Skip to contentSkip to conversation

perf/x86/intel/uncore: Fix the index of PCU.3 Broadwell CPUs

PATCH RESEND In review Masayoshi Mizuma 3 participants 9 messages Jul 10, 2018 – Jul 30, 2018

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH RESEND 1/1 perf/x86/intel/uncore: Fix the index of PCU.3 Broadwell CPUs

From: Masayoshi Mizuma <[email protected]>

commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs") introduced PCU.3 for Broadwell CPU. Unfortunately, the driver_data of PCU.3 conflicts to QPI Port 2 filter.

{ /* QPI Port 2 filter */
PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f46), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 2),

{ /* PCU.3 (for Capability registers) */
PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV,
HSWEP_PCI_PCU_3),
// HSWEP_PCI_PCU_3 == 2

As the result, the index for each device gets same, so the pci device is overwrited in uncore_extra_pci_dev[pkg].dev[idx] when it is probed.

static int uncore_pci_probe(struct pci_dev *pdev,... {
...

        if (UNCORE_PCI_DEV_TYPE(id->driver_data) == UNCORE_EXTRA_PCI_DEV) {
                int idx = UNCORE_PCI_DEV_IDX(id->driver_data); // HERE!!

                uncore_extra_pci_dev[pkg].dev[idx] = pdev;     // HERE!!
                pci_set_drvdata(pdev, NULL);
                return 0;
        }

Due to the overwriting, the following warning message are shown while CPU hot-removing.

  WARNING: CPU: 126 PID: 6 at arch/x86/events/intel/uncore.c:988
  uncore_pci_remove+0x10b/0x150
  Call Trace:
   pci_device_remove+0x42/0xd0
   device_release_driver_internal+0x148/0x220
   pci_stop_bus_device+0x76/0xa0
   pci_stop_root_bus+0x44/0x60
   acpi_pci_root_remove+0x1f/0x80
   acpi_bus_trim+0x57/0x90
   acpi_bus_trim+0x2e/0x90
   acpi_device_hotplug+0x2bc/0x4b0
   acpi_hotplug_work_fn+0x1a/0x30
   process_one_work+0x174/0x3a0
   worker_thread+0x4c/0x3d0
   kthread+0xf8/0x130

To avoid the conflict, this patch changes the PCU.3 driver_data. And also increase UNCORE_EXTRA_PCI_DEV_MAX to handle the new index.

Fixes 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs")
---
 arch/x86/events/intel/uncore.h       | 2 +-
 arch/x86/events/intel/uncore_snbep.c | 7 ++++---
 2 files changed, 5 insertions(+), 4 deletions(-)
arch/x86/events/intel/uncore.h +1 -1
index c9e1e0b..e17ab88 100644
@@ -28,7 +28,7 @@
 #define UNCORE_PCI_DEV_TYPE(data)	((data >> 8) & 0xff)
 #define UNCORE_PCI_DEV_IDX(data)	(data & 0xff)
 #define UNCORE_EXTRA_PCI_DEV		0xff
-#define UNCORE_EXTRA_PCI_DEV_MAX	3
+#define UNCORE_EXTRA_PCI_DEV_MAX	4
 
 #define UNCORE_EVENT_CONSTRAINT(c, n) EVENT_CONSTRAINT(c, n, 0xff)
 
arch/x86/events/intel/uncore_snbep.c +4 -3
index 87dc026..62f007c 100644
@@ -1030,6 +1030,7 @@ enum {
 	SNBEP_PCI_QPI_PORT0_FILTER,
 	SNBEP_PCI_QPI_PORT1_FILTER,
 	HSWEP_PCI_PCU_3,
+	BDX_PCI_PCU_3,
 };
 
 static int snbep_qpi_hw_config(struct intel_uncore_box *box, struct perf_event *event)
@@ -3070,11 +3071,11 @@ void bdx_uncore_cpu_init(void)
 	if (boot_cpu_data.x86_model == 86) {
 		uncore_msr_uncores[BDX_MSR_UNCORE_SBOX] = NULL;
 	/* Detect systems with no SBOXes */
-	} else if (uncore_extra_pci_dev[pkg].dev[HSWEP_PCI_PCU_3]) {
+	} else if (uncore_extra_pci_dev[pkg].dev[BDX_PCI_PCU_3]) {
 		struct pci_dev *pdev;
 		u32 capid4;
 
-		pdev = uncore_extra_pci_dev[pkg].dev[HSWEP_PCI_PCU_3];
+		pdev = uncore_extra_pci_dev[pkg].dev[BDX_PCI_PCU_3];
 		pci_read_config_dword(pdev, 0x94, &capid4);
 		if (((capid4 >> 6) & 0x3) == 0)
 			bdx_msr_uncores[BDX_MSR_UNCORE_SBOX] = NULL;
@@ -3299,7 +3300,7 @@ static const struct pci_device_id bdx_uncore_pci_ids[] = {
 	{ /* PCU.3 (for Capability registers) */
 		PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0),
 		.driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV,
-						   HSWEP_PCI_PCU_3),
+						   BDX_PCI_PCU_3),
 	},
 	{ /* end: all zeroes */ }
 };
Signature · 1 line
2.16.1
#1 lore raw
Quoted 19 lines · Masayoshi Mizuma

From: Masayoshi Mizuma <[email protected]>

commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs") introduced PCU.3 for Broadwell CPU. Unfortunately, the driver_data of PCU.3 conflicts to QPI Port 2 filter.

{ /* QPI Port 2 filter */
PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f46), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 2),

{ /* PCU.3 (for Capability registers) */
PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV,
HSWEP_PCI_PCU_3),
// HSWEP_PCI_PCU_3 == 2

arch/x86/events/intel/uncore_snbep.c enum { · line 1034

Quoted 5 lines
arch/x86/events/intel/uncore_snbep.c +1 -0
@@ -1030,6 +1030,7 @@ enum {
 	SNBEP_PCI_QPI_PORT0_FILTER,
 	SNBEP_PCI_QPI_PORT1_FILTER,
 	HSWEP_PCI_PCU_3,
+	BDX_PCI_PCU_3,
 };

So we use a magic '2' enumerator in the 'QPI Port 2 filter', and that overlaps with HSWEP_PCI_PCU_3, right?

Shouldn't we clean up all the enumerators and not use magic numbers, and this fix the conflict?

Thanks,

Ingo

#2 lore raw
Quoted 31 lines · Ingo Molnar
Quoted 19 lines · Masayoshi Mizuma

From: Masayoshi Mizuma <[email protected]>

commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs") introduced PCU.3 for Broadwell CPU. Unfortunately, the driver_data of PCU.3 conflicts to QPI Port 2 filter.

{ /* QPI Port 2 filter */
PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f46), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 2),

{ /* PCU.3 (for Capability registers) */
PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV,
HSWEP_PCI_PCU_3),
// HSWEP_PCI_PCU_3 == 2

arch/x86/events/intel/uncore_snbep.c enum { · line 1034

Quoted 5 lines
arch/x86/events/intel/uncore_snbep.c +1 -0
@@ -1030,6 +1030,7 @@ enum {
  	SNBEP_PCI_QPI_PORT0_FILTER,
  	SNBEP_PCI_QPI_PORT1_FILTER,
  	HSWEP_PCI_PCU_3,
+	BDX_PCI_PCU_3,
  };

So we use a magic '2' enumerator in the 'QPI Port 2 filter', and that overlaps with HSWEP_PCI_PCU_3, right?

Quoted 3 lines

Shouldn't we clean up all the enumerators and not use magic numbers, and this fix the conflict?

Yes, it should fix the conflict. I will clean up the code.

Thanks,
Kan

Quoted 4 lines

Thanks,

Ingo

#3 lore raw
Quoted 34 lines · Kan
Quoted 31 lines · Ingo Molnar
Quoted 19 lines · Masayoshi Mizuma

From: Masayoshi Mizuma <[email protected]>

commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs") introduced PCU.3 for Broadwell CPU. Unfortunately, the driver_data of PCU.3 conflicts to QPI Port 2 filter.

     { /* QPI Port 2 filter  */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f46), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 2),

     { /* PCU.3 (for Capability registers) */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, HSWEP_PCI_PCU_3), // HSWEP_PCI_PCU_3 == 2

arch/x86/events/intel/uncore_snbep.c enum { · line 1034

Quoted 5 lines
arch/x86/events/intel/uncore_snbep.c +1 -0
@@ -1030,6 +1030,7 @@ enum {
       SNBEP_PCI_QPI_PORT0_FILTER,
       SNBEP_PCI_QPI_PORT1_FILTER,
       HSWEP_PCI_PCU_3,
+    BDX_PCI_PCU_3,
   };

So we use a magic '2' enumerator in the 'QPI Port 2 filter', and that overlaps with HSWEP_PCI_PCU_3, right?

Quoted 5 lines
Quoted 3 lines

Shouldn't we clean up all the enumerators and not use magic numbers, and this fix the conflict?

Yes, it should fix the conflict. I will clean up the code.

Thanks a lot!
I would appreciate if you could add CC to me when you post the patch.

Thanks,
Masa

Quoted 9 lines

Thanks,
Kan

Quoted 4 lines

Thanks,

    Ingo

#4 lore raw
Quoted 37 lines · Masayoshi Mizuma
Quoted 34 lines · Kan
Quoted 31 lines · Ingo Molnar
Quoted 19 lines · Masayoshi Mizuma

From: Masayoshi Mizuma <[email protected]>

commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs") introduced PCU.3 for Broadwell CPU. Unfortunately, the driver_data of PCU.3 conflicts to QPI Port 2 filter.

     { /* QPI Port 2 filter  */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f46), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 2),

     { /* PCU.3 (for Capability registers) */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, HSWEP_PCI_PCU_3), // HSWEP_PCI_PCU_3 == 2

arch/x86/events/intel/uncore_snbep.c enum { · line 1034

Quoted 5 lines
arch/x86/events/intel/uncore_snbep.c +1 -0
@@ -1030,6 +1030,7 @@ enum {
       SNBEP_PCI_QPI_PORT0_FILTER,
       SNBEP_PCI_QPI_PORT1_FILTER,
       HSWEP_PCI_PCU_3,
+    BDX_PCI_PCU_3,
   };

So we use a magic '2' enumerator in the 'QPI Port 2 filter', and that overlaps with HSWEP_PCI_PCU_3, right?

Quoted 9 lines
Quoted 5 lines
Quoted 3 lines

Shouldn't we clean up all the enumerators and not use magic numbers, and this fix the conflict?

Yes, it should fix the conflict. I will clean up the code.

Thanks a lot!
I would appreciate if you could add CC to me when you post the patch.

Here is the patch.

Masa, could you please give it a try?

Thanks,
Kan

From 688378a4003ec33156958a52dc822105c18075af Mon Sep 17 00:00:00 2001
From: Kan Liang <[email protected]>
Date: Mon, 16 Jul 2018 04:57:51 -0400
Subject: [PATCH] perf/x86/intel/uncore: Fix hardcode index of Broadwell extra PCI DEV

Masa reports that a warning message is shown while CPU hot-removing on Broadwell server.

  WARNING: CPU: 126 PID: 6 at arch/x86/events/intel/uncore.c:988
  uncore_pci_remove+0x10b/0x150
  Call Trace:
   pci_device_remove+0x42/0xd0
   device_release_driver_internal+0x148/0x220
   pci_stop_bus_device+0x76/0xa0
   pci_stop_root_bus+0x44/0x60
   acpi_pci_root_remove+0x1f/0x80
   acpi_bus_trim+0x57/0x90
   acpi_bus_trim+0x2e/0x90
   acpi_device_hotplug+0x2bc/0x4b0
   acpi_hotplug_work_fn+0x1a/0x30
   process_one_work+0x174/0x3a0
   worker_thread+0x4c/0x3d0
   kthread+0xf8/0x130

This bug was introduced in:

commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for
Broadwell CPUs")

The index of "QPI Port 2 filter" was hardcode to 2. The index of "PCU.3" used enumerator "HSWEP_PCI_PCU_3", which equals to 2 as well.

To fix the conflict, the hardcode index needs to be cleaned up. Introduce a new enumerator "BDX_PCI_QPI_PORT2_FILTER" for "QPI Port 2 filter" on Broadwell, and increase the UNCORE_EXTRA_PCI_DEV_MAX. Clean up hardcode index.

Fixes 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for

Broadwell CPUs")

---
 arch/x86/events/intel/uncore.h       |  2 +-
 arch/x86/events/intel/uncore_snbep.c | 10 +++++++---
 2 files changed, 8 insertions(+), 4 deletions(-)
arch/x86/events/intel/uncore.h +1 -1
index c9e1e0b..e17ab88 100644
@@ -28,7 +28,7 @@
 #define UNCORE_PCI_DEV_TYPE(data)	((data >> 8) & 0xff)
 #define UNCORE_PCI_DEV_IDX(data)	(data & 0xff)
 #define UNCORE_EXTRA_PCI_DEV		0xff
-#define UNCORE_EXTRA_PCI_DEV_MAX	3
+#define UNCORE_EXTRA_PCI_DEV_MAX	4
 
 #define UNCORE_EVENT_CONSTRAINT(c, n) EVENT_CONSTRAINT(c, n, 0xff)
 
arch/x86/events/intel/uncore_snbep.c +7 -3
index 87dc026..51d7c11 100644
@@ -1029,6 +1029,7 @@ void snbep_uncore_cpu_init(void)
 enum {
 	SNBEP_PCI_QPI_PORT0_FILTER,
 	SNBEP_PCI_QPI_PORT1_FILTER,
+	BDX_PCI_QPI_PORT2_FILTER,
 	HSWEP_PCI_PCU_3,
 };
 
@@ -3286,15 +3287,18 @@ static const struct pci_device_id bdx_uncore_pci_ids[] = {
 	},
 	{ /* QPI Port 0 filter  */
 		PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f86),
-		.driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 0),
+		.driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV,
+						   SNBEP_PCI_QPI_PORT0_FILTER),
 	},
 	{ /* QPI Port 1 filter  */
 		PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f96),
-		.driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 1),
+		.driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV,
+						   SNBEP_PCI_QPI_PORT1_FILTER),
 	},
 	{ /* QPI Port 2 filter  */
 		PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f46),
-		.driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 2),
+		.driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV,
+						   BDX_PCI_QPI_PORT2_FILTER),
 	},
 	{ /* PCU.3 (for Capability registers) */
 		PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0),
Signature · 1 line
2.4.11
author Tested-by
#5 lore raw
Quoted 46 lines · Kan
Quoted 43 lines · Masayoshi Mizuma
Quoted 39 lines · Kan
Quoted 31 lines · Ingo Molnar
Quoted 19 lines · Masayoshi Mizuma

From: Masayoshi Mizuma <[email protected]>

commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs") introduced PCU.3 for Broadwell CPU. Unfortunately, the driver_data of PCU.3 conflicts to QPI Port 2 filter.

      { /* QPI Port 2 filter  */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f46), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 2),

      { /* PCU.3 (for Capability registers) */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0), .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, HSWEP_PCI_PCU_3), // HSWEP_PCI_PCU_3 == 2

arch/x86/events/intel/uncore_snbep.c enum { · line 1034

Quoted 5 lines
arch/x86/events/intel/uncore_snbep.c +1 -0
@@ -1030,6 +1030,7 @@ enum {
        SNBEP_PCI_QPI_PORT0_FILTER,
        SNBEP_PCI_QPI_PORT1_FILTER,
        HSWEP_PCI_PCU_3,
+    BDX_PCI_PCU_3,
    };

So we use a magic '2' enumerator in the 'QPI Port 2 filter', and that overlaps with HSWEP_PCI_PCU_3, right?

Quoted 3 lines

Shouldn't we clean up all the enumerators and not use magic numbers, and this fix the conflict?

Yes, it should fix the conflict. I will clean up the code.

Quoted 7 lines
Quoted 3 lines

Thanks a lot!
I would appreciate if you could add CC to me when you post the patch.

Here is the patch.

Masa, could you please give it a try?

Thank you for the patch, it works well!

Please feel free to add:

Thanks,
Masa

Quoted 98 lines

Thanks,
Kan

From 688378a4003ec33156958a52dc822105c18075af Mon Sep 17 00:00:00 2001
From: Kan Liang <[email protected]>
Date: Mon, 16 Jul 2018 04:57:51 -0400
Subject: [PATCH] perf/x86/intel/uncore: Fix hardcode index of Broadwell extra PCI DEV

Masa reports that a warning message is shown while CPU hot-removing on Broadwell server.

  WARNING: CPU: 126 PID: 6 at arch/x86/events/intel/uncore.c:988
  uncore_pci_remove+0x10b/0x150
  Call Trace:
   pci_device_remove+0x42/0xd0
   device_release_driver_internal+0x148/0x220
   pci_stop_bus_device+0x76/0xa0
   pci_stop_root_bus+0x44/0x60
   acpi_pci_root_remove+0x1f/0x80
   acpi_bus_trim+0x57/0x90
   acpi_bus_trim+0x2e/0x90
   acpi_device_hotplug+0x2bc/0x4b0
   acpi_hotplug_work_fn+0x1a/0x30
   process_one_work+0x174/0x3a0
   worker_thread+0x4c/0x3d0
   kthread+0xf8/0x130

This bug was introduced in:

  commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs")

The index of "QPI Port 2 filter" was hardcode to 2. The index of "PCU.3" used enumerator "HSWEP_PCI_PCU_3", which equals to 2 as well.

To fix the conflict, the hardcode index needs to be cleaned up. Introduce a new enumerator "BDX_PCI_QPI_PORT2_FILTER" for "QPI Port 2 filter" on Broadwell, and increase the UNCORE_EXTRA_PCI_DEV_MAX. Clean up hardcode index.

Fixes 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for

Broadwell CPUs")
---
 arch/x86/events/intel/uncore.h       |  2 +- arch/x86/events/intel/uncore_snbep.c | 10 +++++++---

 2 files changed, 8 insertions(+), 4 deletions(-)
arch/x86/events/intel/uncore.h +1 -1
index c9e1e0b..e17ab88 100644
@@ -28,7 +28,7 @@
  #define UNCORE_PCI_DEV_TYPE(data)    ((data >> 8) & 0xff)
  #define UNCORE_PCI_DEV_IDX(data)    (data & 0xff)
  #define UNCORE_EXTRA_PCI_DEV        0xff
-#define UNCORE_EXTRA_PCI_DEV_MAX    3
+#define UNCORE_EXTRA_PCI_DEV_MAX    4
 
  #define UNCORE_EVENT_CONSTRAINT(c, n) EVENT_CONSTRAINT(c, n, 0xff)
 
arch/x86/events/intel/uncore_snbep.c +1 -0
index 87dc026..51d7c11 100644
@@ -1029,6 +1029,7 @@ void snbep_uncore_cpu_init(void)
  enum {
      SNBEP_PCI_QPI_PORT0_FILTER,
      SNBEP_PCI_QPI_PORT1_FILTER,
+    BDX_PCI_QPI_PORT2_FILTER,
      HSWEP_PCI_PCU_3,
  };
 
@@ -3286,15 +3287,18 @@ static const struct pci_device_id bdx_uncore_pci_ids[] = {
      },

     { /* QPI Port 0 filter  */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f86),
-        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 0), +        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, +                           SNBEP_PCI_QPI_PORT0_FILTER), },
     { /* QPI Port 1 filter  */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f96),
-        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 1), +        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, +                           SNBEP_PCI_QPI_PORT1_FILTER), },
     { /* QPI Port 2 filter  */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f46),
-        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 2), +        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, +                           BDX_PCI_QPI_PORT2_FILTER),
     },
     { /* PCU.3 (for Capability registers) */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0),

#6 lore raw

Hi Ingo,

Is the following Kan's patch ready to merge...?

Thanks,
Masa

On 07/17/2018 12:37 PM, Masayoshi Mizuma wrote:
...

Quoted 111 lines
Quoted 3 lines

Here is the patch.

Masa, could you please give it a try?

Thank you for the patch, it works well!

Please feel free to add:

Thanks,
Masa

Quoted 98 lines

Thanks,
Kan

From 688378a4003ec33156958a52dc822105c18075af Mon Sep 17 00:00:00 2001
From: Kan Liang <[email protected]>
Date: Mon, 16 Jul 2018 04:57:51 -0400
Subject: [PATCH] perf/x86/intel/uncore: Fix hardcode index of Broadwell extra PCI DEV

Masa reports that a warning message is shown while CPU hot-removing on Broadwell server.

  WARNING: CPU: 126 PID: 6 at arch/x86/events/intel/uncore.c:988
  uncore_pci_remove+0x10b/0x150
  Call Trace:
   pci_device_remove+0x42/0xd0
   device_release_driver_internal+0x148/0x220
   pci_stop_bus_device+0x76/0xa0
   pci_stop_root_bus+0x44/0x60
   acpi_pci_root_remove+0x1f/0x80
   acpi_bus_trim+0x57/0x90
   acpi_bus_trim+0x2e/0x90
   acpi_device_hotplug+0x2bc/0x4b0
   acpi_hotplug_work_fn+0x1a/0x30
   process_one_work+0x174/0x3a0
   worker_thread+0x4c/0x3d0
   kthread+0xf8/0x130

This bug was introduced in:

  commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs")

The index of "QPI Port 2 filter" was hardcode to 2. The index of "PCU.3" used enumerator "HSWEP_PCI_PCU_3", which equals to 2 as well.

To fix the conflict, the hardcode index needs to be cleaned up. Introduce a new enumerator "BDX_PCI_QPI_PORT2_FILTER" for "QPI Port 2 filter" on Broadwell, and increase the UNCORE_EXTRA_PCI_DEV_MAX. Clean up hardcode index.

Fixes 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for

Broadwell CPUs")
---
 arch/x86/events/intel/uncore.h       |  2 +- arch/x86/events/intel/uncore_snbep.c | 10 +++++++---

 2 files changed, 8 insertions(+), 4 deletions(-)
arch/x86/events/intel/uncore.h +1 -1
index c9e1e0b..e17ab88 100644
@@ -28,7 +28,7 @@
  #define UNCORE_PCI_DEV_TYPE(data)    ((data >> 8) & 0xff)
  #define UNCORE_PCI_DEV_IDX(data)    (data & 0xff)
  #define UNCORE_EXTRA_PCI_DEV        0xff
-#define UNCORE_EXTRA_PCI_DEV_MAX    3
+#define UNCORE_EXTRA_PCI_DEV_MAX    4
 
  #define UNCORE_EVENT_CONSTRAINT(c, n) EVENT_CONSTRAINT(c, n, 0xff)
 
arch/x86/events/intel/uncore_snbep.c +1 -0
index 87dc026..51d7c11 100644
@@ -1029,6 +1029,7 @@ void snbep_uncore_cpu_init(void)
  enum {
      SNBEP_PCI_QPI_PORT0_FILTER,
      SNBEP_PCI_QPI_PORT1_FILTER,
+    BDX_PCI_QPI_PORT2_FILTER,
      HSWEP_PCI_PCU_3,
  };
 
@@ -3286,15 +3287,18 @@ static const struct pci_device_id bdx_uncore_pci_ids[] = {
      },

     { /* QPI Port 0 filter  */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f86),
-        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 0), +        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, +                           SNBEP_PCI_QPI_PORT0_FILTER), },
     { /* QPI Port 1 filter  */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f96),
-        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 1), +        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, +                           SNBEP_PCI_QPI_PORT1_FILTER), },
     { /* QPI Port 2 filter  */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6f46),
-        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, 2), +        .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, +                           BDX_PCI_QPI_PORT2_FILTER),
     },
     { /* PCU.3 (for Capability registers) */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x6fc0),

#7 lore raw
Quoted 3 lines · Masayoshi Mizuma

Hi Ingo,

Is the following Kan's patch ready to merge...?

Looks good at first sight - but it was whitespace damaged here so I couldn't apply it.

Kan, mind re-sending it properly as a standalone patch, with Masayoshi's Reported-by, Debugged-by and Tested-by tags added?

Thanks,

Ingo

#8 lore raw
Quoted 12 lines · Ingo Molnar
Quoted 3 lines · Masayoshi Mizuma

Hi Ingo,

Is the following Kan's patch ready to merge...?

Looks good at first sight - but it was whitespace damaged here so I couldn't apply it.

Kan, mind re-sending it properly as a standalone patch, with Masayoshi's Reported-by, Debugged-by and Tested-by tags added?

Sure, I will re-send the patch.

Thanks,
Kan