* [PATCH 2/4] drm/xe: Save and restore PCI state
2024-04-02 8:58 [PATCH v2 0/4] drm/xe: Support PCIe FLR Aravind Iddamsetty
@ 2024-04-02 8:58 ` Aravind Iddamsetty
0 siblings, 0 replies; 11+ messages in thread
From: Aravind Iddamsetty @ 2024-04-02 8:58 UTC (permalink / raw)
To: intel-xe, thomas.hellstrom, rodrigo.vivi, lucas.demarchi; +Cc: dri-devel
Save and restore PCI states where ever needed.
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@linux.intel.com>
---
drivers/gpu/drm/xe/xe_device_types.h | 3 ++
drivers/gpu/drm/xe/xe_pci.c | 48 ++++++++++++++++++++++++++--
drivers/gpu/drm/xe/xe_pci.h | 4 ++-
3 files changed, 51 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
index 1df3dcc17d75..3bfde4b59284 100644
--- a/drivers/gpu/drm/xe/xe_device_types.h
+++ b/drivers/gpu/drm/xe/xe_device_types.h
@@ -455,6 +455,9 @@ struct xe_device {
/** @needs_flr_on_fini: requests function-reset on fini */
bool needs_flr_on_fini;
+ /** @pci_state: PCI state of device */
+ struct pci_saved_state *pci_state;
+
/* private: */
#if IS_ENABLED(CONFIG_DRM_XE_DISPLAY)
diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
index c401d4890386..e9e10f8d5f2b 100644
--- a/drivers/gpu/drm/xe/xe_pci.c
+++ b/drivers/gpu/drm/xe/xe_pci.c
@@ -383,6 +383,41 @@ MODULE_DEVICE_TABLE(pci, pciidlist);
#undef INTEL_VGA_DEVICE
+static bool xe_save_pci_state(struct pci_dev *pdev)
+{
+ struct xe_device *xe = pci_get_drvdata(pdev);
+
+ if (pci_save_state(pdev))
+ return false;
+
+ kfree(xe->pci_state);
+
+ xe->pci_state = pci_store_saved_state(pdev);
+ if (!xe->pci_state) {
+ drm_err(&xe->drm, "Failed to store PCI saved state\n");
+ return false;
+ }
+
+ return true;
+}
+
+void xe_load_pci_state(struct pci_dev *pdev)
+{
+ struct xe_device *xe = pci_get_drvdata(pdev);
+ int ret;
+
+ if (!xe->pci_state)
+ return;
+
+ ret = pci_load_saved_state(pdev, xe->pci_state);
+ if (ret) {
+ drm_warn(&xe->drm, "Failed to load PCI state err:%d\n", ret);
+ return;
+ }
+
+ pci_restore_state(pdev);
+}
+
/* is device_id present in comma separated list of ids */
static bool device_id_in_list(u16 device_id, const char *devices, bool negative)
{
@@ -688,6 +723,8 @@ static void xe_pci_remove(struct pci_dev *pdev)
xe_device_remove(xe);
xe_pm_runtime_fini(xe);
+
+ kfree(xe->pci_state);
pci_set_drvdata(pdev, NULL);
}
@@ -786,6 +823,9 @@ static int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
drm_dbg(&xe->drm, "d3cold: capable=%s\n",
str_yes_no(xe->d3cold.capable));
+ if (xe_save_pci_state(pdev))
+ pci_restore_state(pdev);
+
return 0;
}
@@ -833,7 +873,7 @@ static int xe_pci_suspend(struct device *dev)
*/
d3cold_toggle(pdev, D3COLD_ENABLE);
- pci_save_state(pdev);
+ xe_save_pci_state(pdev);
pci_disable_device(pdev);
return 0;
@@ -857,6 +897,8 @@ static int xe_pci_resume(struct device *dev)
pci_set_master(pdev);
+ xe_load_pci_state(pdev);
+
err = xe_pm_resume(pdev_to_xe_device(pdev));
if (err)
return err;
@@ -874,7 +916,7 @@ static int xe_pci_runtime_suspend(struct device *dev)
if (err)
return err;
- pci_save_state(pdev);
+ xe_save_pci_state(pdev);
if (xe->d3cold.allowed) {
d3cold_toggle(pdev, D3COLD_ENABLE);
@@ -899,7 +941,7 @@ static int xe_pci_runtime_resume(struct device *dev)
if (err)
return err;
- pci_restore_state(pdev);
+ xe_load_pci_state(pdev);
if (xe->d3cold.allowed) {
err = pci_enable_device(pdev);
diff --git a/drivers/gpu/drm/xe/xe_pci.h b/drivers/gpu/drm/xe/xe_pci.h
index 611c1209b14c..73b90a430d1f 100644
--- a/drivers/gpu/drm/xe/xe_pci.h
+++ b/drivers/gpu/drm/xe/xe_pci.h
@@ -6,7 +6,9 @@
#ifndef _XE_PCI_H_
#define _XE_PCI_H_
+struct pci_dev;
+
int xe_register_pci_driver(void);
void xe_unregister_pci_driver(void);
-
+void xe_load_pci_state(struct pci_dev *pdev);
#endif
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH 2/4] drm/xe: Save and restore PCI state
2024-04-17 8:41 [PATCH v2 0/4] drm/xe: Support PCIe FLR Aravind Iddamsetty
@ 2024-04-17 8:41 ` Aravind Iddamsetty
0 siblings, 0 replies; 11+ messages in thread
From: Aravind Iddamsetty @ 2024-04-17 8:41 UTC (permalink / raw)
To: dri-devel, daniel, maarten.lankhorst, airlied, mripard,
tzimmermann, rodrigo.vivi
Cc: intel-xe, Lucas De Marchi
Save and restore PCI states where ever needed.
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@linux.intel.com>
---
drivers/gpu/drm/xe/xe_device_types.h | 3 ++
drivers/gpu/drm/xe/xe_pci.c | 48 ++++++++++++++++++++++++++--
drivers/gpu/drm/xe/xe_pci.h | 4 ++-
3 files changed, 51 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
index 8244b177a6a3..0a66555229e9 100644
--- a/drivers/gpu/drm/xe/xe_device_types.h
+++ b/drivers/gpu/drm/xe/xe_device_types.h
@@ -462,6 +462,9 @@ struct xe_device {
/** @needs_flr_on_fini: requests function-reset on fini */
bool needs_flr_on_fini;
+ /** @pci_state: PCI state of device */
+ struct pci_saved_state *pci_state;
+
/* private: */
#if IS_ENABLED(CONFIG_DRM_XE_DISPLAY)
diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
index fa2cc80a08a3..f9957cd16356 100644
--- a/drivers/gpu/drm/xe/xe_pci.c
+++ b/drivers/gpu/drm/xe/xe_pci.c
@@ -391,6 +391,41 @@ MODULE_DEVICE_TABLE(pci, pciidlist);
#undef INTEL_VGA_DEVICE
+static bool xe_save_pci_state(struct pci_dev *pdev)
+{
+ struct xe_device *xe = pci_get_drvdata(pdev);
+
+ if (pci_save_state(pdev))
+ return false;
+
+ kfree(xe->pci_state);
+
+ xe->pci_state = pci_store_saved_state(pdev);
+ if (!xe->pci_state) {
+ drm_err(&xe->drm, "Failed to store PCI saved state\n");
+ return false;
+ }
+
+ return true;
+}
+
+void xe_load_pci_state(struct pci_dev *pdev)
+{
+ struct xe_device *xe = pci_get_drvdata(pdev);
+ int ret;
+
+ if (!xe->pci_state)
+ return;
+
+ ret = pci_load_saved_state(pdev, xe->pci_state);
+ if (ret) {
+ drm_warn(&xe->drm, "Failed to load PCI state err:%d\n", ret);
+ return;
+ }
+
+ pci_restore_state(pdev);
+}
+
/* is device_id present in comma separated list of ids */
static bool device_id_in_list(u16 device_id, const char *devices, bool negative)
{
@@ -696,6 +731,8 @@ static void xe_pci_remove(struct pci_dev *pdev)
xe_device_remove(xe);
xe_pm_runtime_fini(xe);
+
+ kfree(xe->pci_state);
pci_set_drvdata(pdev, NULL);
}
@@ -792,6 +829,9 @@ static int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
drm_dbg(&xe->drm, "d3cold: capable=%s\n",
str_yes_no(xe->d3cold.capable));
+ if (xe_save_pci_state(pdev))
+ pci_restore_state(pdev);
+
return 0;
}
@@ -839,7 +879,7 @@ static int xe_pci_suspend(struct device *dev)
*/
d3cold_toggle(pdev, D3COLD_ENABLE);
- pci_save_state(pdev);
+ xe_save_pci_state(pdev);
pci_disable_device(pdev);
return 0;
@@ -863,6 +903,8 @@ static int xe_pci_resume(struct device *dev)
pci_set_master(pdev);
+ xe_load_pci_state(pdev);
+
err = xe_pm_resume(pdev_to_xe_device(pdev));
if (err)
return err;
@@ -880,7 +922,7 @@ static int xe_pci_runtime_suspend(struct device *dev)
if (err)
return err;
- pci_save_state(pdev);
+ xe_save_pci_state(pdev);
if (xe->d3cold.allowed) {
d3cold_toggle(pdev, D3COLD_ENABLE);
@@ -905,7 +947,7 @@ static int xe_pci_runtime_resume(struct device *dev)
if (err)
return err;
- pci_restore_state(pdev);
+ xe_load_pci_state(pdev);
if (xe->d3cold.allowed) {
err = pci_enable_device(pdev);
diff --git a/drivers/gpu/drm/xe/xe_pci.h b/drivers/gpu/drm/xe/xe_pci.h
index 611c1209b14c..73b90a430d1f 100644
--- a/drivers/gpu/drm/xe/xe_pci.h
+++ b/drivers/gpu/drm/xe/xe_pci.h
@@ -6,7 +6,9 @@
#ifndef _XE_PCI_H_
#define _XE_PCI_H_
+struct pci_dev;
+
int xe_register_pci_driver(void);
void xe_unregister_pci_driver(void);
-
+void xe_load_pci_state(struct pci_dev *pdev);
#endif
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH v3 0/4] drm/xe: Support PCIe FLR
@ 2024-04-19 8:58 Aravind Iddamsetty
2024-04-19 8:58 ` [PATCH v2 1/4] drm: add devm release action Aravind Iddamsetty
` (3 more replies)
0 siblings, 4 replies; 11+ messages in thread
From: Aravind Iddamsetty @ 2024-04-19 8:58 UTC (permalink / raw)
To: dri-devel, daniel, maarten.lankhorst, airlied, mripard,
tzimmermann, rodrigo.vivi
Cc: intel-xe, Lucas De Marchi
PCI subsystem provides callbacks to inform the driver about a request to
do function level reset by user, initiated by writing to sysfs entry
/sys/bus/pci/devices/.../reset. This will allow the driver to handle FLR
without the need to do unbind and rebind as the driver needs to
reinitialize the device afresh post FLR.
v3: Trivial fix
v2:
1. Directly expose the devm_drm_dev_release_action instead of introducing
a helper (Rodrigo)
2. separate out gt idle and pci save/restore to a separate patch (Lucas)
3. Fixed the warnings seen around xe_guc_submit_stop, xe_guc_puc_fini
Rodrigo, retaining your r-b's since there are no functional changes.
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
dmesg snip showing FLR recovery:
[ 590.486336] xe 0000:4d:00.0: enabling device (0140 -> 0142)
[ 590.506933] xe 0000:4d:00.0: [drm] Using GuC firmware from
xe/pvc_guc_70.20.0.bin version 70.20.0
[ 590.542355] xe 0000:4d:00.0: [drm] Using GuC firmware from
xe/pvc_guc_70.20.0.bin version 70.20.0
[ 590.578532] xe 0000:4d:00.0: [drm] VISIBLE VRAM: 0x0000202000000000,
0x0000002000000000
[ 590.578556] xe 0000:4d:00.0: [drm] VRAM[0, 0]: Actual physical size
0x0000001000000000, usable size exclude stolen 0x0000000fff000000, CPU
accessible size 0x0000000fff000000
[ 590.578560] xe 0000:4d:00.0: [drm] VRAM[0, 0]: DPA range:
[0x0000000000000000-1000000000], io range:
[0x0000202000000000-202fff000000]
[ 590.578585] xe 0000:4d:00.0: [drm] VRAM[1, 1]: Actual physical size
0x0000001000000000, usable size exclude stolen 0x0000000fff000000, CPU
accessible size 0x0000000fff000000
[ 590.578589] xe 0000:4d:00.0: [drm] VRAM[1, 1]: DPA range:
[0x0000001000000000-2000000000], io range:
[0x0000203000000000-203fff000000]
[ 590.578592] xe 0000:4d:00.0: [drm] Total VRAM: 0x0000202000000000,
0x0000002000000000
[ 590.578594] xe 0000:4d:00.0: [drm] Available VRAM:
0x0000202000000000, 0x0000001ffe000000
[ 590.738899] xe 0000:4d:00.0: [drm] GT0: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 590.889991] xe 0000:4d:00.0: [drm] GT1: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 590.892835] [drm] Initialized xe 1.1.0 20201103 for 0000:4d:00.0 on
minor 1
[ 590.900215] xe 0000:9a:00.0: enabling device (0140 -> 0142)
[ 590.915991] xe 0000:9a:00.0: [drm] Using GuC firmware from
xe/pvc_guc_70.20.0.bin version 70.20.0
[ 590.957450] xe 0000:9a:00.0: [drm] Using GuC firmware from
xe/pvc_guc_70.20.0.bin version 70.20.0
[ 590.989863] xe 0000:9a:00.0: [drm] VISIBLE VRAM: 0x000020e000000000,
0x0000002000000000
[ 590.989888] xe 0000:9a:00.0: [drm] VRAM[0, 0]: Actual physical size
0x0000001000000000, usable size exclude stolen 0x0000000fff000000, CPU
accessible size 0x0000000fff000000
[ 590.989893] xe 0000:9a:00.0: [drm] VRAM[0, 0]: DPA range:
[0x0000000000000000-1000000000], io range:
[0x000020e000000000-20efff000000]
[ 590.989918] xe 0000:9a:00.0: [drm] VRAM[1, 1]: Actual physical size
0x0000001000000000, usable size exclude stolen 0x0000000fff000000, CPU
accessible size 0x0000000fff000000
[ 590.989921] xe 0000:9a:00.0: [drm] VRAM[1, 1]: DPA range:
[0x0000001000000000-2000000000], io range:
[0x000020f000000000-20ffff000000]
[ 590.989924] xe 0000:9a:00.0: [drm] Total VRAM: 0x000020e000000000,
0x0000002000000000
[ 590.989927] xe 0000:9a:00.0: [drm] Available VRAM:
0x000020e000000000, 0x0000001ffe000000
[ 591.142061] xe 0000:9a:00.0: [drm] GT0: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 591.293505] xe 0000:9a:00.0: [drm] GT1: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 591.295487] [drm] Initialized xe 1.1.0 20201103 for 0000:9a:00.0 on
minor 2
[ 610.685993] Console: switching to colour dummy device 80x25
[ 610.686118] [IGT] xe_exec_basic: executing
[ 610.755398] xe 0000:4d:00.0: [drm] GT0: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 610.771783] xe 0000:4d:00.0: [drm] GT1: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 610.773542] [IGT] xe_exec_basic: starting subtest once-basic
[ 610.960251] [IGT] xe_exec_basic: finished subtest once-basic, SUCCESS
[ 610.962741] [IGT] xe_exec_basic: exiting, ret=0
[ 610.977203] Console: switching to colour frame buffer device 128x48
[ 611.006675] xe_exec_basic (3237) used greatest stack depth: 11128
bytes left
[ 644.682201] xe 0000:4d:00.0: [drm] GT0: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 644.699060] xe 0000:4d:00.0: [drm] GT1: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 644.699118] xe 0000:4d:00.0: preparing for PCIe FLR reset
[ 644.699149] xe 0000:4d:00.0: [drm] removing device access to
userspace
[ 644.928577] xe 0000:4d:00.0: PCI device went through FLR, reenabling
the device
[ 656.104233] xe 0000:4d:00.0: [drm] Using GuC firmware from
xe/pvc_guc_70.20.0.bin version 70.20.0
[ 656.149525] xe 0000:4d:00.0: [drm] Using GuC firmware from
xe/pvc_guc_70.20.0.bin version 70.20.0
[ 656.182711] xe 0000:4d:00.0: [drm] VISIBLE VRAM: 0x0000202000000000,
0x0000002000000000
[ 656.182737] xe 0000:4d:00.0: [drm] VRAM[0, 0]: Actual physical size
0x0000001000000000, usable size exclude stolen 0x0000000fff000000, CPU
accessible size 0x0000000fff000000
[ 656.182742] xe 0000:4d:00.0: [drm] VRAM[0, 0]: DPA range:
[0x0000000000000000-1000000000], io range:
[0x0000202000000000-202fff000000]
[ 656.182768] xe 0000:4d:00.0: [drm] VRAM[1, 1]: Actual physical size
0x0000001000000000, usable size exclude stolen 0x0000000fff000000, CPU
accessible size 0x0000000fff000000
[ 656.182772] xe 0000:4d:00.0: [drm] VRAM[1, 1]: DPA range:
[0x0000001000000000-2000000000], io range:
[0x0000203000000000-203fff000000]
[ 656.182775] xe 0000:4d:00.0: [drm] Total VRAM: 0x0000202000000000,
0x0000002000000000
[ 656.182778] xe 0000:4d:00.0: [drm] Available VRAM:
0x0000202000000000, 0x0000001ffe000000
[ 656.348657] xe 0000:4d:00.0: [drm] GT0: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 656.507619] xe 0000:4d:00.0: [drm] GT1: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 656.510848] [drm] Initialized xe 1.1.0 20201103 for 0000:4d:00.0 on
minor 1
[ 665.754402] Console: switching to colour dummy device 80x25
[ 665.754484] [IGT] xe_exec_basic: executing
[ 665.805853] xe 0000:4d:00.0: [drm] GT0: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 665.819825] xe 0000:4d:00.0: [drm] GT1: CCS_MODE=0 config:00400000,
num_engines:1, num_slices:4
[ 665.820359] [IGT] xe_exec_basic: starting subtest once-basic
[ 665.968899] [IGT] xe_exec_basic: finished subtest once-basic, SUCCESS
[ 665.969534] [IGT] xe_exec_basic: exiting, ret=0
[ 665.981027] Console: switching to colour frame buffer device 128x48
Aravind Iddamsetty (4):
drm: add devm release action
drm/xe: Save and restore PCI state
drm/xe: Extract xe_gt_idle() helper
drm/xe/FLR: Support PCIe FLR
drivers/gpu/drm/drm_drv.c | 6 ++
drivers/gpu/drm/xe/Makefile | 1 +
drivers/gpu/drm/xe/xe_device_types.h | 6 ++
drivers/gpu/drm/xe/xe_gt.c | 28 +++++----
drivers/gpu/drm/xe/xe_gt.h | 1 +
drivers/gpu/drm/xe/xe_guc_pc.c | 4 ++
drivers/gpu/drm/xe/xe_pci.c | 57 ++++++++++++++++--
drivers/gpu/drm/xe/xe_pci.h | 6 +-
drivers/gpu/drm/xe/xe_pci_err.c | 88 ++++++++++++++++++++++++++++
drivers/gpu/drm/xe/xe_pci_err.h | 13 ++++
include/drm/drm_drv.h | 2 +
11 files changed, 197 insertions(+), 15 deletions(-)
create mode 100644 drivers/gpu/drm/xe/xe_pci_err.c
create mode 100644 drivers/gpu/drm/xe/xe_pci_err.h
--
2.25.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* [PATCH v2 1/4] drm: add devm release action
2024-04-19 8:58 [PATCH v3 0/4] drm/xe: Support PCIe FLR Aravind Iddamsetty
@ 2024-04-19 8:58 ` Aravind Iddamsetty
2024-04-19 9:11 ` Maxime Ripard
2024-04-19 8:58 ` [PATCH 2/4] drm/xe: Save and restore PCI state Aravind Iddamsetty
` (2 subsequent siblings)
3 siblings, 1 reply; 11+ messages in thread
From: Aravind Iddamsetty @ 2024-04-19 8:58 UTC (permalink / raw)
To: dri-devel, daniel, maarten.lankhorst, airlied, mripard,
tzimmermann, rodrigo.vivi
Cc: intel-xe, Thomas Hellstr_m
In scenarios where drm_dev_put is directly called by driver we want to
release devm_drm_dev_init_release action associated with struct
drm_device.
v2: Directly expose the original function, instead of introducing a
helper (Rodrigo)
Cc: Thomas Hellstr_m <thomas.hellstrom@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@linux.intel.com>
---
drivers/gpu/drm/drm_drv.c | 6 ++++++
include/drm/drm_drv.h | 2 ++
2 files changed, 8 insertions(+)
diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
index 243cacb3575c..ba60cbb0725f 100644
--- a/drivers/gpu/drm/drm_drv.c
+++ b/drivers/gpu/drm/drm_drv.c
@@ -714,6 +714,12 @@ static int devm_drm_dev_init(struct device *parent,
devm_drm_dev_init_release, dev);
}
+void devm_drm_dev_release_action(struct drm_device *dev)
+{
+ devm_release_action(dev->dev, devm_drm_dev_init_release, dev);
+}
+EXPORT_SYMBOL(devm_drm_dev_release_action);
+
void *__devm_drm_dev_alloc(struct device *parent,
const struct drm_driver *driver,
size_t size, size_t offset)
diff --git a/include/drm/drm_drv.h b/include/drm/drm_drv.h
index 8878260d7529..fa9123684874 100644
--- a/include/drm/drm_drv.h
+++ b/include/drm/drm_drv.h
@@ -444,6 +444,8 @@ struct drm_driver {
const struct file_operations *fops;
};
+void devm_drm_dev_release_action(struct drm_device *dev);
+
void *__devm_drm_dev_alloc(struct device *parent,
const struct drm_driver *driver,
size_t size, size_t offset);
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH 2/4] drm/xe: Save and restore PCI state
2024-04-19 8:58 [PATCH v3 0/4] drm/xe: Support PCIe FLR Aravind Iddamsetty
2024-04-19 8:58 ` [PATCH v2 1/4] drm: add devm release action Aravind Iddamsetty
@ 2024-04-19 8:58 ` Aravind Iddamsetty
2024-04-19 8:58 ` [PATCH v2 3/4] drm/xe: Extract xe_gt_idle() helper Aravind Iddamsetty
2024-04-19 8:58 ` [PATCH v3 4/4] drm/xe/FLR: Support PCIe FLR Aravind Iddamsetty
3 siblings, 0 replies; 11+ messages in thread
From: Aravind Iddamsetty @ 2024-04-19 8:58 UTC (permalink / raw)
To: dri-devel, daniel, maarten.lankhorst, airlied, mripard,
tzimmermann, rodrigo.vivi
Cc: intel-xe, Lucas De Marchi
Save and restore PCI states where ever needed.
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@linux.intel.com>
---
drivers/gpu/drm/xe/xe_device_types.h | 3 ++
drivers/gpu/drm/xe/xe_pci.c | 48 ++++++++++++++++++++++++++--
drivers/gpu/drm/xe/xe_pci.h | 4 ++-
3 files changed, 51 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
index 8244b177a6a3..0a66555229e9 100644
--- a/drivers/gpu/drm/xe/xe_device_types.h
+++ b/drivers/gpu/drm/xe/xe_device_types.h
@@ -462,6 +462,9 @@ struct xe_device {
/** @needs_flr_on_fini: requests function-reset on fini */
bool needs_flr_on_fini;
+ /** @pci_state: PCI state of device */
+ struct pci_saved_state *pci_state;
+
/* private: */
#if IS_ENABLED(CONFIG_DRM_XE_DISPLAY)
diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
index 65ea44672cab..fd45f235b35c 100644
--- a/drivers/gpu/drm/xe/xe_pci.c
+++ b/drivers/gpu/drm/xe/xe_pci.c
@@ -391,6 +391,41 @@ MODULE_DEVICE_TABLE(pci, pciidlist);
#undef INTEL_VGA_DEVICE
+static bool xe_save_pci_state(struct pci_dev *pdev)
+{
+ struct xe_device *xe = pci_get_drvdata(pdev);
+
+ if (pci_save_state(pdev))
+ return false;
+
+ kfree(xe->pci_state);
+
+ xe->pci_state = pci_store_saved_state(pdev);
+ if (!xe->pci_state) {
+ drm_err(&xe->drm, "Failed to store PCI saved state\n");
+ return false;
+ }
+
+ return true;
+}
+
+void xe_load_pci_state(struct pci_dev *pdev)
+{
+ struct xe_device *xe = pci_get_drvdata(pdev);
+ int ret;
+
+ if (!xe->pci_state)
+ return;
+
+ ret = pci_load_saved_state(pdev, xe->pci_state);
+ if (ret) {
+ drm_warn(&xe->drm, "Failed to load PCI state err:%d\n", ret);
+ return;
+ }
+
+ pci_restore_state(pdev);
+}
+
/* is device_id present in comma separated list of ids */
static bool device_id_in_list(u16 device_id, const char *devices, bool negative)
{
@@ -696,6 +731,8 @@ static void xe_pci_remove(struct pci_dev *pdev)
xe_device_remove(xe);
xe_pm_runtime_fini(xe);
+
+ kfree(xe->pci_state);
pci_set_drvdata(pdev, NULL);
}
@@ -796,6 +833,9 @@ static int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
drm_dbg(&xe->drm, "d3cold: capable=%s\n",
str_yes_no(xe->d3cold.capable));
+ if (xe_save_pci_state(pdev))
+ pci_restore_state(pdev);
+
return 0;
err_driver_cleanup:
@@ -847,7 +887,7 @@ static int xe_pci_suspend(struct device *dev)
*/
d3cold_toggle(pdev, D3COLD_ENABLE);
- pci_save_state(pdev);
+ xe_save_pci_state(pdev);
pci_disable_device(pdev);
return 0;
@@ -871,6 +911,8 @@ static int xe_pci_resume(struct device *dev)
pci_set_master(pdev);
+ xe_load_pci_state(pdev);
+
err = xe_pm_resume(pdev_to_xe_device(pdev));
if (err)
return err;
@@ -888,7 +930,7 @@ static int xe_pci_runtime_suspend(struct device *dev)
if (err)
return err;
- pci_save_state(pdev);
+ xe_save_pci_state(pdev);
if (xe->d3cold.allowed) {
d3cold_toggle(pdev, D3COLD_ENABLE);
@@ -913,7 +955,7 @@ static int xe_pci_runtime_resume(struct device *dev)
if (err)
return err;
- pci_restore_state(pdev);
+ xe_load_pci_state(pdev);
if (xe->d3cold.allowed) {
err = pci_enable_device(pdev);
diff --git a/drivers/gpu/drm/xe/xe_pci.h b/drivers/gpu/drm/xe/xe_pci.h
index 611c1209b14c..73b90a430d1f 100644
--- a/drivers/gpu/drm/xe/xe_pci.h
+++ b/drivers/gpu/drm/xe/xe_pci.h
@@ -6,7 +6,9 @@
#ifndef _XE_PCI_H_
#define _XE_PCI_H_
+struct pci_dev;
+
int xe_register_pci_driver(void);
void xe_unregister_pci_driver(void);
-
+void xe_load_pci_state(struct pci_dev *pdev);
#endif
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH v2 3/4] drm/xe: Extract xe_gt_idle() helper
2024-04-19 8:58 [PATCH v3 0/4] drm/xe: Support PCIe FLR Aravind Iddamsetty
2024-04-19 8:58 ` [PATCH v2 1/4] drm: add devm release action Aravind Iddamsetty
2024-04-19 8:58 ` [PATCH 2/4] drm/xe: Save and restore PCI state Aravind Iddamsetty
@ 2024-04-19 8:58 ` Aravind Iddamsetty
2024-04-23 4:21 ` Ghimiray, Himal Prasad
2024-04-19 8:58 ` [PATCH v3 4/4] drm/xe/FLR: Support PCIe FLR Aravind Iddamsetty
3 siblings, 1 reply; 11+ messages in thread
From: Aravind Iddamsetty @ 2024-04-19 8:58 UTC (permalink / raw)
To: dri-devel, daniel, maarten.lankhorst, airlied, mripard,
tzimmermann, rodrigo.vivi
Cc: intel-xe, Lucas De Marchi, Michal Wajdeczko,
Himal Prasad Ghimiray
This would be used in other places outside of gt_reset path.
v2:
1. Add kernel doc for xe_gt_idle (Michal)
2. fix return as no actual error is reported by xe_uc_stop (Himal)
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@linux.intel.com>
---
drivers/gpu/drm/xe/xe_gt.c | 28 ++++++++++++++++++----------
drivers/gpu/drm/xe/xe_gt.h | 1 +
2 files changed, 19 insertions(+), 10 deletions(-)
diff --git a/drivers/gpu/drm/xe/xe_gt.c b/drivers/gpu/drm/xe/xe_gt.c
index 491d0413de15..a24340219c30 100644
--- a/drivers/gpu/drm/xe/xe_gt.c
+++ b/drivers/gpu/drm/xe/xe_gt.c
@@ -629,6 +629,23 @@ static int do_gt_restart(struct xe_gt *gt)
return 0;
}
+/**
+ * xe_gt_idle: Idle the GT.
+ * @gt: The gt object
+ *
+ * Typically called before initiating any resets.
+ *
+ */
+void xe_gt_idle(struct xe_gt *gt)
+{
+ xe_gt_sanitize(gt);
+ xe_uc_gucrc_disable(>->uc);
+ xe_uc_stop_prepare(>->uc);
+ xe_gt_pagefault_reset(gt);
+ xe_uc_stop(>->uc);
+ xe_gt_tlb_invalidation_reset(gt);
+}
+
static int gt_reset(struct xe_gt *gt)
{
int err;
@@ -645,21 +662,12 @@ static int gt_reset(struct xe_gt *gt)
}
xe_pm_runtime_get(gt_to_xe(gt));
- xe_gt_sanitize(gt);
err = xe_force_wake_get(gt_to_fw(gt), XE_FORCEWAKE_ALL);
if (err)
goto err_msg;
- xe_uc_gucrc_disable(>->uc);
- xe_uc_stop_prepare(>->uc);
- xe_gt_pagefault_reset(gt);
-
- err = xe_uc_stop(>->uc);
- if (err)
- goto err_out;
-
- xe_gt_tlb_invalidation_reset(gt);
+ xe_gt_idle(gt);
err = do_gt_reset(gt);
if (err)
diff --git a/drivers/gpu/drm/xe/xe_gt.h b/drivers/gpu/drm/xe/xe_gt.h
index ed6ea8057e35..dcc1023d20e8 100644
--- a/drivers/gpu/drm/xe/xe_gt.h
+++ b/drivers/gpu/drm/xe/xe_gt.h
@@ -43,6 +43,7 @@ int xe_gt_resume(struct xe_gt *gt);
void xe_gt_reset_async(struct xe_gt *gt);
void xe_gt_sanitize(struct xe_gt *gt);
void xe_gt_remove(struct xe_gt *gt);
+void xe_gt_idle(struct xe_gt *gt);
/**
* xe_gt_any_hw_engine_by_reset_domain - scan the list of engines and return the
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH v3 4/4] drm/xe/FLR: Support PCIe FLR
2024-04-19 8:58 [PATCH v3 0/4] drm/xe: Support PCIe FLR Aravind Iddamsetty
` (2 preceding siblings ...)
2024-04-19 8:58 ` [PATCH v2 3/4] drm/xe: Extract xe_gt_idle() helper Aravind Iddamsetty
@ 2024-04-19 8:58 ` Aravind Iddamsetty
3 siblings, 0 replies; 11+ messages in thread
From: Aravind Iddamsetty @ 2024-04-19 8:58 UTC (permalink / raw)
To: dri-devel, daniel, maarten.lankhorst, airlied, mripard,
tzimmermann, rodrigo.vivi
Cc: intel-xe, Lucas De Marchi, Michal Wajdeczko
PCI subsystem provides callbacks to inform the driver about a request to
do function level reset by user, initiated by writing to sysfs entry
/sys/bus/pci/devices/.../reset. This will allow the driver to handle FLR
without the need to do unbind and rebind as the driver needs to
reinitialize the device afresh post FLR.
v2:
1. separate out gt idle and pci save/restore to a separate patch (Lucas)
2. Fixed the warnings seen around xe_guc_submit_stop, xe_guc_puc_fini
v3: declare xe_pci_err_handlers as static(Michal)
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@linux.intel.com>
---
drivers/gpu/drm/xe/Makefile | 1 +
drivers/gpu/drm/xe/xe_device_types.h | 3 +
drivers/gpu/drm/xe/xe_guc_pc.c | 4 ++
drivers/gpu/drm/xe/xe_pci.c | 9 ++-
drivers/gpu/drm/xe/xe_pci.h | 2 +
drivers/gpu/drm/xe/xe_pci_err.c | 88 ++++++++++++++++++++++++++++
drivers/gpu/drm/xe/xe_pci_err.h | 13 ++++
7 files changed, 119 insertions(+), 1 deletion(-)
create mode 100644 drivers/gpu/drm/xe/xe_pci_err.c
create mode 100644 drivers/gpu/drm/xe/xe_pci_err.h
diff --git a/drivers/gpu/drm/xe/Makefile b/drivers/gpu/drm/xe/Makefile
index 8bc62bfbc679..693971a1fac0 100644
--- a/drivers/gpu/drm/xe/Makefile
+++ b/drivers/gpu/drm/xe/Makefile
@@ -117,6 +117,7 @@ xe-y += xe_bb.o \
xe_module.o \
xe_pat.o \
xe_pci.o \
+ xe_pci_err.o \
xe_pcode.o \
xe_pm.o \
xe_preempt_fence.o \
diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
index 0a66555229e9..8c749b378a92 100644
--- a/drivers/gpu/drm/xe/xe_device_types.h
+++ b/drivers/gpu/drm/xe/xe_device_types.h
@@ -465,6 +465,9 @@ struct xe_device {
/** @pci_state: PCI state of device */
struct pci_saved_state *pci_state;
+ /** @pci_device_is_reset: device went through PCIe FLR */
+ bool pci_device_is_reset;
+
/* private: */
#if IS_ENABLED(CONFIG_DRM_XE_DISPLAY)
diff --git a/drivers/gpu/drm/xe/xe_guc_pc.c b/drivers/gpu/drm/xe/xe_guc_pc.c
index 509649d0e65e..efba0fbe2f5c 100644
--- a/drivers/gpu/drm/xe/xe_guc_pc.c
+++ b/drivers/gpu/drm/xe/xe_guc_pc.c
@@ -902,6 +902,10 @@ static void xe_guc_pc_fini(struct drm_device *drm, void *arg)
return;
}
+ /* We already have done this before going through a reset, so skip here */
+ if (xe->pci_device_is_reset)
+ return;
+
XE_WARN_ON(xe_force_wake_get(gt_to_fw(pc_to_gt(pc)), XE_FORCEWAKE_ALL));
XE_WARN_ON(xe_guc_pc_gucrc_disable(pc));
XE_WARN_ON(xe_guc_pc_stop(pc));
diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
index fd45f235b35c..98170773a9e1 100644
--- a/drivers/gpu/drm/xe/xe_pci.c
+++ b/drivers/gpu/drm/xe/xe_pci.c
@@ -23,6 +23,7 @@
#include "xe_macros.h"
#include "xe_mmio.h"
#include "xe_module.h"
+#include "xe_pci_err.h"
#include "xe_pci_types.h"
#include "xe_pm.h"
#include "xe_sriov.h"
@@ -736,7 +737,7 @@ static void xe_pci_remove(struct pci_dev *pdev)
pci_set_drvdata(pdev, NULL);
}
-static int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
+int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
{
const struct xe_device_desc *desc = (const void *)ent->driver_data;
const struct xe_subplatform_desc *subplatform_desc;
@@ -984,6 +985,11 @@ static const struct dev_pm_ops xe_pm_ops = {
};
#endif
+static const struct pci_error_handlers xe_pci_err_handlers = {
+ .reset_prepare = xe_pci_reset_prepare,
+ .reset_done = xe_pci_reset_done,
+};
+
static struct pci_driver xe_pci_driver = {
.name = DRIVER_NAME,
.id_table = pciidlist,
@@ -993,6 +999,7 @@ static struct pci_driver xe_pci_driver = {
#ifdef CONFIG_PM_SLEEP
.driver.pm = &xe_pm_ops,
#endif
+ .err_handler = &xe_pci_err_handlers,
};
int xe_register_pci_driver(void)
diff --git a/drivers/gpu/drm/xe/xe_pci.h b/drivers/gpu/drm/xe/xe_pci.h
index 73b90a430d1f..9faf5380a09e 100644
--- a/drivers/gpu/drm/xe/xe_pci.h
+++ b/drivers/gpu/drm/xe/xe_pci.h
@@ -7,8 +7,10 @@
#define _XE_PCI_H_
struct pci_dev;
+struct pci_device_id;
int xe_register_pci_driver(void);
void xe_unregister_pci_driver(void);
void xe_load_pci_state(struct pci_dev *pdev);
+int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent);
#endif
diff --git a/drivers/gpu/drm/xe/xe_pci_err.c b/drivers/gpu/drm/xe/xe_pci_err.c
new file mode 100644
index 000000000000..5306925ea2fa
--- /dev/null
+++ b/drivers/gpu/drm/xe/xe_pci_err.c
@@ -0,0 +1,88 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2024 Intel Corporation
+ */
+
+#include <linux/pci.h>
+#include <drm/drm_drv.h>
+
+#include "xe_device.h"
+#include "xe_gt.h"
+#include "xe_gt_printk.h"
+#include "xe_pci.h"
+#include "xe_pci_err.h"
+#include "xe_pm.h"
+#include "xe_uc.h"
+
+/**
+ * xe_pci_reset_prepare - Called when user issued a PCIe reset
+ * via /sys/bus/pci/devices/.../reset.
+ * @pdev: PCI device struct
+ */
+void xe_pci_reset_prepare(struct pci_dev *pdev)
+{
+ struct xe_device *xe = pci_get_drvdata(pdev);
+ struct xe_gt *gt;
+ int id, err;
+
+ pci_warn(pdev, "preparing for PCIe reset\n");
+
+ drm_warn(&xe->drm, "removing device access to userspace\n");
+ drm_dev_unplug(&xe->drm);
+
+ xe_pm_runtime_get(xe);
+ /* idle the GTs */
+ for_each_gt(gt, xe, id) {
+ err = xe_force_wake_get(gt_to_fw(gt), XE_FORCEWAKE_ALL);
+ if (err)
+ goto reset;
+ xe_uc_reset_prepare(>->uc);
+ xe_gt_idle(gt);
+ err = xe_force_wake_put(gt_to_fw(gt), XE_FORCEWAKE_ALL);
+ XE_WARN_ON(err);
+ }
+ xe_pm_runtime_put(xe);
+
+reset:
+ pci_disable_device(pdev);
+}
+
+/**
+ * xe_pci_reset_done - Called when PCIe reset is done.
+ * @pdev: PCI device struct
+ */
+void xe_pci_reset_done(struct pci_dev *pdev)
+{
+ const struct pci_device_id *ent = pci_match_id(pdev->driver->id_table, pdev);
+ struct xe_device *xe = pci_get_drvdata(pdev);
+
+ dev_info(&pdev->dev,
+ "device went through PCIe reset, reenabling the device\n");
+
+ if (pci_enable_device(pdev)) {
+ dev_err(&pdev->dev,
+ "Cannot re-enable PCI device after reset\n");
+ return;
+ }
+ pci_set_master(pdev);
+ xe_load_pci_state(pdev);
+
+ xe->pci_device_is_reset = true;
+ /*
+ * We want to completely clean the driver and even destroy
+ * the xe private data and reinitialize afresh similar to
+ * probe
+ */
+ pdev->driver->remove(pdev);
+ if (pci_dev_msi_enabled(pdev))
+ pci_free_irq_vectors(pdev);
+
+ devm_drm_dev_release_action(&xe->drm);
+ pci_disable_device(pdev);
+
+ /*
+ * if this fails the driver might be in a stale state, only option is
+ * to unbind and rebind
+ */
+ xe_pci_probe(pdev, ent);
+}
diff --git a/drivers/gpu/drm/xe/xe_pci_err.h b/drivers/gpu/drm/xe/xe_pci_err.h
new file mode 100644
index 000000000000..95a4c8ce9cf1
--- /dev/null
+++ b/drivers/gpu/drm/xe/xe_pci_err.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2024 Intel Corporation
+ */
+
+#ifndef _XE_PCI_ERR_H_
+#define _XE_PCI_ERR_H_
+
+struct pci_dev;
+
+void xe_pci_reset_prepare(struct pci_dev *pdev);
+void xe_pci_reset_done(struct pci_dev *pdev);
+#endif
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* Re: [PATCH v2 1/4] drm: add devm release action
2024-04-19 8:58 ` [PATCH v2 1/4] drm: add devm release action Aravind Iddamsetty
@ 2024-04-19 9:11 ` Maxime Ripard
2024-04-20 8:12 ` Aravind Iddamsetty
0 siblings, 1 reply; 11+ messages in thread
From: Maxime Ripard @ 2024-04-19 9:11 UTC (permalink / raw)
To: Aravind Iddamsetty
Cc: dri-devel, daniel, maarten.lankhorst, airlied, tzimmermann,
rodrigo.vivi, intel-xe, Thomas Hellstr_m
[-- Attachment #1: Type: text/plain, Size: 1236 bytes --]
On Fri, Apr 19, 2024 at 02:28:23PM +0530, Aravind Iddamsetty wrote:
> In scenarios where drm_dev_put is directly called by driver we want to
> release devm_drm_dev_init_release action associated with struct
> drm_device.
>
> v2: Directly expose the original function, instead of introducing a
> helper (Rodrigo)
>
> Cc: Thomas Hellstr_m <thomas.hellstrom@linux.intel.com>
> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
>
> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
> Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@linux.intel.com>
> ---
> drivers/gpu/drm/drm_drv.c | 6 ++++++
> include/drm/drm_drv.h | 2 ++
> 2 files changed, 8 insertions(+)
>
> diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
> index 243cacb3575c..ba60cbb0725f 100644
> --- a/drivers/gpu/drm/drm_drv.c
> +++ b/drivers/gpu/drm/drm_drv.c
> @@ -714,6 +714,12 @@ static int devm_drm_dev_init(struct device *parent,
> devm_drm_dev_init_release, dev);
> }
>
> +void devm_drm_dev_release_action(struct drm_device *dev)
> +{
> + devm_release_action(dev->dev, devm_drm_dev_init_release, dev);
> +}
> +EXPORT_SYMBOL(devm_drm_dev_release_action);
Again, this needs to be documented.
Maxime
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 273 bytes --]
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH v2 1/4] drm: add devm release action
2024-04-19 9:11 ` Maxime Ripard
@ 2024-04-20 8:12 ` Aravind Iddamsetty
0 siblings, 0 replies; 11+ messages in thread
From: Aravind Iddamsetty @ 2024-04-20 8:12 UTC (permalink / raw)
To: Maxime Ripard
Cc: dri-devel, daniel, maarten.lankhorst, airlied, tzimmermann,
rodrigo.vivi, intel-xe, Thomas Hellstr_m
On 19/04/24 14:41, Maxime Ripard wrote:
> On Fri, Apr 19, 2024 at 02:28:23PM +0530, Aravind Iddamsetty wrote:
>> In scenarios where drm_dev_put is directly called by driver we want to
>> release devm_drm_dev_init_release action associated with struct
>> drm_device.
>>
>> v2: Directly expose the original function, instead of introducing a
>> helper (Rodrigo)
>>
>> Cc: Thomas Hellstr_m <thomas.hellstrom@linux.intel.com>
>> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
>>
>> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
>> Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@linux.intel.com>
>> ---
>> drivers/gpu/drm/drm_drv.c | 6 ++++++
>> include/drm/drm_drv.h | 2 ++
>> 2 files changed, 8 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
>> index 243cacb3575c..ba60cbb0725f 100644
>> --- a/drivers/gpu/drm/drm_drv.c
>> +++ b/drivers/gpu/drm/drm_drv.c
>> @@ -714,6 +714,12 @@ static int devm_drm_dev_init(struct device *parent,
>> devm_drm_dev_init_release, dev);
>> }
>>
>> +void devm_drm_dev_release_action(struct drm_device *dev)
>> +{
>> + devm_release_action(dev->dev, devm_drm_dev_init_release, dev);
>> +}
>> +EXPORT_SYMBOL(devm_drm_dev_release_action);
> Again, this needs to be documented.
sorry I missed your earlier email, will address this.
Thanks,
Aravind.
>
> Maxime
^ permalink raw reply [flat|nested] 11+ messages in thread
* [PATCH 2/4] drm/xe: Save and restore PCI state
2024-04-22 6:57 [PATCH v4 0/4] drm/xe: " Aravind Iddamsetty
@ 2024-04-22 6:57 ` Aravind Iddamsetty
0 siblings, 0 replies; 11+ messages in thread
From: Aravind Iddamsetty @ 2024-04-22 6:57 UTC (permalink / raw)
To: dri-devel, daniel, maarten.lankhorst, airlied, mripard,
tzimmermann, rodrigo.vivi
Cc: intel-xe, Lucas De Marchi
Save and restore PCI states where ever needed.
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@linux.intel.com>
---
drivers/gpu/drm/xe/xe_device_types.h | 3 ++
drivers/gpu/drm/xe/xe_pci.c | 48 ++++++++++++++++++++++++++--
drivers/gpu/drm/xe/xe_pci.h | 4 ++-
3 files changed, 51 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
index 8244b177a6a3..0a66555229e9 100644
--- a/drivers/gpu/drm/xe/xe_device_types.h
+++ b/drivers/gpu/drm/xe/xe_device_types.h
@@ -462,6 +462,9 @@ struct xe_device {
/** @needs_flr_on_fini: requests function-reset on fini */
bool needs_flr_on_fini;
+ /** @pci_state: PCI state of device */
+ struct pci_saved_state *pci_state;
+
/* private: */
#if IS_ENABLED(CONFIG_DRM_XE_DISPLAY)
diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
index fb20c9828563..a62300990e19 100644
--- a/drivers/gpu/drm/xe/xe_pci.c
+++ b/drivers/gpu/drm/xe/xe_pci.c
@@ -393,6 +393,41 @@ MODULE_DEVICE_TABLE(pci, pciidlist);
#undef INTEL_VGA_DEVICE
+static bool xe_save_pci_state(struct pci_dev *pdev)
+{
+ struct xe_device *xe = pci_get_drvdata(pdev);
+
+ if (pci_save_state(pdev))
+ return false;
+
+ kfree(xe->pci_state);
+
+ xe->pci_state = pci_store_saved_state(pdev);
+ if (!xe->pci_state) {
+ drm_err(&xe->drm, "Failed to store PCI saved state\n");
+ return false;
+ }
+
+ return true;
+}
+
+void xe_load_pci_state(struct pci_dev *pdev)
+{
+ struct xe_device *xe = pci_get_drvdata(pdev);
+ int ret;
+
+ if (!xe->pci_state)
+ return;
+
+ ret = pci_load_saved_state(pdev, xe->pci_state);
+ if (ret) {
+ drm_warn(&xe->drm, "Failed to load PCI state err:%d\n", ret);
+ return;
+ }
+
+ pci_restore_state(pdev);
+}
+
/* is device_id present in comma separated list of ids */
static bool device_id_in_list(u16 device_id, const char *devices, bool negative)
{
@@ -698,6 +733,8 @@ static void xe_pci_remove(struct pci_dev *pdev)
xe_device_remove(xe);
xe_pm_runtime_fini(xe);
+
+ kfree(xe->pci_state);
pci_set_drvdata(pdev, NULL);
}
@@ -798,6 +835,9 @@ static int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
drm_dbg(&xe->drm, "d3cold: capable=%s\n",
str_yes_no(xe->d3cold.capable));
+ if (xe_save_pci_state(pdev))
+ pci_restore_state(pdev);
+
return 0;
err_driver_cleanup:
@@ -849,7 +889,7 @@ static int xe_pci_suspend(struct device *dev)
*/
d3cold_toggle(pdev, D3COLD_ENABLE);
- pci_save_state(pdev);
+ xe_save_pci_state(pdev);
pci_disable_device(pdev);
return 0;
@@ -873,6 +913,8 @@ static int xe_pci_resume(struct device *dev)
pci_set_master(pdev);
+ xe_load_pci_state(pdev);
+
err = xe_pm_resume(pdev_to_xe_device(pdev));
if (err)
return err;
@@ -890,7 +932,7 @@ static int xe_pci_runtime_suspend(struct device *dev)
if (err)
return err;
- pci_save_state(pdev);
+ xe_save_pci_state(pdev);
if (xe->d3cold.allowed) {
d3cold_toggle(pdev, D3COLD_ENABLE);
@@ -915,7 +957,7 @@ static int xe_pci_runtime_resume(struct device *dev)
if (err)
return err;
- pci_restore_state(pdev);
+ xe_load_pci_state(pdev);
if (xe->d3cold.allowed) {
err = pci_enable_device(pdev);
diff --git a/drivers/gpu/drm/xe/xe_pci.h b/drivers/gpu/drm/xe/xe_pci.h
index 611c1209b14c..73b90a430d1f 100644
--- a/drivers/gpu/drm/xe/xe_pci.h
+++ b/drivers/gpu/drm/xe/xe_pci.h
@@ -6,7 +6,9 @@
#ifndef _XE_PCI_H_
#define _XE_PCI_H_
+struct pci_dev;
+
int xe_register_pci_driver(void);
void xe_unregister_pci_driver(void);
-
+void xe_load_pci_state(struct pci_dev *pdev);
#endif
--
2.25.1
^ permalink raw reply related [flat|nested] 11+ messages in thread
* Re: [PATCH v2 3/4] drm/xe: Extract xe_gt_idle() helper
2024-04-19 8:58 ` [PATCH v2 3/4] drm/xe: Extract xe_gt_idle() helper Aravind Iddamsetty
@ 2024-04-23 4:21 ` Ghimiray, Himal Prasad
0 siblings, 0 replies; 11+ messages in thread
From: Ghimiray, Himal Prasad @ 2024-04-23 4:21 UTC (permalink / raw)
To: Aravind Iddamsetty, dri-devel, daniel, maarten.lankhorst, airlied,
mripard, tzimmermann, rodrigo.vivi
Cc: intel-xe, Lucas De Marchi, Michal Wajdeczko
[-- Attachment #1: Type: text/plain, Size: 2596 bytes --]
On 19-04-2024 14:28, Aravind Iddamsetty wrote:
> This would be used in other places outside of gt_reset path.
>
> v2:
> 1. Add kernel doc for xe_gt_idle (Michal)
> 2. fix return as no actual error is reported by xe_uc_stop (Himal)
>
> Cc: Lucas De Marchi<lucas.demarchi@intel.com>
> Cc: Michal Wajdeczko<michal.wajdeczko@intel.com>
> Cc: Himal Prasad Ghimiray<himal.prasad.ghimiray@intel.com>
>
> Reviewed-by: Rodrigo Vivi<rodrigo.vivi@intel.com>
> Signed-off-by: Aravind Iddamsetty<aravind.iddamsetty@linux.intel.com>
> ---
> drivers/gpu/drm/xe/xe_gt.c | 28 ++++++++++++++++++----------
> drivers/gpu/drm/xe/xe_gt.h | 1 +
> 2 files changed, 19 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/gpu/drm/xe/xe_gt.c b/drivers/gpu/drm/xe/xe_gt.c
> index 491d0413de15..a24340219c30 100644
> --- a/drivers/gpu/drm/xe/xe_gt.c
> +++ b/drivers/gpu/drm/xe/xe_gt.c
> @@ -629,6 +629,23 @@ static int do_gt_restart(struct xe_gt *gt)
> return 0;
> }
>
> +/**
> + * xe_gt_idle: Idle the GT.
> + * @gt: The gt object
> + *
> + * Typically called before initiating any resets.
> + *
> + */
> +void xe_gt_idle(struct xe_gt *gt)
> +{
> + xe_gt_sanitize(gt);
> + xe_uc_gucrc_disable(>->uc);
> + xe_uc_stop_prepare(>->uc);
> + xe_gt_pagefault_reset(gt);
> + xe_uc_stop(>->uc);
The cleanup of xe_uc_stop to return void needs to be addressed
independently of the patch.
This patch LGTM
Reviewed-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
> + xe_gt_tlb_invalidation_reset(gt);
> +}
> +
> static int gt_reset(struct xe_gt *gt)
> {
> int err;
> @@ -645,21 +662,12 @@ static int gt_reset(struct xe_gt *gt)
> }
>
> xe_pm_runtime_get(gt_to_xe(gt));
> - xe_gt_sanitize(gt);
>
> err = xe_force_wake_get(gt_to_fw(gt), XE_FORCEWAKE_ALL);
> if (err)
> goto err_msg;
>
> - xe_uc_gucrc_disable(>->uc);
> - xe_uc_stop_prepare(>->uc);
> - xe_gt_pagefault_reset(gt);
> -
> - err = xe_uc_stop(>->uc);
> - if (err)
> - goto err_out;
> -
> - xe_gt_tlb_invalidation_reset(gt);
> + xe_gt_idle(gt);
>
> err = do_gt_reset(gt);
> if (err)
> diff --git a/drivers/gpu/drm/xe/xe_gt.h b/drivers/gpu/drm/xe/xe_gt.h
> index ed6ea8057e35..dcc1023d20e8 100644
> --- a/drivers/gpu/drm/xe/xe_gt.h
> +++ b/drivers/gpu/drm/xe/xe_gt.h
> @@ -43,6 +43,7 @@ int xe_gt_resume(struct xe_gt *gt);
> void xe_gt_reset_async(struct xe_gt *gt);
> void xe_gt_sanitize(struct xe_gt *gt);
> void xe_gt_remove(struct xe_gt *gt);
> +void xe_gt_idle(struct xe_gt *gt);
>
> /**
> * xe_gt_any_hw_engine_by_reset_domain - scan the list of engines and return the
[-- Attachment #2: Type: text/html, Size: 39872 bytes --]
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2024-04-23 4:22 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-04-19 8:58 [PATCH v3 0/4] drm/xe: Support PCIe FLR Aravind Iddamsetty
2024-04-19 8:58 ` [PATCH v2 1/4] drm: add devm release action Aravind Iddamsetty
2024-04-19 9:11 ` Maxime Ripard
2024-04-20 8:12 ` Aravind Iddamsetty
2024-04-19 8:58 ` [PATCH 2/4] drm/xe: Save and restore PCI state Aravind Iddamsetty
2024-04-19 8:58 ` [PATCH v2 3/4] drm/xe: Extract xe_gt_idle() helper Aravind Iddamsetty
2024-04-23 4:21 ` Ghimiray, Himal Prasad
2024-04-19 8:58 ` [PATCH v3 4/4] drm/xe/FLR: Support PCIe FLR Aravind Iddamsetty
-- strict thread matches above, loose matches on Subject: below --
2024-04-22 6:57 [PATCH v4 0/4] drm/xe: " Aravind Iddamsetty
2024-04-22 6:57 ` [PATCH 2/4] drm/xe: Save and restore PCI state Aravind Iddamsetty
2024-04-17 8:41 [PATCH v2 0/4] drm/xe: Support PCIe FLR Aravind Iddamsetty
2024-04-17 8:41 ` [PATCH 2/4] drm/xe: Save and restore PCI state Aravind Iddamsetty
2024-04-02 8:58 [PATCH v2 0/4] drm/xe: Support PCIe FLR Aravind Iddamsetty
2024-04-02 8:58 ` [PATCH 2/4] drm/xe: Save and restore PCI state Aravind Iddamsetty
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox