			printk(KERN_ERR PREFIX
+			    "Invalid LSAPIC with Device type processor (SAPIC ID %#x)\n",
+			    tmp);
+		else if (lsapic->uid == acpi_id)
+			goto found;
+	/* Processor statement declaration type */
+	} else if (lsapic->processor_id == acpi_id)
+		goto found;
+
+	return 0;
+found:
+	*apic_id = tmp;
+	return 1;
+}
+
+static int map_madt_entry(int type, u32 acpi_id)
+{
+	unsigned long madt_end, entry;
+	int apic_id = -1;
+
+	if (!madt)
+		return apic_id;
+
+	entry = (unsigned long)madt;
+	madt_end = entry + madt->header.length;
+
+	/* Parse all entries looking for a match. */
+
+	entry += sizeof(struct acpi_table_madt);
+	while (entry + sizeof(struct acpi_subtable_header) < madt_end) {
+		struct acpi_subtable_header *header =
+			(struct acpi_subtable_header *)entry;
+		if (header->type == ACPI_MADT_TYPE_LOCAL_APIC) {
+			if (map_lapic_id(header, acpi_id, &apic_id))
+				break;
+		} else if (header->type == ACPI_MADT_TYPE_LOCAL_X2APIC) {
+			if (map_x2apic_id(header, type, acpi_id, &apic_id))
+				break;
+		} else if (header->type == ACPI_MADT_TYPE_LOCAL_SAPIC) {
+			if (map_lsapic_id(header, type, acpi_id, &apic_id))
+				break;
+		}
+		entry += header->length;
+	}
+	return apic_id;
+}
+
+static int map_mat_entry(acpi_handle handle, int type, u32 acpi_id)
+{
+	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
+	union acpi_object *obj;
+	struct acpi_subtable_header *header;
+	int apic_id = -1;
+
+	if (ACPI_FAILURE(acpi_evaluate_object(handle, "_MAT", NULL, &buffer)))
+		goto exit;
+
+	if (!buffer.length || !buffer.pointer)
+		goto exit;
+
+	obj = buffer.pointer;
+	if (obj->type != ACPI_TYPE_BUFFER ||
+	    obj->buffer.length < sizeof(struct acpi_subtable_header)) {
+		goto exit;
+	}
+
+	header = (struct acpi_subtable_header *)obj->buffer.pointer;
+	if (header->type == ACPI_MADT_TYPE_LOCAL_APIC) {
+		map_lapic_id(header, acpi_id, &apic_id);
+	} else if (header->type == ACPI_MADT_TYPE_LOCAL_SAPIC) {
+		map_lsapic_id(header, type, acpi_id, &apic_id);
+	}
+
+exit:
+	if (buffer.pointer)
+		kfree(buffer.pointer);
+	return apic_id;
+}
+
+int acpi_get_cpuid(acpi_handle handle, int type, u32 acpi_id)
+{
+	int i;
+	int apic_id = -1;
+
+	apic_id = map_mat_entry(handle, type, acpi_id);
+	if (apic_id == -1)
+		apic_id = map_madt_entry(type, acpi_id);
+	if (apic_id == -1)
+		return apic_id;
+
+	for_each_possible_cpu(i) {
+		if (cpu_physical_id(i) == apic_id)
+			return i;
+	}
+	return -1;
+}
+EXPORT_SYMBOL_GPL(acpi_get_cpuid);
+#endif
+
+
 static void acpi_set_pdc_bits(u32 *buf)
 {
 	buf[0] = ACPI_PDC_REVISION_ID;
@@ -188,6 +341,13 @@ early_init_pdc(acpi_handle handle, u32 lvl, void *context, void **rv)
 
 void acpi_early_processor_set_pdc(void)
 {
+
+#ifdef CONFIG_SMP
+	if (ACPI_FAILURE(acpi_get_table(ACPI_SIG_MADT, 0,
+				(struct acpi_table_header **)&madt)))
+		madt = NULL;
+#endif
+
 	/*
 	 * Check whether the system is DMI table. If yes, OSPM
 	 * should not use mwait for CPU-states.
drivers/acpi/processor_driver.c +0 -159

171 lines not on this page · raw message

#5 lore raw

PATCH 5/12 ACPI: processor: add internal processor_physically_present()

In reply to Alex Chiang

Detect if a processor is physically present before evaluating _PDC.

We want this because some BIOS will provide a _PDC even for processors that are not present. These bogus _PDC methods then attempt to load non-existent tables, which causes problems.

Avoid those bogus landmines.

---

 drivers/acpi/processor_core.c |   42 +++++++++++++++++++++++++++++++++++++++++
 1 files changed, 42 insertions(+), 0 deletions(-)
drivers/acpi/processor_core.c +42 -0
index ca12d70..e8b1246 100644
@@ -194,6 +194,45 @@ int acpi_get_cpuid(acpi_handle handle, int type, u32 acpi_id)
 EXPORT_SYMBOL_GPL(acpi_get_cpuid);
 #endif
 
+static bool processor_physically_present(acpi_handle handle)
+{
+	int cpuid, type;
+	u32 acpi_id;
+	acpi_status status;
+	acpi_object_type acpi_type;
+	unsigned long long tmp;
+	union acpi_object object = { 0 };
+	struct acpi_buffer buffer = { sizeof(union acpi_object), &object };
+
+	status = acpi_get_type(handle, &acpi_type);
+	if (ACPI_FAILURE(status))
+		return false;
+
+	switch (acpi_type) {
+	case ACPI_TYPE_PROCESSOR:
+		status = acpi_evaluate_object(handle, NULL, NULL, &buffer);
+		if (ACPI_FAILURE(status))
+			return false;
+		acpi_id = object.processor.proc_id;
+		break;
+	case ACPI_TYPE_DEVICE:
+		status = acpi_evaluate_integer(handle, "_UID", NULL, &tmp);
+		if (ACPI_FAILURE(status))
+			return false;
+		acpi_id = tmp;
+		break;
+	default:
+		return false;
+	}
+
+	type = (acpi_type == ACPI_TYPE_DEVICE) ? 1 : 0;
+	cpuid = acpi_get_cpuid(handle, type, acpi_id);
+
+	if (cpuid == -1)
+		return false;
+
+	return true;
+}
 
 static void acpi_set_pdc_bits(u32 *buf)
 {
@@ -335,6 +374,9 @@ static struct dmi_system_id __cpuinitdata early_pdc_optin_table[] = {
 static acpi_status
 early_init_pdc(acpi_handle handle, u32 lvl, void *context, void **rv)
 {
+	if (processor_physically_present(handle) == false)
+		return AE_OK;
+
 	acpi_processor_set_pdc(handle);
 	return AE_OK;
 }
#6 lore raw

PATCH 6/12 ACPI: processor: remove early _PDC optin quirks

In reply to Alex Chiang

Now that we check for physically present processors before blindly evaluating _PDC, we no longer need to maintain a DMI opt-in table nor a kernel param.

Cc Venkatesh Pallipadi <[email protected]>

---

 Documentation/kernel-parameters.txt |    4 ----
 drivers/acpi/processor_core.c       |   37 -----------------------------------
 2 files changed, 0 insertions(+), 41 deletions(-)
Documentation/kernel-parameters.txt +0 -4
index 826b6e1..736d456 100644
@@ -199,10 +199,6 @@ and is between 256 and 4096 characters. It is defined in the file
 			acpi_display_output=video
 			See above.
 
-	acpi_early_pdc_eval	[HW,ACPI] Evaluate processor _PDC methods
-				early. Needed on some platforms to properly
-				initialize the EC.
-
 	acpi_irq_balance [HW,ACPI]
 			ACPI will balance active IRQs
 			default in APIC mode
drivers/acpi/processor_core.c +0 -37
index e8b1246..05874ba 100644
@@ -341,36 +341,6 @@ void acpi_processor_set_pdc(acpi_handle handle)
 }
 EXPORT_SYMBOL_GPL(acpi_processor_set_pdc);
 
-static int early_pdc_optin;
-static int set_early_pdc_optin(const struct dmi_system_id *id)
-{
-	early_pdc_optin = 1;
-	return 0;
-}
-
-static int param_early_pdc_optin(char *s)
-{
-	early_pdc_optin = 1;
-	return 1;
-}
-__setup("acpi_early_pdc_eval", param_early_pdc_optin);
-
-static struct dmi_system_id __cpuinitdata early_pdc_optin_table[] = {
-	{
-	set_early_pdc_optin, "HP Envy", {
-	DMI_MATCH(DMI_BIOS_VENDOR, "Hewlett-Packard"),
-	DMI_MATCH(DMI_PRODUCT_NAME, "HP Envy") }, NULL},
-	{
-	set_early_pdc_optin, "HP Pavilion dv6", {
-	DMI_MATCH(DMI_BIOS_VENDOR, "Hewlett-Packard"),
-	DMI_MATCH(DMI_PRODUCT_NAME, "HP Pavilion dv6") }, NULL},
-	{
-	set_early_pdc_optin, "HP Pavilion dv7", {
-	DMI_MATCH(DMI_BIOS_VENDOR, "Hewlett-Packard"),
-	DMI_MATCH(DMI_PRODUCT_NAME, "HP Pavilion dv7") }, NULL},
-	{},
-};
-
 static acpi_status
 early_init_pdc(acpi_handle handle, u32 lvl, void *context, void **rv)
 {
@@ -396,13 +366,6 @@ void acpi_early_processor_set_pdc(void)
 	 */
 	dmi_check_system(processor_idle_dmi_table);
 
-	/*
-	 * Allow systems to opt-in to early _PDC evaluation.
-	 */
-	dmi_check_system(early_pdc_optin_table);
-	if (!early_pdc_optin)
-		return;
-
 	acpi_walk_namespace(ACPI_TYPE_PROCESSOR, ACPI_ROOT_OBJECT,
 			    ACPI_UINT32_MAX,
 			    early_init_pdc, NULL, NULL, NULL);
#7 lore raw

PATCH 7/12 ACPI: processor: driver doesn't need to evaluate _PDC

In reply to Alex Chiang

Now that the early _PDC evaluation path knows how to correctly evaluate _PDC on only physically present processors, there's no need for the processor driver to evaluate it later when it loads.

To cover the hotplug case, push _PDC evaluation down into the hotplug paths.

---

 arch/ia64/kernel/acpi.c         |    3 +++
 arch/x86/kernel/acpi/boot.c     |    3 +++
 drivers/acpi/processor_core.c   |    7 -------
 drivers/acpi/processor_driver.c |    3 ---
 4 files changed, 6 insertions(+), 10 deletions(-)
arch/ia64/kernel/acpi.c +3 -0
index 40574ae..394742a 100644
@@ -44,6 +44,7 @@
 #include <linux/efi.h>
 #include <linux/mmzone.h>
 #include <linux/nodemask.h>
+#include <acpi/processor.h>
 #include <asm/io.h>
 #include <asm/iosapic.h>
 #include <asm/machvec.h>
@@ -936,6 +937,8 @@ int acpi_map_lsapic(acpi_handle handle, int *pcpu)
 	cpu_set(cpu, cpu_present_map);
 	ia64_cpu_to_sapicid[cpu] = physid;
 
+	acpi_processor_set_pdc(handle);
+
 	*pcpu = cpu;
 	return (0);
 }
arch/x86/kernel/acpi/boot.c +3 -0
index fb1035c..baa2280 100644
@@ -481,6 +481,7 @@ int acpi_register_gsi(struct device *dev, u32 gsi, int trigger, int polarity)
  *  ACPI based hotplug support for CPU
  */
 #ifdef CONFIG_ACPI_HOTPLUG_CPU
+#include <acpi/processor.h>
 
 static int __cpuinit _acpi_map_lsapic(acpi_handle handle, int *pcpu)
 {
@@ -539,6 +540,8 @@ static int __cpuinit _acpi_map_lsapic(acpi_handle handle, int *pcpu)
 		goto free_new_map;
 	}
 
+	acpi_processor_set_pdc(handle);
+
 	cpu = cpumask_first(new_map);
 
 	*pcpu = cpu;
drivers/acpi/processor_core.c +0 -7
index 05874ba..1fc06c6 100644
@@ -317,8 +317,6 @@ acpi_processor_eval_pdc(acpi_handle handle, struct acpi_object_list *pdc_in)
 	return status;
 }
 
-static int early_pdc_done;
-
 void acpi_processor_set_pdc(acpi_handle handle)
 {
 	struct acpi_object_list *obj_list;
@@ -326,9 +324,6 @@ void acpi_processor_set_pdc(acpi_handle handle)
 	if (arch_has_acpi_pdc() == false)
 		return;
 
-	if (early_pdc_done)
-		return;
-
 	obj_list = acpi_processor_alloc_pdc();
 	if (!obj_list)
 		return;
@@ -369,6 +364,4 @@ void acpi_early_processor_set_pdc(void)
 	acpi_walk_namespace(ACPI_TYPE_PROCESSOR, ACPI_ROOT_OBJECT,
 			    ACPI_UINT32_MAX,
 			    early_init_pdc, NULL, NULL, NULL);
-
-	early_pdc_done = 1;
 }
drivers/acpi/processor_driver.c +0 -3
index 777d046..7e8e1ba 100644
@@ -606,9 +606,6 @@ static int __cpuinit acpi_processor_add(struct acpi_device *device)
 		goto err_remove_fs;
 	}
 
-	/* _PDC call should be done before doing anything else (if reqd.). */
-	acpi_processor_set_pdc(pr->handle);
-
 #ifdef CONFIG_CPU_FREQ
 	acpi_processor_ppc_has_changed(pr, 0);
 #endif
#8 lore raw

PATCH 8/12 ACPI: processor: refactor internal map_lapic_id()

In reply to Alex Chiang

Untangle the if() statement a little for readability.

Cc Venkatesh Pallipadi <[email protected]>

---

 drivers/acpi/processor_core.c |   15 +++++++++------
 1 files changed, 9 insertions(+), 6 deletions(-)
drivers/acpi/processor_core.c +9 -6
index 1fc06c6..bebd411 100644
@@ -50,12 +50,15 @@ static int map_lapic_id(struct acpi_subtable_header *entry,
 {
 	struct acpi_madt_local_apic *lapic =
 		(struct acpi_madt_local_apic *)entry;
-	if ((lapic->lapic_flags & ACPI_MADT_ENABLED) &&
-	    lapic->processor_id == acpi_id) {
-		*apic_id = lapic->id;
-		return 1;
-	}
-	return 0;
+
+	if (!(lapic->lapic_flags & ACPI_MADT_ENABLED))
+		return 0;
+
+	if (lapic->processor_id != acpi_id)
+		return 0;
+
+	*apic_id = lapic->id;
+	return 1;
 }
 
 static int map_x2apic_id(struct acpi_subtable_header *entry,
#9 lore raw

PATCH 9/12 ACPI: processor: refactor internal map_x2apic_id()

In reply to Alex Chiang

Untangle the nested if conditions to make this function look more similar to the other map_*apic_id() functions.

---

 drivers/acpi/processor_core.c |   12 +++---------
 1 files changed, 3 insertions(+), 9 deletions(-)
drivers/acpi/processor_core.c +3 -9
index bebd411..9213f95 100644
@@ -66,22 +66,16 @@ static int map_x2apic_id(struct acpi_subtable_header *entry,
 {
 	struct acpi_madt_local_x2apic *apic =
 		(struct acpi_madt_local_x2apic *)entry;
-	u32 tmp = apic->local_apic_id;
 
-	/* Only check enabled APICs*/
 	if (!(apic->lapic_flags & ACPI_MADT_ENABLED))
 		return 0;
 
-	/* Device statement declaration type */
-	if (device_declaration) {
-		if (apic->uid == acpi_id)
-			goto found;
+	if (device_declaration && (apic->uid == acpi_id)) {
+		*apic_id = apic->local_apic_id;
+		return 1;
 	}
 
 	return 0;
-found:
-	*apic_id = tmp;
-	return 1;
 }
 
 static int map_lsapic_id(struct acpi_subtable_header *entry,
#10 lore raw

PATCH 10/12 ACPI: processor: refactor internal map_lsapic_id()

In reply to Alex Chiang

Un-nest the if statements for readability.

Remove comments that re-state the obvious.

Change the control flow so that we no longer need a temp variable.

---

 drivers/acpi/processor_core.c |   20 +++++---------------
 1 files changed, 5 insertions(+), 15 deletions(-)
drivers/acpi/processor_core.c +5 -15
index 9213f95..ec5989c 100644
@@ -83,27 +83,17 @@ static int map_lsapic_id(struct acpi_subtable_header *entry,
 {
 	struct acpi_madt_local_sapic *lsapic =
 		(struct acpi_madt_local_sapic *)entry;
-	u32 tmp = (lsapic->id << 8) | lsapic->eid;
 
-	/* Only check enabled APICs*/
 	if (!(lsapic->lapic_flags & ACPI_MADT_ENABLED))
 		return 0;
 
-	/* Device statement declaration type */
 	if (device_declaration) {
-		if (entry->length < 16)
-			printk(KERN_ERR PREFIX
-			    "Invalid LSAPIC with Device type processor (SAPIC ID %#x)\n",
-			    tmp);
-		else if (lsapic->uid == acpi_id)
-			goto found;
-	/* Processor statement declaration type */
-	} else if (lsapic->processor_id == acpi_id)
-		goto found;
+		if ((entry->length < 16) || (lsapic->uid != acpi_id))
+			return 0;
+	} else if (lsapic->processor_id != acpi_id)
+		return 0;
 
-	return 0;
-found:
-	*apic_id = tmp;
+	*apic_id = (lsapic->id << 8) | lsapic->eid;
 	return 1;
 }
 
#11 lore raw

Re: [PATCH 10/12] ACPI: processor: refactor internal map_lsapic_id()

Quoted 10 lines · Alex Chiang
drivers/acpi/processor_core.c +0 -1
index 9213f95..ec5989c 100644
@@ -83,27 +83,17 @@ static int map_lsapic_id(struct acpi_subtable_header *entry,
 {
 	struct acpi_madt_local_sapic *lsapic =
 		(struct acpi_madt_local_sapic *)entry;
-	u32 tmp = (lsapic->id << 8) | lsapic->eid;
 

drivers/acpi/processor_core.c static int map_lsapic_id(struct acpi_subtable_header *entry, · line 97

Quoted 10 lines
drivers/acpi/processor_core.c +0 -6
-	/* Only check enabled APICs*/
 	if (!(lsapic->lapic_flags & ACPI_MADT_ENABLED))
 		return 0;
 
-	/* Device statement declaration type */
 	if (device_declaration) {
-		if (entry->length < 16)
-			printk(KERN_ERR PREFIX
-			    "Invalid LSAPIC with Device type processor (SAPIC ID %#x)\n",
-			    tmp);

Why drop this error message?

Quoted 17 lines
drivers/acpi/processor_core.c +5 -8
-		else if (lsapic->uid == acpi_id)
-			goto found;
-	/* Processor statement declaration type */
-	} else if (lsapic->processor_id == acpi_id)
-		goto found;
+		if ((entry->length < 16) || (lsapic->uid != acpi_id))
+			return 0;
+	} else if (lsapic->processor_id != acpi_id)
+		return 0;
 
-	return 0;
-found:
-	*apic_id = tmp;
+	*apic_id = (lsapic->id << 8) | lsapic->eid;
 	return 1;
 }
 
#12 lore raw

Re: [PATCH 10/12] ACPI: processor: refactor internal map_lsapic_id()

Hi David,

* David Rientjes <[email protected]>:

Quoted 16 lines · Alex Chiang
Quoted 15 lines
drivers/acpi/processor_core.c +0 -2
index 9213f95..ec5989c 100644
@@ -83,27 +83,17 @@ static int map_lsapic_id(struct acpi_subtable_header *entry,
 {
 	struct acpi_madt_local_sapic *lsapic =
 		(struct acpi_madt_local_sapic *)entry;
-	u32 tmp = (lsapic->id << 8) | lsapic->eid;
 
-	/* Only check enabled APICs*/
 	if (!(lsapic->lapic_flags & ACPI_MADT_ENABLED))
 		return 0;
 
Quoted 8 lines

drivers/acpi/processor_core.c static int map_lsapic_id(struct acpi_subtable_header *entry, · line 97

Quoted 6 lines
drivers/acpi/processor_core.c +0 -5
-	/* Device statement declaration type */
 	if (device_declaration) {
-		if (entry->length < 16)
-			printk(KERN_ERR PREFIX
-			    "Invalid LSAPIC with Device type processor (SAPIC ID %#x)\n",
-			    tmp);

Why drop this error message?

It didn't seem to add any value.

If BIOS declares a processor with the Device statement in ACPI and it has an invalid length, there's nothing that a user can do about it.

Arguably, the message might be useful during internal platform bringup (or new firmware validation), but even that's a hard sell to me, since one of the first things a bringup engineer would notice would be missing CPUs, and it wouldn't take that long to realize that firmware was broken; why make upstream carry around extra baggage for internal firmware debugging?

But I don't have a strong opinion here. If folks think that message is useful, I can respin leaving it in.

Thanks,
/ac

#13 lore raw

PATCH 11/12 ACPI: processor: push file static MADT pointer into internal map_madt_entry()

In reply to Alex Chiang

There's no real need for a pointer to the MADT to be global. The only function who uses it is map_madt_entry.

This allows us to remove some more ugly #ifdefs.

---

 drivers/acpi/processor_core.c |   18 +++++++++---------
 1 files changed, 9 insertions(+), 9 deletions(-)
drivers/acpi/processor_core.c +9 -9
index ec5989c..ae69f6d 100644
@@ -43,8 +43,6 @@ static struct dmi_system_id __cpuinitdata processor_idle_dmi_table[] = {
 };
 
 #ifdef CONFIG_SMP
-static struct acpi_table_madt *madt;
-
 static int map_lapic_id(struct acpi_subtable_header *entry,
 		 u32 acpi_id, int *apic_id)
 {
@@ -100,8 +98,17 @@ static int map_lsapic_id(struct acpi_subtable_header *entry,
 static int map_madt_entry(int type, u32 acpi_id)
 {
 	unsigned long madt_end, entry;
+	static struct acpi_table_madt *madt;
+	static int read_madt;
 	int apic_id = -1;
 
+	if (!read_madt) {
+		if (ACPI_FAILURE(acpi_get_table(ACPI_SIG_MADT, 0,
+					(struct acpi_table_header **)&madt)))
+			madt = NULL;
+		read_madt++;
+	}
+
 	if (!madt)
 		return apic_id;
 
@@ -335,13 +342,6 @@ early_init_pdc(acpi_handle handle, u32 lvl, void *context, void **rv)
 
 void acpi_early_processor_set_pdc(void)
 {
-
-#ifdef CONFIG_SMP
-	if (ACPI_FAILURE(acpi_get_table(ACPI_SIG_MADT, 0,
-				(struct acpi_table_header **)&madt)))
-		madt = NULL;
-#endif
-
 	/*
 	 * Check whether the system is DMI table. If yes, OSPM
 	 * should not use mwait for CPU-states.
#14 lore raw

PATCH 12/12 ACPI: processor core: style and sparse cleanups

In reply to Alex Chiang

Remove useless comments
Remove useless braces
Cleanup local variable declarations
Remove shadow declarations

---

 drivers/acpi/processor_core.c |   39 +++++++++++++++++----------------------
 1 files changed, 17 insertions(+), 22 deletions(-)
drivers/acpi/processor_core.c +17 -22
index ae69f6d..4c34ec0 100644
@@ -46,9 +46,9 @@ static struct dmi_system_id __cpuinitdata processor_idle_dmi_table[] = {
 static int map_lapic_id(struct acpi_subtable_header *entry,
 		 u32 acpi_id, int *apic_id)
 {
-	struct acpi_madt_local_apic *lapic =
-		(struct acpi_madt_local_apic *)entry;
+	struct acpi_madt_local_apic *lapic;
 
+	lapic = (struct acpi_madt_local_apic *)entry;
 	if (!(lapic->lapic_flags & ACPI_MADT_ENABLED))
 		return 0;
 
@@ -62,14 +62,14 @@ static int map_lapic_id(struct acpi_subtable_header *entry,
 static int map_x2apic_id(struct acpi_subtable_header *entry,
 			 int device_declaration, u32 acpi_id, int *apic_id)
 {
-	struct acpi_madt_local_x2apic *apic =
-		(struct acpi_madt_local_x2apic *)entry;
+	struct acpi_madt_local_x2apic *lx2apic;
 
-	if (!(apic->lapic_flags & ACPI_MADT_ENABLED))
+	lx2apic = (struct acpi_madt_local_x2apic *)entry;
+	if (!(lx2apic->lapic_flags & ACPI_MADT_ENABLED))
 		return 0;
 
-	if (device_declaration && (apic->uid == acpi_id)) {
-		*apic_id = apic->local_apic_id;
+	if (device_declaration && (lx2apic->uid == acpi_id)) {
+		*apic_id = lx2apic->local_apic_id;
 		return 1;
 	}
 
@@ -79,9 +79,9 @@ static int map_x2apic_id(struct acpi_subtable_header *entry,
 static int map_lsapic_id(struct acpi_subtable_header *entry,
 		int device_declaration, u32 acpi_id, int *apic_id)
 {
-	struct acpi_madt_local_sapic *lsapic =
-		(struct acpi_madt_local_sapic *)entry;
+	struct acpi_madt_local_sapic *lsapic;
 
+	lsapic = (struct acpi_madt_local_sapic *)entry;
 	if (!(lsapic->lapic_flags & ACPI_MADT_ENABLED))
 		return 0;
 
@@ -98,6 +98,7 @@ static int map_lsapic_id(struct acpi_subtable_header *entry,
 static int map_madt_entry(int type, u32 acpi_id)
 {
 	unsigned long madt_end, entry;
+	struct acpi_subtable_header *header;
 	static struct acpi_table_madt *madt;
 	static int read_madt;
 	int apic_id = -1;
@@ -115,12 +116,10 @@ static int map_madt_entry(int type, u32 acpi_id)
 	entry = (unsigned long)madt;
 	madt_end = entry + madt->header.length;
 
-	/* Parse all entries looking for a match. */
-
 	entry += sizeof(struct acpi_table_madt);
 	while (entry + sizeof(struct acpi_subtable_header) < madt_end) {
-		struct acpi_subtable_header *header =
-			(struct acpi_subtable_header *)entry;
+		header = (struct acpi_subtable_header *)entry;
+
 		if (header->type == ACPI_MADT_TYPE_LOCAL_APIC) {
 			if (map_lapic_id(header, acpi_id, &apic_id))
 				break;
@@ -151,20 +150,17 @@ static int map_mat_entry(acpi_handle handle, int type, u32 acpi_id)
 
 	obj = buffer.pointer;
 	if (obj->type != ACPI_TYPE_BUFFER ||
-	    obj->buffer.length < sizeof(struct acpi_subtable_header)) {
+	    obj->buffer.length < sizeof(struct acpi_subtable_header))
 		goto exit;
-	}
 
 	header = (struct acpi_subtable_header *)obj->buffer.pointer;
-	if (header->type == ACPI_MADT_TYPE_LOCAL_APIC) {
+	if (header->type == ACPI_MADT_TYPE_LOCAL_APIC)
 		map_lapic_id(header, acpi_id, &apic_id);
-	} else if (header->type == ACPI_MADT_TYPE_LOCAL_SAPIC) {
+	else if (header->type == ACPI_MADT_TYPE_LOCAL_SAPIC)
 		map_lsapic_id(header, type, acpi_id, &apic_id);
-	}
 
 exit:
-	if (buffer.pointer)
-		kfree(buffer.pointer);
+	kfree(buffer.pointer);
 	return apic_id;
 }
 
@@ -179,10 +175,9 @@ int acpi_get_cpuid(acpi_handle handle, int type, u32 acpi_id)
 	if (apic_id == -1)
 		return apic_id;
 
-	for_each_possible_cpu(i) {
+	for_each_possible_cpu(i)
 		if (cpu_physical_id(i) == apic_id)
 			return i;
-	}
 	return -1;
 }
 EXPORT_SYMBOL_GPL(acpi_get_cpuid);
#15 lore raw
In reply to Alex Chiang

Hi Venki,

Do you have any opinions on this patchset? If so, I'd like to address them.

Thanks,
/ac

* Alex Chiang <[email protected]>:

Quoted 53 lines

This series cleans up some of the mess I made when introducing early _PDC.

The major change is renaming processor_core.c to processor_driver.c, and then renaming processor_pdc.c to processor_core.c.

The idea is that the code in processor_core.c will always be built statically into the kernel (as long as ACPI is configured), while allowing the ACPI processor driver to remain modular (if so desired).

We do this because part of the cleanups involves teaching the early _PDC evaluation code how to determine if a processor is physically present or not -- aka enumeration -- and that sort of code doesn't really belong in a file named processor_pdc.

There are quite a few checkpatch errors in the first rename patch, and I'll look at cleaning those up in a later series.

/ac

---

Alex Chiang (12):
ACPI: processor: mv processor_core.c processor_driver.c
ACPI: processor: mv processor_pdc.c processor_core.c
ACPI: processor: export acpi_get_cpuid()
ACPI: processor: move acpi_get_cpuid into processor_core.c
ACPI: processor: add internal processor_physically_present()
ACPI: processor: remove early _PDC optin quirks
ACPI: processor: driver doesn't need to evaluate _PDC
ACPI: processor: refactor internal map_lapic_id()
ACPI: processor: refactor internal map_x2apic_id()
ACPI: processor: refactor internal map_lsapic_id()
ACPI: processor: push file static MADT pointer into internal map_madt_entry()
ACPI: processor core: style and sparse cleanups

 Documentation/kernel-parameters.txt |    4 
 arch/ia64/kernel/acpi.c             |    3 
 arch/x86/kernel/acpi/boot.c         |    3 
 drivers/acpi/Makefile               |    4 
 drivers/acpi/processor_core.c       | 1149 +++++------------------------------
 drivers/acpi/processor_driver.c     |  976 ++++++++++++++++++++++++++++++
 drivers/acpi/processor_pdc.c        |  209 ------
 include/acpi/processor.h            |   10 
 8 files changed, 1172 insertions(+), 1186 deletions(-)
 create mode 100644 drivers/acpi/processor_driver.c
 delete mode 100644 drivers/acpi/processor_pdc.c

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#16 lore raw

Hi Alex,

Sorry for not responding on this one. Just finished up looking through all the patches in the series. It makes things very clean and patches are neatly split up. Thanks for doing this.

Only thing I wanted to mention was, it would have been a bit cleaner to keep _pdc stuff in a separate file. But, it is OK as you have done it as well.

Quoted 62 lines · Alex Chiang

Hi Venki,

Do you have any opinions on this patchset? If so, I'd like to address them.

Thanks,
/ac

* Alex Chiang <[email protected]>:

Quoted 53 lines

This series cleans up some of the mess I made when introducing early _PDC.

The major change is renaming processor_core.c to processor_driver.c, and then renaming processor_pdc.c to processor_core.c.

The idea is that the code in processor_core.c will always be built statically into the kernel (as long as ACPI is configured), while allowing the ACPI processor driver to remain modular (if so desired).

We do this because part of the cleanups involves teaching the early _PDC evaluation code how to determine if a processor is physically present or not -- aka enumeration -- and that sort of code doesn't really belong in a file named processor_pdc.

There are quite a few checkpatch errors in the first rename patch, and I'll look at cleaning those up in a later series.

/ac

---

Alex Chiang (12):
ACPI: processor: mv processor_core.c processor_driver.c
ACPI: processor: mv processor_pdc.c processor_core.c
ACPI: processor: export acpi_get_cpuid()
ACPI: processor: move acpi_get_cpuid into processor_core.c
ACPI: processor: add internal processor_physically_present()
ACPI: processor: remove early _PDC optin quirks
ACPI: processor: driver doesn't need to evaluate _PDC
ACPI: processor: refactor internal map_lapic_id()
ACPI: processor: refactor internal map_x2apic_id()
ACPI: processor: refactor internal map_lsapic_id()
ACPI: processor: push file static MADT pointer into internal map_madt_entry()
ACPI: processor core: style and sparse cleanups

 Documentation/kernel-parameters.txt |    4 
 arch/ia64/kernel/acpi.c             |    3 
 arch/x86/kernel/acpi/boot.c         |    3 
 drivers/acpi/Makefile               |    4 
 drivers/acpi/processor_core.c       | 1149 +++++------------------------------
 drivers/acpi/processor_driver.c     |  976 ++++++++++++++++++++++++++++++
 drivers/acpi/processor_pdc.c        |  209 ------
 include/acpi/processor.h            |   10 
 8 files changed, 1172 insertions(+), 1186 deletions(-)
 create mode 100644 drivers/acpi/processor_driver.c
 delete mode 100644 drivers/acpi/processor_pdc.c

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html