Add new debug infrastructure for configfs
Subsystems (2)
cover Add new debug infrastructure for configfs
Add a new configfs debug group. The intent of this structure is to allow us to separate ABI facing configfs entries from those which are purely for debug purposes. And it allows us more flexibility in how we define those parameters used for debug.
Add a new infrastructure to this debug configfs group that lets us easily define the parameters in a quick list. This is primarily useful for simple, single-type parameters such as enable/disable features or simple values passed. For more complex parameters, we will still need to define these separately.
Pull the GuC target related changes from [1] to fit within this new structure and add a new definition for guc_log_level on top of the existing module parameter (to ensure we aren't impacting existing users of the module parameter).
Note that the debug parameters here are all to be used "at your own risk". Without having in depth knowledge of how these impact the software and hardware, there could be unforeseen consequences of setting them. As such, they are all wrapped in a CONFIG_DRM_XE_DEBUG configfs option.
In terms of the patches here, I'm also making a few organizational changes like naming and sorting. I also know we have a few other module parameters that could migrate here, but I didn't want to overload this series too much, so the focus for now is on the existing configfs entries and demonstrating the new structures with the GuC log level and target parameters.
I used GitHub Copilot with Claude pretty extensively through the process here and attributed as such. Happy to answer any questions around this.
I also decided to drop John Harrison from the NPK patch. It has been modified quite a bit from the original, but more importantly John is no longer with Intel and that email address isn't available any more. If it makes a difference here, John and I had both separately implemented this same change at different occasions for debug. The one I used to start that initial series was cherry-picked from his latest variant.
Tested on BMG.
[1] https://patchwork.freedesktop.org/series/162087/
Stuart Summers (9):
drm/xe: Rename survivability_mode to enable_survivability_mode drm/xe: Sort xe_config_device fields and defaults alphabetically drm/xe: Split out configfs data structures
drm/xe: Add a new debug focused configfs group drm/xe: Move debug configfs entries to xe_configfs_debug.c drm/xe/guc: Add configfs support for guc_log_level drm/xe/guc: Add support for NPK as a GuC log target drm/xe: Add infrastructure for debug configfs parameters drm/xe: Migrate existing debug configfs entries to params
infrastructure
drivers/gpu/drm/xe/Makefile | 2 + drivers/gpu/drm/xe/abi/guc_log_abi.h | 8 + drivers/gpu/drm/xe/xe_configfs.c | 952 ++---------------- drivers/gpu/drm/xe/xe_configfs.h | 23 - drivers/gpu/drm/xe/xe_configfs_debug.c | 900 +++++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug.h | 56 ++ drivers/gpu/drm/xe/xe_configfs_debug_params.c | 112 +++ drivers/gpu/drm/xe/xe_configfs_debug_params.h | 127 +++ drivers/gpu/drm/xe/xe_configfs_types.h | 73 ++ drivers/gpu/drm/xe/xe_defaults.h | 5 + drivers/gpu/drm/xe/xe_guc.c | 12 +- drivers/gpu/drm/xe/xe_guc_log.c | 3 +- drivers/gpu/drm/xe/xe_hw_engine.c | 1 + drivers/gpu/drm/xe/xe_lrc.c | 1 + drivers/gpu/drm/xe/xe_pci.c | 1 + drivers/gpu/drm/xe/xe_psmi.c | 1 + drivers/gpu/drm/xe/xe_rtp.c | 1 + drivers/gpu/drm/xe/xe_survivability_mode.c | 3 +- 18 files changed, 1369 insertions(+), 912 deletions(-) create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug.c create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug.h create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug_params.c create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug_params.h create mode 100644 drivers/gpu/drm/xe/xe_configfs_types.h
Signature · 1 line
2.43.0
PATCH 1/9 drm/xe: Rename survivability_mode to enable_survivability_mode
Rename the survivability_mode field in struct xe_config_device to enable_survivability_mode to be consistent with the enable_psmi naming convention used for similar boolean config flags.
--- drivers/gpu/drm/xe/xe_configfs.c | 39 +++++++++++----------- drivers/gpu/drm/xe/xe_configfs.h | 4 +-- drivers/gpu/drm/xe/xe_survivability_mode.c | 2 +- 3 files changed, 23 insertions(+), 22 deletions(-)
drivers/gpu/drm/xe/xe_configfs.c +20 -19
@@ -58,7 +58,7 @@* │ └── ... * : * └── 0000:03:00.0 -* ├── survivability_mode +* ├── enable_survivability_mode * ├── gt_types_allowed * ├── engines_allowed * └── enable_psmi
@@ -79,7 +79,7 @@* Enable survivability mode on supported cards. This setting only takes * effect when probing the device. Example to enable it:: * -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/survivability_mode +* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode * * This attribute can only be set before binding to the device. *
@@ -260,7 +260,7 @@ struct xe_config_group_device {u64 engines_allowed;struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ];struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; -bool survivability_mode; +bool enable_survivability_mode;bool enable_psmi;struct {unsigned int max_vfs;
@@ -279,7 +279,7 @@ struct xe_config_group_device {static const struct xe_config_device device_defaults = { .gt_types_allowed =U64_MAX , .engines_allowed =U64_MAX , - .survivability_mode = false, + .enable_survivability_mode = false, .enable_psmi = false, .sriov = { .max_vfs =XE_DEFAULT_MAX_VFS ,
@@ -357,20 +357,21 @@ static bool is_bound(struct xe_config_group_device *dev)return ret; } -static ssize_t survivability_mode_show (struct config_item *item,char *page) +static ssize_t enable_survivability_mode_show (struct config_item *item,char *page) {struct xe_config_device *dev =to_xe_config_device (item); -return sprintf (page,"%d\n" , dev->survivability_mode); +return sprintf (page,"%d\n" , dev->enable_survivability_mode); } -static ssize_t survivability_mode_store (struct config_item *item,const char *page,size_t len) +static ssize_t enable_survivability_mode_store (struct config_item *item,const char *page, +size_t len) {struct xe_config_group_device *dev =to_xe_config_group_device (item); -bool survivability_mode; +bool enable_survivability_mode;int ret; - ret =kstrtobool (page, &survivability_mode); + ret =kstrtobool (page, &enable_survivability_mode);if (ret)return ret;
@@ -378,7 +379,7 @@ static ssize_t survivability_mode_store(struct config_item *item, const char *paif (is_bound (dev))return -EBUSY ; - dev->config.survivability_mode = survivability_mode; + dev->config.enable_survivability_mode = enable_survivability_mode;return len; }
@@ -815,7 +816,7 @@ CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, survivability_mode); +CONFIGFS_ATTR(, enable_survivability_mode);static struct configfs_attribute *xe_config_device_attrs[] = { &attr_ctx_restore_mid_bb,
@@ -823,7 +824,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = { &attr_enable_psmi, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_survivability_mode, + &attr_enable_survivability_mode, NULL, };
@@ -847,7 +848,7 @@ static bool xe_config_device_is_visible(struct config_item *item, {struct xe_config_group_device *dev =to_xe_config_group_device (item); -if (attr == &attr_survivability_mode) { +if (attr == &attr_enable_survivability_mode) {if (!dev->desc->is_dgfx || dev->desc->platform <XE_BATTLEMAGE )return false; }
@@ -1098,7 +1099,7 @@ static void dump_custom_dev_config(struct pci_dev *pdev,PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed);PRI_CUSTOM_ATTR ("%llx" , engines_allowed);PRI_CUSTOM_ATTR ("%d" , enable_psmi); -PRI_CUSTOM_ATTR ("%d" , survivability_mode); +PRI_CUSTOM_ATTR ("%d" , enable_survivability_mode);PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);#undef PRI_CUSTOM_ATTR
@@ -1128,20 +1129,20 @@ void xe_configfs_check_device(struct pci_dev *pdev) }/** -* xe_configfs_get_survivability_mode - get configfs survivability mode attribute +* xe_configfs_get_enable_survivability_mode - get configfs survivability mode attribute * @pdev: pci device * -* Return: survivability_mode attribute in configfs +* Return: enable_survivability_mode attribute in configfs */ -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) {struct xe_config_group_device *dev =find_xe_config_group_device (pdev);bool mode;if (!dev) -return device_defaults.survivability_mode; +return device_defaults.enable_survivability_mode; - mode = dev->config.survivability_mode; + mode = dev->config.enable_survivability_mode;config_group_put (&dev->group);return mode;
drivers/gpu/drm/xe/xe_configfs.h +2 -2
@@ -18,7 +18,7 @@ struct pci_dev;int xe_configfs_init (void );void xe_configfs_exit (void );void xe_configfs_check_device (struct pci_dev *pdev); -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev); +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev);bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev);bool xe_configfs_media_gt_allowed (struct pci_dev *pdev);u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev);
@@ -37,7 +37,7 @@ bool xe_configfs_admin_only_pf(struct pci_dev *pdev);static inline int xe_configfs_init (void ) {return 0 ; }static inline void xe_configfs_exit (void ) { }static inline void xe_configfs_check_device (struct pci_dev *pdev) { } -static inline bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) {return false; } +static inline bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) {return false; }static inline bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev) {return true; }static inline bool xe_configfs_media_gt_allowed (struct pci_dev *pdev) {return true; }static inline u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev) {return U64_MAX ; }
drivers/gpu/drm/xe/xe_survivability_mode.c +1 -1
@@ -377,7 +377,7 @@ bool xe_survivability_mode_is_requested(struct xe_device *xe)if (!IS_DGFX (xe) ||IS_SRIOV_VF (xe) || xe->info.platform <XE_BATTLEMAGE )return false; - survivability_mode =xe_configfs_get_survivability_mode (pdev); + survivability_mode =xe_configfs_get_enable_survivability_mode (pdev);/* Enable survivability mode if set via configfs */ if (survivability_mode)return true;
Signature · 1 line
2.43.0
Re: [PATCH 1/9] drm/xe: Rename survivability_mode to enable_survivability_mode
Quoted 26 lines · Stuart Summers
Rename the
survivability_modefield in structxe_config_devicetoenable_survivability_modeto be consistent with theenable_psminaming convention used for similar boolean config flags.Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7--- drivers/gpu/drm/xe/xe_configfs.c | 39 +++++++++++----------- drivers/gpu/drm/xe/xe_configfs.h | 4 +-- drivers/gpu/drm/xe/xe_survivability_mode.c | 2 +- 3 files changed, 23 insertions(+), 22 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -58,7 +58,7 @@* │ └── ... * : * └── 0000:03:00.0 -* ├── survivability_mode +* ├── enable_survivability_mode * ├── gt_types_allowed * ├── engines_allowed * └── enable_psmi
drivers/gpu/drm/xe/xe_configfs.c line 82
Quoted 5 lines
drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -79,7 +79,7 @@* Enable survivability mode on supported cards. This setting only takes * effect when probing the device. Example to enable it:: * -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/survivability_mode +* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode
I think we have not documented our configfs structure in the kernel ABI documentation. Are we supposed to do that? Is that not yet done because we do not consider our configfs stable enough to be classified as proper ABI?
In this example, we would be potentially changing what some would call ABI, hence my questions above.
--
Gustavo Sousa
Quoted 156 lines
drivers/gpu/drm/xe/xe_configfs.c +18
-17* * This attribute can only be set before binding to the device. * @@ -260,7 +260,7 @@ struct xe_config_group_device {u64 engines_allowed;struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ];struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; -bool survivability_mode; +bool enable_survivability_mode;bool enable_psmi;struct {unsigned int max_vfs;@@ -279,7 +279,7 @@ struct xe_config_group_device {static const struct xe_config_device device_defaults = { .gt_types_allowed =U64_MAX , .engines_allowed =U64_MAX , - .survivability_mode = false, + .enable_survivability_mode = false, .enable_psmi = false, .sriov = { .max_vfs =XE_DEFAULT_MAX_VFS ,@@ -357,20 +357,21 @@ static bool is_bound(struct xe_config_group_device *dev)return ret; } -static ssize_t survivability_mode_show (struct config_item *item,char *page) +static ssize_t enable_survivability_mode_show (struct config_item *item,char *page) {struct xe_config_device *dev =to_xe_config_device (item); -return sprintf (page,"%d\n" , dev->survivability_mode); +return sprintf (page,"%d\n" , dev->enable_survivability_mode); } -static ssize_t survivability_mode_store (struct config_item *item,const char *page,size_t len) +static ssize_t enable_survivability_mode_store (struct config_item *item,const char *page, +size_t len) {struct xe_config_group_device *dev =to_xe_config_group_device (item); -bool survivability_mode; +bool enable_survivability_mode;int ret; - ret =kstrtobool (page, &survivability_mode); + ret =kstrtobool (page, &enable_survivability_mode);if (ret)return ret;@@ -378,7 +379,7 @@ static ssize_t survivability_mode_store(struct config_item *item, const char *paif (is_bound (dev))return -EBUSY ; - dev->config.survivability_mode = survivability_mode; + dev->config.enable_survivability_mode = enable_survivability_mode;return len; }@@ -815,7 +816,7 @@ CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, survivability_mode); +CONFIGFS_ATTR(, enable_survivability_mode);static struct configfs_attribute *xe_config_device_attrs[] = { &attr_ctx_restore_mid_bb,@@ -823,7 +824,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = { &attr_enable_psmi, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_survivability_mode, + &attr_enable_survivability_mode, NULL, };@@ -847,7 +848,7 @@ static bool xe_config_device_is_visible(struct config_item *item, {struct xe_config_group_device *dev =to_xe_config_group_device (item); -if (attr == &attr_survivability_mode) { +if (attr == &attr_enable_survivability_mode) {if (!dev->desc->is_dgfx || dev->desc->platform <XE_BATTLEMAGE )return false; }@@ -1098,7 +1099,7 @@ static void dump_custom_dev_config(struct pci_dev *pdev,PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed);PRI_CUSTOM_ATTR ("%llx" , engines_allowed);PRI_CUSTOM_ATTR ("%d" , enable_psmi); -PRI_CUSTOM_ATTR ("%d" , survivability_mode); +PRI_CUSTOM_ATTR ("%d" , enable_survivability_mode);PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);#undef PRI_CUSTOM_ATTR@@ -1128,20 +1129,20 @@ void xe_configfs_check_device(struct pci_dev *pdev) }/** -* xe_configfs_get_survivability_mode - get configfs survivability mode attribute +* xe_configfs_get_enable_survivability_mode - get configfs survivability mode attribute * @pdev: pci device * -* Return: survivability_mode attribute in configfs +* Return: enable_survivability_mode attribute in configfs */ -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) {struct xe_config_group_device *dev =find_xe_config_group_device (pdev);bool mode;if (!dev) -return device_defaults.survivability_mode; +return device_defaults.enable_survivability_mode; - mode = dev->config.survivability_mode; + mode = dev->config.enable_survivability_mode;config_group_put (&dev->group);return mode;drivers/gpu/drm/xe/xe_configfs.h +2
-2@@ -18,7 +18,7 @@ struct pci_dev;int xe_configfs_init (void );void xe_configfs_exit (void );void xe_configfs_check_device (struct pci_dev *pdev); -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev); +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev);bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev);bool xe_configfs_media_gt_allowed (struct pci_dev *pdev);u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev);@@ -37,7 +37,7 @@ bool xe_configfs_admin_only_pf(struct pci_dev *pdev);static inline int xe_configfs_init (void ) {return 0 ; }static inline void xe_configfs_exit (void ) { }static inline void xe_configfs_check_device (struct pci_dev *pdev) { } -static inline bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) {return false; } +static inline bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) {return false; }static inline bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev) {return true; }static inline bool xe_configfs_media_gt_allowed (struct pci_dev *pdev) {return true; }static inline u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev) {return U64_MAX ; }drivers/gpu/drm/xe/xe_survivability_mode.c +1
-1@@ -377,7 +377,7 @@ bool xe_survivability_mode_is_requested(struct xe_device *xe)if (!IS_DGFX (xe) ||IS_SRIOV_VF (xe) || xe->info.platform <XE_BATTLEMAGE )return false; - survivability_mode =xe_configfs_get_survivability_mode (pdev); + survivability_mode =xe_configfs_get_enable_survivability_mode (pdev);/* Enable survivability mode if set via configfs */ if (survivability_mode)return true;Signature · 1 line
2.43.0
Re: [PATCH 1/9] drm/xe: Rename survivability_mode to enable_survivability_mode
Quoted 47 lines · Gustavo Sousa
Quoted 30 lines · Stuart Summers
Rename the
survivability_modefield in structxe_config_devicetoenable_survivability_modeto be consistent with theenable_psminaming
convention used for similar boolean config flags.Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 39 +++++++++++-------
----
drivers/gpu/drm/xe/xe_configfs.h | 4 +--drivers/gpu/drm/xe/xe_survivability_mode.c| 2 +-3 files changed, 23 insertions(+), 22 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -58,7 +58,7 @@ * │ └── ... * : * └── 0000:03:00.0 - * ├── survivability_mode + * ├── enable_survivability_mode * ├── gt_types_allowed * ├── engines_allowed * └── enable_psmi@@ -79,7 +79,7 @@ * Enable survivability mode on supported cards. This setting onlyQuoted 7 lines
takes
* effect when probing the device. Example to enable it::
*
- * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/survivability_mode
+ * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_modeI think we have not documented our configfs structure in the kernel ABI
documentation. Are we supposed to do that? Is that not yet done because we do not consider our configfs stable enough to be classified
as proper ABI?
Quoted 3 lines
In this example, we would be potentially changing what some would call
ABI, hence my questions above.
Yeah this is something I wanted to discuss here too. I kind of just blindly alphabetized things here, but we might want to sort by data type or something else. also, we only have a few parameters here, but I'd like to set this up to support a much larger set of parameters on the debug side - we need much finer grained support for things like power-on for instance. And I'd like a way that easily scales here.
Also I agree on the documentation.. I'll make sure to add that in the next patch once we conclude on the approach here.
Thanks,
Stuart
Quoted 196 lines
--
Gustavo SousaQuoted 192 lines
*
* This attribute can only be set before binding to the device. *drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -260,7 +260,7 @@ struct xe_config_group_device { u64 engines_allowed;struct
wa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
- boolsurvivability_mode;
+ boolenable_survivability_mode;
boolenable_psmi;
struct {
unsigned intmax_vfs;drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -279,7 +279,7 @@ struct xe_config_group_device { static const struct xe_config_device device_defaults = { .gt_types_allowed = U64_MAX, .engines_allowed = U64_MAX, - .survivability_mode = false, + .enable_survivability_mode = false, .enable_psmi = false, .sriov = { .max_vfs = XE_DEFAULT_MAX_VFS,@@ -357,20 +357,21 @@ static bool is_bound(struct xe_config_group_device *dev) return ret; }
-staticssize_tsurvivability_mode_show(struct config_item *item, char *page)
+staticssize_tenable_survivability_mode_show(struct config_item *item, char *page)
{
structxe_config_device*dev =to_xe_config_device(item);
- returnsprintf(page, "%d\n", dev->survivability_mode); + return sprintf(page, "%d\n", dev-Quoted 1 line
enable_survivability_mode);}
-staticssize_tsurvivability_mode_store(struct config_item *item, const char *page, size_t len)
+staticssize_tenable_survivability_mode_store(structconfig_item*item, const char *page,
+size_tlen) {
structxe_config_group_device*dev =to_xe_config_group_device(item);
- boolsurvivability_mode;
+ boolenable_survivability_mode;
int ret;
- ret =kstrtobool(page, &survivability_mode);
+ ret =kstrtobool(page, &enable_survivability_mode); if (ret)
return ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -378,7 +379,7 @@ static ssize_t survivability_mode_store(struct
config_item*item, const char *pa
if (is_bound(dev))
return-EBUSY;
- dev->config.survivability_mode =survivability_mode; + dev->config.enable_survivability_mode =enable_survivability_mode;
return len;
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -815,7 +816,7 @@ CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, survivability_mode); +CONFIGFS_ATTR(, enable_survivability_mode);
static structconfigfs_attribute*xe_config_device_attrs[] = {&attr_ctx_restore_mid_bb,drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -823,7 +824,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = { &attr_enable_psmi, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_survivability_mode, + &attr_enable_survivability_mode, NULL, };drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -847,7 +848,7 @@ static bool xe_config_device_is_visible(struct config_item *item, {struct
xe_config_group_device*dev =to_xe_config_group_device(item);
- if (attr ==&attr_survivability_mode) {
+ if (attr ==&attr_enable_survivability_mode) { if (!dev->desc->is_dgfx ||dev->desc->platform<XE_BATTLEMAGE)
return false;
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -1098,7 +1099,7 @@ static void dump_custom_dev_config(structpci_dev *pdev, PRI_CUSTOM_ATTR("%llx", gt_types_allowed); PRI_CUSTOM_ATTR("%llx", engines_allowed); PRI_CUSTOM_ATTR("%d", enable_psmi); - PRI_CUSTOM_ATTR("%d", survivability_mode); + PRI_CUSTOM_ATTR("%d", enable_survivability_mode); PRI_CUSTOM_ATTR("%u", sriov.admin_only_pf);
#undefPRI_CUSTOM_ATTRdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1128,20 +1129,20 @@ void xe_configfs_check_device(struct pci_dev *pdev) }
/**
- *xe_configfs_get_survivability_mode- get configfs survivability mode attribute
+ *xe_configfs_get_enable_survivability_mode- get configfs survivability mode attribute
* @pdev: pci device
*
- * Return:survivability_modeattribute in configfs
+ * Return:enable_survivability_modeattribute in configfs */drivers/gpu/drm/xe/xe_configfs.c +1
-1-bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) +bool xe_configfs_get_enable_survivability_mode(struct pci_dev*pdev)
{
structxe_config_group_device*dev =find_xe_config_group_device(pdev);
bool mode;
if (!dev)
- return device_defaults.survivability_mode; + return device_defaults.enable_survivability_mode;
- mode =dev->config.survivability_mode;
+ mode =dev->config.enable_survivability_mode;config_group_put(&dev->group);
return mode;drivers/gpu/drm/xe/xe_configfs.h +1
-1@@ -18,7 +18,7 @@ struct pci_dev; int xe_configfs_init(void); void xe_configfs_exit(void); void xe_configfs_check_device(struct pci_dev *pdev); -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev); +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev); bool xe_configfs_primary_gt_allowed(struct pci_dev *pdev); bool xe_configfs_media_gt_allowed(struct pci_dev *pdev);u64
xe_configfs_get_engines_allowed(struct pci_dev *pdev);drivers/gpu/drm/xe/xe_configfs.h +1
-1@@ -37,7 +37,7 @@ bool xe_configfs_admin_only_pf(struct pci_dev *pdev); static inline int xe_configfs_init(void) { return 0; } static inline void xe_configfs_exit(void) { } static inline void xe_configfs_check_device(struct pci_dev *pdev) { } -static inline bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) { return false; } +static inline bool
xe_configfs_get_enable_survivability_mode(struct pci_dev *pdev){ return false; }
static inline boolxe_configfs_primary_gt_allowed(struct pci_dev *pdev){ return true; }
static inline boolxe_configfs_media_gt_allowed(struct pci_dev *pdev){ return true; }
static inline u64xe_configfs_get_engines_allowed(struct pci_dev *pdev){ returnU64_MAX; }drivers/gpu/drm/xe/xe_survivability_mode.c +0
-0@@ -377,7 +377,7 @@ bool xe_survivability_mode_is_requested(struct xe_device *xe)if (!IS_DGFX(xe) ||
IS_SRIOV_VF(xe)||xe->info.platform<XE_BATTLEMAGE)
return false;
- survivability_mode =xe_configfs_get_survivability_mode(pdev);
+ survivability_mode =xe_configfs_get_enable_survivability_mode(pdev);
/* Enable survivability mode if set via configfs */ if (survivability_mode)
return true;Signature · 1 line
2.43.0
Re: [PATCH 1/9] drm/xe: Rename survivability_mode to enable_survivability_mode
Quoted 61 lines · Stuart
Quoted 49 lines · Gustavo Sousa
Quoted 31 lines · Stuart Summers
Rename the
survivability_modefield in structxe_config_devicetoenable_survivability_modeto be consistent with theenable_psminaming
convention used for similar boolean config flags.Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 39 +++++++++++----- --
----
drivers/gpu/drm/xe/xe_configfs.h | 4 +--drivers/gpu/drm/xe/xe_survivability_mode.c| 2 +-3 files changed, 23 insertions(+), 22 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -58,7 +58,7 @@ * │ └── ... * : * └── 0000:03:00.0 - * ├── survivability_mode + * ├── enable_survivability_mode * ├── gt_types_allowed * ├── engines_allowed * └── enable_psmi@@ -79,7 +79,7 @@ * Enable survivability mode on supported cards. This settingQuoted 8 lines
only
takes
* effect when probing the device. Example to enable it::
*
- * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/survivability_mode
+ * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_modeI think we have not documented our configfs structure in the kernel ABI
documentation. Are we supposed to do that? Is that not yet done because we do not consider our configfs stable enough to be classified
as proper ABI?Quoted 3 lines
In this example, we would be potentially changing what some would call
ABI, hence my questions above.Yeah this is something I wanted to discuss here too. I kind of just blindly alphabetized things here, but we might want to sort by data type or something else. also, we only have a few parameters here, but I'd like to set this up to support a much larger set of parameters on the debug side - we need much finer grained support for things like power-on for instance. And I'd like a way that easily scales here.
Quoted 2 lines
Also I agree on the documentation.. I'll make sure to add that in the next patch once we conclude on the approach here.
And I'm realizing I responded to the wrong patch here, sorry! :(
-Stuart
Quoted 209 lines
Thanks,
StuartQuoted 204 lines
--
Gustavo SousaQuoted 200 lines
*
* This attribute can only be set before binding to the device. *drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -260,7 +260,7 @@ struct xe_config_group_device { u64 engines_allowed;struct
wa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
- boolsurvivability_mode;
+ boolenable_survivability_mode;
boolenable_psmi;
struct {
unsigned intmax_vfs;drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -279,7 +279,7 @@ struct xe_config_group_device { static const struct xe_config_device device_defaults = { .gt_types_allowed = U64_MAX, .engines_allowed = U64_MAX, - .survivability_mode = false, + .enable_survivability_mode = false, .enable_psmi = false, .sriov = { .max_vfs = XE_DEFAULT_MAX_VFS,@@ -357,20 +357,21 @@ static bool is_bound(struct xe_config_group_device *dev) return ret; }
-staticssize_tsurvivability_mode_show(struct config_item *item, char *page)
+staticssize_tenable_survivability_mode_show(struct config_item *item, char *page)
{
structxe_config_device*dev =to_xe_config_device(item);
- returnsprintf(page, "%d\n", dev->survivability_mode); + return sprintf(page, "%d\n", dev-Quoted 1 line
enable_survivability_mode);}
-staticssize_tsurvivability_mode_store(structconfig_item*item,
const char *page,size_tlen)
+staticssize_tenable_survivability_mode_store(structconfig_item
*item, const char *page,
+size_tlen) {
structxe_config_group_device*dev =to_xe_config_group_device(item);
- boolsurvivability_mode;
+ boolenable_survivability_mode;
int ret;
- ret =kstrtobool(page, &survivability_mode); + ret =kstrtobool(page, &enable_survivability_mode); if (ret)
return ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -378,7 +379,7 @@ static ssize_tsurvivability_mode_store(struct
config_item*item, const char *pa
if (is_bound(dev))
return-EBUSY;
- dev->config.survivability_mode =survivability_mode; + dev->config.enable_survivability_mode =enable_survivability_mode;
return len;
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -815,7 +816,7 @@ CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, survivability_mode); +CONFIGFS_ATTR(, enable_survivability_mode);
static structconfigfs_attribute*xe_config_device_attrs[] = {&attr_ctx_restore_mid_bb,drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -823,7 +824,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = { &attr_enable_psmi, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_survivability_mode, + &attr_enable_survivability_mode, NULL, };drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -847,7 +848,7 @@ static boolxe_config_device_is_visible(struct
config_item*item,
{
structxe_config_group_device*dev =to_xe_config_group_device(item);
- if (attr ==&attr_survivability_mode) {
+ if (attr ==&attr_enable_survivability_mode) { if (!dev->desc->is_dgfx ||dev->desc->platform<XE_BATTLEMAGE)
return false;
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -1098,7 +1099,7 @@ static void dump_custom_dev_config(structpci_dev *pdev, PRI_CUSTOM_ATTR("%llx", gt_types_allowed); PRI_CUSTOM_ATTR("%llx", engines_allowed); PRI_CUSTOM_ATTR("%d", enable_psmi); - PRI_CUSTOM_ATTR("%d", survivability_mode); + PRI_CUSTOM_ATTR("%d", enable_survivability_mode); PRI_CUSTOM_ATTR("%u", sriov.admin_only_pf);
#undefPRI_CUSTOM_ATTRdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1128,20 +1129,20 @@ void xe_configfs_check_device(struct pci_dev *pdev) }
/**
- *xe_configfs_get_survivability_mode- get configfs survivability
mode attribute
+ *xe_configfs_get_enable_survivability_mode- get configfs survivability mode attribute
* @pdev: pci device
*
- * Return:survivability_modeattribute in configfs
+ * Return:enable_survivability_modeattribute in configfs */drivers/gpu/drm/xe/xe_configfs.c +1
-1-bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) +bool xe_configfs_get_enable_survivability_mode(struct pci_dev*pdev)
{
structxe_config_group_device*dev =find_xe_config_group_device(pdev);
bool mode;
if (!dev)
- return device_defaults.survivability_mode; + return device_defaults.enable_survivability_mode;
- mode =dev->config.survivability_mode;
+ mode =dev->config.enable_survivability_mode;config_group_put(&dev->group);
return mode;drivers/gpu/drm/xe/xe_configfs.h +1
-1@@ -18,7 +18,7 @@ struct pci_dev; int xe_configfs_init(void); void xe_configfs_exit(void); void xe_configfs_check_device(struct pci_dev *pdev); -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev); +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev); bool xe_configfs_primary_gt_allowed(struct pci_dev *pdev); bool xe_configfs_media_gt_allowed(struct pci_dev *pdev);u64
xe_configfs_get_engines_allowed(struct pci_dev *pdev);drivers/gpu/drm/xe/xe_configfs.h +0
-0@@ -37,7 +37,7 @@ bool xe_configfs_admin_only_pf(struct pci_dev *pdev); static inline int xe_configfs_init(void) { return 0; } static inline void xe_configfs_exit(void) { }static inline void
xe_configfs_check_device(struct pci_dev *pdev)
{ }
-static inline boolxe_configfs_get_survivability_mode(struct pci_dev *pdev){ return false; }
+static inline boolxe_configfs_get_enable_survivability_mode(struct pci_dev *pdev){ return false; }
static inline boolxe_configfs_primary_gt_allowed(struct pci_dev *pdev){ return true; }
static inline boolxe_configfs_media_gt_allowed(struct pci_dev *pdev){ return true; }
static inline u64xe_configfs_get_engines_allowed(struct pci_dev *pdev){ returnU64_MAX; }drivers/gpu/drm/xe/xe_survivability_mode.c +0
-0@@ -377,7 +377,7 @@ boolxe_survivability_mode_is_requested(struct
xe_device*xe)
if (!IS_DGFX(xe) ||IS_SRIOV_VF(xe)||xe->info.platform<XE_BATTLEMAGE)
return false;
- survivability_mode =xe_configfs_get_survivability_mode(pdev);
+ survivability_mode =xe_configfs_get_enable_survivability_mode(pdev);
/* Enable survivability mode if set via configfs */ if (survivability_mode)
return true;Signature · 1 line
2.43.0
Re: [PATCH 1/9] drm/xe: Rename survivability_mode to enable_survivability_mode
In reply to Gustavo SousaQuoted 47 lines · Gustavo Sousa
Quoted 30 lines · Stuart Summers
Rename the
survivability_modefield in structxe_config_devicetoenable_survivability_modeto be consistent with theenable_psminaming
convention used for similar boolean config flags.Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 39 +++++++++++-------
----
drivers/gpu/drm/xe/xe_configfs.h | 4 +--drivers/gpu/drm/xe/xe_survivability_mode.c| 2 +-3 files changed, 23 insertions(+), 22 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -58,7 +58,7 @@ * │ └── ... * : * └── 0000:03:00.0 - * ├── survivability_mode + * ├── enable_survivability_mode * ├── gt_types_allowed * ├── engines_allowed * └── enable_psmi@@ -79,7 +79,7 @@ * Enable survivability mode on supported cards. This setting onlyQuoted 7 lines
takes
* effect when probing the device. Example to enable it::
*
- * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/survivability_mode
+ * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_modeI think we have not documented our configfs structure in the kernel ABI
documentation. Are we supposed to do that? Is that not yet done because we do not consider our configfs stable enough to be classified
as proper ABI?
Quoted 3 lines
In this example, we would be potentially changing what some would call
ABI, hence my questions above.
So yeah the purpose of this series really is to set us up for ABI. But I want to get concensus that doing this makes sense and won't break any existing use cases - these parameters are already here of course. The guc_log_level module parameter (later in the series) is another one I want to eventually drop and replace with configfs, but that has probably the most active users today.
Thanks,
Stuart
Quoted 196 lines
--
Gustavo SousaQuoted 192 lines
*
* This attribute can only be set before binding to the device. *drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -260,7 +260,7 @@ struct xe_config_group_device { u64 engines_allowed;struct
wa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
- boolsurvivability_mode;
+ boolenable_survivability_mode;
boolenable_psmi;
struct {
unsigned intmax_vfs;drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -279,7 +279,7 @@ struct xe_config_group_device { static const struct xe_config_device device_defaults = { .gt_types_allowed = U64_MAX, .engines_allowed = U64_MAX, - .survivability_mode = false, + .enable_survivability_mode = false, .enable_psmi = false, .sriov = { .max_vfs = XE_DEFAULT_MAX_VFS,@@ -357,20 +357,21 @@ static bool is_bound(struct xe_config_group_device *dev) return ret; }
-staticssize_tsurvivability_mode_show(struct config_item *item, char *page)
+staticssize_tenable_survivability_mode_show(struct config_item *item, char *page)
{
structxe_config_device*dev =to_xe_config_device(item);
- returnsprintf(page, "%d\n", dev->survivability_mode); + return sprintf(page, "%d\n", dev-Quoted 1 line
enable_survivability_mode);}
-staticssize_tsurvivability_mode_store(struct config_item *item, const char *page, size_t len)
+staticssize_tenable_survivability_mode_store(structconfig_item*item, const char *page,
+size_tlen) {
structxe_config_group_device*dev =to_xe_config_group_device(item);
- boolsurvivability_mode;
+ boolenable_survivability_mode;
int ret;
- ret =kstrtobool(page, &survivability_mode);
+ ret =kstrtobool(page, &enable_survivability_mode); if (ret)
return ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -378,7 +379,7 @@ static ssize_t survivability_mode_store(struct
config_item*item, const char *pa
if (is_bound(dev))
return-EBUSY;
- dev->config.survivability_mode =survivability_mode; + dev->config.enable_survivability_mode =enable_survivability_mode;
return len;
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -815,7 +816,7 @@ CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, survivability_mode); +CONFIGFS_ATTR(, enable_survivability_mode);
static structconfigfs_attribute*xe_config_device_attrs[] = {&attr_ctx_restore_mid_bb,drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -823,7 +824,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = { &attr_enable_psmi, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_survivability_mode, + &attr_enable_survivability_mode, NULL, };drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -847,7 +848,7 @@ static bool xe_config_device_is_visible(struct config_item *item, {struct
xe_config_group_device*dev =to_xe_config_group_device(item);
- if (attr ==&attr_survivability_mode) {
+ if (attr ==&attr_enable_survivability_mode) { if (!dev->desc->is_dgfx ||dev->desc->platform<XE_BATTLEMAGE)
return false;
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -1098,7 +1099,7 @@ static void dump_custom_dev_config(structpci_dev *pdev, PRI_CUSTOM_ATTR("%llx", gt_types_allowed); PRI_CUSTOM_ATTR("%llx", engines_allowed); PRI_CUSTOM_ATTR("%d", enable_psmi); - PRI_CUSTOM_ATTR("%d", survivability_mode); + PRI_CUSTOM_ATTR("%d", enable_survivability_mode); PRI_CUSTOM_ATTR("%u", sriov.admin_only_pf);
#undefPRI_CUSTOM_ATTRdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1128,20 +1129,20 @@ void xe_configfs_check_device(struct pci_dev *pdev) }
/**
- *xe_configfs_get_survivability_mode- get configfs survivability mode attribute
+ *xe_configfs_get_enable_survivability_mode- get configfs survivability mode attribute
* @pdev: pci device
*
- * Return:survivability_modeattribute in configfs
+ * Return:enable_survivability_modeattribute in configfs */drivers/gpu/drm/xe/xe_configfs.c +1
-1-bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) +bool xe_configfs_get_enable_survivability_mode(struct pci_dev*pdev)
{
structxe_config_group_device*dev =find_xe_config_group_device(pdev);
bool mode;
if (!dev)
- return device_defaults.survivability_mode; + return device_defaults.enable_survivability_mode;
- mode =dev->config.survivability_mode;
+ mode =dev->config.enable_survivability_mode;config_group_put(&dev->group);
return mode;drivers/gpu/drm/xe/xe_configfs.h +1
-1@@ -18,7 +18,7 @@ struct pci_dev; int xe_configfs_init(void); void xe_configfs_exit(void); void xe_configfs_check_device(struct pci_dev *pdev); -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev); +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev); bool xe_configfs_primary_gt_allowed(struct pci_dev *pdev); bool xe_configfs_media_gt_allowed(struct pci_dev *pdev);u64
xe_configfs_get_engines_allowed(struct pci_dev *pdev);drivers/gpu/drm/xe/xe_configfs.h +1
-1@@ -37,7 +37,7 @@ bool xe_configfs_admin_only_pf(struct pci_dev *pdev); static inline int xe_configfs_init(void) { return 0; } static inline void xe_configfs_exit(void) { } static inline void xe_configfs_check_device(struct pci_dev *pdev) { } -static inline bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) { return false; } +static inline bool
xe_configfs_get_enable_survivability_mode(struct pci_dev *pdev){ return false; }
static inline boolxe_configfs_primary_gt_allowed(struct pci_dev *pdev){ return true; }
static inline boolxe_configfs_media_gt_allowed(struct pci_dev *pdev){ return true; }
static inline u64xe_configfs_get_engines_allowed(struct pci_dev *pdev){ returnU64_MAX; }drivers/gpu/drm/xe/xe_survivability_mode.c +0
-0@@ -377,7 +377,7 @@ bool xe_survivability_mode_is_requested(struct xe_device *xe)if (!IS_DGFX(xe) ||
IS_SRIOV_VF(xe)||xe->info.platform<XE_BATTLEMAGE)
return false;
- survivability_mode =xe_configfs_get_survivability_mode(pdev);
+ survivability_mode =xe_configfs_get_enable_survivability_mode(pdev);
/* Enable survivability mode if set via configfs */ if (survivability_mode)
return true;Signature · 1 line
2.43.0
Rodrigo Vivi
maintainerRe: [PATCH 1/9] drm/xe: Rename survivability_mode to enable_survivability_mode
In reply to Gustavo SousaQuoted 5 lines · Gustavo Sousa
Quoted 3 lines · Stuart Summers
Rename the
survivability_modefield in structxe_config_devicetoenable_survivability_modeto be consistent with theenable_psminaming convention used for similar boolean config flags.
I don't believe we need this level of consistency.
I'd like to keep the survivability_mode as is because perhaps we could perhaps even expand to multiple different modes later instead of only enable disable flag?!
Quoted 196 lines
Quoted 23 lines
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7--- drivers/gpu/drm/xe/xe_configfs.c | 39 +++++++++++----------- drivers/gpu/drm/xe/xe_configfs.h | 4 +-- drivers/gpu/drm/xe/xe_survivability_mode.c | 2 +- 3 files changed, 23 insertions(+), 22 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -58,7 +58,7 @@* │ └── ... * : * └── 0000:03:00.0 -* ├── survivability_mode +* ├── enable_survivability_mode * ├── gt_types_allowed * ├── engines_allowed * └── enable_psmi drivers/gpu/drm/xe/xe_configfs.c line 82
Quoted 5 lines
drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -79,7 +79,7 @@* Enable survivability mode on supported cards. This setting only takes * effect when probing the device. Example to enable it:: * -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/survivability_mode +* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode I think we have not documented our configfs structure in the kernel ABI documentation. Are we supposed to do that? Is that not yet done because we do not consider our configfs stable enough to be classified as proper ABI?
In this example, we would be potentially changing what some would call ABI, hence my questions above.
--
Gustavo SousaQuoted 156 lines
drivers/gpu/drm/xe/xe_configfs.c +18
-17* * This attribute can only be set before binding to the device. * @@ -260,7 +260,7 @@ struct xe_config_group_device {u64 engines_allowed;struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ];struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; -bool survivability_mode; +bool enable_survivability_mode;bool enable_psmi;struct {unsigned int max_vfs;@@ -279,7 +279,7 @@ struct xe_config_group_device {static const struct xe_config_device device_defaults = { .gt_types_allowed =U64_MAX , .engines_allowed =U64_MAX , - .survivability_mode = false, + .enable_survivability_mode = false, .enable_psmi = false, .sriov = { .max_vfs =XE_DEFAULT_MAX_VFS ,@@ -357,20 +357,21 @@ static bool is_bound(struct xe_config_group_device *dev)return ret; } -static ssize_t survivability_mode_show (struct config_item *item,char *page) +static ssize_t enable_survivability_mode_show (struct config_item *item,char *page) {struct xe_config_device *dev =to_xe_config_device (item); -return sprintf (page,"%d\n" , dev->survivability_mode); +return sprintf (page,"%d\n" , dev->enable_survivability_mode); } -static ssize_t survivability_mode_store (struct config_item *item,const char *page,size_t len) +static ssize_t enable_survivability_mode_store (struct config_item *item,const char *page, +size_t len) {struct xe_config_group_device *dev =to_xe_config_group_device (item); -bool survivability_mode; +bool enable_survivability_mode;int ret; - ret =kstrtobool (page, &survivability_mode); + ret =kstrtobool (page, &enable_survivability_mode);if (ret)return ret;@@ -378,7 +379,7 @@ static ssize_t survivability_mode_store(struct config_item *item, const char *paif (is_bound (dev))return -EBUSY ; - dev->config.survivability_mode = survivability_mode; + dev->config.enable_survivability_mode = enable_survivability_mode;return len; }@@ -815,7 +816,7 @@ CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, survivability_mode); +CONFIGFS_ATTR(, enable_survivability_mode);static struct configfs_attribute *xe_config_device_attrs[] = { &attr_ctx_restore_mid_bb,@@ -823,7 +824,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = { &attr_enable_psmi, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_survivability_mode, + &attr_enable_survivability_mode, NULL, };@@ -847,7 +848,7 @@ static bool xe_config_device_is_visible(struct config_item *item, {struct xe_config_group_device *dev =to_xe_config_group_device (item); -if (attr == &attr_survivability_mode) { +if (attr == &attr_enable_survivability_mode) {if (!dev->desc->is_dgfx || dev->desc->platform <XE_BATTLEMAGE )return false; }@@ -1098,7 +1099,7 @@ static void dump_custom_dev_config(struct pci_dev *pdev,PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed);PRI_CUSTOM_ATTR ("%llx" , engines_allowed);PRI_CUSTOM_ATTR ("%d" , enable_psmi); -PRI_CUSTOM_ATTR ("%d" , survivability_mode); +PRI_CUSTOM_ATTR ("%d" , enable_survivability_mode);PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);#undef PRI_CUSTOM_ATTR@@ -1128,20 +1129,20 @@ void xe_configfs_check_device(struct pci_dev *pdev) }/** -* xe_configfs_get_survivability_mode - get configfs survivability mode attribute +* xe_configfs_get_enable_survivability_mode - get configfs survivability mode attribute * @pdev: pci device * -* Return: survivability_mode attribute in configfs +* Return: enable_survivability_mode attribute in configfs */ -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) {struct xe_config_group_device *dev =find_xe_config_group_device (pdev);bool mode;if (!dev) -return device_defaults.survivability_mode; +return device_defaults.enable_survivability_mode; - mode = dev->config.survivability_mode; + mode = dev->config.enable_survivability_mode;config_group_put (&dev->group);return mode;drivers/gpu/drm/xe/xe_configfs.h +2
-2@@ -18,7 +18,7 @@ struct pci_dev;int xe_configfs_init (void );void xe_configfs_exit (void );void xe_configfs_check_device (struct pci_dev *pdev); -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev); +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev);bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev);bool xe_configfs_media_gt_allowed (struct pci_dev *pdev);u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev);@@ -37,7 +37,7 @@ bool xe_configfs_admin_only_pf(struct pci_dev *pdev);static inline int xe_configfs_init (void ) {return 0 ; }static inline void xe_configfs_exit (void ) { }static inline void xe_configfs_check_device (struct pci_dev *pdev) { } -static inline bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) {return false; } +static inline bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) {return false; }static inline bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev) {return true; }static inline bool xe_configfs_media_gt_allowed (struct pci_dev *pdev) {return true; }static inline u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev) {return U64_MAX ; }drivers/gpu/drm/xe/xe_survivability_mode.c +1
-1@@ -377,7 +377,7 @@ bool xe_survivability_mode_is_requested(struct xe_device *xe)if (!IS_DGFX (xe) ||IS_SRIOV_VF (xe) || xe->info.platform <XE_BATTLEMAGE )return false; - survivability_mode =xe_configfs_get_survivability_mode (pdev); + survivability_mode =xe_configfs_get_enable_survivability_mode (pdev);/* Enable survivability mode if set via configfs */ if (survivability_mode)return true;Signature · 1 line
2.43.0
Re: [PATCH 1/9] drm/xe: Rename survivability_mode to enable_survivability_mode
Quoted 8 lines · Rodrigo Vivi
Quoted 6 lines · Gustavo Sousa
Quoted 4 lines · Stuart Summers
Rename the
survivability_modefield in structxe_config_devicetoenable_survivability_modeto be consistent with theenable_psminaming
convention used for similar boolean config flags.
Quoted 8 lines
I don't believe we need this level of consistency.
I'd like to keep the
survivability_modeas is because perhaps we could
perhaps even expand to multiple different modes later instead of only enable disable flag?!Cc Riana Tauro <[email protected]>
Yeah no problem. I probably should have made these patches all RFC.. I'll drop this one for now. I agree with your comments here.
Thanks,
Stuart
Quoted 241 lines
Quoted 241 lines
Quoted 27 lines
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 39 +++++++++++-----
------
drivers/gpu/drm/xe/xe_configfs.h | 4 +--drivers/gpu/drm/xe/xe_survivability_mode.c| 2 +-3 files changed, 23 insertions(+), 22 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -58,7 +58,7 @@ * │ └── ... * : * └── 0000:03:00.0 - * ├── survivability_mode + * ├── enable_survivability_mode * ├── gt_types_allowed * ├── engines_allowed * └── enable_psmi@@ -79,7 +79,7 @@ * Enable survivability mode on supported cards. This settingQuoted 7 lines
only takes
* effect when probing the device. Example to enable it::
*
- * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/survivability_mode
+ * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_modeI think we have not documented our configfs structure in the kernel ABI
documentation. Are we supposed to do that? Is that not yet done because we do not consider our configfs stable enough to be classified
as proper ABI?In this example, we would be potentially changing what some would call
ABI, hence my questions above.--
Gustavo SousaQuoted 192 lines
*
* This attribute can only be set before binding to the device. *drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -260,7 +260,7 @@ struct xe_config_group_device { u64 engines_allowed;struct
wa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
- boolsurvivability_mode;
+ boolenable_survivability_mode;
boolenable_psmi;
struct {
unsigned intmax_vfs;drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -279,7 +279,7 @@ struct xe_config_group_device { static const struct xe_config_device device_defaults = { .gt_types_allowed = U64_MAX, .engines_allowed = U64_MAX, - .survivability_mode = false, + .enable_survivability_mode = false, .enable_psmi = false, .sriov = { .max_vfs = XE_DEFAULT_MAX_VFS,@@ -357,20 +357,21 @@ static bool is_bound(struct xe_config_group_device *dev) return ret; }
-staticssize_tsurvivability_mode_show(struct config_item *item, char *page)
+staticssize_tenable_survivability_mode_show(struct config_item *item, char *page)
{
structxe_config_device*dev =to_xe_config_device(item);
- returnsprintf(page, "%d\n", dev->survivability_mode); + return sprintf(page, "%d\n", dev-Quoted 1 line
enable_survivability_mode);}
-staticssize_tsurvivability_mode_store(struct config_item *item, const char *page, size_t len)
+staticssize_tenable_survivability_mode_store(structconfig_item*item, const char *page,
+size_tlen) {
structxe_config_group_device*dev =to_xe_config_group_device(item);
- boolsurvivability_mode;
+ boolenable_survivability_mode;
int ret;
- ret =kstrtobool(page, &survivability_mode); + ret =kstrtobool(page, &enable_survivability_mode); if (ret)
return ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -378,7 +379,7 @@ static ssize_tsurvivability_mode_store(struct
config_item*item, const char *pa if (is_bound(dev))
return-EBUSY;
- dev->config.survivability_mode =survivability_mode; + dev->config.enable_survivability_mode =enable_survivability_mode;
return len;
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -815,7 +816,7 @@ CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, survivability_mode); +CONFIGFS_ATTR(, enable_survivability_mode);
static structconfigfs_attribute*xe_config_device_attrs[] = {&attr_ctx_restore_mid_bb,drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -823,7 +824,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = { &attr_enable_psmi, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_survivability_mode, + &attr_enable_survivability_mode, NULL, };drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -847,7 +848,7 @@ static bool xe_config_device_is_visible(struct config_item *item, {struct
xe_config_group_device*dev =to_xe_config_group_device(item);
- if (attr ==&attr_survivability_mode) {
+ if (attr ==&attr_enable_survivability_mode) { if (!dev->desc->is_dgfx ||dev->desc->platform<XE_BATTLEMAGE)
return false;
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -1098,7 +1099,7 @@ static void dump_custom_dev_config(structpci_dev *pdev, PRI_CUSTOM_ATTR("%llx", gt_types_allowed); PRI_CUSTOM_ATTR("%llx", engines_allowed); PRI_CUSTOM_ATTR("%d", enable_psmi); - PRI_CUSTOM_ATTR("%d", survivability_mode); + PRI_CUSTOM_ATTR("%d", enable_survivability_mode); PRI_CUSTOM_ATTR("%u", sriov.admin_only_pf);
#undefPRI_CUSTOM_ATTRdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1128,20 +1129,20 @@ void xe_configfs_check_device(struct pci_dev *pdev) }
/**
- *xe_configfs_get_survivability_mode- get configfs survivability mode attribute
+ *xe_configfs_get_enable_survivability_mode- get configfs survivability mode attribute
* @pdev: pci device
*
- * Return:survivability_modeattribute in configfs
+ * Return:enable_survivability_modeattribute in configfs */drivers/gpu/drm/xe/xe_configfs.c +1
-1-bool xe_configfs_get_survivability_mode (struct pci_dev *pdev) +bool xe_configfs_get_enable_survivability_mode(struct pci_dev*pdev)
{
structxe_config_group_device*dev =find_xe_config_group_device(pdev);
bool mode;
if (!dev)
- return device_defaults.survivability_mode; + return device_defaults.enable_survivability_mode;
- mode =dev->config.survivability_mode;
+ mode =dev->config.enable_survivability_mode;config_group_put(&dev->group);
return mode;drivers/gpu/drm/xe/xe_configfs.h +1
-1@@ -18,7 +18,7 @@ struct pci_dev; int xe_configfs_init(void); void xe_configfs_exit(void); void xe_configfs_check_device(struct pci_dev *pdev); -bool xe_configfs_get_survivability_mode (struct pci_dev *pdev); +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev); bool xe_configfs_primary_gt_allowed(struct pci_dev *pdev); bool xe_configfs_media_gt_allowed(struct pci_dev *pdev);u64
xe_configfs_get_engines_allowed(struct pci_dev *pdev);drivers/gpu/drm/xe/xe_configfs.h +0
-0@@ -37,7 +37,7 @@ bool xe_configfs_admin_only_pf(struct pci_dev *pdev); static inline int xe_configfs_init(void) { return 0; } static inline void xe_configfs_exit(void) { }static inline void
xe_configfs_check_device(struct pci_dev *pdev){ }
-static inline boolxe_configfs_get_survivability_mode(struct pci_dev *pdev){ return false; }
+static inline boolxe_configfs_get_enable_survivability_mode(struct pci_dev *pdev){ return false; }
static inline boolxe_configfs_primary_gt_allowed(struct pci_dev *pdev){ return true; }
static inline boolxe_configfs_media_gt_allowed(struct pci_dev *pdev){ return true; }
static inline u64xe_configfs_get_engines_allowed(struct pci_dev *pdev){ returnU64_MAX; }drivers/gpu/drm/xe/xe_survivability_mode.c +0
-0@@ -377,7 +377,7 @@ bool xe_survivability_mode_is_requested(struct xe_device *xe)if (!IS_DGFX(xe) ||
IS_SRIOV_VF(xe)||xe->info.platform<XE_BATTLEMAGE)
return false;
- survivability_mode =xe_configfs_get_survivability_mode(pdev);
+ survivability_mode =xe_configfs_get_enable_survivability_mode(pdev);
/* Enable survivability mode if set via configfs */ if (survivability_mode)
return true;Signature · 1 line
2.43.0
PATCH 2/9 drm/xe: Sort xe_config_device fields and defaults alphabetically
In reply to Stuart SummersSort the fields in struct xe_config_device and the corresponding device_defaults initializer in alphabetical order. Non-SRIOV fields come first (alphabetically), followed by the SRIOV sub-struct with its own fields in alphabetical order. Also sort the PRI_CUSTOM_ATTR calls in dump_custom_dev_config() to match and the documentation for each of the fields.
--- drivers/gpu/drm/xe/xe_configfs.c | 192 +++++++++++++++---------------- 1 file changed, 96 insertions(+), 96 deletions(-)
drivers/gpu/drm/xe/xe_configfs.c +96 -96
@@ -73,53 +73,79 @@* Configure Attributes * ==================== * -* Survivability mode: -* ------------------- +* Context restore BB +* ------------------ * -* Enable survivability mode on supported cards. This setting only takes -* effect when probing the device. Example to enable it:: +* Allow to execute a batch buffer during any context switches. When the +* GPU is restoring the context, it executes additional commands. It's useful +* for testing additional workarounds and validating certain HW behaviors: it's +* not intended for normal execution and will taint the kernel with TAINT_TEST +* when used. * -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode +* The syntax allows to pass straight instructions to be executed by the engine +* in a batch buffer or set specific registers. * -* This attribute can only be set before binding to the device. +* #. Generic instruction:: * -* Allowed GT types: -* ----------------- +* <engine-class> cmd <instr> [[dword0] [dword1] [...]] * -* Allow only specific types of GTs to be detected and initialized by the -* driver. Any combination of GT types can be enabled/disabled, although -* some settings will cause the device to fail to probe. +* #. Simple register setting:: * -* Writes support both comma- and newline-separated input format. Reads -* will always return one GT type per line. "primary" and "media" are the -* GT type names supported by this interface. +* <engine-class> reg <address> <value> * -* This attribute can only be set before binding to the device. +* Commands are saved per engine class: all instances of that class will execute +* those commands during context switch. The instruction, dword arguments, +* addresses and values are in hex format like in the examples below. * -* Examples: +* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 after the +* normal context restore:: * -* Allow both primary and media GTs to be initialized and used. This matches -* the driver's default behavior:: +* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ +* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb * -* # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed +* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 at the +* beginning of the context restore:: * -* Allow only the primary GT of each tile to be initialized and used, -* effectively disabling the media GT if it exists on the platform:: +* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ +* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb + +* #. Load certain values in a couple of registers (it can be used as a simpler +* alternative to the `cmd`) action:: * -* # echo 'primary' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed +* # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb <<EOF +* rcs reg 4F100 DEADBEEF +* rcs reg 4F104 FFFFFFFF +* EOF * -* Allow only the media GT of each tile to be initialized and used, -* effectively disabling the primary GT. **This configuration will cause -* device probe failure on all current platforms, but may be allowed on -* igpu platforms in the future**:: +* .. note:: * -* # echo 'media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed +* When using multiple lines, make sure to use a command that is +* implemented with a single write syscall, like HEREDOC. * -* Disable all GTs. Only other GPU IP (such as display) is potentially usable. -* **This configuration will cause device probe failure on all current -* platforms, but may be allowed on igpu platforms in the future**:: +* Currently this is implemented only for post and mid context restore and +* these attributes can only be set before binding to the device. * -* # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed +* PSMI +* ---- +* +* Enable extra debugging capabilities to trace engine execution. Only useful +* during early platform enabling and requires additional hardware connected. +* Once it's enabled, additionals WAs are added and runtime configuration is +* done via debugfs. Example to enable it:: +* +* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi +* +* This attribute can only be set before binding to the device. +* +* Survivability mode: +* ------------------- +* +* Enable survivability mode on supported cards. This setting only takes +* effect when probing the device. Example to enable it:: +* +* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode +* +* This attribute can only be set before binding to the device. * * Allowed engines: * ----------------
@@ -147,69 +173,43 @@* * This attribute can only be set before binding to the device. * -* PSMI -* ---- +* Allowed GT types: +* ----------------- * -* Enable extra debugging capabilities to trace engine execution. Only useful -* during early platform enabling and requires additional hardware connected. -* Once it's enabled, additionals WAs are added and runtime configuration is -* done via debugfs. Example to enable it:: +* Allow only specific types of GTs to be detected and initialized by the +* driver. Any combination of GT types can be enabled/disabled, although +* some settings will cause the device to fail to probe. * -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi +* Writes support both comma- and newline-separated input format. Reads +* will always return one GT type per line. "primary" and "media" are the +* GT type names supported by this interface. * * This attribute can only be set before binding to the device. * -* Context restore BB -* ------------------ -* -* Allow to execute a batch buffer during any context switches. When the -* GPU is restoring the context, it executes additional commands. It's useful -* for testing additional workarounds and validating certain HW behaviors: it's -* not intended for normal execution and will taint the kernel with TAINT_TEST -* when used. -* -* The syntax allows to pass straight instructions to be executed by the engine -* in a batch buffer or set specific registers. -* -* #. Generic instruction:: -* -* <engine-class> cmd <instr> [[dword0] [dword1] [...]] -* -* #. Simple register setting:: -* -* <engine-class> reg <address> <value> -* -* Commands are saved per engine class: all instances of that class will execute -* those commands during context switch. The instruction, dword arguments, -* addresses and values are in hex format like in the examples below. +* Examples: * -* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 after the -* normal context restore:: +* Allow both primary and media GTs to be initialized and used. This matches +* the driver's default behavior:: * -* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ -* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb +* # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed * -* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 at the -* beginning of the context restore:: +* Allow only the primary GT of each tile to be initialized and used, +* effectively disabling the media GT if it exists on the platform:: * -* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ -* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb - -* #. Load certain values in a couple of registers (it can be used as a simpler -* alternative to the `cmd`) action:: +* # echo 'primary' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed * -* # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb <<EOF -* rcs reg 4F100 DEADBEEF -* rcs reg 4F104 FFFFFFFF -* EOF +* Allow only the media GT of each tile to be initialized and used, +* effectively disabling the primary GT. **This configuration will cause +* device probe failure on all current platforms, but may be allowed on +* igpu platforms in the future**:: * -* .. note:: +* # echo 'media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed * -* When using multiple lines, make sure to use a command that is -* implemented with a single write syscall, like HEREDOC. +* Disable all GTs. Only other GPU IP (such as display) is potentially usable. +* **This configuration will cause device probe failure on all current +* platforms, but may be allowed on igpu platforms in the future**:: * -* Currently this is implemented only for post and mid context restore and -* these attributes can only be set before binding to the device. +* # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed * * Max SR-IOV Virtual Functions * ----------------------------
@@ -256,15 +256,15 @@ struct xe_config_group_device {struct config_group sriov;struct xe_config_device { -u64 gt_types_allowed; -u64 engines_allowed; -struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ];struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; -bool enable_survivability_mode; +struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ];bool enable_psmi; +bool enable_survivability_mode; +u64 engines_allowed; +u64 gt_types_allowed;struct { -unsigned int max_vfs;bool admin_only_pf; +unsigned int max_vfs; } sriov; } config;
@@ -277,13 +277,13 @@ struct xe_config_group_device { };static const struct xe_config_device device_defaults = { - .gt_types_allowed =U64_MAX , - .engines_allowed =U64_MAX , - .enable_survivability_mode = false, .enable_psmi = false, + .enable_survivability_mode = false, + .engines_allowed =U64_MAX , + .gt_types_allowed =U64_MAX , .sriov = { - .max_vfs =XE_DEFAULT_MAX_VFS , .admin_only_pf =XE_DEFAULT_ADMIN_ONLY_PF , + .max_vfs =XE_DEFAULT_MAX_VFS , }, };
@@ -814,17 +814,17 @@ static ssize_t ctx_restore_post_bb_store(struct config_item *item, CONFIGFS_ATTR(, ctx_restore_mid_bb); CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); +CONFIGFS_ATTR(, enable_survivability_mode); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, enable_survivability_mode);static struct configfs_attribute *xe_config_device_attrs[] = { &attr_ctx_restore_mid_bb, &attr_ctx_restore_post_bb, &attr_enable_psmi, + &attr_enable_survivability_mode, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_enable_survivability_mode,NULL , };
@@ -929,12 +929,12 @@ static ssize_t sriov_admin_only_pf_store(struct config_item *item, const char *preturn len; } -CONFIGFS_ATTR (sriov_, max_vfs);CONFIGFS_ATTR (sriov_, admin_only_pf); +CONFIGFS_ATTR (sriov_, max_vfs);static struct configfs_attribute *xe_config_sriov_attrs[] = { - &sriov_attr_max_vfs, &sriov_attr_admin_only_pf, + &sriov_attr_max_vfs,NULL , };
@@ -1096,10 +1096,10 @@ static void dump_custom_dev_config(struct pci_dev *pdev, dev->config.attr_); \ }while (0 ) -PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed); -PRI_CUSTOM_ATTR ("%llx" , engines_allowed);PRI_CUSTOM_ATTR ("%d" , enable_psmi);PRI_CUSTOM_ATTR ("%d" , enable_survivability_mode); +PRI_CUSTOM_ATTR ("%llx" , engines_allowed); +PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed);PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);#undef PRI_CUSTOM_ATTR
Signature · 1 line
2.43.0
Re: [PATCH 2/9] drm/xe: Sort xe_config_device fields and defaults alphabetically
Quoted 6 lines · Stuart Summers
Sort the fields in struct
xe_config_deviceand the correspondingdevice_defaultsinitializer in alphabetical order. Non-SRIOV fields come first (alphabetically), followed by the SRIOV sub-struct with its own fields in alphabetical order. Also sort thePRI_CUSTOM_ATTRcalls indump_custom_dev_config()to match and the documentation for each of the fields.
I think the sorting alphabetically seems fine, but I'll defer it to others to weigh in here on this decision though.
I guess we will not have many instances of that struct, so I guess potential padding resulting from the sorting requirement should be fine.
We probably want to add a note somewhere in the code asking to keep that stuff ordered alphabetically.
By the way, for other people looking at this, I find it easier to review this with: git show --color-moved --histogram.
Code-wise, looks good to me, so if this is the direction we take,
Quoted 305 lines
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7--- drivers/gpu/drm/xe/xe_configfs.c | 192 +++++++++++++++---------------- 1 file changed, 96 insertions(+), 96 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +96
-96@@ -73,53 +73,79 @@* Configure Attributes * ==================== * -* Survivability mode: -* ------------------- +* Context restore BB +* ------------------ * -* Enable survivability mode on supported cards. This setting only takes -* effect when probing the device. Example to enable it:: +* Allow to execute a batch buffer during any context switches. When the +* GPU is restoring the context, it executes additional commands. It's useful +* for testing additional workarounds and validating certain HW behaviors: it's +* not intended for normal execution and will taint the kernel with TAINT_TEST +* when used. * -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode +* The syntax allows to pass straight instructions to be executed by the engine +* in a batch buffer or set specific registers. * -* This attribute can only be set before binding to the device. +* #. Generic instruction:: * -* Allowed GT types: -* ----------------- +* <engine-class> cmd <instr> [[dword0] [dword1] [...]] * -* Allow only specific types of GTs to be detected and initialized by the -* driver. Any combination of GT types can be enabled/disabled, although -* some settings will cause the device to fail to probe. +* #. Simple register setting:: * -* Writes support both comma- and newline-separated input format. Reads -* will always return one GT type per line. "primary" and "media" are the -* GT type names supported by this interface. +* <engine-class> reg <address> <value> * -* This attribute can only be set before binding to the device. +* Commands are saved per engine class: all instances of that class will execute +* those commands during context switch. The instruction, dword arguments, +* addresses and values are in hex format like in the examples below. * -* Examples: +* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 after the +* normal context restore:: * -* Allow both primary and media GTs to be initialized and used. This matches -* the driver's default behavior:: +* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ +* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb * -* # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed +* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 at the +* beginning of the context restore:: * -* Allow only the primary GT of each tile to be initialized and used, -* effectively disabling the media GT if it exists on the platform:: +* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ +* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb + +* #. Load certain values in a couple of registers (it can be used as a simpler +* alternative to the `cmd`) action:: * -* # echo 'primary' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed +* # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb <<EOF +* rcs reg 4F100 DEADBEEF +* rcs reg 4F104 FFFFFFFF +* EOF * -* Allow only the media GT of each tile to be initialized and used, -* effectively disabling the primary GT. **This configuration will cause -* device probe failure on all current platforms, but may be allowed on -* igpu platforms in the future**:: +* .. note:: * -* # echo 'media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed +* When using multiple lines, make sure to use a command that is +* implemented with a single write syscall, like HEREDOC. * -* Disable all GTs. Only other GPU IP (such as display) is potentially usable. -* **This configuration will cause device probe failure on all current -* platforms, but may be allowed on igpu platforms in the future**:: +* Currently this is implemented only for post and mid context restore and +* these attributes can only be set before binding to the device. * -* # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed +* PSMI +* ---- +* +* Enable extra debugging capabilities to trace engine execution. Only useful +* during early platform enabling and requires additional hardware connected. +* Once it's enabled, additionals WAs are added and runtime configuration is +* done via debugfs. Example to enable it:: +* +* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi +* +* This attribute can only be set before binding to the device. +* +* Survivability mode: +* ------------------- +* +* Enable survivability mode on supported cards. This setting only takes +* effect when probing the device. Example to enable it:: +* +* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode +* +* This attribute can only be set before binding to the device. * * Allowed engines: * ---------------- @@ -147,69 +173,43 @@* * This attribute can only be set before binding to the device. * -* PSMI -* ---- +* Allowed GT types: +* ----------------- * -* Enable extra debugging capabilities to trace engine execution. Only useful -* during early platform enabling and requires additional hardware connected. -* Once it's enabled, additionals WAs are added and runtime configuration is -* done via debugfs. Example to enable it:: +* Allow only specific types of GTs to be detected and initialized by the +* driver. Any combination of GT types can be enabled/disabled, although +* some settings will cause the device to fail to probe. * -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi +* Writes support both comma- and newline-separated input format. Reads +* will always return one GT type per line. "primary" and "media" are the +* GT type names supported by this interface. * * This attribute can only be set before binding to the device. * -* Context restore BB -* ------------------ -* -* Allow to execute a batch buffer during any context switches. When the -* GPU is restoring the context, it executes additional commands. It's useful -* for testing additional workarounds and validating certain HW behaviors: it's -* not intended for normal execution and will taint the kernel with TAINT_TEST -* when used. -* -* The syntax allows to pass straight instructions to be executed by the engine -* in a batch buffer or set specific registers. -* -* #. Generic instruction:: -* -* <engine-class> cmd <instr> [[dword0] [dword1] [...]] -* -* #. Simple register setting:: -* -* <engine-class> reg <address> <value> -* -* Commands are saved per engine class: all instances of that class will execute -* those commands during context switch. The instruction, dword arguments, -* addresses and values are in hex format like in the examples below. +* Examples: * -* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 after the -* normal context restore:: +* Allow both primary and media GTs to be initialized and used. This matches +* the driver's default behavior:: * -* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ -* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb +* # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed * -* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 at the -* beginning of the context restore:: +* Allow only the primary GT of each tile to be initialized and used, +* effectively disabling the media GT if it exists on the platform:: * -* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ -* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb - -* #. Load certain values in a couple of registers (it can be used as a simpler -* alternative to the `cmd`) action:: +* # echo 'primary' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed * -* # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb <<EOF -* rcs reg 4F100 DEADBEEF -* rcs reg 4F104 FFFFFFFF -* EOF +* Allow only the media GT of each tile to be initialized and used, +* effectively disabling the primary GT. **This configuration will cause +* device probe failure on all current platforms, but may be allowed on +* igpu platforms in the future**:: * -* .. note:: +* # echo 'media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed * -* When using multiple lines, make sure to use a command that is -* implemented with a single write syscall, like HEREDOC. +* Disable all GTs. Only other GPU IP (such as display) is potentially usable. +* **This configuration will cause device probe failure on all current +* platforms, but may be allowed on igpu platforms in the future**:: * -* Currently this is implemented only for post and mid context restore and -* these attributes can only be set before binding to the device. +* # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed * * Max SR-IOV Virtual Functions * ---------------------------- @@ -256,15 +256,15 @@ struct xe_config_group_device {struct config_group sriov;struct xe_config_device { -u64 gt_types_allowed; -u64 engines_allowed; -struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ];struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; -bool enable_survivability_mode; +struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ];bool enable_psmi; +bool enable_survivability_mode; +u64 engines_allowed; +u64 gt_types_allowed;struct { -unsigned int max_vfs;bool admin_only_pf; +unsigned int max_vfs; } sriov; } config;@@ -277,13 +277,13 @@ struct xe_config_group_device { };static const struct xe_config_device device_defaults = { - .gt_types_allowed =U64_MAX , - .engines_allowed =U64_MAX , - .enable_survivability_mode = false, .enable_psmi = false, + .enable_survivability_mode = false, + .engines_allowed =U64_MAX , + .gt_types_allowed =U64_MAX , .sriov = { - .max_vfs =XE_DEFAULT_MAX_VFS , .admin_only_pf =XE_DEFAULT_ADMIN_ONLY_PF , + .max_vfs =XE_DEFAULT_MAX_VFS , }, };@@ -814,17 +814,17 @@ static ssize_t ctx_restore_post_bb_store(struct config_item *item, CONFIGFS_ATTR(, ctx_restore_mid_bb); CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); +CONFIGFS_ATTR(, enable_survivability_mode); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, enable_survivability_mode);static struct configfs_attribute *xe_config_device_attrs[] = { &attr_ctx_restore_mid_bb, &attr_ctx_restore_post_bb, &attr_enable_psmi, + &attr_enable_survivability_mode, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_enable_survivability_mode,NULL , };@@ -929,12 +929,12 @@ static ssize_t sriov_admin_only_pf_store(struct config_item *item, const char *preturn len; } -CONFIGFS_ATTR (sriov_, max_vfs);CONFIGFS_ATTR (sriov_, admin_only_pf); +CONFIGFS_ATTR (sriov_, max_vfs);static struct configfs_attribute *xe_config_sriov_attrs[] = { - &sriov_attr_max_vfs, &sriov_attr_admin_only_pf, + &sriov_attr_max_vfs,NULL , };@@ -1096,10 +1096,10 @@ static void dump_custom_dev_config(struct pci_dev *pdev, dev->config.attr_); \ }while (0 ) -PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed); -PRI_CUSTOM_ATTR ("%llx" , engines_allowed);PRI_CUSTOM_ATTR ("%d" , enable_psmi);PRI_CUSTOM_ATTR ("%d" , enable_survivability_mode); +PRI_CUSTOM_ATTR ("%llx" , engines_allowed); +PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed);PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);#undef PRI_CUSTOM_ATTRSignature · 1 line
2.43.0
Re: [PATCH 2/9] drm/xe: Sort xe_config_device fields and defaults alphabetically
Quoted 21 lines · Gustavo Sousa
Quoted 7 lines · Stuart Summers
Sort the fields in struct
xe_config_deviceand the correspondingdevice_defaultsinitializer in alphabetical order. Non-SRIOV fields come first (alphabetically), followed by the SRIOV sub-struct with its own fields in alphabetical order. Also sort thePRI_CUSTOM_ATTRcalls indump_custom_dev_config()to match and the documentation for
each of the fields.I think the sorting alphabetically seems fine, but I'll defer it to others to weigh in here on this decision though.
I guess we will not have many instances of that struct, so I guess potential padding resulting from the sorting requirement should be fine.
We probably want to add a note somewhere in the code asking to keep that
stuff ordered alphabetically.
Quoted 7 lines
By the way, for other people looking at this, I find it easier to review
this with: git show --color-moved --histogram.Code-wise, looks good to me, so if this is the direction we take,
Reviewed-by Gustavo Sousa <[email protected]>
I had accidentally responded to the wrong patch with this... Also thanks for the review!
Yeah this is something I wanted to discuss here too. I kind of just blindly alphabetized things here, but we might want to sort by data type or something else. also, we only have a few parameters here, but I'd like to set this up to support a much larger set of parameters on the debug side - we need much finer grained support for things like power-on for instance. And I'd like a way that easily scales here.
Also I agree on the documentation.. I'll make sure to add that in the next patch once we conclude on the approach here.
Thanks,
Stuart
Quoted 390 lines
Quoted 390 lines
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 192 +++++++++++++++------------
----1 file changed, 96 insertions(+), 96 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +2
-3@@ -73,53 +73,79 @@ * Configure Attributes * ==================== * - * Survivability mode: - * ------------------- + * Context restore BB + * ------------------ * - * Enable survivability mode on supported cards. This setting onlytakes
- * effect when probing the device. Example to enable it::
+ * Allow to execute a batch buffer during any context switches. When the
+ * GPU is restoring the context, it executes additional commands. It's useful
+ * for testing additional workarounds and validating certain HW
behaviors: it's
+ * not intended for normal execution and will taint the kernel withTAINT_TEST
+ * when used.
*
- * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode
+ * The syntax allows to pass straight instructions to be executed by the engine
+ * in a batch buffer or set specific registers.
*
- * This attribute can only be set before binding to the device.
+ * #. Generic instruction::
*
- * Allowed GT types:
- * -----------------
+ * <engine-class> cmd <instr> [[dword0] [dword1] [...]] *
- * Allow only specific types of GTs to be detected and initialized by the
- * driver. Any combination of GT types can be enabled/disabled, although
- * some settings will cause the device to fail to probe.
+ * #. Simple register setting::
*
- * Writes support both comma- and newline-separated input format. Reads
- * will always return one GT type per line. "primary" and "media" are the
- * GT type names supported by this interface.
+ * <engine-class> reg <address> <value>
*
- * This attribute can only be set before binding to the device.
+ * Commands are saved per engine class: all instances of that class will execute
+ * those commands during context switch. The instruction, dword arguments,
+ * addresses and values are in hex format like in the examples below.
*
- * Examples:
+ * #. Execute a LRI command to write0xDEADBEEFto register0x4f10after the
+ * normal context restore::
*
- * Allow both primary and media GTs to be initialized and used. This matches
- * the driver's default behavior::
+ * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
+ * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb
*
- * # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * #. Execute a LRI command to write0xDEADBEEFto register0x4f10at the
+ * beginning of the context restore::
*
- * Allow only the primary GT of each tile to be initialized and used,
- * effectively disabling the media GT if it exists on the platform::
+ * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
+ * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb
+
+ * #. Load certain values in a couple of registers (it can be used as a simpler
+ * alternative to thecmd) action::
*
- * # echo 'primary' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb <<EOF
+ * rcs reg 4F100 DEADBEEF
+ * rcs reg 4F104 FFFFFFFF
+ * EOF
*
- * Allow only the media GT of each tile to be initialized and used,
- * effectively disabling the primary GT. **This configuration will cause
- * device probe failure on all current platforms, but may be allowed on
- * igpu platforms in the future**::
+ * .. note::
*
- * # echo 'media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * When using multiple lines, make sure to use a command that is
+ * implemented with a single write syscall, like HEREDOC. *
- * Disable all GTs. Only other GPU IP (such as display) is potentially usable.
- * **This configuration will cause device probe failure on all current
- * platforms, but may be allowed on igpu platforms in the future**::
+ * Currently this is implemented only for post and mid context restore and
+ * these attributes can only be set before binding to the device. *
- * # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * PSMI
+ * ----
+ *
+ * Enable extra debugging capabilities to trace engine execution. Only useful
+ * during early platform enabling and requires additional hardware connected.
+ * Once it's enabled, additionals WAs are added and runtime configuration is
+ * done via debugfs. Example to enable it::
+ *
+ * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi
+ *
+ * This attribute can only be set before binding to the device.
+ *
+ * Survivability mode:
+ * -------------------
+ *
+ * Enable survivability mode on supported cards. This setting only takes
+ * effect when probing the device. Example to enable it::
+ *
+ * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode
+ *
+ * This attribute can only be set before binding to the device. *
* Allowed engines:
* ----------------drivers/gpu/drm/xe/xe_configfs.c +2
-3@@ -147,69 +173,43 @@ * * This attribute can only be set before binding to the device. * - *PSMI - * ---- + * Allowed GT types: + * ----------------- * - * Enable extra debugging capabilities to trace engine execution.Only useful
- * during early platform enabling and requires additional hardware connected.
- * Once it's enabled, additionals WAs are added and runtime configuration is
- * done via debugfs. Example to enable it::
+ * Allow only specific types of GTs to be detected and initialized by the
+ * driver. Any combination of GT types can be enabled/disabled, although
+ * some settings will cause the device to fail to probe. *
- * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi
+ * Writes support both comma- and newline-separated input format. Reads
+ * will always return one GT type per line. "primary" and "media" are the
+ * GT type names supported by this interface.
*
* This attribute can only be set before binding to the device. *
- * Context restore BB
- * ------------------
- *
- * Allow to execute a batch buffer during any context switches. When the
- * GPU is restoring the context, it executes additional commands. It's useful
- * for testing additional workarounds and validating certain HW
behaviors: it's
- * not intended for normal execution and will taint the kernel withTAINT_TEST
- * when used.
- *
- * The syntax allows to pass straight instructions to be executed by the engine
- * in a batch buffer or set specific registers.
- *
- * #. Generic instruction::
- *
- * <engine-class> cmd <instr> [[dword0] [dword1] [...]]
- *
- * #. Simple register setting::
- *
- * <engine-class> reg <address> <value>
- *
- * Commands are saved per engine class: all instances of that class will execute
- * those commands during context switch. The instruction, dword arguments,
- * addresses and values are in hex format like in the examples below.
+ * Examples:
*
- * #. Execute a LRI command to write0xDEADBEEFto register0x4f10after the
- * normal context restore::
+ * Allow both primary and media GTs to be initialized and used. This matches
+ * the driver's default behavior::
*
- * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
- * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb
+ * # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
- * #. Execute a LRI command to write0xDEADBEEFto register0x4f10at the
- * beginning of the context restore::
+ * Allow only the primary GT of each tile to be initialized and used,
+ * effectively disabling the media GT if it exists on the platform::
*
- * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
- * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb
-
- * #. Load certain values in a couple of registers (it can be used as a simpler
- * alternative to thecmd) action::
+ * # echo 'primary' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
- * # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb <<EOF
- * rcs reg 4F100 DEADBEEF
- * rcs reg 4F104 FFFFFFFF
- * EOF
+ * Allow only the media GT of each tile to be initialized and used,
+ * effectively disabling the primary GT. **This configuration will cause
+ * device probe failure on all current platforms, but may be allowed on
+ * igpu platforms in the future**::
*
- * .. note::
+ * # echo 'media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
- * When using multiple lines, make sure to use a command that is
- * implemented with a single write syscall, like HEREDOC.
+ * Disable all GTs. Only other GPU IP (such as display) is potentially usable.
+ * **This configuration will cause device probe failure on all current
+ * platforms, but may be allowed on igpu platforms in the future**::
*
- * Currently this is implemented only for post and mid context restore and
- * these attributes can only be set before binding to the device.
+ * # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
* Max SR-IOV Virtual Functions
* ----------------------------drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -256,15 +256,15 @@ struct xe_config_group_device { struct config_group sriov;
structxe_config_device{
- u64gt_types_allowed;
- u64engines_allowed;
- structwa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
- boolenable_survivability_mode;
+ structwa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
boolenable_psmi;
+ boolenable_survivability_mode;
+ u64engines_allowed;
+ u64gt_types_allowed;
struct {
- unsigned intmax_vfs;
booladmin_only_pf;
+ unsigned intmax_vfs;
} sriov;
} config;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -277,13 +277,13 @@ struct xe_config_group_device { };
static const structxe_config_devicedevice_defaults= { - .gt_types_allowed =U64_MAX,
- .engines_allowed =U64_MAX,
- .enable_survivability_mode = false, .enable_psmi = false,
+ .enable_survivability_mode = false,
+ .engines_allowed =U64_MAX,
+ .gt_types_allowed =U64_MAX,
.sriov = {
- .max_vfs =XE_DEFAULT_MAX_VFS, .admin_only_pf =XE_DEFAULT_ADMIN_ONLY_PF, + .max_vfs =XE_DEFAULT_MAX_VFS,
},
};drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -814,17 +814,17 @@ static ssize_t ctx_restore_post_bb_store(struct config_item *item, CONFIGFS_ATTR(, ctx_restore_mid_bb); CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); +CONFIGFS_ATTR(, enable_survivability_mode); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, enable_survivability_mode);
static structconfigfs_attribute*xe_config_device_attrs[] = {&attr_ctx_restore_mid_bb,&attr_ctx_restore_post_bb,
&attr_enable_psmi,
+ &attr_enable_survivability_mode,&attr_engines_allowed,&attr_gt_types_allowed,
- &attr_enable_survivability_mode,
NULL,
};drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -929,12 +929,12 @@ static ssize_tsriov_admin_only_pf_store(struct
config_item*item, const char *p return len;
}
-CONFIGFS_ATTR(sriov_,max_vfs);
CONFIGFS_ATTR(sriov_,admin_only_pf); +CONFIGFS_ATTR(sriov_,max_vfs);
static structconfigfs_attribute*xe_config_sriov_attrs[] = { - &sriov_attr_max_vfs,
&sriov_attr_admin_only_pf,
+ &sriov_attr_max_vfs,
NULL,
};drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1096,10 +1096,10 @@ static void dump_custom_dev_config(structpci_dev *pdev, dev->config.attr_); \ } while (0)
- PRI_CUSTOM_ATTR("%llx",gt_types_allowed); - PRI_CUSTOM_ATTR("%llx",engines_allowed);PRI_CUSTOM_ATTR("%d", enable_psmi);PRI_CUSTOM_ATTR("%d", enable_survivability_mode); + PRI_CUSTOM_ATTR("%llx",engines_allowed); + PRI_CUSTOM_ATTR("%llx",gt_types_allowed);PRI_CUSTOM_ATTR("%u", sriov.admin_only_pf);
#undefPRI_CUSTOM_ATTRSignature · 1 line
2.43.0
RE: [PATCH 2/9] drm/xe: Sort xe_config_device fields and defaults alphabetically
In reply to Stuart SummersQuoted 261 lines · Stuart Summers
Sort the fields in struct
xe_config_deviceand the correspondingdevice_defaultsinitializer in alphabetical order. Non-SRIOV fields come first (alphabetically), followed by the SRIOV sub-struct with its own fields in alphabetical order. Also sort thePRI_CUSTOM_ATTRcalls indump_custom_dev_config()to match and the documentation for each of the fields.Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7--- drivers/gpu/drm/xe/xe_configfs.c | 192 +++++++++++++++---------------- 1 file changed, 96 insertions(+), 96 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +11
-5@@ -73,53 +73,79 @@* Configure Attributes * ==================== * -* Survivability mode: -* ------------------- +* Context restore BB +* ------------------ * -* Enable survivability mode on supported cards. This setting only takes -* effect when probing the device. Example to enable it:: +* Allow to execute a batch buffer during any context switches. When +the +* GPU is restoring the context, it executes additional commands. It's +useful +* for testing additional workarounds and validating certain HW +behaviors: it's +* not intended for normal execution and will taint the kernel with +TAINT_TEST +* when used. * -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode
+ * The syntax allows to pass straight instructions to be executed by
+ the engine
+ * in a batch buffer or set specific registers.
*
- * This attribute can only be set before binding to the device.
+ * #. Generic instruction::
*
- * Allowed GT types:
- * -----------------
+ * <engine-class> cmd <instr> [[dword0] [dword1] [...]]
*
- * Allow only specific types of GTs to be detected and initialized by the
- * driver. Any combination of GT types can be enabled/disabled, although
- * some settings will cause the device to fail to probe.
+ * #. Simple register setting::
*
- * Writes support both comma- and newline-separated input format. Reads
- * will always return one GT type per line. "primary" and "media" are the
- * GT type names supported by this interface.
+ * <engine-class> reg <address> <value>
*
- * This attribute can only be set before binding to the device.
+ * Commands are saved per engine class: all instances of that class
+ will execute
+ * those commands during context switch. The instruction, dword
+ arguments,
+ * addresses and values are in hex format like in the examples below. *
- * Examples:
+ * #. Execute a LRI command to write0xDEADBEEFto register0x4f10after the
+ * normal context restore::
*
- * Allow both primary and media GTs to be initialized and used. This matches
- * the driver's default behavior::
+ * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
+ * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb
*
- * # echo 'primary,media' >
/sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * #. Execute a LRI command to write0xDEADBEEFto register0x4f10at the
+ * beginning of the context restore::
*
- * Allow only the primary GT of each tile to be initialized and used,
- * effectively disabling the media GT if it exists on the platform::
+ * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
+ * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb
+
+ * #. Load certain values in a couple of registers (it can be used as a simpler
+ * alternative to thecmd) action::
*
- * # echo 'primary' >
/sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb
<<EOF
+ * rcs reg 4F100 DEADBEEF
+ * rcs reg 4F104 FFFFFFFF
+ * EOF
*
- * Allow only the media GT of each tile to be initialized and used,
- * effectively disabling the primary GT. **This configuration will cause
- * device probe failure on all current platforms, but may be allowed on
- * igpu platforms in the future**::
+ * .. note::
*
- * # echo 'media' >
/sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * When using multiple lines, make sure to use a command that is
+ * implemented with a single write syscall, like HEREDOC.
*
- * Disable all GTs. Only other GPU IP (such as display) is potentially usable.
- * **This configuration will cause device probe failure on all current
- * platforms, but may be allowed on igpu platforms in the future**::
+ * Currently this is implemented only for post and mid context restore
+ and
+ * these attributes can only be set before binding to the device. *
- * # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * PSMI
+ * ----
+ *
+ * Enable extra debugging capabilities to trace engine execution. Only
+ useful
+ * during early platform enabling and requires additional hardware connected.
+ * Once it's enabled, additionals WAs are added and runtime
+ configuration is
+ * done via debugfs. Example to enable it::
+ *
+ * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi
+ *
+ * This attribute can only be set before binding to the device.
+ *
+ * Survivability mode:
+ * -------------------
+ *
+ * Enable survivability mode on supported cards. This setting only
+ takes
+ * effect when probing the device. Example to enable it::
+ *
+ * # echo 1 >
/sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode
+ *
+ * This attribute can only be set before binding to the device.
*
* Allowed engines:
* ----------------drivers/gpu/drm/xe/xe_configfs.c +16
-34@@ -147,69 +173,43 @@* * This attribute can only be set before binding to the device. * -* PSMI -* ---- +* Allowed GT types: +* ----------------- * -* Enable extra debugging capabilities to trace engine execution. Only useful -* during early platform enabling and requires additional hardware connected. -* Once it's enabled, additionals WAs are added and runtime configuration is -* done via debugfs. Example to enable it:: +* Allow only specific types of GTs to be detected and initialized by +the +* driver. Any combination of GT types can be enabled/disabled, +although +* some settings will cause the device to fail to probe. * -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi +* Writes support both comma- and newline-separated input format. Reads +* will always return one GT type per line. "primary" and "media" are +the +* GT type names supported by this interface. * * This attribute can only be set before binding to the device. * -* Context restore BB -* ------------------ -* -* Allow to execute a batch buffer during any context switches. When the -* GPU is restoring the context, it executes additional commands. It's useful -* for testing additional workarounds and validating certain HW behaviors: it's -* not intended for normal execution and will taint the kernel with TAINT_TEST -* when used. -* -* The syntax allows to pass straight instructions to be executed by the engine -* in a batch buffer or set specific registers. -* -* #. Generic instruction:: -* -* <engine-class> cmd <instr> [[dword0] [dword1] [...]] -* -* #. Simple register setting:: -* -* <engine-class> reg <address> <value> -* -* Commands are saved per engine class: all instances of that class will execute -* those commands during context switch. The instruction, dword arguments, -* addresses and values are in hex format like in the examples below. +* Examples: * -* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 after the -* normal context restore:: +* Allow both primary and media GTs to be initialized and used. This +matches +* the driver's default behavior:: * -* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ -* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb +* # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
- * #. Execute a LRI command to write0xDEADBEEFto register0x4f10at the
- * beginning of the context restore::
+ * Allow only the primary GT of each tile to be initialized and used,
+ * effectively disabling the media GT if it exists on the platform::
*
- * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
- * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb
-
- * #. Load certain values in a couple of registers (it can be used as a simpler
- * alternative to thecmd) action::
+ * # echo 'primary' >
/sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
- * # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb
<<EOF
- * rcs reg 4F100 DEADBEEF
- * rcs reg 4F104 FFFFFFFF
- * EOF
+ * Allow only the media GT of each tile to be initialized and used,
+ * effectively disabling the primary GT. **This configuration will
+ cause
+ * device probe failure on all current platforms, but may be allowed on
+ * igpu platforms in the future**::
*
- * .. note::
+ * # echo 'media' >
/sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
- * When using multiple lines, make sure to use a command that is
- * implemented with a single write syscall, like HEREDOC.
+ * Disable all GTs. Only other GPU IP (such as display) is potentially usable.
+ * **This configuration will cause device probe failure on all current
+ * platforms, but may be allowed on igpu platforms in the future**::
*
- * Currently this is implemented only for post and mid context restore and
- * these attributes can only be set before binding to the device.
+ * # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
* Max SR-IOV Virtual Functions
* ----------------------------drivers/gpu/drm/xe/xe_configfs.c +1
-4@@ -256,15 +256,15 @@ struct xe_config_group_device {struct config_group sriov;struct xe_config_device { -u64 gt_types_allowed; -u64 engines_allowed; -struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ];struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; -bool enable_survivability_mode; +struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ];
drivers/gpu/drm/xe/xe_configfs.c struct xe_config_group_device { · line 269
Quoted 10 lines
drivers/gpu/drm/xe/xe_configfs.c +4
-1bool enable_psmi; +bool enable_survivability_mode; +u64 engines_allowed; +u64 gt_types_allowed;struct { -unsigned int max_vfs;bool admin_only_pf; +unsigned int max_vfs; } sriov; } config;
Here is the comment from [1], which seems reasonable to me.
"
Does sorting struct members alphabetically here introduce unnecessary memory padding?
Placing the bool fields enable_psmi and enable_survivability_mode immediately before the u64 fields engines_allowed and gt_types_allowed introduces 6 bytes of padding. In the sriov sub-struct, placing the bool admin_only_pf before the unsigned int max_vfs introduces 3 bytes of padding.
Could these fields be ordered by size to minimize padding instead of being sorted alphabetically?
"
If we decide to sort them, we could add comments like below for the flags in struct intel_device_info. " * Keep all flags below alphabetically sorted"
[1]: https://sashiko.dev/#/patchset/20260504044348.209625-1-stuart.summers%40intel.com
Shuicheng
Quoted 69 lines
drivers/gpu/drm/xe/xe_configfs.c +6
-6@@ -277,13 +277,13 @@ struct xe_config_group_device { };static const struct xe_config_device device_defaults = { - .gt_types_allowed =U64_MAX , - .engines_allowed =U64_MAX , - .enable_survivability_mode = false, .enable_psmi = false, + .enable_survivability_mode = false, + .engines_allowed =U64_MAX , + .gt_types_allowed =U64_MAX , .sriov = { - .max_vfs =XE_DEFAULT_MAX_VFS , .admin_only_pf =XE_DEFAULT_ADMIN_ONLY_PF , + .max_vfs =XE_DEFAULT_MAX_VFS , }, };@@ -814,17 +814,17 @@ static ssize_t ctx_restore_post_bb_store(struct config_item *item, CONFIGFS_ATTR(, ctx_restore_mid_bb); CONFIGFS_ATTR(, ctx_restore_post_bb); CONFIGFS_ATTR(, enable_psmi); +CONFIGFS_ATTR(, enable_survivability_mode); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(, enable_survivability_mode);static struct configfs_attribute *xe_config_device_attrs[] = { &attr_ctx_restore_mid_bb, &attr_ctx_restore_post_bb, &attr_enable_psmi, + &attr_enable_survivability_mode, &attr_engines_allowed, &attr_gt_types_allowed, - &attr_enable_survivability_mode,NULL , };@@ -929,12 +929,12 @@ static ssize_t sriov_admin_only_pf_store(struct
config_item*item, const char *pdrivers/gpu/drm/xe/xe_configfs.c +4
-4return len; } -CONFIGFS_ATTR (sriov_, max_vfs);CONFIGFS_ATTR (sriov_, admin_only_pf); +CONFIGFS_ATTR (sriov_, max_vfs);static struct configfs_attribute *xe_config_sriov_attrs[] = { - &sriov_attr_max_vfs, &sriov_attr_admin_only_pf, + &sriov_attr_max_vfs,NULL , };@@ -1096,10 +1096,10 @@ static void dump_custom_dev_config(struct pci_dev *pdev, dev->config.attr_); \ }while (0 ) -PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed); -PRI_CUSTOM_ATTR ("%llx" , engines_allowed);PRI_CUSTOM_ATTR ("%d" , enable_psmi);PRI_CUSTOM_ATTR ("%d" , enable_survivability_mode); +PRI_CUSTOM_ATTR ("%llx" , engines_allowed); +PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed);PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);#undef
PRI_CUSTOM_ATTR
--
2.43.0
Re: [PATCH 2/9] drm/xe: Sort xe_config_device fields and defaults alphabetically
Quoted 358 lines · Shuicheng
Quoted 331 lines · Stuart Summers
Sort the fields in struct
xe_config_deviceand the correspondingdevice_defaultsinitializer in alphabetical order. Non-SRIOV fields come first
(alphabetically), followed by the SRIOV sub-struct with its own fields in
alphabetical order. Also sort thePRI_CUSTOM_ATTRcalls indump_custom_dev_config()to match and the documentation for each of the
fields.Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 192 +++++++++++++++------------
----1 file changed, 96 insertions(+), 96 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +2
-3@@ -73,53 +73,79 @@ * Configure Attributes * ==================== * - * Survivability mode: - * ------------------- + * Context restore BB + * ------------------ * - * Enable survivability mode on supported cards. This setting onlytakes
- * effect when probing the device. Example to enable it::
+ * Allow to execute a batch buffer during any context switches. When
+ the
+ * GPU is restoring the context, it executes additional commands. It's
+ useful
+ * for testing additional workarounds and validating certain HW
+ behaviors: it's
+ * not intended for normal execution and will taint the kernel with
+TAINT_TEST
+ * when used.
*
- * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode
+ * The syntax allows to pass straight instructions to be executed by
+ the engine
+ * in a batch buffer or set specific registers.
*
- * This attribute can only be set before binding to the device.
+ * #. Generic instruction::
*
- * Allowed GT types:
- * -----------------
+ * <engine-class> cmd <instr> [[dword0] [dword1] [...]] *
- * Allow only specific types of GTs to be detected and initialized by the
- * driver. Any combination of GT types can be enabled/disabled, although
- * some settings will cause the device to fail to probe.
+ * #. Simple register setting::
*
- * Writes support both comma- and newline-separated input format. Reads
- * will always return one GT type per line. "primary" and "media" are the
- * GT type names supported by this interface.
+ * <engine-class> reg <address> <value>
*
- * This attribute can only be set before binding to the device.
+ * Commands are saved per engine class: all instances of that class
+ will execute
+ * those commands during context switch. The instruction, dword
+ arguments,
+ * addresses and values are in hex format like in the examples below.
*
- * Examples:
+ * #. Execute a LRI command to write0xDEADBEEFto register0x4f10after the
+ * normal context restore::
*
- * Allow both primary and media GTs to be initialized and used. This matches
- * the driver's default behavior::
+ * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
+ * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb
*
- * # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * #. Execute a LRI command to write0xDEADBEEFto register0x4f10at the
+ * beginning of the context restore::
*
- * Allow only the primary GT of each tile to be initialized and used,
- * effectively disabling the media GT if it exists on the platform::
+ * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
+ * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb
+
+ * #. Load certain values in a couple of registers (it can be used as a simpler
+ * alternative to thecmd) action::
*
- * # echo 'primary' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb <<EOF
+ * rcs reg 4F100 DEADBEEF
+ * rcs reg 4F104 FFFFFFFF
+ * EOF
*
- * Allow only the media GT of each tile to be initialized and used,
- * effectively disabling the primary GT. **This configuration will cause
- * device probe failure on all current platforms, but may be allowed on
- * igpu platforms in the future**::
+ * .. note::
*
- * # echo 'media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * When using multiple lines, make sure to use a command that is
+ * implemented with a single write syscall, like HEREDOC. *
- * Disable all GTs. Only other GPU IP (such as display) is potentially usable.
- * **This configuration will cause device probe failure on all current
- * platforms, but may be allowed on igpu platforms in the future**::
+ * Currently this is implemented only for post and mid context restore
+ and
+ * these attributes can only be set before binding to the device. *
- * # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
+ * PSMI
+ * ----
+ *
+ * Enable extra debugging capabilities to trace engine execution. Only
+ useful
+ * during early platform enabling and requires additional hardware connected.
+ * Once it's enabled, additionals WAs are added and runtime
+ configuration is
+ * done via debugfs. Example to enable it::
+ *
+ * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi
+ *
+ * This attribute can only be set before binding to the device.
+ *
+ * Survivability mode:
+ * -------------------
+ *
+ * Enable survivability mode on supported cards. This setting only
+ takes
+ * effect when probing the device. Example to enable it::
+ *
+ * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode
+ *
+ * This attribute can only be set before binding to the device. *
* Allowed engines:
* ----------------drivers/gpu/drm/xe/xe_configfs.c +2
-3@@ -147,69 +173,43 @@ * * This attribute can only be set before binding to the device. * - *PSMI - * ---- + * Allowed GT types: + * ----------------- * - * Enable extra debugging capabilities to trace engine execution.Only useful
- * during early platform enabling and requires additional hardware connected.
- * Once it's enabled, additionals WAs are added and runtime configuration is
- * done via debugfs. Example to enable it::
+ * Allow only specific types of GTs to be detected and initialized by
+ the
+ * driver. Any combination of GT types can be enabled/disabled,
+ although
+ * some settings will cause the device to fail to probe. *
- * # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi
+ * Writes support both comma- and newline-separated input format. Reads
+ * will always return one GT type per line. "primary" and "media" are
+ the
+ * GT type names supported by this interface.
*
* This attribute can only be set before binding to the device. *
- * Context restore BB
- * ------------------
- *
- * Allow to execute a batch buffer during any context switches. When the
- * GPU is restoring the context, it executes additional commands. It's useful
- * for testing additional workarounds and validating certain HW
behaviors: it's
- * not intended for normal execution and will taint the kernel withTAINT_TEST
- * when used.
- *
- * The syntax allows to pass straight instructions to be executed by the engine
- * in a batch buffer or set specific registers.
- *
- * #. Generic instruction::
- *
- * <engine-class> cmd <instr> [[dword0] [dword1] [...]]
- *
- * #. Simple register setting::
- *
- * <engine-class> reg <address> <value>
- *
- * Commands are saved per engine class: all instances of that class will execute
- * those commands during context switch. The instruction, dword arguments,
- * addresses and values are in hex format like in the examples below.
+ * Examples:
*
- * #. Execute a LRI command to write0xDEADBEEFto register0x4f10after the
- * normal context restore::
+ * Allow both primary and media GTs to be initialized and used. This
+ matches
+ * the driver's default behavior::
*
- * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
- * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb
+ * # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
- * #. Execute a LRI command to write0xDEADBEEFto register0x4f10at the
- * beginning of the context restore::
+ * Allow only the primary GT of each tile to be initialized and used,
+ * effectively disabling the media GT if it exists on the platform::
*
- * # echo 'rcs cmd 11000001 4F100 DEADBEEF' \
- * > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb
-
- * #. Load certain values in a couple of registers (it can be used as a simpler
- * alternative to thecmd) action::
+ * # echo 'primary' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
- * # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb <<EOF
- * rcs reg 4F100 DEADBEEF
- * rcs reg 4F104 FFFFFFFF
- * EOF
+ * Allow only the media GT of each tile to be initialized and used,
+ * effectively disabling the primary GT. **This configuration will
+ cause
+ * device probe failure on all current platforms, but may be allowed on
+ * igpu platforms in the future**::
*
- * .. note::
+ * # echo 'media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
- * When using multiple lines, make sure to use a command that is
- * implemented with a single write syscall, like HEREDOC.
+ * Disable all GTs. Only other GPU IP (such as display) is potentially usable.
+ * **This configuration will cause device probe failure on all current
+ * platforms, but may be allowed on igpu platforms in the future**::
*
- * Currently this is implemented only for post and mid context restore and
- * these attributes can only be set before binding to the device.
+ * # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed
*
* Max SR-IOV Virtual Functions
* ----------------------------drivers/gpu/drm/xe/xe_configfs.c +0
-3@@ -256,15 +256,15 @@ struct xe_config_group_device { struct config_group sriov; struct xe_config_device { - u64 gt_types_allowed; - u64 engines_allowed; - struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];struct
wa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
- boolenable_survivability_mode;
+ structwa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
boolenable_psmi;
+ boolenable_survivability_mode;Quoted 8 lines
+ u64
engines_allowed;
+ u64gt_types_allowed;
struct {
- unsigned intmax_vfs;
booladmin_only_pf;
+ unsigned intmax_vfs;
} sriov;
} config;Here is the comment from [1], which seems reasonable to me. "
Does sorting struct members alphabetically here introduce unnecessary memory
padding?
Placing the bool fieldsenable_psmiandenable_survivability_modeimmediately
before the u64 fieldsengines_allowedandgt_types_allowedintroduces 6 bytes
of padding. In the sriov sub-struct, placing the booladmin_only_pfbefore the
unsigned intmax_vfsintroduces 3 bytes of padding.
Could these fields be ordered by size to minimize padding instead of being
sorted alphabetically?
"
Quoted 3 lines
If we decide to sort them, we could add comments like below for the flags in struct
intel_device_info.
" * Keep all flags below alphabetically sorted"
Oh thanks for pointing the sashiko review out - I wasn't getting emails about this so I hadn't looked there yet. I'll go through all that feedback and update where it makes sense.
I agree with what you said above too.
Thanks,
Stuart
Quoted 79 lines
[1]:
https://sashiko.dev/#/patchset/20260504044348.209625-1-stuart.summers%40intel.comShuicheng
Quoted 72 lines
drivers/gpu/drm/xe/xe_configfs.c +4
-4@@ -277,13 +277,13 @@ struct xe_config_group_device { }; static const struct xe_config_device device_defaults = { - .gt_types_allowed = U64_MAX, - .engines_allowed = U64_MAX, - .enable_survivability_mode = false, .enable_psmi = false, + .enable_survivability_mode = false, + .engines_allowed = U64_MAX, + .gt_types_allowed = U64_MAX, .sriov = { - .max_vfs = XE_DEFAULT_MAX_VFS, .admin_only_pf = XE_DEFAULT_ADMIN_ONLY_PF, + .max_vfs = XE_DEFAULT_MAX_VFS, }, };@@ -814,17 +814,17 @@ static ssize_tctx_restore_post_bb_store(struct
config_item*item,CONFIGFS_ATTR(, ctx_restore_mid_bb);CONFIGFS_ATTR(, ctx_restore_post_bb);CONFIGFS_ATTR(, enable_psmi);
+CONFIGFS_ATTR(,enable_survivability_mode);CONFIGFS_ATTR(, engines_allowed);CONFIGFS_ATTR(, gt_types_allowed); -CONFIGFS_ATTR(,enable_survivability_mode);static struct
configfs_attribute*xe_config_device_attrs[] = {&attr_ctx_restore_mid_bb,&attr_ctx_restore_post_bb,
&attr_enable_psmi,
+ &attr_enable_survivability_mode,&attr_engines_allowed,&attr_gt_types_allowed,
- &attr_enable_survivability_mode,
NULL,
};drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -929,12 +929,12 @@ static ssize_tsriov_admin_only_pf_store(struct
config_item*item, const char *p
return len;
}-CONFIGFS_ATTR(sriov_,
max_vfs);
CONFIGFS_ATTR(sriov_,admin_only_pf);
+CONFIGFS_ATTR(sriov_,max_vfs);static struct
configfs_attribute*xe_config_sriov_attrs[] = { - &sriov_attr_max_vfs,&sriov_attr_admin_only_pf,
+ &sriov_attr_max_vfs,
NULL,
};drivers/gpu/drm/xe/xe_configfs.c +2
-2@@ -1096,10 +1096,10 @@ static void dump_custom_dev_config(structpci_dev *pdev, dev->config.attr_); \ } while (0) - PRI_CUSTOM_ATTR("%llx", gt_types_allowed); - PRI_CUSTOM_ATTR("%llx", engines_allowed); PRI_CUSTOM_ATTR("%d", enable_psmi); PRI_CUSTOM_ATTR("%d", enable_survivability_mode); + PRI_CUSTOM_ATTR("%llx", engines_allowed); + PRI_CUSTOM_ATTR("%llx", gt_types_allowed); PRI_CUSTOM_ATTR("%u", sriov.admin_only_pf);#undef
PRI_CUSTOM_ATTR
--
2.43.0
PATCH 3/9 drm/xe: Split out configfs data structures
In reply to Stuart SummersSplit the configfs data structures into their own _types.h file.
--- drivers/gpu/drm/xe/xe_configfs.c | 85 +++++++------------------- drivers/gpu/drm/xe/xe_configfs_types.h | 59 ++++++++++++++++++ 2 files changed, 80 insertions(+), 64 deletions(-) create mode 100644 drivers/gpu/drm/xe/xe_configfs_types.h
drivers/gpu/drm/xe/xe_configfs.c +21 -64
@@ -4,7 +4,6 @@*/ #include <linux/bitops.h> -#include <linux/ctype.h> #include <linux/configfs.h> #include <linux/cleanup.h> #include <linux/find.h>
@@ -15,12 +14,10 @@#include "instructions/xe_mi_commands.h" #include "xe_configfs.h" +#include "xe_configfs_types.h" #include "xe_defaults.h" #include "xe_gt_types.h" -#include "xe_hw_engine_types.h" #include "xe_module.h" -#include "xe_pci_types.h" -#include "xe_sriov_types.h" /** * DOC: Xe Configfs
@@ -245,36 +242,6 @@* # rmdir /sys/kernel/config/xe/0000:03:00.0/ */ -/* Similar to struct xe_bb, but not tied to HW (yet) */ -struct wa_bb { -u32 *cs; -u32 len;/* in dwords */ -}; - -struct xe_config_group_device { -struct config_group group; -struct config_group sriov; - -struct xe_config_device { -struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; -struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ]; -bool enable_psmi; -bool enable_survivability_mode; -u64 engines_allowed; -u64 gt_types_allowed; -struct { -bool admin_only_pf; -unsigned int max_vfs; - } sriov; - } config; - -/* protects attributes */ -struct mutex lock; -/* matching descriptor */ -const struct xe_device_desc *desc; -/* tentative SR-IOV mode */ -enum xe_sriov_mode mode; -};static const struct xe_config_device device_defaults = { .enable_psmi = false,
@@ -322,16 +289,6 @@ static const struct { { .name = "media", .type = XE_GT_TYPE_MEDIA }, }; -static struct xe_config_group_device *to_xe_config_group_device (struct config_item *item) -{ - return container_of(to_config_group(item), struct xe_config_group_device, group); -} - -static struct xe_config_device *to_xe_config_device (struct config_item *item) -{ -return &to_xe_config_group_device (item)->config; -} -static bool is_bound (struct xe_config_group_device *dev) {unsigned int domain, bus, slot, function;
@@ -359,7 +316,7 @@ static bool is_bound(struct xe_config_group_device *dev)static ssize_t enable_survivability_mode_show (struct config_item *item,char *page) { -struct xe_config_device *dev =to_xe_config_device (item); +struct xe_config_device *dev =xe_configfs_to_device (item);return sprintf (page,"%d\n" , dev->enable_survivability_mode); }
@@ -367,7 +324,7 @@ static ssize_t enable_survivability_mode_show(struct config_item *item, char *pastatic ssize_t enable_survivability_mode_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);bool enable_survivability_mode;int ret;
@@ -386,7 +343,7 @@ static ssize_t enable_survivability_mode_store(struct config_item *item, const cstatic ssize_t gt_types_allowed_show (struct config_item *item,char *page) { -struct xe_config_device *dev =to_xe_config_device (item); +struct xe_config_device *dev =xe_configfs_to_device (item);char *p = page;for (size_t i =0 ; i <ARRAY_SIZE (gt_types); i++)
@@ -399,7 +356,7 @@ static ssize_t gt_types_allowed_show(struct config_item *item, char *page)static ssize_t gt_types_allowed_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item); char *buf __free(kfree) = kstrdup(page, GFP_KERNEL);char *p = buf;u64 typemask =0 ;
@@ -437,7 +394,7 @@ static ssize_t gt_types_allowed_store(struct config_item *item, const char *pagestatic ssize_t engines_allowed_show (struct config_item *item,char *page) { -struct xe_config_device *dev =to_xe_config_device (item); +struct xe_config_device *dev =xe_configfs_to_device (item);char *p = page;for (size_t i =0 ; i <ARRAY_SIZE (engine_info); i++) {
@@ -529,7 +486,7 @@ static int parse_engine(const char *s, const char *end_chars, u64 *mask,static ssize_t engines_allowed_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);ssize_t patternlen, p;u64 mask, val =0 ;
@@ -552,14 +509,14 @@ static ssize_t engines_allowed_store(struct config_item *item, const char *page,static ssize_t enable_psmi_show (struct config_item *item,char *page) { -struct xe_config_device *dev =to_xe_config_device (item); +struct xe_config_device *dev =xe_configfs_to_device (item);return sprintf (page,"%d\n" , dev->enable_psmi); }static ssize_t enable_psmi_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);bool val;int ret;
@@ -634,14 +591,14 @@ static ssize_t wa_bb_show(struct xe_config_group_device *dev,static ssize_t ctx_restore_mid_bb_show (struct config_item *item,char *page) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);return wa_bb_show (dev, dev->config.ctx_restore_mid_bb, page,SZ_4K ); }static ssize_t ctx_restore_post_bb_show (struct config_item *item,char *page) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);return wa_bb_show (dev, dev->config.ctx_restore_post_bb, page,SZ_4K ); }
@@ -798,7 +755,7 @@ static ssize_t wa_bb_store(struct wa_bb wa_bb[static XE_ENGINE_CLASS_MAX],static ssize_t ctx_restore_mid_bb_store (struct config_item *item,const char *data,size_t sz) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);return wa_bb_store (dev->config.ctx_restore_mid_bb, dev, data, sz); }
@@ -806,7 +763,7 @@ static ssize_t ctx_restore_mid_bb_store(struct config_item *item,static ssize_t ctx_restore_post_bb_store (struct config_item *item,const char *data,size_t sz) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);return wa_bb_store (dev->config.ctx_restore_post_bb, dev, data, sz); }
@@ -830,7 +787,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = {static void xe_config_device_release (struct config_item *item) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);mutex_destroy (&dev->lock);
@@ -846,7 +803,7 @@ static struct configfs_item_operations xe_config_device_ops = {static bool xe_config_device_is_visible (struct config_item *item,struct configfs_attribute *attr,int n) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);if (attr == &attr_enable_survivability_mode) {if (!dev->desc->is_dgfx || dev->desc->platform <XE_BATTLEMAGE )
@@ -869,7 +826,7 @@ static const struct config_item_type xe_config_device_type = {static ssize_t sriov_max_vfs_show (struct config_item *item,char *page) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);guard (mutex)(&dev->lock);
@@ -881,7 +838,7 @@ static ssize_t sriov_max_vfs_show(struct config_item *item, char *page)static ssize_t sriov_max_vfs_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);unsigned int max_vfs;int ret;
@@ -903,7 +860,7 @@ static ssize_t sriov_max_vfs_store(struct config_item *item, const char *page, sstatic ssize_t sriov_admin_only_pf_show (struct config_item *item,char *page) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);guard (mutex)(&dev->lock);
@@ -912,7 +869,7 @@ static ssize_t sriov_admin_only_pf_show(struct config_item *item, char *page)static ssize_t sriov_admin_only_pf_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);bool admin_only_pf;int ret;
@@ -941,7 +898,7 @@ static struct configfs_attribute *xe_config_sriov_attrs[] = {static bool xe_config_sriov_is_visible (struct config_item *item,struct configfs_attribute *attr,int n) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);if (attr == &sriov_attr_max_vfs && dev->mode !=XE_SRIOV_MODE_PF )return false;
@@ -1084,7 +1041,7 @@ static struct xe_config_group_device *find_xe_config_group_device(struct pci_devif (!item)return NULL ; -return to_xe_config_group_device (item); +return xe_configfs_to_group_device (item); } static void dump_custom_dev_config(struct pci_dev *pdev,
drivers/gpu/drm/xe/xe_configfs_types.h new file +59 -0
59 lines not on this page · raw message
Signature · 1 line
2.43.0
Re: [PATCH 3/9] drm/xe: Split out configfs data structures
Quoted 401 lines · Stuart Summers
Split the configfs data structures into their own
_types.hfile.Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 85 +++++++----------------- --
drivers/gpu/drm/xe/xe_configfs_types.h | 59 ++++++++++++++++++2 files changed, 80 insertions(+), 64 deletions(-)create mode 100644
drivers/gpu/drm/xe/xe_configfs_types.hdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -4,7 +4,6 @@*/
#include <linux/bitops.h>
-#include <linux/ctype.h>
#include <linux/configfs.h>
#include <linux/cleanup.h>
#include <linux/find.h>drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -15,12 +14,10 @@
#include "instructions/xe_mi_commands.h"
#include "xe_configfs.h"
+#include "xe_configfs_types.h"
#include "xe_defaults.h"
#include "xe_gt_types.h"
-#include "xe_hw_engine_types.h"
#include "xe_module.h"drivers/gpu/drm/xe/xe_configfs.c +0
-2-#include "xe_pci_types.h" -#include "xe_sriov_types.h"
/**
* DOC: Xe Configfsdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -245,36 +242,6 @@* # rmdir /sys/kernel/config/xe/0000:03:00.0/ */ drivers/gpu/drm/xe/xe_configfs.c +0
-13-/* Similar to struct xe_bb, but not tied to HW (yet) */ -struct wa_bb { - u32 *cs; - u32 len; /* in dwords */ -}; - -struct xe_config_group_device { - struct config_group group; - struct config_group sriov; - - struct xe_config_device { - struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX]; - struct wa_bbctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
- boolenable_psmi;
- boolenable_survivability_mode;
- u64engines_allowed;
- u64gt_types_allowed;
- struct {
- booladmin_only_pf;
- unsigned intmax_vfs;
- } sriov;
- } config;drivers/gpu/drm/xe/xe_configfs.c +0
-8- - /* protects attributes */ - struct mutex lock; - /* matching descriptor */ - const struct xe_device_desc *desc; - /* tentative SR-IOV mode */ - enum xe_sriov_mode mode; -};
static const structxe_config_devicedevice_defaults= { .enable_psmi = false,drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -322,16 +289,6 @@ static const struct { { .name = "media", .type = XE_GT_TYPE_MEDIA }, };
-static structxe_config_group_device*to_xe_config_group_device(structconfig_item*item) -{
- returncontainer_of(to_config_group(item), struct xe_config_group_device, group);drivers/gpu/drm/xe/xe_configfs.c +0
-3-} - -static struct xe_config_device *to_xe_config_device(struct
config_item*item)drivers/gpu/drm/xe/xe_configfs.c +0
-4-{ - return &to_xe_config_group_device(item)->config; -} -static bool
is_bound(struct xe_config_group_device *dev){
unsigned int domain, bus, slot, function;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -359,7 +316,7 @@ static bool is_bound(struct xe_config_group_device *dev)
staticssize_tenable_survivability_mode_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item);
returnsprintf(page, "%d\n", dev->enable_survivability_mode); }drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -367,7 +324,7 @@ static ssize_tenable_survivability_mode_show(struct
config_item*item, char *pa staticssize_tenable_survivability_mode_store(structconfig_item*item, const char *page,
size_tlen) {
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
boolenable_survivability_mode;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -386,7 +343,7 @@ static ssize_tenable_survivability_mode_store(struct
config_item*item, const c
staticssize_tgt_types_allowed_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item); char *p = page;
for (size_ti = 0; i <ARRAY_SIZE(gt_types); i++)drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -399,7 +356,7 @@ static ssize_t gt_types_allowed_show(struct config_item *item, char *page) static ssize_t gt_types_allowed_store(struct config_item *item,const char *page, size_t len) { - struct xe_config_group_device *dev =to_xe_config_group_device (item); + struct xe_config_group_device *dev =
xe_configfs_to_group_device(item);
char *buf__free(kfree)=kstrdup(page, GFP_KERNEL); char *p = buf;
u64 typemask = 0;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -437,7 +394,7 @@ static ssize_t gt_types_allowed_store(struct
config_item*item, const char *page
staticssize_tengines_allowed_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item); char *p = page;
for (size_ti = 0; i <ARRAY_SIZE(engine_info); i++) {drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -529,7 +486,7 @@ static int parse_engine(const char *s, const char *end_chars, u64 *mask,static
ssize_tengines_allowed_store(structconfig_item*item, const char *page,
size_tlen)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
ssize_t patternlen, p;
u64 mask, val = 0;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -552,14 +509,14 @@ static ssize_t engines_allowed_store(struct config_item *item, const char *page,
staticssize_tenable_psmi_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item);
returnsprintf(page, "%d\n", dev->enable_psmi); }
staticssize_tenable_psmi_store(struct config_item *item, const char *page, size_t len)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
bool val;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -634,14 +591,14 @@ static ssize_t wa_bb_show(struct xe_config_group_device *dev,
staticssize_tctx_restore_mid_bb_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_show(dev, dev->config.ctx_restore_mid_bb, page, SZ_4K);
}
staticssize_tctx_restore_post_bb_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_show(dev, dev->config.ctx_restore_post_bb, page, SZ_4K);
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -798,7 +755,7 @@ static ssize_t wa_bb_store(struct wa_bb wa_bb[static XE_ENGINE_CLASS_MAX], static ssize_t ctx_restore_mid_bb_store(struct config_item *item, const char *data, size_t sz) { - struct xe_config_group_device *dev =to_xe_config_group_device (item); + struct xe_config_group_device *dev =xe_configfs_to_group_device (item);
returnwa_bb_store(dev->config.ctx_restore_mid_bb, dev, data, sz);
}drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -806,7 +763,7 @@ static ssize_t ctx_restore_mid_bb_store(structconfig_item *item, static ssize_t ctx_restore_post_bb_store(struct config_item *item, const char *data, size_t sz) { - struct xe_config_group_device *dev = to_xe_config_group_device(item); + struct xe_config_group_device *dev =xe_configfs_to_group_device (item);
returnwa_bb_store(dev->config.ctx_restore_post_bb, dev, data, sz);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -830,7 +787,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = {
static voidxe_config_device_release(struct config_item *item){
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
mutex_destroy(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -846,7 +803,7 @@ static struct configfs_item_operations xe_config_device_ops = { static bool xe_config_device_is_visible(struct config_item *item,struct
configfs_attribute*attr, int n)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
if (attr ==&attr_enable_survivability_mode) { if (!dev->desc->is_dgfx ||dev->desc->platform<XE_BATTLEMAGE)drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -869,7 +826,7 @@ static const struct config_item_type xe_config_device_type = {
staticssize_tsriov_max_vfs_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
guard(mutex)(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -881,7 +838,7 @@ static ssize_t sriov_max_vfs_show(struct config_item *item, char *page)
staticssize_tsriov_max_vfs_store(struct config_item *item, const char *page, size_t len)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
unsigned intmax_vfs;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -903,7 +860,7 @@ static ssize_t sriov_max_vfs_store(struct
config_item*item, const char *page, s
staticssize_tsriov_admin_only_pf_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
guard(mutex)(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -912,7 +869,7 @@ static ssize_t sriov_admin_only_pf_show(struct config_item *item, char *page)
staticssize_tsriov_admin_only_pf_store(struct config_item *item, const char *page, size_t len)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
booladmin_only_pf;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -941,7 +898,7 @@ static struct configfs_attribute *xe_config_sriov_attrs[] = { static bool xe_config_sriov_is_visible(struct config_item *item,struct
configfs_attribute*attr, int n)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
if (attr ==&sriov_attr_max_vfs&&dev->mode!=XE_SRIOV_MODE_PF)
return false;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1084,7 +1041,7 @@ static struct xe_config_group_device *find_xe_config_group_device(struct pci_dev if (!item) return NULL;
- returnto_xe_config_group_device(item); + returnxe_configfs_to_group_device(item); }
static void dump_custom_dev_config(structpci_dev*pdev,drivers/gpu/drm/xe/xe_configfs_types.h new file +30
-0@@ -0,0 +1,59 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_TYPES_H_ +#define _XE_CONFIGFS_TYPES_H_ + +#include <linux/configfs.h> +#include <linux/mutex.h> +#include <linux/types.h> + +#include "xe_hw_engine_types.h" +#include "xe_pci_types.h" +#include "xe_sriov_types.h" + +struct config_item; + +/* Similar to struct xe_bb, but not tied to HW (yet) */ +struct wa_bb { + u32 *cs; + u32 len; /* in dwords */ +}; + +struct xe_config_group_device { + struct config_group group; + struct config_group sriov; + + struct xe_config_device { + struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX]; + struct wa_bbctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
+ boolenable_psmi;
+ boolenable_survivability_mode;
+ u64engines_allowed;
+ u64gt_types_allowed;
+ struct {
+ booladmin_only_pf;
+ unsigned intmax_vfs;
+ } sriov;
+ } config;drivers/gpu/drm/xe/xe_configfs_types.h new file +10
-0+ + /* protects attributes */ + struct mutex lock; + /* matching descriptor */ + const struct xe_device_desc *desc; + /* tentative SR-IOV mode */ + enum xe_sriov_mode mode; +}; + +static inline struct xe_config_group_device
Quoted 1 line
*xe_configfs_to_group_device(struct
config_item*item)
Sorry about that, I'll split these up into multiple lines so we aren't going too long here...
-Stuart
Quoted 12 lines
+{
+ returncontainer_of(to_config_group(item), struct xe_config_group_device, group);drivers/gpu/drm/xe/xe_configfs_types.h new file +3
-0+} + +static inline struct xe_config_device *xe_configfs_to_device(struct
config_item*item)drivers/gpu/drm/xe/xe_configfs_types.h new file +5
-0+{ + return &xe_configfs_to_group_device(item)->config; +} + +#endif /* _XE_CONFIGFS_TYPES_H_ */
Re: [PATCH 3/9] drm/xe: Split out configfs data structures
In reply to Stuart SummersQuoted 1 line · Stuart Summers
Split the configfs data structures into their own
_types.hfile.
Why? The commit message must always answer the question, "why".
The obvious downside here is that you expose the types that were previously hidden in xe_configfs.c to anyone who includes xe_configfs_types.h. And that header depends on a ton of other headers, making the header interdepencies worse overall.
Maybe you need that header later for something, but please spell that out here.
BR,
Jani.
Quoted 332 lines
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7--- drivers/gpu/drm/xe/xe_configfs.c | 85 +++++++------------------- drivers/gpu/drm/xe/xe_configfs_types.h | 59 ++++++++++++++++++ 2 files changed, 80 insertions(+), 64 deletions(-) create mode 100644 drivers/gpu/drm/xe/xe_configfs_types.hdrivers/gpu/drm/xe/xe_configfs.c +21
-64@@ -4,7 +4,6 @@*/ #include <linux/bitops.h> -#include <linux/ctype.h> #include <linux/configfs.h> #include <linux/cleanup.h> #include <linux/find.h> @@ -15,12 +14,10 @@#include "instructions/xe_mi_commands.h" #include "xe_configfs.h" +#include "xe_configfs_types.h" #include "xe_defaults.h" #include "xe_gt_types.h" -#include "xe_hw_engine_types.h" #include "xe_module.h" -#include "xe_pci_types.h" -#include "xe_sriov_types.h" /** * DOC: Xe Configfs @@ -245,36 +242,6 @@* # rmdir /sys/kernel/config/xe/0000:03:00.0/ */ -/* Similar to struct xe_bb, but not tied to HW (yet) */ -struct wa_bb { -u32 *cs; -u32 len;/* in dwords */ -}; - -struct xe_config_group_device { -struct config_group group; -struct config_group sriov; - -struct xe_config_device { -struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; -struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ]; -bool enable_psmi; -bool enable_survivability_mode; -u64 engines_allowed; -u64 gt_types_allowed; -struct { -bool admin_only_pf; -unsigned int max_vfs; - } sriov; - } config; - -/* protects attributes */ -struct mutex lock; -/* matching descriptor */ -const struct xe_device_desc *desc; -/* tentative SR-IOV mode */ -enum xe_sriov_mode mode; -};static const struct xe_config_device device_defaults = { .enable_psmi = false,@@ -322,16 +289,6 @@ static const struct { { .name = "media", .type = XE_GT_TYPE_MEDIA }, }; -static struct xe_config_group_device *to_xe_config_group_device (struct config_item *item) -{ - return container_of(to_config_group(item), struct xe_config_group_device, group); -} - -static struct xe_config_device *to_xe_config_device (struct config_item *item) -{ -return &to_xe_config_group_device (item)->config; -} -static bool is_bound (struct xe_config_group_device *dev) {unsigned int domain, bus, slot, function;@@ -359,7 +316,7 @@ static bool is_bound(struct xe_config_group_device *dev)static ssize_t enable_survivability_mode_show (struct config_item *item,char *page) { -struct xe_config_device *dev =to_xe_config_device (item); +struct xe_config_device *dev =xe_configfs_to_device (item);return sprintf (page,"%d\n" , dev->enable_survivability_mode); }@@ -367,7 +324,7 @@ static ssize_t enable_survivability_mode_show(struct config_item *item, char *pastatic ssize_t enable_survivability_mode_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);bool enable_survivability_mode;int ret;@@ -386,7 +343,7 @@ static ssize_t enable_survivability_mode_store(struct config_item *item, const cstatic ssize_t gt_types_allowed_show (struct config_item *item,char *page) { -struct xe_config_device *dev =to_xe_config_device (item); +struct xe_config_device *dev =xe_configfs_to_device (item);char *p = page;for (size_t i =0 ; i <ARRAY_SIZE (gt_types); i++)@@ -399,7 +356,7 @@ static ssize_t gt_types_allowed_show(struct config_item *item, char *page)static ssize_t gt_types_allowed_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item); char *buf __free(kfree) = kstrdup(page, GFP_KERNEL);char *p = buf;u64 typemask =0 ;@@ -437,7 +394,7 @@ static ssize_t gt_types_allowed_store(struct config_item *item, const char *pagestatic ssize_t engines_allowed_show (struct config_item *item,char *page) { -struct xe_config_device *dev =to_xe_config_device (item); +struct xe_config_device *dev =xe_configfs_to_device (item);char *p = page;for (size_t i =0 ; i <ARRAY_SIZE (engine_info); i++) {@@ -529,7 +486,7 @@ static int parse_engine(const char *s, const char *end_chars, u64 *mask,static ssize_t engines_allowed_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);ssize_t patternlen, p;u64 mask, val =0 ;@@ -552,14 +509,14 @@ static ssize_t engines_allowed_store(struct config_item *item, const char *page,static ssize_t enable_psmi_show (struct config_item *item,char *page) { -struct xe_config_device *dev =to_xe_config_device (item); +struct xe_config_device *dev =xe_configfs_to_device (item);return sprintf (page,"%d\n" , dev->enable_psmi); }static ssize_t enable_psmi_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);bool val;int ret;@@ -634,14 +591,14 @@ static ssize_t wa_bb_show(struct xe_config_group_device *dev,static ssize_t ctx_restore_mid_bb_show (struct config_item *item,char *page) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);return wa_bb_show (dev, dev->config.ctx_restore_mid_bb, page,SZ_4K ); }static ssize_t ctx_restore_post_bb_show (struct config_item *item,char *page) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);return wa_bb_show (dev, dev->config.ctx_restore_post_bb, page,SZ_4K ); }@@ -798,7 +755,7 @@ static ssize_t wa_bb_store(struct wa_bb wa_bb[static XE_ENGINE_CLASS_MAX],static ssize_t ctx_restore_mid_bb_store (struct config_item *item,const char *data,size_t sz) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);return wa_bb_store (dev->config.ctx_restore_mid_bb, dev, data, sz); }@@ -806,7 +763,7 @@ static ssize_t ctx_restore_mid_bb_store(struct config_item *item,static ssize_t ctx_restore_post_bb_store (struct config_item *item,const char *data,size_t sz) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);return wa_bb_store (dev->config.ctx_restore_post_bb, dev, data, sz); }@@ -830,7 +787,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = {static void xe_config_device_release (struct config_item *item) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);mutex_destroy (&dev->lock);@@ -846,7 +803,7 @@ static struct configfs_item_operations xe_config_device_ops = {static bool xe_config_device_is_visible (struct config_item *item,struct configfs_attribute *attr,int n) { -struct xe_config_group_device *dev =to_xe_config_group_device (item); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item);if (attr == &attr_enable_survivability_mode) {if (!dev->desc->is_dgfx || dev->desc->platform <XE_BATTLEMAGE )@@ -869,7 +826,7 @@ static const struct config_item_type xe_config_device_type = {static ssize_t sriov_max_vfs_show (struct config_item *item,char *page) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);guard (mutex)(&dev->lock);@@ -881,7 +838,7 @@ static ssize_t sriov_max_vfs_show(struct config_item *item, char *page)static ssize_t sriov_max_vfs_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);unsigned int max_vfs;int ret;@@ -903,7 +860,7 @@ static ssize_t sriov_max_vfs_store(struct config_item *item, const char *page, sstatic ssize_t sriov_admin_only_pf_show (struct config_item *item,char *page) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);guard (mutex)(&dev->lock);@@ -912,7 +869,7 @@ static ssize_t sriov_admin_only_pf_show(struct config_item *item, char *page)static ssize_t sriov_admin_only_pf_store (struct config_item *item,const char *page,size_t len) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);bool admin_only_pf;int ret;@@ -941,7 +898,7 @@ static struct configfs_attribute *xe_config_sriov_attrs[] = {static bool xe_config_sriov_is_visible (struct config_item *item,struct configfs_attribute *attr,int n) { -struct xe_config_group_device *dev =to_xe_config_group_device (item->ci_parent); +struct xe_config_group_device *dev =xe_configfs_to_group_device (item->ci_parent);if (attr == &sriov_attr_max_vfs && dev->mode !=XE_SRIOV_MODE_PF )return false;@@ -1084,7 +1041,7 @@ static struct xe_config_group_device *find_xe_config_group_device(struct pci_devif (!item)return NULL ; -return to_xe_config_group_device (item); +return xe_configfs_to_group_device (item); } static void dump_custom_dev_config(struct pci_dev *pdev,drivers/gpu/drm/xe/xe_configfs_types.h new file +59
-0@@ -0,0 +1,59 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_TYPES_H_ +#define _XE_CONFIGFS_TYPES_H_ + +#include <linux/configfs.h> +#include <linux/mutex.h> +#include <linux/types.h> + +#include "xe_hw_engine_types.h" +#include "xe_pci_types.h" +#include "xe_sriov_types.h" + +struct config_item ; + +/* Similar to struct xe_bb, but not tied to HW (yet) */ +struct wa_bb { +u32 *cs; +u32 len;/* in dwords */ +}; + +struct xe_config_group_device { +struct config_group group; +struct config_group sriov; + +struct xe_config_device { +struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; +struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ]; +bool enable_psmi; +bool enable_survivability_mode; +u64 engines_allowed; +u64 gt_types_allowed; +struct { +bool admin_only_pf; +unsigned int max_vfs; + } sriov; + } config; + +/* protects attributes */ +struct mutex lock; +/* matching descriptor */ +const struct xe_device_desc *desc; +/* tentative SR-IOV mode */ +enum xe_sriov_mode mode; +}; + +static inline struct xe_config_group_device *xe_configfs_to_group_device (struct config_item *item) +{ + return container_of(to_config_group(item), struct xe_config_group_device, group); +} + +static inline struct xe_config_device *xe_configfs_to_device (struct config_item *item) +{ +return &xe_configfs_to_group_device (item)->config; +} + +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
Jani Nikula, Intel
Re: [PATCH 3/9] drm/xe: Split out configfs data structures
Quoted 5 lines · Jani Nikula
Quoted 1 line · Stuart Summers
Split the configfs data structures into their own
_types.hfile.Why? The commit message must always answer the question, "why".
Quoted 8 lines
The obvious downside here is that you expose the types that were previously hidden in
xe_configfs.cto anyone who includesxe_configfs_types.h. And that header depends on a ton of other headers,
making the header interdepencies worse overall.Maybe you need that header later for something, but please spell that out here.
Yeah this is exactly true.. I'll do a better job with the commit messages here going forward. Really appreciate the feedback Jani!
-Stuart
Quoted 424 lines
BR,
Jani.Quoted 418 lines
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 85 +++++++---------------
----
drivers/gpu/drm/xe/xe_configfs_types.h | 59 ++++++++++++++++++2 files changed, 80 insertions(+), 64 deletions(-)create mode 100644
drivers/gpu/drm/xe/xe_configfs_types.hdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -4,7 +4,6 @@*/
#include <linux/bitops.h>
-#include <linux/ctype.h>
#include <linux/configfs.h>
#include <linux/cleanup.h>
#include <linux/find.h>drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -15,12 +14,10 @@
#include "instructions/xe_mi_commands.h"
#include "xe_configfs.h"
+#include "xe_configfs_types.h"
#include "xe_defaults.h"
#include "xe_gt_types.h"
-#include "xe_hw_engine_types.h"
#include "xe_module.h"drivers/gpu/drm/xe/xe_configfs.c +0
-2-#include "xe_pci_types.h" -#include "xe_sriov_types.h"
/**
* DOC: Xe Configfsdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -245,36 +242,6 @@* # rmdir /sys/kernel/config/xe/0000:03:00.0/ */ drivers/gpu/drm/xe/xe_configfs.c +0
-12-/* Similar to struct xe_bb, but not tied to HW (yet) */ -struct wa_bb { - u32 *cs; - u32 len; /* in dwords */ -}; - -struct xe_config_group_device { - struct config_group group; - struct config_group sriov; - - struct xe_config_device { - struct wa_bbctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
- structwa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
- boolenable_psmi;
- boolenable_survivability_mode;
- u64engines_allowed;
- u64gt_types_allowed;
- struct {
- booladmin_only_pf;
- unsigned intmax_vfs;
- } sriov;
- } config;drivers/gpu/drm/xe/xe_configfs.c +0
-8- - /* protects attributes */ - struct mutex lock; - /* matching descriptor */ - const struct xe_device_desc *desc; - /* tentative SR-IOV mode */ - enum xe_sriov_mode mode; -};
static const structxe_config_devicedevice_defaults= { .enable_psmi = false,drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -322,16 +289,6 @@ static const struct { { .name = "media", .type = XE_GT_TYPE_MEDIA }, };
-static structxe_config_group_device*to_xe_config_group_device(structconfig_item*item) -{
- returncontainer_of(to_config_group(item), struct xe_config_group_device, group);drivers/gpu/drm/xe/xe_configfs.c +0
-3-} - -static struct xe_config_device *to_xe_config_device(struct
config_item*item)drivers/gpu/drm/xe/xe_configfs.c +0
-4-{ - return &to_xe_config_group_device(item)->config; -} -static bool
is_bound(struct xe_config_group_device *dev){
unsigned int domain, bus, slot, function;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -359,7 +316,7 @@ static bool is_bound(struct xe_config_group_device *dev)
staticssize_tenable_survivability_mode_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item);
return sprintf(page, "%d\n", dev-Quoted 1 line
enable_survivability_mode);}
drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -367,7 +324,7 @@ static ssize_tenable_survivability_mode_show(struct
config_item*item, char *pa staticssize_tenable_survivability_mode_store(structconfig_item*item, const char *page,
size_tlen) {
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
boolenable_survivability_mode;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -386,7 +343,7 @@ static ssize_tenable_survivability_mode_store(struct
config_item*item, const c
staticssize_tgt_types_allowed_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item); char *p = page;
for (size_ti = 0; i <ARRAY_SIZE(gt_types); i++)drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -399,7 +356,7 @@ static ssize_t gt_types_allowed_show(struct config_item *item, char *page) static ssize_t gt_types_allowed_store(struct config_item *item,const char *page, size_t len) { - struct xe_config_group_device *dev =to_xe_config_group_device (item); + struct xe_config_group_device *dev =
xe_configfs_to_group_device(item);
char *buf__free(kfree)=kstrdup(page, GFP_KERNEL); char *p = buf;
u64 typemask = 0;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -437,7 +394,7 @@ static ssize_t gt_types_allowed_store(struct
config_item*item, const char *page
staticssize_tengines_allowed_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item); char *p = page;
for (size_ti = 0; i <ARRAY_SIZE(engine_info); i++) {drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -529,7 +486,7 @@ static int parse_engine(const char *s, const char *end_chars, u64 *mask, static ssize_t engines_allowed_store(struct config_item *item, const char *page, size_t len) { - struct xe_config_group_device *dev = to_xe_config_group_device(item); + struct xe_config_group_device *dev =
xe_configfs_to_group_device(item);
ssize_t patternlen, p;
u64 mask, val = 0;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -552,14 +509,14 @@ static ssize_t engines_allowed_store(struct config_item *item, const char *page,
staticssize_tenable_psmi_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item);
returnsprintf(page, "%d\n", dev->enable_psmi); }
staticssize_tenable_psmi_store(struct config_item *item, const char *page, size_t len)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
bool val;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -634,14 +591,14 @@ static ssize_t wa_bb_show(struct xe_config_group_device *dev,
staticssize_tctx_restore_mid_bb_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_show(dev, dev->config.ctx_restore_mid_bb, page, SZ_4K);
}
staticssize_tctx_restore_post_bb_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_show(dev, dev->config.ctx_restore_post_bb, page, SZ_4K);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -798,7 +755,7 @@ static ssize_t wa_bb_store(struct wa_bb wa_bb[static XE_ENGINE_CLASS_MAX], static ssize_t ctx_restore_mid_bb_store(struct config_item *item,const char *data,
size_tsz)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_store(dev->config.ctx_restore_mid_bb, dev, data, sz);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -806,7 +763,7 @@ static ssize_t ctx_restore_mid_bb_store(struct config_item *item, static ssize_t ctx_restore_post_bb_store(struct config_item *item,const char *data,
size_tsz)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_store(dev->config.ctx_restore_post_bb, dev, data, sz);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -830,7 +787,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = {
static voidxe_config_device_release(struct config_item *item){
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
mutex_destroy(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -846,7 +803,7 @@ static struct configfs_item_operations xe_config_device_ops = { static bool xe_config_device_is_visible(struct config_item *item,struct
configfs_attribute*attr, int n)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
if (attr ==&attr_enable_survivability_mode) { if (!dev->desc->is_dgfx ||dev->desc->platform<XE_BATTLEMAGE)drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -869,7 +826,7 @@ static const struct config_item_type xe_config_device_type = {
staticssize_tsriov_max_vfs_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
guard(mutex)(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -881,7 +838,7 @@ static ssize_t sriov_max_vfs_show(struct config_item *item, char *page)
staticssize_tsriov_max_vfs_store(struct config_item *item, const char *page, size_t len)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
unsigned intmax_vfs;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -903,7 +860,7 @@ static ssize_t sriov_max_vfs_store(struct
config_item*item, const char *page, s
staticssize_tsriov_admin_only_pf_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
guard(mutex)(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -912,7 +869,7 @@ static ssize_t sriov_admin_only_pf_show(struct config_item *item, char *page)
staticssize_tsriov_admin_only_pf_store(struct config_item *item, const char *page, size_t len)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
booladmin_only_pf;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -941,7 +898,7 @@ static struct configfs_attribute *xe_config_sriov_attrs[] = { static bool xe_config_sriov_is_visible(struct config_item *item,struct
configfs_attribute*attr, int n)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
if (attr ==&sriov_attr_max_vfs&&dev->mode!=XE_SRIOV_MODE_PF)
return false;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1084,7 +1041,7 @@ static struct xe_config_group_device *find_xe_config_group_device(struct pci_dev if (!item) return NULL;
- returnto_xe_config_group_device(item); + returnxe_configfs_to_group_device(item); }
static void dump_custom_dev_config(structpci_dev*pdev,drivers/gpu/drm/xe/xe_configfs_types.h new file +29
-0@@ -0,0 +1,59 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_TYPES_H_ +#define _XE_CONFIGFS_TYPES_H_ + +#include <linux/configfs.h> +#include <linux/mutex.h> +#include <linux/types.h> + +#include "xe_hw_engine_types.h" +#include "xe_pci_types.h" +#include "xe_sriov_types.h" + +struct config_item; + +/* Similar to struct xe_bb, but not tied to HW (yet) */ +struct wa_bb { + u32 *cs; + u32 len; /* in dwords */ +}; + +struct xe_config_group_device { + struct config_group group; + struct config_group sriov; + + struct xe_config_device { + struct wa_bbctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
+ structwa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
+ boolenable_psmi;
+ boolenable_survivability_mode;
+ u64engines_allowed;
+ u64gt_types_allowed;
+ struct {
+ booladmin_only_pf;
+ unsigned intmax_vfs;
+ } sriov;
+ } config;drivers/gpu/drm/xe/xe_configfs_types.h new file +10
-0+ + /* protects attributes */ + struct mutex lock; + /* matching descriptor */ + const struct xe_device_desc *desc; + /* tentative SR-IOV mode */ + enum xe_sriov_mode mode; +}; + +static inline struct xe_config_group_device*xe_configfs_to_group_device(struct
config_item*item) +{
+ returncontainer_of(to_config_group(item), struct xe_config_group_device, group);drivers/gpu/drm/xe/xe_configfs_types.h new file +3
-0+} + +static inline struct xe_config_device*xe_configfs_to_device(struct
config_item*item)drivers/gpu/drm/xe/xe_configfs_types.h new file +5
-0+{ + return &xe_configfs_to_group_device(item)->config; +} + +#endif /* _XE_CONFIGFS_TYPES_H_ */
Matthew Brost
maintainerRe: [PATCH 3/9] drm/xe: Split out configfs data structures
Quoted 6 lines · Stuart
Quoted 5 lines · Jani Nikula
Quoted 1 line · Stuart Summers
Split the configfs data structures into their own
_types.hfile.Why? The commit message must always answer the question, "why".
This is a good point. In general, we should be moving public structs that have no business being in public headers (we do quite a bit of this in Xe) into private structs in C files. This patch does the opposite, so unless there is a very good reason to make this public (e.g, a struct is intended to embedded), my initial reaction is not to do this.
Private structures enforce good layering and component isolation, and prevent vectors for abuse (i.e., another layer messing with a struct owned by a different layer).
Matt
Quoted 439 lines
Quoted 8 lines
The obvious downside here is that you expose the types that were previously hidden in
xe_configfs.cto anyone who includesxe_configfs_types.h. And that header depends on a ton of other headers,
making the header interdepencies worse overall.Maybe you need that header later for something, but please spell that out here.
Yeah this is exactly true.. I'll do a better job with the commit messages here going forward. Really appreciate the feedback Jani!
-Stuart
Quoted 424 lines
BR,
Jani.Quoted 418 lines
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 85 +++++++---------------
----
drivers/gpu/drm/xe/xe_configfs_types.h | 59 ++++++++++++++++++2 files changed, 80 insertions(+), 64 deletions(-)create mode 100644
drivers/gpu/drm/xe/xe_configfs_types.hdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -4,7 +4,6 @@*/
#include <linux/bitops.h>
-#include <linux/ctype.h>
#include <linux/configfs.h>
#include <linux/cleanup.h>
#include <linux/find.h>drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -15,12 +14,10 @@
#include "instructions/xe_mi_commands.h"
#include "xe_configfs.h"
+#include "xe_configfs_types.h"
#include "xe_defaults.h"
#include "xe_gt_types.h"
-#include "xe_hw_engine_types.h"
#include "xe_module.h"drivers/gpu/drm/xe/xe_configfs.c +0
-2-#include "xe_pci_types.h" -#include "xe_sriov_types.h"
/**
* DOC: Xe Configfsdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -245,36 +242,6 @@* # rmdir /sys/kernel/config/xe/0000:03:00.0/ */ drivers/gpu/drm/xe/xe_configfs.c +0
-12-/* Similar to struct xe_bb, but not tied to HW (yet) */ -struct wa_bb { - u32 *cs; - u32 len; /* in dwords */ -}; - -struct xe_config_group_device { - struct config_group group; - struct config_group sriov; - - struct xe_config_device { - struct wa_bbctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
- structwa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
- boolenable_psmi;
- boolenable_survivability_mode;
- u64engines_allowed;
- u64gt_types_allowed;
- struct {
- booladmin_only_pf;
- unsigned intmax_vfs;
- } sriov;
- } config;drivers/gpu/drm/xe/xe_configfs.c +0
-8- - /* protects attributes */ - struct mutex lock; - /* matching descriptor */ - const struct xe_device_desc *desc; - /* tentative SR-IOV mode */ - enum xe_sriov_mode mode; -};
static const structxe_config_devicedevice_defaults= { .enable_psmi = false,drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -322,16 +289,6 @@ static const struct { { .name = "media", .type = XE_GT_TYPE_MEDIA }, };
-static structxe_config_group_device*to_xe_config_group_device(structconfig_item*item) -{
- returncontainer_of(to_config_group(item), struct xe_config_group_device, group);drivers/gpu/drm/xe/xe_configfs.c +0
-3-} - -static struct xe_config_device *to_xe_config_device(struct
config_item*item)drivers/gpu/drm/xe/xe_configfs.c +0
-4-{ - return &to_xe_config_group_device(item)->config; -} -static bool
is_bound(struct xe_config_group_device *dev){
unsigned int domain, bus, slot, function;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -359,7 +316,7 @@ static bool is_bound(struct xe_config_group_device *dev)
staticssize_tenable_survivability_mode_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item);
return sprintf(page, "%d\n", dev-Quoted 1 line
enable_survivability_mode);}
drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -367,7 +324,7 @@ static ssize_tenable_survivability_mode_show(struct
config_item*item, char *pa staticssize_tenable_survivability_mode_store(structconfig_item*item, const char *page,
size_tlen) {
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
boolenable_survivability_mode;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -386,7 +343,7 @@ static ssize_tenable_survivability_mode_store(struct
config_item*item, const c
staticssize_tgt_types_allowed_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item); char *p = page;
for (size_ti = 0; i <ARRAY_SIZE(gt_types); i++)drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -399,7 +356,7 @@ static ssize_t gt_types_allowed_show(struct config_item *item, char *page) static ssize_t gt_types_allowed_store(struct config_item *item,const char *page, size_t len) { - struct xe_config_group_device *dev =to_xe_config_group_device (item); + struct xe_config_group_device *dev =
xe_configfs_to_group_device(item);
char *buf__free(kfree)=kstrdup(page, GFP_KERNEL); char *p = buf;
u64 typemask = 0;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -437,7 +394,7 @@ static ssize_t gt_types_allowed_store(struct
config_item*item, const char *page
staticssize_tengines_allowed_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item); char *p = page;
for (size_ti = 0; i <ARRAY_SIZE(engine_info); i++) {drivers/gpu/drm/xe/xe_configfs.c +1
-1@@ -529,7 +486,7 @@ static int parse_engine(const char *s, const char *end_chars, u64 *mask, static ssize_t engines_allowed_store(struct config_item *item, const char *page, size_t len) { - struct xe_config_group_device *dev = to_xe_config_group_device(item); + struct xe_config_group_device *dev =
xe_configfs_to_group_device(item);
ssize_t patternlen, p;
u64 mask, val = 0;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -552,14 +509,14 @@ static ssize_t engines_allowed_store(struct config_item *item, const char *page,
staticssize_tenable_psmi_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item); + structxe_config_device*dev =xe_configfs_to_device(item);
returnsprintf(page, "%d\n", dev->enable_psmi); }
staticssize_tenable_psmi_store(struct config_item *item, const char *page, size_t len)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
bool val;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -634,14 +591,14 @@ static ssize_t wa_bb_show(struct xe_config_group_device *dev,
staticssize_tctx_restore_mid_bb_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_show(dev, dev->config.ctx_restore_mid_bb, page, SZ_4K);
}
staticssize_tctx_restore_post_bb_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_show(dev, dev->config.ctx_restore_post_bb, page, SZ_4K);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -798,7 +755,7 @@ static ssize_t wa_bb_store(struct wa_bb wa_bb[static XE_ENGINE_CLASS_MAX], static ssize_t ctx_restore_mid_bb_store(struct config_item *item,const char *data,
size_tsz)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_store(dev->config.ctx_restore_mid_bb, dev, data, sz);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -806,7 +763,7 @@ static ssize_t ctx_restore_mid_bb_store(struct config_item *item, static ssize_t ctx_restore_post_bb_store(struct config_item *item,const char *data,
size_tsz)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_store(dev->config.ctx_restore_post_bb, dev, data, sz);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -830,7 +787,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = {
static voidxe_config_device_release(struct config_item *item){
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
mutex_destroy(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -846,7 +803,7 @@ static struct configfs_item_operations xe_config_device_ops = { static bool xe_config_device_is_visible(struct config_item *item,struct
configfs_attribute*attr, int n)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
if (attr ==&attr_enable_survivability_mode) { if (!dev->desc->is_dgfx ||dev->desc->platform<XE_BATTLEMAGE)drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -869,7 +826,7 @@ static const struct config_item_type xe_config_device_type = {
staticssize_tsriov_max_vfs_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
guard(mutex)(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -881,7 +838,7 @@ static ssize_t sriov_max_vfs_show(struct config_item *item, char *page)
staticssize_tsriov_max_vfs_store(struct config_item *item, const char *page, size_t len)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
unsigned intmax_vfs;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -903,7 +860,7 @@ static ssize_t sriov_max_vfs_store(struct
config_item*item, const char *page, s
staticssize_tsriov_admin_only_pf_show(struct config_item *item, char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
guard(mutex)(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -912,7 +869,7 @@ static ssize_t sriov_admin_only_pf_show(struct config_item *item, char *page)
staticssize_tsriov_admin_only_pf_store(struct config_item *item, const char *page, size_t len)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
booladmin_only_pf;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -941,7 +898,7 @@ static struct configfs_attribute *xe_config_sriov_attrs[] = { static bool xe_config_sriov_is_visible(struct config_item *item,struct
configfs_attribute*attr, int n)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
if (attr ==&sriov_attr_max_vfs&&dev->mode!=XE_SRIOV_MODE_PF)
return false;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1084,7 +1041,7 @@ static struct xe_config_group_device *find_xe_config_group_device(struct pci_dev if (!item) return NULL;
- returnto_xe_config_group_device(item); + returnxe_configfs_to_group_device(item); }
static void dump_custom_dev_config(structpci_dev*pdev,drivers/gpu/drm/xe/xe_configfs_types.h new file +29
-0@@ -0,0 +1,59 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_TYPES_H_ +#define _XE_CONFIGFS_TYPES_H_ + +#include <linux/configfs.h> +#include <linux/mutex.h> +#include <linux/types.h> + +#include "xe_hw_engine_types.h" +#include "xe_pci_types.h" +#include "xe_sriov_types.h" + +struct config_item; + +/* Similar to struct xe_bb, but not tied to HW (yet) */ +struct wa_bb { + u32 *cs; + u32 len; /* in dwords */ +}; + +struct xe_config_group_device { + struct config_group group; + struct config_group sriov; + + struct xe_config_device { + struct wa_bbctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
+ structwa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
+ boolenable_psmi;
+ boolenable_survivability_mode;
+ u64engines_allowed;
+ u64gt_types_allowed;
+ struct {
+ booladmin_only_pf;
+ unsigned intmax_vfs;
+ } sriov;
+ } config;drivers/gpu/drm/xe/xe_configfs_types.h new file +10
-0+ + /* protects attributes */ + struct mutex lock; + /* matching descriptor */ + const struct xe_device_desc *desc; + /* tentative SR-IOV mode */ + enum xe_sriov_mode mode; +}; + +static inline struct xe_config_group_device*xe_configfs_to_group_device(struct
config_item*item) +{
+ returncontainer_of(to_config_group(item), struct xe_config_group_device, group);drivers/gpu/drm/xe/xe_configfs_types.h new file +3
-0+} + +static inline struct xe_config_device*xe_configfs_to_device(struct
config_item*item)drivers/gpu/drm/xe/xe_configfs_types.h new file +5
-0+{ + return &xe_configfs_to_group_device(item)->config; +} + +#endif /* _XE_CONFIGFS_TYPES_H_ */
Re: [PATCH 3/9] drm/xe: Split out configfs data structures
Quoted 18 lines · Matthew Brost
Quoted 8 lines · Stuart
Quoted 7 lines · Jani Nikula
Quoted 2 lines · Stuart Summers
Split the configfs data structures into their own
_types.hfile.Why? The commit message must always answer the question, "why".
This is a good point. In general, we should be moving public structs that have no business being in public headers (we do quite a bit of this
in Xe) into private structs in C files. This patch does the opposite, so
unless there is a very good reason to make this public (e.g, a struct is intended to embedded), my initial reaction is not to do this.
Quoted 3 lines
Private structures enforce good layering and component isolation, and prevent vectors for abuse (i.e., another layer messing with a struct owned by a different layer).
Yeah makes sense. In the series overall I'm creating new xe_configfs_debug.c and xe_configfs_debug_params.c files that each need to access these structs, hence the move to a common header. I could wait to do that move until the patch that actually uses it - just thought the review would be a little easier this way.
Either way I agree the commit message should have been explicit - I'll fix that in the next rev.
Thanks,
Stuart
Quoted 480 lines
Matt
Quoted 477 lines
Quoted 9 lines
The obvious downside here is that you expose the types that were previously hidden in
xe_configfs.cto anyone who includesxe_configfs_types.h. And that header depends on a ton of other headers,
making the header interdepencies worse overall.Maybe you need that header later for something, but please spell that
out here.Yeah this is exactly true.. I'll do a better job with the commit messages here going forward. Really appreciate the feedback Jani!
-Stuart
Quoted 461 lines
BR,
Jani.Quoted 455 lines
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/xe_configfs.c | 85 +++++++-----------
----
----
drivers/gpu/drm/xe/xe_configfs_types.h | 59 ++++++++++++++++++2 files changed, 80 insertions(+), 64 deletions(-)create mode 100644
drivers/gpu/drm/xe/xe_configfs_types.hdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -4,7 +4,6 @@*/
#include <linux/bitops.h>
-#include <linux/ctype.h>
#include <linux/configfs.h>
#include <linux/cleanup.h>
#include <linux/find.h>drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -15,12 +14,10 @@
#include "instructions/xe_mi_commands.h"
#include "xe_configfs.h"
+#include "xe_configfs_types.h"
#include "xe_defaults.h"
#include "xe_gt_types.h"
-#include "xe_hw_engine_types.h"
#include "xe_module.h"drivers/gpu/drm/xe/xe_configfs.c +0
-2-#include "xe_pci_types.h" -#include "xe_sriov_types.h"
/**
* DOC: Xe Configfsdrivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -245,36 +242,6 @@* # rmdir /sys/kernel/config/xe/0000:03:00.0/ */ drivers/gpu/drm/xe/xe_configfs.c +0
-12-/* Similar to struct xe_bb, but not tied to HW (yet) */ -struct wa_bb { - u32 *cs; - u32 len; /* in dwords */ -}; - -struct xe_config_group_device { - struct config_group group; - struct config_group sriov; - - struct xe_config_device { - struct wa_bbctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
- structwa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
- boolenable_psmi;
- boolenable_survivability_mode;
- u64engines_allowed;
- u64gt_types_allowed;
- struct {
- booladmin_only_pf;
- unsigned intmax_vfs;
- } sriov;
- } config;drivers/gpu/drm/xe/xe_configfs.c +0
-8- - /* protects attributes */ - struct mutex lock; - /* matching descriptor */ - const struct xe_device_desc *desc; - /* tentative SR-IOV mode */ - enum xe_sriov_mode mode; -};
static const structxe_config_devicedevice_defaults= { .enable_psmi = false,drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -322,16 +289,6 @@ static const struct { { .name = "media", .type = XE_GT_TYPE_MEDIA }, };
-static structxe_config_group_device*to_xe_config_group_device(structconfig_item*item) -{
- returncontainer_of(to_config_group(item), struct xe_config_group_device, group);drivers/gpu/drm/xe/xe_configfs.c +0
-3-} - -static struct xe_config_device *to_xe_config_device(struct
config_item*item)drivers/gpu/drm/xe/xe_configfs.c +0
-4-{ - return &to_xe_config_group_device(item)->config; -} -static bool
is_bound(struct xe_config_group_device *dev){
unsigned int domain, bus, slot, function;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -359,7 +316,7 @@ static bool is_bound(struct xe_config_group_device *dev)
staticssize_tenable_survivability_mode_show(structconfig_item
*item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item);
+ structxe_config_device*dev =xe_configfs_to_device(item);
return sprintf(page, "%d\n", dev-Quoted 1 line
enable_survivability_mode);}
drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -367,7 +324,7 @@ static ssize_tenable_survivability_mode_show(struct
config_item*item, char *pa
staticssize_tenable_survivability_mode_store(structconfig_item
*item, const char *page,
size_tlen) {
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
boolenable_survivability_mode;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -386,7 +343,7 @@ static ssize_tenable_survivability_mode_store(struct
config_item*item, const c
staticssize_tgt_types_allowed_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item);
+ structxe_config_device*dev =xe_configfs_to_device(item);
char *p = page;
for (size_ti = 0; i <ARRAY_SIZE(gt_types); i++)drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -399,7 +356,7 @@ static ssize_t gt_types_allowed_show(struct config_item *item, char *page)static
ssize_tgt_types_allowed_store(structconfig_item*item,
const char *page,
size_tlen) {
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
char *buf__free(kfree)=kstrdup(page, GFP_KERNEL); char *p = buf;
u64 typemask = 0;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -437,7 +394,7 @@ static ssize_tgt_types_allowed_store(struct
config_item*item, const char *page
staticssize_tengines_allowed_show(structconfig_item*item, char
*page)
{
- structxe_config_device*dev =to_xe_config_device(item);
+ structxe_config_device*dev =xe_configfs_to_device(item);
char *p = page;
for (size_ti = 0; i <ARRAY_SIZE(engine_info); i++) {drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -529,7 +486,7 @@ static int parse_engine(const char *s,const
char*end_chars, u64 *mask,
staticssize_tengines_allowed_store(structconfig_item*item, const char *page,
size_tlen)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
ssize_t patternlen, p;
u64 mask, val = 0;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -552,14 +509,14 @@ static ssize_tengines_allowed_store(struct
config_item*item, const char *page,
staticssize_tenable_psmi_show(struct config_item *item, char *page)
{
- structxe_config_device*dev =to_xe_config_device(item);
+ structxe_config_device*dev =xe_configfs_to_device(item);
returnsprintf(page, "%d\n", dev->enable_psmi); }
staticssize_tenable_psmi_store(structconfig_item*item, const
char *page,size_tlen)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
bool val;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -634,14 +591,14 @@ static ssize_t wa_bb_show(struct xe_config_group_device *dev,
staticssize_tctx_restore_mid_bb_show(structconfig_item*item,
char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_show(dev, dev->config.ctx_restore_mid_bb, page, SZ_4K);
}
staticssize_tctx_restore_post_bb_show(structconfig_item*item,
char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_show(dev, dev->config.ctx_restore_post_bb, page, SZ_4K);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -798,7 +755,7 @@ static ssize_t wa_bb_store(struct wa_bb wa_bb[static XE_ENGINE_CLASS_MAX],static
ssize_tctx_restore_mid_bb_store(structconfig_item*item,
const char *data,size_t
sz)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
returnwa_bb_store(dev->config.ctx_restore_mid_bb, dev, data, sz);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -806,7 +763,7 @@ static ssize_tctx_restore_mid_bb_store(struct
config_item*item,
staticssize_tctx_restore_post_bb_store(structconfig_item*item,
const char *data,size_t
sz)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
return wa_bb_store(dev->config.ctx_restore_post_bb, dev,
data, sz);
}drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -830,7 +787,7 @@ static struct configfs_attribute *xe_config_device_attrs[] = {
static voidxe_config_device_release(struct config_item *item){
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
mutex_destroy(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -846,7 +803,7 @@ static struct configfs_item_operations xe_config_device_ops = {static bool xe_config_device_is_visible(struct
config_item*item,
structconfigfs_attribute
*attr, int n)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item);
if (attr ==&attr_enable_survivability_mode) { if (!dev->desc->is_dgfx ||dev->desc->platform<XE_BATTLEMAGE)drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -869,7 +826,7 @@ static const struct config_item_type xe_config_device_type = {
staticssize_tsriov_max_vfs_show(structconfig_item*item, char
*page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
guard(mutex)(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -881,7 +838,7 @@ static ssize_t sriov_max_vfs_show(struct config_item *item, char *page)
staticssize_tsriov_max_vfs_store(structconfig_item*item, const
char *page,size_tlen)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent); unsigned intmax_vfs;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -903,7 +860,7 @@ static ssize_t sriov_max_vfs_store(struct
config_item*item, const char *page, s
staticssize_tsriov_admin_only_pf_show(structconfig_item*item,
char *page)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
guard(mutex)(&dev->lock);drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -912,7 +869,7 @@ static ssize_tsriov_admin_only_pf_show(struct
config_item*item, char *page)
staticssize_tsriov_admin_only_pf_store(structconfig_item*item,
const char *page,size_tlen)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
booladmin_only_pf;
int ret;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -941,7 +898,7 @@ static struct configfs_attribute *xe_config_sriov_attrs[] = {static bool xe_config_sriov_is_visible(struct
config_item*item,
structconfigfs_attribute
*attr, int n)
{
- structxe_config_group_device*dev =to_xe_config_group_device(item->ci_parent);
+ structxe_config_group_device*dev =xe_configfs_to_group_device(item->ci_parent);
if (attr ==&sriov_attr_max_vfs&&dev->mode!=XE_SRIOV_MODE_PF)
return false;drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1084,7 +1041,7 @@ static struct xe_config_group_device *find_xe_config_group_device(struct pci_dev if (!item) return NULL;
- returnto_xe_config_group_device(item); + returnxe_configfs_to_group_device(item); }
static void dump_custom_dev_config(structpci_dev*pdev,drivers/gpu/drm/xe/xe_configfs_types.h new file +29
-0@@ -0,0 +1,59 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_TYPES_H_ +#define _XE_CONFIGFS_TYPES_H_ + +#include <linux/configfs.h> +#include <linux/mutex.h> +#include <linux/types.h> + +#include "xe_hw_engine_types.h" +#include "xe_pci_types.h" +#include "xe_sriov_types.h" + +struct config_item; + +/* Similar to struct xe_bb, but not tied to HW (yet) */ +struct wa_bb { + u32 *cs; + u32 len; /* in dwords */ +}; + +struct xe_config_group_device { + struct config_group group; + struct config_group sriov; + + struct xe_config_device { + struct wa_bbctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];
+ structwa_bb
ctx_restore_post_bb[XE_ENGINE_CLASS_MAX];
+ boolenable_psmi;
+ boolenable_survivability_mode;
+ u64engines_allowed;
+ u64gt_types_allowed;
+ struct {
+ booladmin_only_pf;
+ unsigned intmax_vfs;
+ } sriov;
+ } config;drivers/gpu/drm/xe/xe_configfs_types.h new file +10
-0+ + /* protects attributes */ + struct mutex lock; + /* matching descriptor */ + const struct xe_device_desc *desc; + /* tentative SR-IOV mode */ + enum xe_sriov_mode mode; +}; + +static inline struct xe_config_group_device*xe_configfs_to_group_device(struct
config_item*item) +{
+ returncontainer_of(to_config_group(item), struct xe_config_group_device, group);drivers/gpu/drm/xe/xe_configfs_types.h new file +3
-0+} + +static inline struct xe_config_device*xe_configfs_to_device(struct
config_item*item)drivers/gpu/drm/xe/xe_configfs_types.h new file +5
-0+{ + return &xe_configfs_to_group_device(item)->config; +} + +#endif /* _XE_CONFIGFS_TYPES_H_ */
PATCH 4/9 drm/xe: Add a new debug focused configfs group
In reply to Stuart SummersAdd the skeleton code for a new debug specific configfs group.
--- drivers/gpu/drm/xe/Makefile | 2 ++ drivers/gpu/drm/xe/xe_configfs.c | 4 ++++ drivers/gpu/drm/xe/xe_configfs_debug.c | 14 ++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++ drivers/gpu/drm/xe/xe_configfs_types.h | 7 +++++++ 5 files changed, 35 insertions(+) create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug.c create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug.h
drivers/gpu/drm/xe/Makefile +2 -0
@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.o xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o +xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y) # graphics virtualization (SR-IOV) support xe-y += \
drivers/gpu/drm/xe/xe_configfs.c +4 -0
@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *groconfig_group_init_type_name (&dev->sriov,"sriov" , &xe_config_sriov_type);configfs_add_default_group (&dev->sriov, &dev->group); } +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +config_group_init_type_name (&dev->debug,"debug" , &xe_configfs_debug_type); +configfs_add_default_group (&dev->debug, &dev->group); +#endif mutex_init (&dev->lock);
drivers/gpu/drm/xe/xe_configfs_debug.c new file +14 -0
@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner =THIS_MODULE , +};
drivers/gpu/drm/xe/xe_configfs_debug.h new file +8 -0
@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */
drivers/gpu/drm/xe/xe_configfs_types.h +7 -0
@@ -24,6 +24,9 @@ struct wa_bb {struct xe_config_group_device {struct config_group group;struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +struct config_group debug; +#endif struct xe_config_device {struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ];
@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_item return &xe_configfs_to_group_device(item)->config; } +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type xe_configfs_debug_type; +#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
Quoted 19 lines · Stuart Summers
Add the skeleton code for a new debug specific configfs group.
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7--- drivers/gpu/drm/xe/Makefile | 2 ++ drivers/gpu/drm/xe/xe_configfs.c | 4 ++++ drivers/gpu/drm/xe/xe_configfs_debug.c | 14 ++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++ drivers/gpu/drm/xe/xe_configfs_types.h | 7 +++++++ 5 files changed, 35 insertions(+) create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug.c create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0
drivers/gpu/drm/xe/Makefile xe-$(CONFIG_HWMON) += xe_hwmon.o · line 165
Quoted 5 lines
drivers/gpu/drm/xe/Makefile +2
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.o xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o +xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y)
What are the downsides of keeping this available for non-debug builds?
I imagine that the debug configfs could be useful for someone not involved in kernel development working together with the developers to debug issues without needing to rebuild their kernel.
Quoted 8 lines
drivers/gpu/drm/xe/Makefile +0
-0# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0
drivers/gpu/drm/xe/xe_configfs.c static struct config_group *xe_config_make_device_group(struct config_group *gro · line 1012
Quoted 7 lines
drivers/gpu/drm/xe/xe_configfs.c +4
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *groconfig_group_init_type_name (&dev->sriov,"sriov" , &xe_config_sriov_type);configfs_add_default_group (&dev->sriov, &dev->group); } +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +config_group_init_type_name (&dev->debug,"debug" , &xe_configfs_debug_type); +configfs_add_default_group (&dev->debug, &dev->group); +#endif
I guess there isn't much we can do about it today, but, based on what I read in configfs documentation a while ago, I suspect it wasn't really designed with arbitrary hierarchies of config items in mind. It would be nice if there was an API for that though.
Basically, a config group is expected to be a directory to contain a set of config items of the same type, which is not the case on our current usage (i.e. the "sriov" and now the "debug" group).
I guess we are able to get our desired behavior (basically just a subdirectory behaving as a config item) by not implementing the "make_item" hook. Not sure if that's an expected use of the API or if that's more of a hack on our side and could come back to bite us in the future.
--
Gustavo Sousa
Quoted 62 lines
drivers/gpu/drm/xe/xe_configfs.c +0
-0mutex_init (&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +14
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner =THIS_MODULE , +};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +7
-0@@ -24,6 +24,9 @@ struct wa_bb {struct xe_config_group_device {struct config_group group;struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +struct config_group debug; +#endif struct xe_config_device {struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ];@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_item return &xe_configfs_to_group_device(item)->config; } +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type xe_configfs_debug_type; +#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */ Signature · 1 line
2.43.0
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
Quoted 32 lines · Gustavo Sousa
Quoted 20 lines · Stuart Summers
Add the skeleton code for a new debug specific configfs group.
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/Makefile | 2 ++
drivers/gpu/drm/xe/xe_configfs.c | 4 ++++drivers/gpu/drm/xe/xe_configfs_debug.c | 14 ++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++ drivers/gpu/drm/xe/xe_configfs_types.h | 7 +++++++ 5 files changed, 35 insertions(+)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.ccreate mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.oQuoted 6 lines
xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o
+xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y)What are the downsides of keeping this available for non-debug builds?
Quoted 4 lines
I imagine that the debug configfs could be useful for someone not involved in kernel development working together with the developers to
debug issues without needing to rebuild their kernel.
So the focal parameter we had in i915 was enable_rc6. Something like this is what we explicitly don't want to expose to a general user as it can have unforeseen consequences (like having very high power usage). But, having the ability to disable rc6 for debug purposes as a developer is extremely useful at narrowing down certain types of bugs at the hardware/software interface.
That's just one example, but there are a bunch of things that could fit this category and I expect we'll have even more use cases over time.
Right now we have to hold these types of changes in a side branch not exposed externally and I really want to converge our processes so we can more easily make these kinds of changes and let us debug directly out of drm-tip.
Quoted 36 lines
Quoted 14 lines
# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *gro config_group_init_type_name(&dev->sriov, "sriov", &xe_config_sriov_type);configfs_add_default_group(&dev->sriov, &dev-
Quoted 7 lines
Quoted 1 line
group);
}
+#ifIS_ENABLED(CONFIG_DRM_XE_DEBUG)
+ config_group_init_type_name(&dev->debug, "debug",&xe_configfs_debug_type);
+ configfs_add_default_group(&dev->debug,&dev->group); +#endifI guess there isn't much we can do about it today, but, based on what I
read in configfs documentation a while ago, I suspect it wasn't really
designed with arbitrary hierarchies of config items in mind. It would
be nice if there was an API for that though.Basically, a config group is expected to be a directory to contain a set
of config items of the same type, which is not the case on our current
usage (i.e. the "sriov" and now the "debug" group).
Quoted 7 lines
I guess we are able to get our desired behavior (basically just a subdirectory behaving as a config item) by not implementing the "
make_item" hook. Not sure if that's an expected use of the API or if
that's more of a hack on our side and could come back to bite us in the
future.
Hm.. ok that's good feedback. I do like bucketing these in groups and subdirectories. It also makes it easy to show or not show based on the Kconfig. But let me go through and make sure there isn't something more suited here. Happy to hear any other feedback as well here.
Thanks,
Stuart
Quoted 71 lines
--
Gustavo SousaQuoted 67 lines
mutex_init(&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +14
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner = THIS_MODULE, +};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +3
-0@@ -24,6 +24,9 @@ struct wa_bb { struct xe_config_group_device { struct config_group group; struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) + struct config_group debug; +#endif
structxe_config_device{
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];drivers/gpu/drm/xe/xe_configfs_types.h +0
-0@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_item return &xe_configfs_to_group_device(item)->config; }drivers/gpu/drm/xe/xe_configfs_types.h +4
-0+#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type xe_configfs_debug_type; +#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
Quoted 38 lines · Stuart
Quoted 32 lines · Gustavo Sousa
Quoted 20 lines · Stuart Summers
Add the skeleton code for a new debug specific configfs group.
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/Makefile | 2 ++
drivers/gpu/drm/xe/xe_configfs.c | 4 ++++drivers/gpu/drm/xe/xe_configfs_debug.c | 14 ++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++ drivers/gpu/drm/xe/xe_configfs_types.h | 7 +++++++ 5 files changed, 35 insertions(+)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.ccreate mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.oQuoted 6 lines
xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o
+xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y)What are the downsides of keeping this available for non-debug builds?
Quoted 4 lines
I imagine that the debug configfs could be useful for someone not involved in kernel development working together with the developers to
debug issues without needing to rebuild their kernel.
Quoted 6 lines
So the focal parameter we had in i915 was
enable_rc6. Something like this is what we explicitly don't want to expose to a general user as it can have unforeseen consequences (like having very high power usage). But, having the ability to disable rc6 for debug purposes as a developer is extremely useful at narrowing down certain types of bugs at the hardware/software interface.
We could probably keep the debug stuff available but cause a taint if they ever get used?
The scenario I had in mind is someone who is not familiar with building their own kernel working together with the development community on a hard-to-reproduce issue and some of the debug configfs attributes being useful to gather more relevant data.
--
Gustavo Sousa
Quoted 132 lines
That's just one example, but there are a bunch of things that could fit this category and I expect we'll have even more use cases over time.
Right now we have to hold these types of changes in a side branch not exposed externally and I really want to converge our processes so we can more easily make these kinds of changes and let us debug directly out of drm-tip.
Quoted 36 lines
Quoted 14 lines
# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *gro config_group_init_type_name(&dev->sriov, "sriov", &xe_config_sriov_type);configfs_add_default_group(&dev->sriov, &dev-
Quoted 7 lines
Quoted 1 line
group);
}
+#ifIS_ENABLED(CONFIG_DRM_XE_DEBUG)
+ config_group_init_type_name(&dev->debug, "debug",&xe_configfs_debug_type);
+ configfs_add_default_group(&dev->debug,&dev->group); +#endifI guess there isn't much we can do about it today, but, based on what I
read in configfs documentation a while ago, I suspect it wasn't really
designed with arbitrary hierarchies of config items in mind. It would
be nice if there was an API for that though.Basically, a config group is expected to be a directory to contain a set
of config items of the same type, which is not the case on our current
usage (i.e. the "sriov" and now the "debug" group).Quoted 7 lines
I guess we are able to get our desired behavior (basically just a subdirectory behaving as a config item) by not implementing the "
make_item" hook. Not sure if that's an expected use of the API or if
that's more of a hack on our side and could come back to bite us in the
future.Hm.. ok that's good feedback. I do like bucketing these in groups and subdirectories. It also makes it easy to show or not show based on the
Kconfig. But let me go through and make sure there isn't something more suited here. Happy to hear any other feedback as well here.Thanks,
StuartQuoted 71 lines
--
Gustavo SousaQuoted 67 lines
mutex_init(&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +14
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner = THIS_MODULE, +};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +3
-0@@ -24,6 +24,9 @@ struct wa_bb { struct xe_config_group_device { struct config_group group; struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) + struct config_group debug; +#endif
structxe_config_device{
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];drivers/gpu/drm/xe/xe_configfs_types.h +0
-0@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_item return &xe_configfs_to_group_device(item)->config; }drivers/gpu/drm/xe/xe_configfs_types.h +4
-0+#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type xe_configfs_debug_type; +#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
Quoted 52 lines · Gustavo Sousa
Quoted 39 lines · Stuart
Quoted 32 lines · Gustavo Sousa
Quoted 20 lines · Stuart Summers
Add the skeleton code for a new debug specific configfs group.
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/Makefile | 2 ++
drivers/gpu/drm/xe/xe_configfs.c | 4 ++++drivers/gpu/drm/xe/xe_configfs_debug.c | 14 ++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++ drivers/gpu/drm/xe/xe_configfs_types.h | 7 +++++++ 5 files changed, 35 insertions(+)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.ccreate mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.oQuoted 6 lines
xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o
+xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y)What are the downsides of keeping this available for non-debug builds?
Quoted 5 lines
I imagine that the debug configfs could be useful for someone not involved in kernel development working together with the developers
to
debug issues without needing to rebuild their kernel.Quoted 10 lines
So the focal parameter we had in i915 was
enable_rc6. Something like
this is what we explicitly don't want to expose to a general user as it
can have unforeseen consequences (like having very high power usage).
But, having the ability to disable rc6 for debug purposes as a developer is extremely useful at narrowing down certain types of bugs
at the hardware/software interface.
Quoted 2 lines
We could probably keep the debug stuff available but cause a taint if they ever get used?
So the tainting is really just for our CI I think. This won't necessarily impact end users - at least on the client side - who might just ignore it (or not even notice it).
Quoted 7 lines
The scenario I had in mind is someone who is not familiar with building
their own kernel working together with the development community on a hard-to-reproduce issue and some of the debug configfs attributes being
useful to gather more relevant data.
Yeah but IMO we really should be able to instruct such a user how to do that - the steps there aren't too difficult if they're already installing kernels etc.
A lot of our debug hooks are already held under various Kconfig options anyway like DRM_XE_DEBUG, DRM_XE_DEBUG_GUC, DRM_XE_DEBUG_VM, etc. And a lot of the information under those Kconfigs are also critical for different debugs.
I'd really prefer to keep these under DRM_XE_DEBUG (or some other Kconfig) if possible so we can maximize the ease of adding new parameters as specific debugs come up.
Thanks,
Stuart
Quoted 149 lines
--
Gustavo SousaQuoted 145 lines
That's just one example, but there are a bunch of things that could fit
this category and I expect we'll have even more use cases over time.Right now we have to hold these types of changes in a side branch not
exposed externally and I really want to converge our processes so we
can more easily make these kinds of changes and let us debug directly
out of drm-tip.Quoted 39 lines
Quoted 15 lines
# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *gro config_group_init_type_name(&dev->sriov,"sriov" , &xe_config_sriov_type);configfs_add_default_group(&dev->sriov, &dev-
Quoted 7 lines
Quoted 1 line
group);
}
+#ifIS_ENABLED(CONFIG_DRM_XE_DEBUG)
+ config_group_init_type_name(&dev->debug, "debug",&xe_configfs_debug_type);
+ configfs_add_default_group(&dev->debug,&dev->group); +#endifI guess there isn't much we can do about it today, but, based on what
I
read in configfs documentation a while ago, I suspect it wasn't really
designed with arbitrary hierarchies of config items in mind. It would
be nice if there was an API for that though.Basically, a config group is expected to be a directory to contain a
set
of config items of the same type, which is not the case on our current
usage (i.e. the "sriov" and now the "debug" group).Quoted 9 lines
I guess we are able to get our desired behavior (basically just a subdirectory behaving as a config item) by not implementing the "
make_item" hook. Not sure if that's an expected use of the API or
if
that's more of a hack on our side and could come back to bite us in
the
future.Hm.. ok that's good feedback. I do like bucketing these in groups and
subdirectories. It also makes it easy to show or not show based on theKconfig. But let me go through and make sure there isn't something more
suited here. Happy to hear any other feedback as well here.Thanks,
StuartQuoted 71 lines
--
Gustavo SousaQuoted 67 lines
mutex_init(&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +14
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner = THIS_MODULE, +};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +3
-0@@ -24,6 +24,9 @@ struct wa_bb { struct xe_config_group_device { struct config_group group; struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) + struct config_group debug; +#endif
structxe_config_device{
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];drivers/gpu/drm/xe/xe_configfs_types.h +0
-0@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_item return &xe_configfs_to_group_device(item)->config; }drivers/gpu/drm/xe/xe_configfs_types.h +4
-0+#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type xe_configfs_debug_type; +#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
Quoted 53 lines · Stuart
Quoted 52 lines · Gustavo Sousa
Quoted 49 lines · Stuart
Quoted 32 lines · Gustavo Sousa
Quoted 20 lines · Stuart Summers
Add the skeleton code for a new debug specific configfs group.
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/Makefile | 2 ++
drivers/gpu/drm/xe/xe_configfs.c | 4 ++++drivers/gpu/drm/xe/xe_configfs_debug.c | 14 ++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++ drivers/gpu/drm/xe/xe_configfs_types.h | 7 +++++++ 5 files changed, 35 insertions(+)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.ccreate mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.oQuoted 6 lines
xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o
+xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y)What are the downsides of keeping this available for non-debug builds?
Quoted 5 lines
I imagine that the debug configfs could be useful for someone not involved in kernel development working together with the developers
to
debug issues without needing to rebuild their kernel.So the focal parameter we had in i915 was
enable_rc6. Something like
this is what we explicitly don't want to expose to a general user as it
can have unforeseen consequences (like having very high power usage).
But, having the ability to disable rc6 for debug purposes as a developer is extremely useful at narrowing down certain types of bugs
at the hardware/software interface.
Quoted 6 lines
Quoted 2 lines
We could probably keep the debug stuff available but cause a taint if they ever get used?
So the tainting is really just for our CI I think. This won't necessarily impact end users - at least on the client side - who might just ignore it (or not even notice it).
I *think* tainting is not only for CI (someone could correct me here). I see that our driver adds a taint when the user is applying some parameters (module parameters or configfs) that alters the "officially supported" behavior of the driver.
Some examples I could find are usages of module_param_named_unsafe() in the module param front, and setup_configfs_post_ctx_restore_bb(), setup_configfs_mid_ctx_restore_bb() in configfs.
Quoted 9 lines
Quoted 8 lines
The scenario I had in mind is someone who is not familiar with building
their own kernel working together with the development community on a hard-to-reproduce issue and some of the debug configfs attributes being
useful to gather more relevant data.
Quoted 8 lines
Yeah but IMO we really should be able to instruct such a user how to do that - the steps there aren't too difficult if they're already installing kernels etc.
A lot of our debug hooks are already held under various
Kconfigoptions anyway likeDRM_XE_DEBUG,DRM_XE_DEBUG_GUC,DRM_XE_DEBUG_VM, etc. And a lot of the information under those Kconfigs are also critical for different debugs.
Are those due to performance or resource usage implications if they are kept enabled? If so, is this the same case here?
Quoted 4 lines
I'd really prefer to keep these under
DRM_XE_DEBUG(or some otherKconfig) if possible so we can maximize the ease of adding new parameters as specific debugs come up.
But why would keeping them under a build flag make it easier to add new parameters as opposed to always keeping the knobs available for use?
--
Gustavo Sousa
Quoted 153 lines
Thanks,
StuartQuoted 149 lines
--
Gustavo SousaQuoted 145 lines
That's just one example, but there are a bunch of things that could fit
this category and I expect we'll have even more use cases over time.Right now we have to hold these types of changes in a side branch not
exposed externally and I really want to converge our processes so we
can more easily make these kinds of changes and let us debug directly
out of drm-tip.Quoted 39 lines
Quoted 15 lines
# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *gro config_group_init_type_name(&dev->sriov,"sriov" , &xe_config_sriov_type);configfs_add_default_group(&dev->sriov, &dev-
Quoted 7 lines
Quoted 1 line
group);
}
+#ifIS_ENABLED(CONFIG_DRM_XE_DEBUG)
+ config_group_init_type_name(&dev->debug, "debug",&xe_configfs_debug_type);
+ configfs_add_default_group(&dev->debug,&dev->group); +#endifI guess there isn't much we can do about it today, but, based on what
I
read in configfs documentation a while ago, I suspect it wasn't really
designed with arbitrary hierarchies of config items in mind. It would
be nice if there was an API for that though.Basically, a config group is expected to be a directory to contain a
set
of config items of the same type, which is not the case on our current
usage (i.e. the "sriov" and now the "debug" group).Quoted 9 lines
I guess we are able to get our desired behavior (basically just a subdirectory behaving as a config item) by not implementing the "
make_item" hook. Not sure if that's an expected use of the API or
if
that's more of a hack on our side and could come back to bite us in
the
future.Hm.. ok that's good feedback. I do like bucketing these in groups and
subdirectories. It also makes it easy to show or not show based on theKconfig. But let me go through and make sure there isn't something more
suited here. Happy to hear any other feedback as well here.Thanks,
StuartQuoted 71 lines
--
Gustavo SousaQuoted 67 lines
mutex_init(&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +14
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner = THIS_MODULE, +};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +3
-0@@ -24,6 +24,9 @@ struct wa_bb { struct xe_config_group_device { struct config_group group; struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) + struct config_group debug; +#endif
structxe_config_device{
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];drivers/gpu/drm/xe/xe_configfs_types.h +0
-0@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_item return &xe_configfs_to_group_device(item)->config; }drivers/gpu/drm/xe/xe_configfs_types.h +4
-0+#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type xe_configfs_debug_type; +#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
Quoted 62 lines · Gustavo Sousa
Quoted 53 lines · Stuart
Quoted 52 lines · Gustavo Sousa
Quoted 49 lines · Stuart
Quoted 32 lines · Gustavo Sousa
Quoted 20 lines · Stuart Summers
Add the skeleton code for a new debug specific configfs group.
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/Makefile | 2 ++
drivers/gpu/drm/xe/xe_configfs.c | 4 ++++drivers/gpu/drm/xe/xe_configfs_debug.c | 14 ++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++ drivers/gpu/drm/xe/xe_configfs_types.h | 7 +++++++ 5 files changed, 35 insertions(+)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.ccreate mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.oQuoted 6 lines
xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o
+xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y)What are the downsides of keeping this available for non-debug builds?
Quoted 5 lines
I imagine that the debug configfs could be useful for someone not involved in kernel development working together with the developers
to
debug issues without needing to rebuild their kernel.So the focal parameter we had in i915 was
enable_rc6. Something like
this is what we explicitly don't want to expose to a general user as it
can have unforeseen consequences (like having very high power usage).
But, having the ability to disable rc6 for debug purposes as a developer is extremely useful at narrowing down certain types of bugs
at the hardware/software interface.Quoted 6 lines
Quoted 2 lines
We could probably keep the debug stuff available but cause a taint if they ever get used?
So the tainting is really just for our CI I think. This won't necessarily impact end users - at least on the client side - who might just ignore it (or not even notice it).
Quoted 8 lines
I *think* tainting is not only for CI (someone could correct me here). I see that our driver adds a taint when the user is applying some parameters (module parameters or configfs) that alters the "officially supported" behavior of the driver.
Some examples I could find are usages of
module_param_named_unsafe()in the module param front, andsetup_configfs_post_ctx_restore_bb(),setup_configfs_mid_ctx_restore_bb()in configfs.
Yeah, there are various reasons to taint the kernel (hardware, software, or behavioral/usage) and they indicate different things; some are more serious than others. See Documentation/admin-guide/tainted-kernels.rst for details. The general idea though is that you've entered a dangerous/risky situation and we can no longer trust the system to behave properly going forward. I believe the kernel community will often ignore or de-prioritize bug reports that arise from tainted kernels because there isn't really any guarantee of proper behavior or support from the community at that point.
Some of our configfs entries in Xe are "official" and end-user focused (e.g., SR-IOV configuration settings and such) so those shouldn't cause any taint. But others are low-level debugging knobs that are effectively handing the user a loaded gun to shoot themselves in the foot with. E.g., the "workaround batchbuffer" configfs stuff is very useful for hardware debugging sessions by expert driver developers, but if used improperly makes it it very easy to mess up both hardware and software internal state in ways that make the system unstable. Tainting on these developer-focused debug facilities is important because we can't trust any bug reports that show up after someone has used those settings. With i915 module parameters we got into situations where a bunch of websites and wikis posted horribly inaccurate suggestions and users would blindly copy those settings without realizing that they were deliberately breaking their systems, and some of the debug knobs we're making available in configfs are even more risky. Honestly, it might be best to even stick them behind CONFIG_EXPERT and/or other build settings.
Matt
Quoted 186 lines
Quoted 10 lines
Quoted 9 lines
The scenario I had in mind is someone who is not familiar with building
their own kernel working together with the development community on a hard-to-reproduce issue and some of the debug configfs attributes being
useful to gather more relevant data.Quoted 8 lines
Yeah but IMO we really should be able to instruct such a user how to do that - the steps there aren't too difficult if they're already installing kernels etc.
A lot of our debug hooks are already held under various
Kconfigoptions anyway likeDRM_XE_DEBUG,DRM_XE_DEBUG_GUC,DRM_XE_DEBUG_VM, etc. And a lot of the information under those Kconfigs are also critical for different debugs.Are those due to performance or resource usage implications if they are kept enabled? If so, is this the same case here?
Quoted 4 lines
I'd really prefer to keep these under
DRM_XE_DEBUG(or some otherKconfig) if possible so we can maximize the ease of adding new parameters as specific debugs come up.But why would keeping them under a build flag make it easier to add new parameters as opposed to always keeping the knobs available for use?
--
Gustavo SousaQuoted 153 lines
Thanks,
StuartQuoted 149 lines
--
Gustavo SousaQuoted 145 lines
That's just one example, but there are a bunch of things that could fit
this category and I expect we'll have even more use cases over time.Right now we have to hold these types of changes in a side branch not
exposed externally and I really want to converge our processes so we
can more easily make these kinds of changes and let us debug directly
out of drm-tip.Quoted 39 lines
Quoted 15 lines
# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *gro config_group_init_type_name(&dev->sriov,"sriov" , &xe_config_sriov_type);configfs_add_default_group(&dev->sriov, &dev-
Quoted 7 lines
Quoted 1 line
group);
}
+#ifIS_ENABLED(CONFIG_DRM_XE_DEBUG)
+ config_group_init_type_name(&dev->debug, "debug",&xe_configfs_debug_type);
+ configfs_add_default_group(&dev->debug,&dev->group); +#endifI guess there isn't much we can do about it today, but, based on what
I
read in configfs documentation a while ago, I suspect it wasn't really
designed with arbitrary hierarchies of config items in mind. It would
be nice if there was an API for that though.Basically, a config group is expected to be a directory to contain a
set
of config items of the same type, which is not the case on our current
usage (i.e. the "sriov" and now the "debug" group).Quoted 9 lines
I guess we are able to get our desired behavior (basically just a subdirectory behaving as a config item) by not implementing the "
make_item" hook. Not sure if that's an expected use of the API or
if
that's more of a hack on our side and could come back to bite us in
the
future.Hm.. ok that's good feedback. I do like bucketing these in groups and
subdirectories. It also makes it easy to show or not show based on theKconfig. But let me go through and make sure there isn't something more
suited here. Happy to hear any other feedback as well here.Thanks,
StuartQuoted 71 lines
--
Gustavo SousaQuoted 67 lines
mutex_init(&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +14
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner = THIS_MODULE, +};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +3
-0@@ -24,6 +24,9 @@ struct wa_bb { struct xe_config_group_device { struct config_group group; struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) + struct config_group debug; +#endif
structxe_config_device{
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];drivers/gpu/drm/xe/xe_configfs_types.h +0
-0@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_item return &xe_configfs_to_group_device(item)->config; }drivers/gpu/drm/xe/xe_configfs_types.h +4
-0+#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type xe_configfs_debug_type; +#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
Signature · 4 lines
Matt Roper Graphics Software Engineer Linux GPU Platform Enablement Intel Corporation
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
Quoted 118 lines · Matt Roper
Quoted 81 lines · Gustavo Sousa
Quoted 63 lines · Stuart
Quoted 62 lines · Gustavo Sousa
Quoted 59 lines · Stuart
Quoted 37 lines · Gustavo Sousa
Quoted 24 lines · Stuart Summers
Add the skeleton code for a new debug specific configfs group.
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/Makefile | 2 ++ drivers/gpu/drm/xe/xe_configfs.c | 4 ++++drivers/gpu/drm/xe/xe_configfs_debug.c| 14
++++++++++++++
drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++drivers/gpu/drm/xe/xe_configfs_types.h| 7 +++++++5 files changed, 35 insertions(+)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.c
create mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.oQuoted 6 lines
xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o
+xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y)What are the downsides of keeping this available for non- debug
builds?Quoted 6 lines
I imagine that the debug configfs could be useful for someone not
involved in kernel development working together with the developers
to
debug issues without needing to rebuild their kernel.So the focal parameter we had in i915 was
enable_rc6. Something
like
this is what we explicitly don't want to expose to a general user
as it
can have unforeseen consequences (like having very high power usage).
But, having the ability to disable rc6 for debug purposes as a
developer is extremely useful at narrowing down certain types of
bugs
at the hardware/software interface.Quoted 8 lines
Quoted 3 lines
We could probably keep the debug stuff available but cause a taint if
they ever get used?So the tainting is really just for our CI I think. This won't necessarily impact end users - at least on the client side - who might
just ignore it (or not even notice it).I *think* tainting is not only for CI (someone could correct me here).
I see that our driver adds a taint when the user is applying some parameters (module parameters or configfs) that alters the "officially
supported" behavior of the driver.Quoted 4 lines
Some examples I could find are usages of
module_param_named_unsafe()in
the module param front, andsetup_configfs_post_ctx_restore_bb(),setup_configfs_mid_ctx_restore_bb()in configfs.Yeah, there are various reasons to taint the kernel (hardware, software,
or behavioral/usage) and they indicate different things; some are more
serious than others. SeeDocumentation/admin-guide/tainted-kernels.rst
for details. The general idea though is that you've entered a dangerous/risky situation and we can no longer trust the system to behave properly going forward. I believe the kernel community will often ignore or de-prioritize bug reports that arise from tainted kernels because there isn't really any guarantee of proper behavior or
support from the community at that point.Some of our configfs entries in Xe are "official" and end-user focused
(e.g., SR-IOV configuration settings and such) so those shouldn't cause
any taint. But others are low-level debugging knobs that are effectively handing the user a loaded gun to shoot themselves in the foot with. E.g., the "workaround batchbuffer" configfs stuff is very useful for hardware debugging sessions by expert driver developers, but
if used improperly makes it it very easy to mess up both hardware and software internal state in ways that make the system unstable. Tainting
on these developer-focused debug facilities is important because we can't trust any bug reports that show up after someone has used those settings. With i915 module parameters we got into situations where a bunch of websites and wikis posted horribly inaccurate suggestions and
Quoted 8 lines
users would blindly copy those settings without realizing that they were
deliberately breaking their systems, and some of the debug knobs we're
making available in configfs are even more risky. Honestly, it might be
best to even stick them behindCONFIG_EXPERTand/or other build settings.
So this was the reason I used CONFIG_DRM_XE_DEBUG... Should we have these behind multiple? Or maybe we should force build CONFIG_EXPORT if CONFIG_DRM_XE_DEBUG is set?
I don't really think tainting the kernel adds much here IMO. Yeah it does tell the user they're doing something that could cause issues, but they're already running in a debug mode if they got to this point, so IMO they already know what they're doing (or should).
Thanks,
Stuart
Quoted 218 lines
Matt
Quoted 213 lines
Quoted 17 lines
Quoted 11 lines
The scenario I had in mind is someone who is not familiar with building
their own kernel working together with the development community on a
hard-to-reproduce issue and some of the debug configfs attributes
being
useful to gather more relevant data.Yeah but IMO we really should be able to instruct such a user how to do
that - the steps there aren't too difficult if they're already installing kernels etc.Quoted 6 lines
A lot of our debug hooks are already held under various
Kconfigoptions
anyway likeDRM_XE_DEBUG,DRM_XE_DEBUG_GUC,DRM_XE_DEBUG_VM, etc. And a
lot of the information under those Kconfigs are also critical for different debugs.Are those due to performance or resource usage implications if they are
kept enabled? If so, is this the same case here?Quoted 4 lines
I'd really prefer to keep these under
DRM_XE_DEBUG(or some otherKconfig) if possible so we can maximize the ease of adding new parameters as specific debugs come up.But why would keeping them under a build flag make it easier to add new
parameters as opposed to always keeping the knobs available for use?--
Gustavo SousaQuoted 172 lines
Thanks,
StuartQuoted 168 lines
--
Gustavo SousaQuoted 164 lines
That's just one example, but there are a bunch of things that could
fit
this category and I expect we'll have even more use cases over
time.Right now we have to hold these types of changes in a side branch
not
exposed externally and I really want to converge our processes so
we
can more easily make these kinds of changes and let us debug directly
out of drm-tip.Quoted 45 lines
Quoted 16 lines
# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *gro config_group_init_type_name(&dev->sriov,"sriov" , &xe_config_sriov_type); configfs_add_default_group(&dev->sriov,&dev-
Quoted 8 lines
Quoted 1 line
group);
}
+#ifIS_ENABLED(CONFIG_DRM_XE_DEBUG)
+ config_group_init_type_name(&dev->debug, "debug",&xe_configfs_debug_type);
+ configfs_add_default_group(&dev->debug, &dev-Quoted 1 line
group);
+#endif
I guess there isn't much we can do about it today, but, based on
what
I
read in configfs documentation a while ago, I suspect it wasn't
really
designed with arbitrary hierarchies of config items in mind. It
would
be nice if there was an API for that though.Basically, a config group is expected to be a directory to contain a
set
of config items of the same type, which is not the case on our
current
usage (i.e. the "sriov" and now the "debug" group).Quoted 13 lines
I guess we are able to get our desired behavior (basically just a
subdirectory behaving as a config item) by not implementing the
"make_item" hook. Not sure if that's an expected use of the API
or
if
that's more of a hack on our side and could come back to bite us
in
the
future.Hm.. ok that's good feedback. I do like bucketing these in groups
and
subdirectories. It also makes it easy to show or not show based on
theKconfig. But let me go through and make sure there isn't something
more
suited here. Happy to hear any other feedback as well here.Thanks,
StuartQuoted 73 lines
--
Gustavo SousaQuoted 69 lines
mutex_init(&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +14
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner = THIS_MODULE, +};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +3
-0@@ -24,6 +24,9 @@ struct wa_bb { struct xe_config_group_device { struct config_group group; struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) + struct config_group debug; +#endif
structxe_config_device{
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];drivers/gpu/drm/xe/xe_configfs_types.h +0
-0@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_itemreturn &xe_configfs_to_group_device(item)-
Quoted 1 line
config;
}
drivers/gpu/drm/xe/xe_configfs_types.h +2
-0+#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type
xe_configfs_debug_type;drivers/gpu/drm/xe/xe_configfs_types.h +2
-0+#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
Quoted 128 lines · Stuart
Quoted 118 lines · Matt Roper
Quoted 81 lines · Gustavo Sousa
Quoted 63 lines · Stuart
Quoted 62 lines · Gustavo Sousa
Quoted 59 lines · Stuart
Quoted 37 lines · Gustavo Sousa
Quoted 24 lines · Stuart Summers
Add the skeleton code for a new debug specific configfs group.
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/Makefile | 2 ++ drivers/gpu/drm/xe/xe_configfs.c | 4 ++++drivers/gpu/drm/xe/xe_configfs_debug.c| 14
++++++++++++++
drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++drivers/gpu/drm/xe/xe_configfs_types.h| 7 +++++++5 files changed, 35 insertions(+)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.c
create mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.oQuoted 6 lines
xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o
+xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y)What are the downsides of keeping this available for non- debug
builds?Quoted 6 lines
I imagine that the debug configfs could be useful for someone not
involved in kernel development working together with the developers
to
debug issues without needing to rebuild their kernel.So the focal parameter we had in i915 was
enable_rc6. Something
like
this is what we explicitly don't want to expose to a general user
as it
can have unforeseen consequences (like having very high power usage).
But, having the ability to disable rc6 for debug purposes as a
developer is extremely useful at narrowing down certain types of
bugs
at the hardware/software interface.Quoted 8 lines
Quoted 3 lines
We could probably keep the debug stuff available but cause a taint if
they ever get used?So the tainting is really just for our CI I think. This won't necessarily impact end users - at least on the client side - who might
just ignore it (or not even notice it).I *think* tainting is not only for CI (someone could correct me here).
I see that our driver adds a taint when the user is applying some parameters (module parameters or configfs) that alters the "officially
supported" behavior of the driver.Quoted 4 lines
Some examples I could find are usages of
module_param_named_unsafe()in
the module param front, andsetup_configfs_post_ctx_restore_bb(),setup_configfs_mid_ctx_restore_bb()in configfs.Yeah, there are various reasons to taint the kernel (hardware, software,
or behavioral/usage) and they indicate different things; some are more
serious than others. SeeDocumentation/admin-guide/tainted-kernels.rst
for details. The general idea though is that you've entered a dangerous/risky situation and we can no longer trust the system to behave properly going forward. I believe the kernel community will often ignore or de-prioritize bug reports that arise from tainted kernels because there isn't really any guarantee of proper behavior or
support from the community at that point.Some of our configfs entries in Xe are "official" and end-user focused
(e.g., SR-IOV configuration settings and such) so those shouldn't cause
any taint. But others are low-level debugging knobs that are effectively handing the user a loaded gun to shoot themselves in the foot with. E.g., the "workaround batchbuffer" configfs stuff is very useful for hardware debugging sessions by expert driver developers, but
if used improperly makes it it very easy to mess up both hardware and software internal state in ways that make the system unstable. Tainting
on these developer-focused debug facilities is important because we can't trust any bug reports that show up after someone has used those settings. With i915 module parameters we got into situations where a bunch of websites and wikis posted horribly inaccurate suggestions andQuoted 8 lines
users would blindly copy those settings without realizing that they were
deliberately breaking their systems, and some of the debug knobs we're
making available in configfs are even more risky. Honestly, it might be
best to even stick them behindCONFIG_EXPERTand/or other build settings.
Quoted 3 lines
So this was the reason I used
CONFIG_DRM_XE_DEBUG... Should we have these behind multiple? Or maybe we should force buildCONFIG_EXPORTifCONFIG_DRM_XE_DEBUGis set?
Probably best to check with the maintainers on exactly what config options we should hide some of these behind.
Quoted 5 lines
I don't really think tainting the kernel adds much here IMO. Yeah it does tell the user they're doing something that could cause issues, but they're already running in a debug mode if they got to this point, so IMO they already know what they're doing (or should).
The taint isn't for the user running the driver in debug mode. It's for the people receiving bug reports and managing the bug tracking databases so that they can adjust their expectations, possibly disregarding or deprioritizing bug reports from users mucking around with hardware-internal or driver-internal state in a way that isn't expected to be safe. And it's not just us that needs this notification (e.g., for gitlab issues), but also distribution maintainers and such who don't really know much about our driver and may not otherwise realize that that the user broke their system with something only intended for low-level debugging.
For example, if someone uses 'ctx_restore_post_bb' to have the driver poke arbitrary chicken bits in the hardware on every context switch that really aren't supposed to be touched, that might be harmless for a while and then suddenly break when something else in the driver changes. We might get bug reports incorrectly indicating that a perfectly good patch is causing a driver regression, and then spend tons of time trying to reproduce and debug if we don't have the taint notification to let us know that the user was doing something unsupported and unsafe.
Matt
Quoted 223 lines
Thanks,
StuartQuoted 218 lines
Matt
Quoted 213 lines
Quoted 17 lines
Quoted 11 lines
The scenario I had in mind is someone who is not familiar with building
their own kernel working together with the development community on a
hard-to-reproduce issue and some of the debug configfs attributes
being
useful to gather more relevant data.Yeah but IMO we really should be able to instruct such a user how to do
that - the steps there aren't too difficult if they're already installing kernels etc.Quoted 6 lines
A lot of our debug hooks are already held under various
Kconfigoptions
anyway likeDRM_XE_DEBUG,DRM_XE_DEBUG_GUC,DRM_XE_DEBUG_VM, etc. And a
lot of the information under those Kconfigs are also critical for different debugs.Are those due to performance or resource usage implications if they are
kept enabled? If so, is this the same case here?Quoted 4 lines
I'd really prefer to keep these under
DRM_XE_DEBUG(or some otherKconfig) if possible so we can maximize the ease of adding new parameters as specific debugs come up.But why would keeping them under a build flag make it easier to add new
parameters as opposed to always keeping the knobs available for use?--
Gustavo SousaQuoted 172 lines
Thanks,
StuartQuoted 168 lines
--
Gustavo SousaQuoted 164 lines
That's just one example, but there are a bunch of things that could
fit
this category and I expect we'll have even more use cases over
time.Right now we have to hold these types of changes in a side branch
not
exposed externally and I really want to converge our processes so
we
can more easily make these kinds of changes and let us debug directly
out of drm-tip.Quoted 45 lines
Quoted 16 lines
# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *gro config_group_init_type_name(&dev->sriov,"sriov" , &xe_config_sriov_type); configfs_add_default_group(&dev->sriov,&dev-
Quoted 8 lines
Quoted 1 line
group);
}
+#ifIS_ENABLED(CONFIG_DRM_XE_DEBUG)
+ config_group_init_type_name(&dev->debug, "debug",&xe_configfs_debug_type);
+ configfs_add_default_group(&dev->debug, &dev-Quoted 1 line
group);
+#endif
I guess there isn't much we can do about it today, but, based on
what
I
read in configfs documentation a while ago, I suspect it wasn't
really
designed with arbitrary hierarchies of config items in mind. It
would
be nice if there was an API for that though.Basically, a config group is expected to be a directory to contain a
set
of config items of the same type, which is not the case on our
current
usage (i.e. the "sriov" and now the "debug" group).Quoted 13 lines
I guess we are able to get our desired behavior (basically just a
subdirectory behaving as a config item) by not implementing the
"make_item" hook. Not sure if that's an expected use of the API
or
if
that's more of a hack on our side and could come back to bite us
in
the
future.Hm.. ok that's good feedback. I do like bucketing these in groups
and
subdirectories. It also makes it easy to show or not show based on
theKconfig. But let me go through and make sure there isn't something
more
suited here. Happy to hear any other feedback as well here.Thanks,
StuartQuoted 73 lines
--
Gustavo SousaQuoted 69 lines
mutex_init(&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +14
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner = THIS_MODULE, +};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +3
-0@@ -24,6 +24,9 @@ struct wa_bb { struct xe_config_group_device { struct config_group group; struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) + struct config_group debug; +#endif
structxe_config_device{
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];drivers/gpu/drm/xe/xe_configfs_types.h +0
-0@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_itemreturn &xe_configfs_to_group_device(item)-
Quoted 1 line
config;
}
drivers/gpu/drm/xe/xe_configfs_types.h +2
-0+#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type
xe_configfs_debug_type;drivers/gpu/drm/xe/xe_configfs_types.h +2
-0+#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
Signature · 4 lines
Matt Roper Graphics Software Engineer Linux GPU Platform Enablement Intel Corporation
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
Quoted 187 lines · Matt Roper
Quoted 154 lines · Stuart
Quoted 144 lines · Matt Roper
Quoted 93 lines · Gustavo Sousa
Quoted 77 lines · Stuart
Quoted 71 lines · Gustavo Sousa
Quoted 60 lines · Stuart
Quoted 41 lines · Gustavo Sousa
Quoted 26 lines · Stuart Summers
Add the skeleton code for a new debug specific
configfs
group.Signed-off-by: Stuart Summers
<[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/Makefile | 2 ++ drivers/gpu/drm/xe/xe_configfs.c | 4 ++++drivers/gpu/drm/xe/xe_configfs_debug.c| 14
++++++++++++++
drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++drivers/gpu/drm/xe/xe_configfs_types.h| 7 +++++++5 files changed, 35 insertions(+)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.c
create mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.oQuoted 7 lines
xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o
+xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-
y)What are the downsides of keeping this available for non-
debug
builds?Quoted 7 lines
I imagine that the debug configfs could be useful for someone not
involved in kernel development working together with the
developers
to
debug issues without needing to rebuild their kernel.So the focal parameter we had in i915 was
enable_rc6. Something
like
this is what we explicitly don't want to expose to a general
user
as it
can have unforeseen consequences (like having very high power
usage).Quoted 8 lines
But, having the ability to disable rc6 for debug purposes as
a
developer is extremely useful at narrowing down certain types
of
bugs
at the hardware/software interface.Quoted 4 lines
We could probably keep the debug stuff available but cause a
taint if
they ever get used?Quoted 5 lines
So the tainting is really just for our CI I think. This won't necessarily impact end users - at least on the client side - who
might
just ignore it (or not even notice it).I *think* tainting is not only for CI (someone could correct me here).
I see that our driver adds a taint when the user is applying some
parameters (module parameters or configfs) that alters the "officially
supported" behavior of the driver.Quoted 5 lines
Some examples I could find are usages of
module_param_named_unsafe()in
the module param front, andsetup_configfs_post_ctx_restore_bb(),setup_configfs_mid_ctx_restore_bb()in configfs.Yeah, there are various reasons to taint the kernel (hardware, software,
or behavioral/usage) and they indicate different things; some are more
serious than others. SeeDocumentation/admin-guide/tainted-kernels.rst
for details. The general idea though is that you've entered a dangerous/risky situation and we can no longer trust the system to
behave properly going forward. I believe the kernel community will
often ignore or de-prioritize bug reports that arise from tainted kernels because there isn't really any guarantee of proper behavior
or
support from the community at that point.Some of our configfs entries in Xe are "official" and end-user focused
(e.g., SR-IOV configuration settings and such) so those shouldn't cause
any taint. But others are low-level debugging knobs that are effectively handing the user a loaded gun to shoot themselves in the
foot with. E.g., the "workaround batchbuffer" configfs stuff is very
useful for hardware debugging sessions by expert driver developers,
but
if used improperly makes it it very easy to mess up both hardware and
software internal state in ways that make the system unstable. Tainting
on these developer-focused debug facilities is important because we
can't trust any bug reports that show up after someone has used those
settings. With i915 module parameters we got into situations where a
bunch of websites and wikis posted horribly inaccurate suggestions
and
users would blindly copy those settings without realizing that theyQuoted 8 lines
were
deliberately breaking their systems, and some of the debug knobs we're
making available in configfs are even more risky. Honestly, it might
be
best to even stick them behindCONFIG_EXPERTand/or other build settings.Quoted 4 lines
So this was the reason I used
CONFIG_DRM_XE_DEBUG... Should we have these behind multiple? Or maybe we should force buildCONFIG_EXPORTifCONFIG_DRM_XE_DEBUGis set?Probably best to check with the maintainers on exactly what config options we should hide some of these behind.
Quoted 8 lines
I don't really think tainting the kernel adds much here IMO. Yeah it
does tell the user they're doing something that could cause issues, but
they're already running in a debug mode if they got to this point, so
IMO they already know what they're doing (or should).The taint isn't for the user running the driver in debug mode. It's for
the people receiving bug reports and managing the bug tracking databases
so that they can adjust their expectations, possibly disregarding or deprioritizing bug reports from users mucking around with hardware-internal or driver-internal state in a way that isn't expected
to be safe. And it's not just us that needs this notification (e.g., for gitlab issues), but also distribution maintainers and such who don't
really know much about our driver and may not otherwise realize that that the user broke their system with something only intended for low-level debugging.
Quoted 13 lines
For example, if someone uses '
ctx_restore_post_bb' to have the driver poke arbitrary chicken bits in the hardware on every context switch that
really aren't supposed to be touched, that might be harmless for a while
and then suddenly break when something else in the driver changes. We
might get bug reports incorrectly indicating that a perfectly good patch
is causing a driver regression, and then spend tons of time trying to reproduce and debug if we don't have the taint notification to
let us know that the user was doing something unsupported and unsafe.
Yeah it's all a good point. And maybe a good reason to add some kind of auto taint like we do with the unsafe modparams... ok let me take a stab here on the next rev.
Thanks,
Stuart
Quoted 255 lines
Matt
Quoted 250 lines
Thanks,
StuartQuoted 245 lines
Matt
Quoted 240 lines
Quoted 20 lines
Quoted 12 lines
The scenario I had in mind is someone who is not familiar with
building
their own kernel working together with the development community on a
hard-to-reproduce issue and some of the debug configfs attributes
being
useful to gather more relevant data.Yeah but IMO we really should be able to instruct such a user how
to do
that - the steps there aren't too difficult if they're already
installing kernels etc.Quoted 9 lines
A lot of our debug hooks are already held under various
Kconfig
options
anyway likeDRM_XE_DEBUG,DRM_XE_DEBUG_GUC,DRM_XE_DEBUG_VM, etc.
And a
lot of the information under those Kconfigs are also critical for
different debugs.Are those due to performance or resource usage implications if they
are
kept enabled? If so, is this the same case here?Quoted 6 lines
I'd really prefer to keep these under
DRM_XE_DEBUG(or some otherKconfig) if possible so we can maximize the ease of adding new
parameters as specific debugs come up.But why would keeping them under a build flag make it easier to add
new
parameters as opposed to always keeping the knobs available for use?--
Gustavo SousaQuoted 189 lines
Thanks,
StuartQuoted 185 lines
--
Gustavo SousaQuoted 181 lines
That's just one example, but there are a bunch of things that
could
fit
this category and I expect we'll have even more use cases over
time.Right now we have to hold these types of changes in a side
branch
not
exposed externally and I really want to converge our processes so
we
can more easily make these kinds of changes and let us debug
directly
out of drm-tip.Quoted 60 lines
Quoted 19 lines
# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *groconfig_group_init_type_name(&dev-
Quoted 1 line
sriov,
"sriov",
&xe_config_sriov_type);
configfs_add_default_group(&dev-Quoted 1 line
sriov,
&dev-
Quoted 1 line
group);
Quoted 8 lines
}
+#ifIS_ENABLED(CONFIG_DRM_XE_DEBUG)
+ config_group_init_type_name(&dev->debug, "debug",&xe_configfs_debug_type);
+ configfs_add_default_group(&dev->debug, &dev-Quoted 1 line
group);
+#endif
I guess there isn't much we can do about it today, but, based on
what
I
read in configfs documentation a while ago, I suspect it
wasn't
really
designed with arbitrary hierarchies of config items in mind. It
would
be nice if there was an API for that though.Basically, a config group is expected to be a directory to
contain a
set
of config items of the same type, which is not the case on
our
current
usage (i.e. the "sriov" and now the "debug" group).I guess we are able to get our desired behavior (basically
just a
subdirectory behaving as a config item) by not implementing
the
"make_item" hook. Not sure if that's an expected use of
the APIQuoted 8 lines
or
if
that's more of a hack on our side and could come back to
bite us
in
the
future.Hm.. ok that's good feedback. I do like bucketing these in
groups
and
subdirectories. It also makes it easy to show or not show based on
theKconfig. But let me go through and make sure there isn't something
more
suited here. Happy to hear any other feedback as well here.Thanks,
StuartQuoted 75 lines
--
Gustavo SousaQuoted 71 lines
mutex_init(&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +12
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type= {
+ .ct_owner =THIS_MODULE,drivers/gpu/drm/xe/xe_configfs_debug.c new file +1
-0+};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +3
-0@@ -24,6 +24,9 @@ struct wa_bb { struct xe_config_group_device { struct config_group group; struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) + struct config_group debug; +#endif
structxe_config_device{
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];drivers/gpu/drm/xe/xe_configfs_types.h +0
-0@@ -56,4 +59,8 @@ static inline struct
xe_config_device*xe_configfs_to_device(structconfig_itemreturn &xe_configfs_to_group_device(item)-Quoted 1 line
config;
}
drivers/gpu/drm/xe/xe_configfs_types.h +2
-0+#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type
xe_configfs_debug_type;drivers/gpu/drm/xe/xe_configfs_types.h +2
-0+#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
Re: [PATCH 4/9] drm/xe: Add a new debug focused configfs group
In reply to Summers, StuartQuoted 101 lines · Stuart
Quoted 32 lines · Gustavo Sousa
Quoted 20 lines · Stuart Summers
Add the skeleton code for a new debug specific configfs group.
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7---
drivers/gpu/drm/xe/Makefile | 2 ++
drivers/gpu/drm/xe/xe_configfs.c | 4 ++++drivers/gpu/drm/xe/xe_configfs_debug.c | 14 ++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug.h | 8 ++++++++ drivers/gpu/drm/xe/xe_configfs_types.h | 7 +++++++ 5 files changed, 35 insertions(+)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.ccreate mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hdrivers/gpu/drm/xe/Makefile +0
-0@@ -161,6 +161,8 @@ xe-$(CONFIG_HWMON) += xe_hwmon.oQuoted 6 lines
xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o
+xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y)What are the downsides of keeping this available for non-debug builds?
Quoted 4 lines
I imagine that the debug configfs could be useful for someone not involved in kernel development working together with the developers to
debug issues without needing to rebuild their kernel.So the focal parameter we had in i915 was
enable_rc6. Something like this is what we explicitly don't want to expose to a general user as it
can have unforeseen consequences (like having very high power usage). But, having the ability to disable rc6 for debug purposes as a developer is extremely useful at narrowing down certain types of bugs at the hardware/software interface.That's just one example, but there are a bunch of things that could fit
this category and I expect we'll have even more use cases over time.Right now we have to hold these types of changes in a side branch not exposed externally and I really want to converge our processes so we can more easily make these kinds of changes and let us debug directly out of drm-tip.
Quoted 38 lines
Quoted 14 lines
# graphics virtualization (SR-IOV) support xe-y += \drivers/gpu/drm/xe/xe_configfs.c +0
-0@@ -1006,6 +1006,10 @@ static struct config_group *xe_config_make_device_group(struct config_group *gro config_group_init_type_name(&dev->sriov, "sriov", &xe_config_sriov_type);configfs_add_default_group(&dev->sriov, &dev-
Quoted 7 lines
Quoted 1 line
group);
}
+#ifIS_ENABLED(CONFIG_DRM_XE_DEBUG)
+ config_group_init_type_name(&dev->debug, "debug",&xe_configfs_debug_type);
+ configfs_add_default_group(&dev->debug,&dev->group); +#endifI guess there isn't much we can do about it today, but, based on what
I
read in configfs documentation a while ago, I suspect it wasn't really
designed with arbitrary hierarchies of config items in mind. It would
be nice if there was an API for that though.Basically, a config group is expected to be a directory to contain a
set
of config items of the same type, which is not the case on our current
usage (i.e. the "sriov" and now the "debug" group).Quoted 7 lines
I guess we are able to get our desired behavior (basically just a subdirectory behaving as a config item) by not implementing the "
make_item" hook. Not sure if that's an expected use of the API or if
that's more of a hack on our side and could come back to bite us in the
future.
Quoted 6 lines
Hm.. ok that's good feedback. I do like bucketing these in groups and subdirectories. It also makes it easy to show or not show based on the
Kconfig. But let me go through and make sure there isn't something more
suited here. Happy to hear any other feedback as well here.
Hey so I don't think it's necessarily a bad idea to use make_item here for the leaf nodes. But make_group does cover. I see other drivers going both directions here. That said, we aren't doing anything new in this patch series with respect to configfs item creation. I'd prefer to keep things consistent in this series if possible. We can look at dividing out into config_item and config_group entries separately.
Thanks,
Stuart
Quoted 76 lines
Thanks,
StuartQuoted 71 lines
--
Gustavo SousaQuoted 67 lines
mutex_init(&dev->lock);drivers/gpu/drm/xe/xe_configfs_debug.c new file +14
-0@@ -0,0 +1,14 @@ +// SPDX-License-Identifier: MIT +/* +* Copyright © 2026 Intel Corporation +*/ + +#include <linux/configfs.h> +#include <linux/module.h> + +#include "xe_configfs_debug.h" +#include "xe_configfs_types.h" + +const struct config_item_type xe_configfs_debug_type = { + .ct_owner = THIS_MODULE, +};drivers/gpu/drm/xe/xe_configfs_debug.h new file +8
-0@@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: MIT */ +/* +* Copyright © 2026 Intel Corporation +*/ +#ifndef _XE_CONFIGFS_DEBUG_H_ +#define _XE_CONFIGFS_DEBUG_H_ + +#endif /* _XE_CONFIGFS_DEBUG_H_ */ drivers/gpu/drm/xe/xe_configfs_types.h +3
-0@@ -24,6 +24,9 @@ struct wa_bb { struct xe_config_group_device { struct config_group group; struct config_group sriov; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) + struct config_group debug; +#endif
structxe_config_device{
structwa_bb
ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX];drivers/gpu/drm/xe/xe_configfs_types.h +0
-0@@ -56,4 +59,8 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_item return &xe_configfs_to_group_device(item)->config; }drivers/gpu/drm/xe/xe_configfs_types.h +4
-0+#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +extern const struct config_item_type xe_configfs_debug_type; +#endif +#endif /* _XE_CONFIGFS_TYPES_H_ */
Signature · 1 line
2.43.0
PATCH 5/9 drm/xe: Move debug configfs entries to xe_configfs_debug.c
In reply to Stuart SummersMove the debug specific configfs attributes into the new xe_configfs_debug.c file under a new debug configfs subdirectory. Ensure these are wrapped in CONFIG_DRM_XE_DEBUG to allow finer grained debug changes outside of more ABI specific configfs entries.
--- drivers/gpu/drm/xe/xe_configfs.c | 889 ++------------------- drivers/gpu/drm/xe/xe_configfs.h | 23 - drivers/gpu/drm/xe/xe_configfs_debug.c | 861 +++++++++++++++++++- drivers/gpu/drm/xe/xe_configfs_debug.h | 36 + drivers/gpu/drm/xe/xe_configfs_types.h | 19 +- drivers/gpu/drm/xe/xe_guc.c | 1 + drivers/gpu/drm/xe/xe_hw_engine.c | 1 + drivers/gpu/drm/xe/xe_lrc.c | 1 + drivers/gpu/drm/xe/xe_pci.c | 1 + drivers/gpu/drm/xe/xe_psmi.c | 1 + drivers/gpu/drm/xe/xe_rtp.c | 1 + drivers/gpu/drm/xe/xe_survivability_mode.c | 1 + 12 files changed, 970 insertions(+), 865 deletions(-)
drivers/gpu/drm/xe/xe_configfs.c +54 -835
@@ -6,13 +6,11 @@#include <linux/bitops.h> #include <linux/configfs.h> #include <linux/cleanup.h> -#include <linux/find.h> #include <linux/init.h> #include <linux/module.h> #include <linux/pci.h> #include <linux/string.h> -#include "instructions/xe_mi_commands.h" #include "xe_configfs.h" #include "xe_configfs_types.h" #include "xe_defaults.h"
@@ -50,15 +48,11 @@* * /sys/kernel/config/xe/ * ├── 0000:00:02.0 -* │ └── ... -* ├── 0000:00:02.1 -* │ └── ... -* : -* └── 0000:03:00.0 -* ├── enable_survivability_mode -* ├── gt_types_allowed -* ├── engines_allowed -* └── enable_psmi +* │ └── sriov/ +* │ ├── admin_only_pf +* │ └── max_vfs +* └── 0000:00:02.1 +* └── ... * * After configuring the attributes as per next section, the device can be * probed with::
@@ -70,144 +64,6 @@* Configure Attributes * ==================== * -* Context restore BB -* ------------------ -* -* Allow to execute a batch buffer during any context switches. When the -* GPU is restoring the context, it executes additional commands. It's useful -* for testing additional workarounds and validating certain HW behaviors: it's -* not intended for normal execution and will taint the kernel with TAINT_TEST -* when used. -* -* The syntax allows to pass straight instructions to be executed by the engine -* in a batch buffer or set specific registers. -* -* #. Generic instruction:: -* -* <engine-class> cmd <instr> [[dword0] [dword1] [...]] -* -* #. Simple register setting:: -* -* <engine-class> reg <address> <value> -* -* Commands are saved per engine class: all instances of that class will execute -* those commands during context switch. The instruction, dword arguments, -* addresses and values are in hex format like in the examples below. -* -* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 after the -* normal context restore:: -* -* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ -* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb -* -* #. Execute a LRI command to write 0xDEADBEEF to register 0x4f10 at the -* beginning of the context restore:: -* -* # echo 'rcs cmd 11000001 4F100 DEADBEEF' \ -* > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_mid_bb - -* #. Load certain values in a couple of registers (it can be used as a simpler -* alternative to the `cmd`) action:: -* -* # cat > /sys/kernel/config/xe/0000:03:00.0/ctx_restore_post_bb <<EOF -* rcs reg 4F100 DEADBEEF -* rcs reg 4F104 FFFFFFFF -* EOF -* -* .. note:: -* -* When using multiple lines, make sure to use a command that is -* implemented with a single write syscall, like HEREDOC. -* -* Currently this is implemented only for post and mid context restore and -* these attributes can only be set before binding to the device. -* -* PSMI -* ---- -* -* Enable extra debugging capabilities to trace engine execution. Only useful -* during early platform enabling and requires additional hardware connected. -* Once it's enabled, additionals WAs are added and runtime configuration is -* done via debugfs. Example to enable it:: -* -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_psmi -* -* This attribute can only be set before binding to the device. -* -* Survivability mode: -* ------------------- -* -* Enable survivability mode on supported cards. This setting only takes -* effect when probing the device. Example to enable it:: -* -* # echo 1 > /sys/kernel/config/xe/0000:03:00.0/enable_survivability_mode -* -* This attribute can only be set before binding to the device. -* -* Allowed engines: -* ---------------- -* -* Allow only a set of engine(s) to be available, disabling the other engines -* even if they are available in hardware. This is applied after HW fuses are -* considered on each tile. Examples: -* -* Allow only one render and one copy engines, nothing else:: -* -* # echo 'rcs0,bcs0' > /sys/kernel/config/xe/0000:03:00.0/engines_allowed -* -* Allow only compute engines and first copy engine:: -* -* # echo 'ccs*,bcs0' > /sys/kernel/config/xe/0000:03:00.0/engines_allowed -* -* Note that the engine names are the per-GT hardware names. On multi-tile -* platforms, writing ``rcs0,bcs0`` to this file would allow the first render -* and copy engines on each tile. -* -* The requested configuration may not be supported by the platform and driver -* may fail to probe. For example: if at least one copy engine is expected to be -* available for migrations, but it's disabled. This is intended for debugging -* purposes only. -* -* This attribute can only be set before binding to the device. -* -* Allowed GT types: -* ----------------- -* -* Allow only specific types of GTs to be detected and initialized by the -* driver. Any combination of GT types can be enabled/disabled, although -* some settings will cause the device to fail to probe. -* -* Writes support both comma- and newline-separated input format. Reads -* will always return one GT type per line. "primary" and "media" are the -* GT type names supported by this interface. -* -* This attribute can only be set before binding to the device. -* -* Examples: -* -* Allow both primary and media GTs to be initialized and used. This matches -* the driver's default behavior:: -* -* # echo 'primary,media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed -* -* Allow only the primary GT of each tile to be initialized and used, -* effectively disabling the media GT if it exists on the platform:: -* -* # echo 'primary' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed -* -* Allow only the media GT of each tile to be initialized and used, -* effectively disabling the primary GT. **This configuration will cause -* device probe failure on all current platforms, but may be allowed on -* igpu platforms in the future**:: -* -* # echo 'media' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed -* -* Disable all GTs. Only other GPU IP (such as display) is potentially usable. -* **This configuration will cause device probe failure on all current -* platforms, but may be allowed on igpu platforms in the future**:: -* -* # echo '' > /sys/kernel/config/xe/0000:03:00.0/gt_types_allowed -* * Max SR-IOV Virtual Functions * ---------------------------- *
@@ -243,11 +99,15 @@*/ -static const struct xe_config_device device_defaults = { - .enable_psmi = false, - .enable_survivability_mode = false, - .engines_allowed =U64_MAX , - .gt_types_allowed =U64_MAX , +const struct xe_config_device xe_configfs_device_defaults = { +#if IS_ENABLED(CONFIG_DRM_XE_DEBUG) + .debug = { + .enable_psmi = false, + .enable_survivability_mode = false, + .engines_allowed =U64_MAX , + .gt_types_allowed =U64_MAX , + }, +#endif .sriov = { .admin_only_pf =XE_DEFAULT_ADMIN_ONLY_PF , .max_vfs =XE_DEFAULT_MAX_VFS ,
@@ -256,40 +116,22 @@ static const struct xe_config_device device_defaults = {static void set_device_defaults (struct xe_config_device *config) { - *config = device_defaults; + *config = xe_configfs_device_defaults;#ifdef CONFIG_PCI_IOV config->sriov.max_vfs = xe_modparam.max_vfs;#endif } -struct engine_info { -const char *cls; -u64 mask; -enum xe_engine_class engine_class; -}; - -/* Some helpful macros to aid on the sizing of buffer allocation when parsing */ -#define MAX_ENGINE_CLASS_CHARS 5 -#define MAX_ENGINE_INSTANCE_CHARS 2 - -static const struct engine_info engine_info[] = { - { .cls ="rcs" , .mask =XE_HW_ENGINE_RCS_MASK , .engine_class =XE_ENGINE_CLASS_RENDER }, - { .cls ="bcs" , .mask =XE_HW_ENGINE_BCS_MASK , .engine_class =XE_ENGINE_CLASS_COPY }, - { .cls ="vcs" , .mask =XE_HW_ENGINE_VCS_MASK , .engine_class =XE_ENGINE_CLASS_VIDEO_DECODE }, - { .cls ="vecs" , .mask =XE_HW_ENGINE_VECS_MASK , .engine_class =XE_ENGINE_CLASS_VIDEO_ENHANCE }, - { .cls ="ccs" , .mask =XE_HW_ENGINE_CCS_MASK , .engine_class =XE_ENGINE_CLASS_COMPUTE }, - { .cls ="gsccs" , .mask =XE_HW_ENGINE_GSCCS_MASK , .engine_class =XE_ENGINE_CLASS_OTHER }, -}; - -static const struct { -const char *name; -enum xe_gt_type type; -} gt_types[] = { - { .name ="primary" , .type =XE_GT_TYPE_MAIN }, - { .name ="media" , .type =XE_GT_TYPE_MEDIA }, -}; - -static bool is_bound (struct xe_config_group_device *dev) +/** +* xe_configfs_is_bound - check whether the matching pci device is bound +* @dev: configfs group device +* +* Caller must hold @dev->lock. +* +* Return: true if the matching pci_dev is already bound to a driver, +* false otherwise. +*/ +bool xe_configfs_is_bound (struct xe_config_group_device *dev) {unsigned int domain, bus, slot, function;struct pci_dev *pdev;
@@ -314,485 +156,14 @@ static bool is_bound(struct xe_config_group_device *dev)return ret; } -static ssize_t enable_survivability_mode_show (struct config_item *item,char *page) -{ -struct xe_config_device *dev =xe_configfs_to_device (item); - -return sprintf (page,"%d\n" , dev->enable_survivability_mode); -} - -static ssize_t enable_survivability_mode_store (struct config_item *item,const char *page, -size_t len) -{ -struct xe_config_group_device *dev =xe_configfs_to_group_device (item); -bool enable_survivability_mode; -int ret; - - ret =kstrtobool (page, &enable_survivability_mode); -if (ret) -return ret; - -guard (mutex)(&dev->lock); -if (is_bound (dev)) -return -EBUSY ; - - dev->config.enable_survivability_mode = enable_survivability_mode; - -return len; -} - -static ssize_t gt_types_allowed_show (struct config_item *item,char *page) -{ -struct xe_config_device *dev =xe_configfs_to_device (item); -char *p = page; - -for (size_t i =0 ; i <ARRAY_SIZE (gt_types); i++) -if (dev->gt_types_allowed &BIT_ULL (gt_types[i].type)) - p +=sprintf (p,"%s\n" , gt_types[i].name); - -return p - page; -} - -static ssize_t gt_types_allowed_store (struct config_item *item,const char *page, -size_t len) -{ -struct xe_config_group_device *dev =xe_configfs_to_group_device (item); - char *buf __free(kfree) = kstrdup(page, GFP_KERNEL); -char *p = buf; -u64 typemask =0 ; - -if (!buf) -return -ENOMEM ; - -while (p) { -char *typename =strsep (&p,",\n" ); -bool matched = false; - -if (typename[0 ] =='\0' ) -continue ; - -for (size_t i =0 ; i <ARRAY_SIZE (gt_types); i++) { -if (strcmp (typename, gt_types[i].name) ==0 ) { - typemask |=BIT (gt_types[i].type); - matched = true; -break ; - } - } - -if (!matched) -return -EINVAL ; - } - -guard (mutex)(&dev->lock); -if (is_bound (dev)) -return -EBUSY ; - - dev->config.gt_types_allowed = typemask; - -return len; -} - -static ssize_t engines_allowed_show (struct config_item *item,char *page) -{ -struct xe_config_device *dev =xe_configfs_to_device (item); -char *p = page; - -for (size_t i =0 ; i <ARRAY_SIZE (engine_info); i++) { -u64 mask = engine_info[i].mask; - -if ((dev->engines_allowed & mask) == mask) { - p +=sprintf (p,"%s*\n" , engine_info[i].cls); - }else if (mask & dev->engines_allowed) { -u16 bit0 =__ffs64 (mask), bit; - - mask &= dev->engines_allowed; - -for_each_set_bit (bit, (const unsigned long *)&mask,64 ) - p +=sprintf (p,"%s%u\n" , engine_info[i].cls, - bit - bit0); - } - } - -return p - page; -} - -/* -* Lookup engine_info. If @mask is not NULL, reduce the mask according to the -* instance in @pattern. -* -* Examples of inputs: -* - lookup_engine_info("rcs0", &mask): return "rcs" entry from @engine_info and -* mask == BIT_ULL(XE_HW_ENGINE_RCS0) -* - lookup_engine_info("rcs*", &mask): return "rcs" entry from @engine_info and -* mask == XE_HW_ENGINE_RCS_MASK -* - lookup_engine_info("rcs", NULL): return "rcs" entry from @engine_info -*/ -static const struct engine_info *lookup_engine_info (const char *pattern,u64 *mask) -{ -for (size_t i =0 ; i <ARRAY_SIZE (engine_info); i++) { -u8 instance; -u16 bit; - -if (!str_has_prefix (pattern, engine_info[i].cls)) -continue ; - - pattern +=strlen (engine_info[i].cls); -if (!mask) -return *pattern ?NULL : &engine_info[i]; - -if (!strcmp (pattern,"*" )) { - *mask = engine_info[i].mask; -return &engine_info[i]; - } - -if (kstrtou8 (pattern,10 , &instance)) -return NULL ; - - bit =__ffs64 (engine_info[i].mask) + instance; -if (bit >=fls64 (engine_info[i].mask)) -return NULL ; - - *mask =BIT_ULL (bit); -return &engine_info[i]; - } - -return NULL ; -} - -static int parse_engine (const char *s,const char *end_chars,u64 *mask, -const struct engine_info **pinfo) -{ -char buf[MAX_ENGINE_CLASS_CHARS +MAX_ENGINE_INSTANCE_CHARS +1 ]; -const struct engine_info *info; -size_t len; - - len =strcspn (s, end_chars); -if (len >=sizeof (buf)) -return -EINVAL ; - -memcpy (buf, s, len); - buf[len] ='\0' ; - - info =lookup_engine_info (buf, mask); -if (!info) -return -ENOENT ; - -if (pinfo) - *pinfo = info; - -return len; -} - -static ssize_t engines_allowed_store (struct config_item *item,const char *page, -size_t len) -{ -struct xe_config_group_device *dev =xe_configfs_to_group_device (item); -ssize_t patternlen, p; -u64 mask, val =0 ; - -for (p =0 ; p < len; p += patternlen +1 ) { - patternlen =parse_engine (page + p,",\n" , &mask,NULL ); -if (patternlen <0 ) -return -EINVAL ; - - val |= mask; - } - -guard (mutex)(&dev->lock); -if (is_bound (dev)) -return -EBUSY ; - - dev->config.engines_allowed = val; - -return len; -} - -static ssize_t enable_psmi_show (struct config_item *item,char *page) -{ -struct xe_config_device *dev =xe_configfs_to_device (item); - -return sprintf (page,"%d\n" , dev->enable_psmi); -} - -static ssize_t enable_psmi_store (struct config_item *item,const char *page,size_t len) -{ -struct xe_config_group_device *dev =xe_configfs_to_group_device (item); -bool val; -int ret; - - ret =kstrtobool (page, &val); -if (ret) -return ret; - -guard (mutex)(&dev->lock); -if (is_bound (dev)) -return -EBUSY ; - - dev->config.enable_psmi = val; - -return len; -} - -static bool wa_bb_read_advance (bool dereference,char **p, -const char *append,size_t len, -size_t *max_size) -{ -if (dereference) { -if (len >= *max_size) -return false; - *max_size -= len; -if (append) -memcpy (*p, append, len); - } - - *p += len; - -return true; -} - -static ssize_t wa_bb_show (struct xe_config_group_device *dev, -struct wa_bb wa_bb[static XE_ENGINE_CLASS_MAX ], -char *data,size_t sz) -{ -char *p = data; - -guard (mutex)(&dev->lock); - -for (size_t i =0 ; i <ARRAY_SIZE (engine_info); i++) { -enum xe_engine_class ec = engine_info[i].engine_class; -size_t len; - -if (!wa_bb[ec].len) -continue ; - - len =snprintf (p, sz,"%s:" , engine_info[i].cls); -if (!wa_bb_read_advance (data, &p,NULL , len, &sz)) -return -ENOBUFS ; - -for (size_t j =0 ; j < wa_bb[ec].len; j++) { - len =snprintf (p, sz," %08x" , wa_bb[ec].cs[j]); -if (!wa_bb_read_advance (data, &p,NULL , len, &sz)) -return -ENOBUFS ; - } - -if (!wa_bb_read_advance (data, &p,"\n" ,1 , &sz)) -return -ENOBUFS ; - } - -if (!wa_bb_read_advance (data, &p,"" ,1 , &sz)) -return -ENOBUFS ; - -/* Reserve one more to match check for '\0' */ -if (!data) - p++; - -return p - data; -} - -static ssize_t ctx_restore_mid_bb_show (struct config_item *item,char *page) -{ -struct xe_config_group_device *dev =xe_configfs_to_group_device (item); - -return wa_bb_show (dev, dev->config.ctx_restore_mid_bb, page,SZ_4K ); -} - -static ssize_t ctx_restore_post_bb_show (struct config_item *item,char *page) -{ -struct xe_config_group_device *dev =xe_configfs_to_group_device (item); - -return wa_bb_show (dev, dev->config.ctx_restore_post_bb, page,SZ_4K ); -} - -static void wa_bb_append (struct wa_bb *wa_bb,u32 val) -{ -if (wa_bb->cs) - wa_bb->cs[wa_bb->len] = val; - - wa_bb->len++; -} - -static ssize_t parse_hex (const char *line,u32 *pval) -{ -char numstr[12 ]; -const char *p; -ssize_t numlen; - - p = line +strspn (line," \t" ); -if (!*p || *p =='\n' ) -return 0 ; - - numlen =strcspn (p," \t\n" ); -if (!numlen || numlen >=sizeof (numstr) -1 ) -return -EINVAL ; - -memcpy (numstr, p, numlen); - numstr[numlen] ='\0' ; - p += numlen; - -if (kstrtou32 (numstr,16 , pval)) -return -EINVAL ; - -return p - line; -} - -/* -* Parse lines with the format -* -* <engine-class> cmd <u32> <u32...> -* <engine-class> reg <u32_addr> <u32_val> -* -* and optionally save them in @wa_bb[i].cs is non-NULL. -* -* Return the number of dwords parsed. -*/ -static ssize_t parse_wa_bb_lines (const char *lines, -struct wa_bb wa_bb[static XE_ENGINE_CLASS_MAX ]) -{ -ssize_t dwords =0 , ret; -const char *p; - -for (p = lines; *p; p++) { -const struct engine_info *info =NULL ; -u32 val, val2; - -/* Also allow empty lines */ - p +=strspn (p," \t\n" ); -if (!*p) -break ; - - ret =parse_engine (p," \t\n" ,NULL , &info); -if (ret <0 ) -return ret; - - p += ret; - p +=strspn (p," \t" ); - -if (str_has_prefix (p,"cmd" )) { -for (p +=strlen ("cmd" ); *p;) { - ret =parse_hex (p, &val); -if (ret <0 ) -return -EINVAL ; -if (!ret) -break ; - - p += ret; - dwords++; -wa_bb_append (&wa_bb[info->engine_class], val); - } - }else if (str_has_prefix (p,"reg" )) { - p +=strlen ("reg" ); - ret =parse_hex (p, &val); -if (ret <=0 ) -return -EINVAL ; - - p += ret; - ret =parse_hex (p, &val2); -if (ret <=0 ) -return -EINVAL ; - - p += ret; - dwords +=3 ; -wa_bb_append (&wa_bb[info->engine_class], -MI_LOAD_REGISTER_IMM |MI_LRI_NUM_REGS (1 )); -wa_bb_append (&wa_bb[info->engine_class], val); -wa_bb_append (&wa_bb[info->engine_class], val2); - }else { -return -EINVAL ; - } - } - -return dwords; -} - -static ssize_t wa_bb_store (struct wa_bb wa_bb[static XE_ENGINE_CLASS_MAX ], -struct xe_config_group_device *dev, -const char *page,size_t len) -{ -/* tmp_wa_bb must match wa_bb's size */ -struct wa_bb tmp_wa_bb[XE_ENGINE_CLASS_MAX ] = { }; -ssize_t count, class; -u32 *tmp; - -/* 1. Count dwords - wa_bb[i].cs is NULL for all classes */ - count =parse_wa_bb_lines (page, tmp_wa_bb); -if (count <0 ) -return count; - -guard (mutex)(&dev->lock); - -if (is_bound (dev)) -return -EBUSY ; - -/* -* 2. Allocate a u32 array and set the pointers to the right positions -* according to the length of each class' wa_bb -*/ - tmp =krealloc (wa_bb[0 ].cs, count *sizeof (u32),GFP_KERNEL ); -if (!tmp) -return -ENOMEM ; - -if (!count) { -memset (wa_bb,0 ,sizeof (tmp_wa_bb)); -return len; - } - -for (class =0 , count =0 ; class <XE_ENGINE_CLASS_MAX ; ++class) { - tmp_wa_bb[class].cs = tmp + count; - count += tmp_wa_bb[class].len; - tmp_wa_bb[class].len =0 ; - } - -/* 3. Parse wa_bb lines again, this time saving the values */ - count =parse_wa_bb_lines (page, tmp_wa_bb); -if (count <0 ) -return count; - -memcpy (wa_bb, tmp_wa_bb,sizeof (tmp_wa_bb)); - -return len; -} - -static ssize_t ctx_restore_mid_bb_store (struct config_item *item, -const char *data,size_t sz) -{ -struct xe_config_group_device *dev =xe_configfs_to_group_device (item); - -return wa_bb_store (dev->config.ctx_restore_mid_bb, dev, data, sz); -} - -static ssize_t ctx_restore_post_bb_store (struct config_item *item, -const char *data,size_t sz) -{ -struct xe_config_group_device *dev =xe_configfs_to_group_device (item); - -return wa_bb_store (dev->config.ctx_restore_post_bb, dev, data, sz); -} - -CONFIGFS_ATTR(, ctx_restore_mid_bb); -CONFIGFS_ATTR(, ctx_restore_post_bb); -CONFIGFS_ATTR(, enable_psmi); -CONFIGFS_ATTR(, enable_survivability_mode); -CONFIGFS_ATTR(, engines_allowed); -CONFIGFS_ATTR(, gt_types_allowed); - -static struct configfs_attribute *xe_config_device_attrs[] = { - &attr_ctx_restore_mid_bb, - &attr_ctx_restore_post_bb, - &attr_enable_psmi, - &attr_enable_survivability_mode, - &attr_engines_allowed, - &attr_gt_types_allowed, -NULL , -}; -static void xe_config_device_release (struct config_item *item) {struct xe_config_group_device *dev =xe_configfs_to_group_device (item);mutex_destroy (&dev->lock); -kfree (dev->config.ctx_restore_mid_bb[0 ].cs); -kfree (dev->config.ctx_restore_post_bb[0 ].cs); +kfree (dev->config.debug.ctx_restore_mid_bb[0 ].cs); +kfree (dev->config.debug.ctx_restore_post_bb[0 ].cs);kfree (dev); }
@@ -800,27 +171,8 @@ static struct configfs_item_operations xe_config_device_ops = { .release = xe_config_device_release, }; -static bool xe_config_device_is_visible (struct config_item *item, -struct configfs_attribute *attr,int n) -{ -struct xe_config_group_device *dev =xe_configfs_to_group_device (item); - -if (attr == &attr_enable_survivability_mode) { -if (!dev->desc->is_dgfx || dev->desc->platform <XE_BATTLEMAGE ) -return false; - } - -return true; -} - -static struct configfs_group_operations xe_config_device_group_ops = { - .is_visible = xe_config_device_is_visible, -}; -static const struct config_item_type xe_config_device_type = { .ct_item_ops = &xe_config_device_ops, - .ct_group_ops = &xe_config_device_group_ops, - .ct_attrs = xe_config_device_attrs, .ct_owner =THIS_MODULE , };
@@ -844,7 +196,7 @@ static ssize_t sriov_max_vfs_store(struct config_item *item, const char *page, sguard (mutex)(&dev->lock); -if (is_bound (dev)) +if (xe_configfs_is_bound (dev))return -EBUSY ; ret =kstrtouint (page,0 , &max_vfs);
@@ -875,7 +227,7 @@ static ssize_t sriov_admin_only_pf_store(struct config_item *item, const char *pguard (mutex)(&dev->lock); -if (is_bound (dev)) +if (xe_configfs_is_bound (dev))return -EBUSY ; ret =kstrtobool (page, &admin_only_pf);
@@ -886,12 +238,12 @@ static ssize_t sriov_admin_only_pf_store(struct config_item *item, const char *preturn len; } -CONFIGFS_ATTR (sriov_, admin_only_pf);CONFIGFS_ATTR (sriov_, max_vfs); +CONFIGFS_ATTR (sriov_, admin_only_pf);static struct configfs_attribute *xe_config_sriov_attrs[] = { - &sriov_attr_admin_only_pf, &sriov_attr_max_vfs, + &sriov_attr_admin_only_pf,NULL , };
@@ -1048,19 +400,36 @@ static struct xe_config_group_device *find_xe_config_group_device(struct pci_devreturn xe_configfs_to_group_device (item); } +/** +* xe_configfs_find_device - find the configfs group device for a pci_dev +* @pdev: pci device +* +* Look up the &xe_config_group_device associated with @pdev. On success, +* an additional reference is held on the returned group; the caller must +* drop it with config_group_put() when done. +* +* Return: pointer to &xe_config_group_device, or %NULL if no group exists. +*/ +struct xe_config_group_device *xe_configfs_find_device (struct pci_dev *pdev) +{ +return find_xe_config_group_device (pdev); +} +static void dump_custom_dev_config (struct pci_dev *pdev,struct xe_config_group_device *dev) {#define PRI_CUSTOM_ATTR (fmt_, attr_) do { \ - if (dev->config.attr_ != device_defaults.attr_) \ + if (dev->config.attr_ != xe_configfs_device_defaults.attr_) \ pci_info(pdev, "configfs: " __stringify(attr_) " = " fmt_ "\n", \ dev->config.attr_); \ } while (0) -PRI_CUSTOM_ATTR ("%d" , enable_psmi); -PRI_CUSTOM_ATTR ("%d" , enable_survivability_mode); -PRI_CUSTOM_ATTR ("%llx" , engines_allowed); -PRI_CUSTOM_ATTR ("%llx" , gt_types_allowed); +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) +PRI_CUSTOM_ATTR ("%d" , debug.enable_psmi); +PRI_CUSTOM_ATTR ("%d" , debug.enable_survivability_mode); +PRI_CUSTOM_ATTR ("%llx" , debug.engines_allowed); +PRI_CUSTOM_ATTR ("%llx" , debug.gt_types_allowed); +#endif PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);#undef PRI_CUSTOM_ATTR
@@ -1081,7 +450,7 @@ void xe_configfs_check_device(struct pci_dev *pdev)return ;/* memcmp here is safe as both are zero-initialized */ -if (memcmp (&dev->config, &device_defaults,sizeof (dev->config))) { +if (memcmp (&dev->config, &xe_configfs_device_defaults,sizeof (dev->config))) {pci_info (pdev,"Found custom settings in configfs\n" );dump_custom_dev_config (pdev, dev); }
@@ -1089,156 +458,6 @@ void xe_configfs_check_device(struct pci_dev *pdev)config_group_put (&dev->group); } -/** -* xe_configfs_get_enable_survivability_mode - get configfs survivability mode attribute -* @pdev: pci device -* -* Return: enable_survivability_mode attribute in configfs -*/ -bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) -{ -struct xe_config_group_device *dev =find_xe_config_group_device (pdev); -bool mode; - -if (!dev) -return device_defaults.enable_survivability_mode; - - mode = dev->config.enable_survivability_mode; -config_group_put (&dev->group); - -return mode; -} - -static u64 get_gt_types_allowed (struct pci_dev *pdev) -{ -struct xe_config_group_device *dev =find_xe_config_group_device (pdev); -u64 mask; - -if (!dev) -return device_defaults.gt_types_allowed; - - mask = dev->config.gt_types_allowed; -config_group_put (&dev->group); - -return mask; -} - -/** -* xe_configfs_primary_gt_allowed - determine whether primary GTs are supported -* @pdev: pci device -* -* Return: True if primary GTs are enabled, false if they have been disabled via -* configfs. -*/ -bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev) -{ -return get_gt_types_allowed (pdev) &BIT_ULL (XE_GT_TYPE_MAIN ); -} - -/** -* xe_configfs_media_gt_allowed - determine whether media GTs are supported -* @pdev: pci device -* -* Return: True if the media GTs are enabled, false if they have been disabled -* via configfs. -*/ -bool xe_configfs_media_gt_allowed (struct pci_dev *pdev) -{ -return get_gt_types_allowed (pdev) &BIT_ULL (XE_GT_TYPE_MEDIA ); -} - -/** -* xe_configfs_get_engines_allowed - get engine allowed mask from configfs -* @pdev: pci device -* -* Return: engine mask with allowed engines set in configfs -*/ -u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev) -{ -struct xe_config_group_device *dev =find_xe_config_group_device (pdev); -u64 engines_allowed; - -if (!dev) -return device_defaults.engines_allowed; - - engines_allowed = dev->config.engines_allowed; -config_group_put (&dev->group); - -return engines_allowed; -} - -/** -* xe_configfs_get_psmi_enabled - get configfs enable_psmi setting -* @pdev: pci device -* -* Return: enable_psmi setting in configfs -*/ -bool xe_configfs_get_psmi_enabled (struct pci_dev *pdev) -{ -struct xe_config_group_device *dev =find_xe_config_group_device (pdev); -bool ret; - -if (!dev) -return false; - - ret = dev->config.enable_psmi; -config_group_put (&dev->group); - -return ret; -} - -/** -* xe_configfs_get_ctx_restore_mid_bb - get configfs ctx_restore_mid_bb setting -* @pdev: pci device -* @class: hw engine class -* @cs: pointer to the bb to use - only valid during probe -* -* Return: Number of dwords used in the mid_ctx_restore setting in configfs -*/ -u32 xe_configfs_get_ctx_restore_mid_bb (struct pci_dev *pdev, -enum xe_engine_class class, -const u32 **cs) -{ -struct xe_config_group_device *dev =find_xe_config_group_device (pdev); -u32 len; - -if (!dev) -return 0 ; - -if (cs) - *cs = dev->config.ctx_restore_mid_bb[class].cs; - - len = dev->config.ctx_restore_mid_bb[class].len; -config_group_put (&dev->group); - -return len; -} - -/** -* xe_configfs_get_ctx_restore_post_bb - get configfs ctx_restore_post_bb setting -* @pdev: pci device -* @class: hw engine class -* @cs: pointer to the bb to use - only valid during probe -* -* Return: Number of dwords used in the post_ctx_restore setting in configfs -*/ -u32 xe_configfs_get_ctx_restore_post_bb (struct pci_dev *pdev, -enum xe_engine_class class, -const u32 **cs) -{ -struct xe_config_group_device *dev =find_xe_config_group_device (pdev); -u32 len; - -if (!dev) -return 0 ; - - *cs = dev->config.ctx_restore_post_bb[class].cs; - len = dev->config.ctx_restore_post_bb[class].len; -config_group_put (&dev->group); - -return len; -} -#ifdef CONFIG_PCI_IOV /** * xe_configfs_admin_only_pf() - Get PF's operational mode.
drivers/gpu/drm/xe/xe_configfs.h +0 -23
@@ -9,7 +9,6 @@#include <linux/types.h> #include "xe_defaults.h" -#include "xe_hw_engine_types.h" #include "xe_module.h" struct pci_dev ;
@@ -18,17 +17,6 @@ struct pci_dev;int xe_configfs_init (void );void xe_configfs_exit (void );void xe_configfs_check_device (struct pci_dev *pdev); -bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev); -bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev); -bool xe_configfs_media_gt_allowed (struct pci_dev *pdev); -u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev); -bool xe_configfs_get_psmi_enabled (struct pci_dev *pdev); -u32 xe_configfs_get_ctx_restore_mid_bb (struct pci_dev *pdev, -enum xe_engine_class class, -const u32 **cs); -u32 xe_configfs_get_ctx_restore_post_bb (struct pci_dev *pdev, -enum xe_engine_class class, -const u32 **cs);#ifdef CONFIG_PCI_IOV unsigned int xe_configfs_get_max_vfs (struct pci_dev *pdev);bool xe_configfs_admin_only_pf (struct pci_dev *pdev);
@@ -37,17 +25,6 @@ bool xe_configfs_admin_only_pf(struct pci_dev *pdev);static inline int xe_configfs_init (void ) {return 0 ; }static inline void xe_configfs_exit (void ) { }static inline void xe_configfs_check_device (struct pci_dev *pdev) { } -static inline bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) {return false; } -static inline bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev) {return true; } -static inline bool xe_configfs_media_gt_allowed (struct pci_dev *pdev) {return true; } -static inline u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev) {return U64_MAX ; } -static inline bool xe_configfs_get_psmi_enabled (struct pci_dev *pdev) {return false; } -static inline u32 xe_configfs_get_ctx_restore_mid_bb (struct pci_dev *pdev, -enum xe_engine_class class, -const u32 **cs) {return 0 ; } -static inline u32 xe_configfs_get_ctx_restore_post_bb (struct pci_dev *pdev, -enum xe_engine_class class, -const u32 **cs) {return 0 ; }#ifdef CONFIG_PCI_IOV static inline unsigned int xe_configfs_get_max_vfs (struct pci_dev *pdev) {
drivers/gpu/drm/xe/xe_configfs_debug.c +860 -1
872 lines not on this page · raw message
drivers/gpu/drm/xe/xe_configfs_debug.h +36 -0
@@ -5,4 +5,40 @@#ifndef _XE_CONFIGFS_DEBUG_H_#define _XE_CONFIGFS_DEBUG_H_ +#include <linux/types.h> + +#include "xe_hw_engine_types.h" + +struct pci_dev ; + +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) &&IS_ENABLED (CONFIG_CONFIGFS_FS ) +bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev); +bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev); +bool xe_configfs_media_gt_allowed (struct pci_dev *pdev); +u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev); +bool xe_configfs_get_psmi_enabled (struct pci_dev *pdev); +u32 xe_configfs_get_ctx_restore_mid_bb (struct pci_dev *pdev, +enum xe_engine_class class, +const u32 **cs); +u32 xe_configfs_get_ctx_restore_post_bb (struct pci_dev *pdev, +enum xe_engine_class class, +const u32 **cs); +#else +/* +* Dummy values here are not used since these function accesses are always +* wrapped in CONFIG_DRM_XE_DEBUG. +*/ +static inline bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) {return false; } +static inline bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev) {return true; } +static inline bool xe_configfs_media_gt_allowed (struct pci_dev *pdev) {return true; } +static inline u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev) {return U64_MAX ; } +static inline bool xe_configfs_get_psmi_enabled (struct pci_dev *pdev) {return false; } +static inline u32 xe_configfs_get_ctx_restore_mid_bb (struct pci_dev *pdev, +enum xe_engine_class class, +const u32 **cs) {return 0 ; } +static inline u32 xe_configfs_get_ctx_restore_post_bb (struct pci_dev *pdev, +enum xe_engine_class class, +const u32 **cs) {return 0 ; } +#endif +#endif /* _XE_CONFIGFS_DEBUG_H_ */
drivers/gpu/drm/xe/xe_configfs_types.h +13 -6
@@ -14,6 +14,7 @@#include "xe_sriov_types.h" struct config_item ; +struct pci_dev ;/* Similar to struct xe_bb, but not tied to HW (yet) */ struct wa_bb {
@@ -29,12 +30,14 @@ struct xe_config_group_device {#endif struct xe_config_device { -struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; -struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ]; -bool enable_psmi; -bool enable_survivability_mode; -u64 engines_allowed; -u64 gt_types_allowed; +struct { +struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ]; +struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ]; +bool enable_psmi; +bool enable_survivability_mode; +u64 engines_allowed; +u64 gt_types_allowed; + } debug;struct {bool admin_only_pf;unsigned int max_vfs;
@@ -59,6 +62,10 @@ static inline struct xe_config_device *xe_configfs_to_device(struct config_itemreturn &xe_configfs_to_group_device (item)->config; } +bool xe_configfs_is_bound (struct xe_config_group_device *dev); +struct xe_config_group_device *xe_configfs_find_device (struct pci_dev *pdev); +extern const struct xe_config_device xe_configfs_device_defaults; +#if IS_ENABLED (CONFIG_DRM_XE_DEBUG )extern const struct config_item_type xe_configfs_debug_type;#endif
drivers/gpu/drm/xe/xe_guc.c +1 -0
@@ -18,6 +18,7 @@#include "regs/xe_irq_regs.h" #include "xe_bo.h" #include "xe_configfs.h" +#include "xe_configfs_debug.h" #include "xe_device.h" #include "xe_force_wake.h" #include "xe_gt.h"
drivers/gpu/drm/xe/xe_hw_engine.c +1 -0
@@ -18,6 +18,7 @@#include "xe_assert.h" #include "xe_bo.h" #include "xe_configfs.h" +#include "xe_configfs_debug.h" #include "xe_device.h" #include "xe_execlist.h" #include "xe_force_wake.h"
drivers/gpu/drm/xe/xe_lrc.c +1 -0
@@ -19,6 +19,7 @@#include "xe_bb.h" #include "xe_bo.h" #include "xe_configfs.h" +#include "xe_configfs_debug.h" #include "xe_device.h" #include "xe_drm_client.h" #include "xe_exec_queue_types.h"
drivers/gpu/drm/xe/xe_pci.c +1 -0
@@ -19,6 +19,7 @@#include "regs/xe_gt_regs.h" #include "regs/xe_regs.h" #include "xe_configfs.h" +#include "xe_configfs_debug.h" #include "xe_device.h" #include "xe_drv.h" #include "xe_gt.h"
drivers/gpu/drm/xe/xe_psmi.c +1 -0
@@ -8,6 +8,7 @@#include "xe_bo.h" #include "xe_device_types.h" #include "xe_configfs.h" +#include "xe_configfs_debug.h" #include "xe_psmi.h" /*
drivers/gpu/drm/xe/xe_rtp.c +1 -0
@@ -10,6 +10,7 @@#include <uapi/drm/xe_drm.h> #include "xe_configfs.h" +#include "xe_configfs_debug.h" #include "xe_device.h" #include "xe_gt.h" #include "xe_gt_topology.h"
drivers/gpu/drm/xe/xe_survivability_mode.c +1 -0
@@ -11,6 +11,7 @@#include <linux/sysfs.h> #include "xe_configfs.h" +#include "xe_configfs_debug.h" #include "xe_device.h" #include "xe_heci_gsc.h" #include "xe_i2c.h"
Signature · 1 line
2.43.0
PATCH 6/9 drm/xe/guc: Add configfs support for guc_log_level
In reply to Stuart SummersAllow the GuC log level to be selected per-device via configfs in addition to the existing 'guc_log_level' module parameter. The configfs attribute lives under the new 'debug' configfs subdirectory:
/sys/kernel/config/xe/<bdf>/debug/guc_log_level
When the configfs attribute is set to a valid level (0-5), it overrides the module parameter for that device. The default value is -1 (XE_GUC_LOG_LEVEL_UNSET), which means 'unset' and falls back to the 'guc_log_level' module parameter, ensuring existing users that rely solely on the module parameter are unaffected by this change.
The module parameter implementation itself is untouched.
Note that the expectation is a user will only set the log level when CONFIG_DRM_XE_DEBUG is set. Otherwise the expectation is a user will always just use the default (1).
--- drivers/gpu/drm/xe/xe_configfs.c | 2 + drivers/gpu/drm/xe/xe_configfs_debug.c | 84 +++++++++++++++++++++++++- drivers/gpu/drm/xe/xe_configfs_debug.h | 7 +++ drivers/gpu/drm/xe/xe_configfs_types.h | 1 + drivers/gpu/drm/xe/xe_defaults.h | 3 + drivers/gpu/drm/xe/xe_guc_log.c | 3 +- 6 files changed, 98 insertions(+), 2 deletions(-)
drivers/gpu/drm/xe/xe_configfs.c +2 -0
@@ -106,6 +106,7 @@ const struct xe_config_device xe_configfs_device_defaults = { .enable_survivability_mode = false, .engines_allowed = U64_MAX, .gt_types_allowed = U64_MAX, + .guc_log_level = XE_GUC_LOG_LEVEL_UNSET, },#endif .sriov = {
@@ -429,6 +430,7 @@ static void dump_custom_dev_config(struct pci_dev *pdev,PRI_CUSTOM_ATTR ("%d" , debug.enable_survivability_mode);PRI_CUSTOM_ATTR ("%llx" , debug.engines_allowed);PRI_CUSTOM_ATTR ("%llx" , debug.gt_types_allowed); +PRI_CUSTOM_ATTR ("%d" , debug.guc_log_level);#endif PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);
drivers/gpu/drm/xe/xe_configfs_debug.c +83 -1
126 lines not on this page · raw message
drivers/gpu/drm/xe/xe_configfs_debug.h +7 -0
@@ -7,7 +7,9 @@#include <linux/types.h> +#include "xe_defaults.h" #include "xe_hw_engine_types.h" +#include "xe_module.h" struct pci_dev ;
@@ -17,6 +19,7 @@ bool xe_configfs_primary_gt_allowed(struct pci_dev *pdev);bool xe_configfs_media_gt_allowed (struct pci_dev *pdev);u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev);bool xe_configfs_get_psmi_enabled (struct pci_dev *pdev); +int xe_configfs_get_guc_log_level (struct pci_dev *pdev);u32 xe_configfs_get_ctx_restore_mid_bb (struct pci_dev *pdev,enum xe_engine_class class,const u32 **cs);
@@ -33,6 +36,10 @@ static inline bool xe_configfs_primary_gt_allowed(struct pci_dev *pdev) { returnstatic inline bool xe_configfs_media_gt_allowed (struct pci_dev *pdev) {return true; }static inline u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev) {return U64_MAX ; }static inline bool xe_configfs_get_psmi_enabled (struct pci_dev *pdev) {return false; } +static inline int xe_configfs_get_guc_log_level (struct pci_dev *pdev) +{ +return xe_modparam.guc_log_level; +}static inline u32 xe_configfs_get_ctx_restore_mid_bb (struct pci_dev *pdev,enum xe_engine_class class,const u32 **cs) {return 0 ; }
drivers/gpu/drm/xe/xe_configfs_types.h +1 -0
@@ -37,6 +37,7 @@ struct xe_config_group_device {bool enable_survivability_mode;u64 engines_allowed;u64 gt_types_allowed; +int guc_log_level; } debug;struct {bool admin_only_pf;
drivers/gpu/drm/xe/xe_defaults.h +3 -0
@@ -13,6 +13,9 @@#define XE_DEFAULT_GUC_LOG_LEVEL 1#endif +/* Sentinel value for guc_log_level configfs: not set, fall back to module param */ +#define XE_GUC_LOG_LEVEL_UNSET -1 +#define XE_DEFAULT_PROBE_DISPLAY IS_ENABLED(CONFIG_DRM_XE_DISPLAY)#define XE_DEFAULT_VRAM_BAR_SIZE 0#define XE_DEFAULT_FORCE_PROBE CONFIG_DRM_XE_FORCE_PROBE
drivers/gpu/drm/xe/xe_guc_log.c +2 -1
@@ -13,6 +13,7 @@#include "abi/guc_lfd_abi.h" #include "regs/xe_guc_regs.h" #include "xe_bo.h" +#include "xe_configfs_debug.h" #include "xe_devcoredump.h" #include "xe_force_wake.h" #include "xe_gt_printk.h"
@@ -637,7 +638,7 @@ int xe_guc_log_init(struct xe_guc_log *log)xe_map_memset (xe, &bo->vmap,0 ,0 ,xe_bo_size (bo)); log->bo = bo; - log->level = xe_modparam.guc_log_level; + log->level =xe_configfs_get_guc_log_level (to_pci_dev (xe->drm.dev));return 0 ; }
Signature · 1 line
2.43.0
Re: [PATCH 6/9] drm/xe/guc: Add configfs support for guc_log_level
Quoted 12 lines · Stuart Summers
Allow the GuC log level to be selected per-device via configfs in addition to the existing '
guc_log_level' module parameter. The configfs attribute lives under the new 'debug' configfs subdirectory:/sys/kernel/config/xe/<bdf>/debug/guc_log_level
When the configfs attribute is set to a valid level (0-5), it overrides the module parameter for that device. The default value is -1 (
XE_GUC_LOG_LEVEL_UNSET), which means 'unset' and falls back to the 'guc_log_level' module parameter, ensuring existing users that rely solely on the module parameter are unaffected by this change.
Quoted 5 lines
The module parameter implementation itself is untouched.
Note that the expectation is a user will only set the log level when
CONFIG_DRM_XE_DEBUGis set. Otherwise the expectation is a user will always just use the default (1).
Why is this the expectation? if a customer is seeing a bug that we're not able to reproduce, it might be useful to ask them to repro with a higher GuC verbosity level without having to rebuild the kernel in debug mode.
Daniele
Quoted 249 lines
Signed-off-by Stuart Summers <[email protected]>Assisted-by Copilot:claude-opus-4.7--- drivers/gpu/drm/xe/xe_configfs.c | 2 + drivers/gpu/drm/xe/xe_configfs_debug.c | 84 +++++++++++++++++++++++++- drivers/gpu/drm/xe/xe_configfs_debug.h | 7 +++ drivers/gpu/drm/xe/xe_configfs_types.h | 1 + drivers/gpu/drm/xe/xe_defaults.h | 3 + drivers/gpu/drm/xe/xe_guc_log.c | 3 +- 6 files changed, 98 insertions(+), 2 deletions(-)drivers/gpu/drm/xe/xe_configfs.c +2
-0@@ -106,6 +106,7 @@ const struct xe_config_device xe_configfs_device_defaults = { .enable_survivability_mode = false, .engines_allowed = U64_MAX, .gt_types_allowed = U64_MAX, + .guc_log_level = XE_GUC_LOG_LEVEL_UNSET, },#endif .sriov = {@@ -429,6 +430,7 @@ static void dump_custom_dev_config(struct pci_dev *pdev,PRI_CUSTOM_ATTR ("%d" , debug.enable_survivability_mode);PRI_CUSTOM_ATTR ("%llx" , debug.engines_allowed);PRI_CUSTOM_ATTR ("%llx" , debug.gt_types_allowed); +PRI_CUSTOM_ATTR ("%d" , debug.guc_log_level);#endif PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);drivers/gpu/drm/xe/xe_configfs_debug.c +83
-1@@ -15,6 +15,7 @@#include "xe_configfs_debug.h" #include "xe_configfs_types.h" #include "xe_gt_types.h" +#include "xe_guc_log.h" #include "xe_hw_engine_types.h" #include "xe_pci_types.h" @@ -43,7 +44,8 @@* ├── enable_psmi * ├── enable_survivability_mode * ├── engines_allowed -* └── gt_types_allowed +* ├── gt_types_allowed +* └── guc_log_level * * Configure Attributes * ==================== @@ -186,6 +188,27 @@* * # echo '' > /sys/kernel/config/xe/0000:03:00.0/debug/gt_types_allowed * +* GuC log level: +* -------------- +* +* Set the GuC firmware logging verbosity for this device. Accepted values +* match the ``guc_log_level`` module parameter: +* +* - 0: disable +* - 1: normal (non-verbose) +* - 2..%GUC_LOG_LEVEL_MAX: verbose levels +* +* Example:: +* +* # echo 3 > /sys/kernel/config/xe/0000:03:00.0/debug/guc_log_level +* +* The default value is %XE_GUC_LOG_LEVEL_UNSET (-1), which means the value +* of the ``guc_log_level`` module parameter is used. Any value greater +* than -1 written to this attribute overrides the module parameter for +* this device. +* +* This attribute can only be set before binding to the device. +* */ struct engine_info {@@ -325,6 +348,34 @@ bool xe_configfs_get_psmi_enabled(struct pci_dev *pdev)return ret; } +/** +* xe_configfs_get_guc_log_level - get configfs guc_log_level setting +* @pdev: pci device +* +* Returns the guc_log_level value configured via configfs. If the configfs +* value is negative (the default is %XE_GUC_LOG_LEVEL_UNSET, -1), the value +* of the ``guc_log_level`` module parameter is returned instead, allowing +* the configfs entry to override the module parameter without affecting +* users that rely solely on the module parameter. +* +* Return: GuC log level to use for this device. +*/ +int xe_configfs_get_guc_log_level (struct pci_dev *pdev) +{ +struct xe_config_group_device *dev =xe_configfs_find_device (pdev); +int level = xe_modparam.guc_log_level; + +if (!dev) + goto out; + +if (dev->config.debug.guc_log_level >=0 ) + level = dev->config.debug.guc_log_level; + +config_group_put (&dev->group); +out: +return level; +} +/** * xe_configfs_get_ctx_restore_mid_bb - get configfs ctx_restore_mid_bb setting * @pdev: pci device @@ -596,6 +647,35 @@ static ssize_t enable_psmi_store(struct config_item *item, const char *page, sizreturn len; } +static ssize_t guc_log_level_show (struct config_item *item,char *page) +{ +struct xe_config_device *dev =debug_to_device (item); + +return sprintf (page,"%d\n" , dev->debug.guc_log_level); +} + +static ssize_t guc_log_level_store (struct config_item *item,const char *page,size_t len) +{ +struct xe_config_group_device *dev =debug_to_group_device (item); +int val; +int ret; + + ret =kstrtoint (page,0 , &val); +if (ret) +return ret; + +if (val >GUC_LOG_LEVEL_MAX ) +return -EINVAL ; + +guard (mutex)(&dev->lock); +if (xe_configfs_is_bound (dev)) +return -EBUSY ; + + dev->config.debug.guc_log_level = val; + +return len; +} +static bool wa_bb_read_advance (bool dereference,char **p,const char *append,size_t len,size_t *max_size)@@ -837,6 +917,7 @@ CONFIGFS_ATTR(, enable_psmi); CONFIGFS_ATTR(, enable_survivability_mode); CONFIGFS_ATTR(, engines_allowed); CONFIGFS_ATTR(, gt_types_allowed); +CONFIGFS_ATTR(, guc_log_level); static bool xe_configfs_debug_is_visible(struct config_item *item, struct configfs_attribute *attr,@@ -863,6 +944,7 @@ static struct configfs_attribute *xe_configfs_debug_attrs[] = { &attr_enable_survivability_mode, &attr_engines_allowed, &attr_gt_types_allowed, + &attr_guc_log_level, NULL, };drivers/gpu/drm/xe/xe_configfs_debug.h +7
-0@@ -7,7 +7,9 @@#include <linux/types.h> +#include "xe_defaults.h" #include "xe_hw_engine_types.h" +#include "xe_module.h" struct pci_dev ;@@ -17,6 +19,7 @@ bool xe_configfs_primary_gt_allowed(struct pci_dev *pdev);bool xe_configfs_media_gt_allowed (struct pci_dev *pdev);u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev);bool xe_configfs_get_psmi_enabled (struct pci_dev *pdev); +int xe_configfs_get_guc_log_level (struct pci_dev *pdev);u32 xe_configfs_get_ctx_restore_mid_bb (struct pci_dev *pdev,enum xe_engine_class class,const u32 **cs);@@ -33,6 +36,10 @@ static inline bool xe_configfs_primary_gt_allowed(struct pci_dev *pdev) { returnstatic inline bool xe_configfs_media_gt_allowed (struct pci_dev *pdev) {return true; }static inline u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev) {return U64_MAX ; }static inline bool xe_configfs_get_psmi_enabled (struct pci_dev *pdev) {return false; } +static inline int xe_configfs_get_guc_log_level (struct pci_dev *pdev) +{ +return xe_modparam.guc_log_level; +}static inline u32 xe_configfs_get_ctx_restore_mid_bb (struct pci_dev *pdev,enum xe_engine_class class,const u32 **cs) {return 0 ; }drivers/gpu/drm/xe/xe_configfs_types.h +1
-0@@ -37,6 +37,7 @@ struct xe_config_group_device {bool enable_survivability_mode;u64 engines_allowed;u64 gt_types_allowed; +int guc_log_level; } debug;struct {bool admin_only_pf;drivers/gpu/drm/xe/xe_defaults.h +3
-0@@ -13,6 +13,9 @@#define XE_DEFAULT_GUC_LOG_LEVEL 1#endif +/* Sentinel value for guc_log_level configfs: not set, fall back to module param */ +#define XE_GUC_LOG_LEVEL_UNSET -1 +#define XE_DEFAULT_PROBE_DISPLAY IS_ENABLED(CONFIG_DRM_XE_DISPLAY)#define XE_DEFAULT_VRAM_BAR_SIZE 0#define XE_DEFAULT_FORCE_PROBE CONFIG_DRM_XE_FORCE_PROBEdrivers/gpu/drm/xe/xe_guc_log.c +2
-1@@ -13,6 +13,7 @@#include "abi/guc_lfd_abi.h" #include "regs/xe_guc_regs.h" #include "xe_bo.h" +#include "xe_configfs_debug.h" #include "xe_devcoredump.h" #include "xe_force_wake.h" #include "xe_gt_printk.h" @@ -637,7 +638,7 @@ int xe_guc_log_init(struct xe_guc_log *log)xe_map_memset (xe, &bo->vmap,0 ,0 ,xe_bo_size (bo)); log->bo = bo; - log->level = xe_modparam.guc_log_level; + log->level =xe_configfs_get_guc_log_level (to_pci_dev (xe->drm.dev));return 0 ; }
PATCH 7/9 drm/xe/guc: Add support for NPK as a GuC log target
In reply to Stuart SummersGuC provides the ability to gather logs through a hardware interface called NPK. For certain debugging scenarios this can be advantageous over getting logs from memory (or in addition to).
Add a hook for this alternate debugging mode via a configfs. This translates into a parameter passed to GuC during load time.
v2: Convert to configfs from modparam (Matt)
v3: Configfs documentation formatting (Shuicheng)
Kerneldoc/comment add + configfs entry ordering Only set the guc_log_target when GuC log is enabled (Daniele)
--- drivers/gpu/drm/xe/abi/guc_log_abi.h | 8 ++++ drivers/gpu/drm/xe/xe_configfs.c | 2 + drivers/gpu/drm/xe/xe_configfs_debug.c | 64 +++++++++++++++++++++++++- drivers/gpu/drm/xe/xe_configfs_debug.h | 5 ++ drivers/gpu/drm/xe/xe_configfs_types.h | 1 + drivers/gpu/drm/xe/xe_defaults.h | 2 + drivers/gpu/drm/xe/xe_guc.c | 11 +++-- 7 files changed, 89 insertions(+), 4 deletions(-)
drivers/gpu/drm/xe/abi/guc_log_abi.h +8 -0
@@ -51,6 +51,14 @@ enum guc_log_type {#define GUC_LOG_BUFFER_TYPE_MAX 3 +enum guc_log_target { +GUC_LOG_TARGET_MEM =0 , +GUC_LOG_TARGET_NPK , +GUC_LOG_TARGET_MEM_AND_NPK , +}; + +#define GUC_LOG_TARGET_MAX GUC_LOG_TARGET_MEM_AND_NPK +/** * struct guc_log_buffer_state - GuC log buffer state *
drivers/gpu/drm/xe/xe_configfs.c +2 -0
@@ -107,6 +107,7 @@ const struct xe_config_device xe_configfs_device_defaults = { .engines_allowed = U64_MAX, .gt_types_allowed = U64_MAX, .guc_log_level = XE_GUC_LOG_LEVEL_UNSET, + .guc_log_target = XE_DEFAULT_GUC_LOG_TARGET, },#endif .sriov = {
@@ -431,6 +432,7 @@ static void dump_custom_dev_config(struct pci_dev *pdev,PRI_CUSTOM_ATTR ("%llx" , debug.engines_allowed);PRI_CUSTOM_ATTR ("%llx" , debug.gt_types_allowed);PRI_CUSTOM_ATTR ("%d" , debug.guc_log_level); +PRI_CUSTOM_ATTR ("%d" , debug.guc_log_target);#endif PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);
drivers/gpu/drm/xe/xe_configfs_debug.c +63 -1
100 lines not on this page · raw message
drivers/gpu/drm/xe/xe_configfs_debug.h +5 -0
@@ -15,6 +15,7 @@ struct pci_dev;#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) &&IS_ENABLED (CONFIG_CONFIGFS_FS )bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev); +u8 xe_configfs_get_guc_log_target (struct pci_dev *pdev);bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev);bool xe_configfs_media_gt_allowed (struct pci_dev *pdev);u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev);
@@ -32,6 +33,10 @@ u32 xe_configfs_get_ctx_restore_post_bb(struct pci_dev *pdev,* wrapped in CONFIG_DRM_XE_DEBUG. */ static inline bool xe_configfs_get_enable_survivability_mode (struct pci_dev *pdev) {return false; } +static inline u8 xe_configfs_get_guc_log_target (struct pci_dev *pdev) +{ +return XE_DEFAULT_GUC_LOG_TARGET ; +}static inline bool xe_configfs_primary_gt_allowed (struct pci_dev *pdev) {return true; }static inline bool xe_configfs_media_gt_allowed (struct pci_dev *pdev) {return true; }static inline u64 xe_configfs_get_engines_allowed (struct pci_dev *pdev) {return U64_MAX ; }
drivers/gpu/drm/xe/xe_configfs_types.h +1 -0
@@ -38,6 +38,7 @@ struct xe_config_group_device {u64 engines_allowed;u64 gt_types_allowed;int guc_log_level; +u8 guc_log_target; } debug;struct {bool admin_only_pf;
drivers/gpu/drm/xe/xe_defaults.h +2 -0
@@ -5,6 +5,7 @@#ifndef _XE_DEFAULTS_H_#define _XE_DEFAULTS_H_ +#include "abi/guc_log_abi.h" #include "xe_device_types.h" #if IS_ENABLED (CONFIG_DRM_XE_DEBUG )
@@ -12,6 +13,7 @@#else #define XE_DEFAULT_GUC_LOG_LEVEL 1#endif +#define XE_DEFAULT_GUC_LOG_TARGET GUC_LOG_TARGET_MEM/* Sentinel value for guc_log_level configfs: not set, fall back to module param */ #define XE_GUC_LOG_LEVEL_UNSET -1
drivers/gpu/drm/xe/xe_guc.c +8 -3
@@ -74,13 +74,18 @@ static u32 guc_bo_ggtt_addr(struct xe_guc *guc,static u32 guc_ctl_debug_flags (struct xe_guc *guc) { +struct pci_dev *pdev =to_pci_dev (guc_to_xe (guc)->drm.dev);u32 level =xe_guc_log_get_level (&guc->log);u32 flags =0 ; -if (!GUC_LOG_LEVEL_IS_VERBOSE (level)) +if (!GUC_LOG_LEVEL_IS_VERBOSE (level)) { flags |=GUC_LOG_DISABLED ; -else - flags |=FIELD_PREP (GUC_LOG_VERBOSITY ,GUC_LOG_LEVEL_TO_VERBOSITY (level)); + }else { + flags |=FIELD_PREP (GUC_LOG_VERBOSITY , +GUC_LOG_LEVEL_TO_VERBOSITY (level)); + flags |=FIELD_PREP (GUC_LOG_DESTINATION , +xe_configfs_get_guc_log_target (pdev)); + }return flags; }
Signature · 1 line
2.43.0
PATCH 8/9 drm/xe: Add infrastructure for debug configfs parameters
In reply to Stuart SummersBorrow the X-macro pattern from i915_debugfs_params.c and adapt it to the Xe debug configfs interface, so future debug-only configfs entries can be added with a single line.
Add a new header xe_configfs_debug_params.h that defines:
XE_CONFIGFS_DEBUG_PARAMS_FOR_EACH(param)
invoked as 'param(type, name, default, validator)' for each debug configfs parameter. The matching backing struct, default initializer fragment, configfs_attribute objects, show/store handlers and entries in xe_configfs_debug_attrs[] are all generated from this single list.
Supported type tokens are int, bool, u8, u16, u32 and u64. Sized unsigned types are used directly rather than mirroring i915's int/unsigned int (which exists there only because those debugfs entries shadow module_param values, and module_param itself does not support sized types).
The validator slot is a function-like macro that takes the parsed typed value and returns 0 or a negative errno. XE_PARAM_VALIDATE_NONE is provided for parameters whose full type range is acceptable. Function-like macros (rather than function pointers) let the compiler fold trivial validators away entirely.
The auto-generated handlers and configfs_attribute objects live in a new xe_configfs_debug_params.c so they do not crowd the hand-written entries in xe_configfs_debug.c. The attribute objects are non-static (declared 'extern' in the header) so xe_configfs_debug.c can splice them into xe_configfs_debug_attrs[] and reference them from is_visible().
The parameter list is intentionally empty in this patch; converting existing entries (or adding new ones) is left to follow-up changes so this infrastructure can be reviewed in isolation.
Note that the expectation is a user will only set these debug parameters when CONFIG_DRM_XE_DEBUG is set; the entire facility lives behind that Kconfig.
--- drivers/gpu/drm/xe/Makefile | 2 +- drivers/gpu/drm/xe/xe_configfs.c | 7 ++ drivers/gpu/drm/xe/xe_configfs_debug.c | 2 + drivers/gpu/drm/xe/xe_configfs_debug_params.c | 112 +++++++++++++++++ drivers/gpu/drm/xe/xe_configfs_debug_params.h | 117 ++++++++++++++++++ drivers/gpu/drm/xe/xe_configfs_types.h | 2 + 6 files changed, 241 insertions(+), 1 deletion(-) create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug_params.c create mode 100644 drivers/gpu/drm/xe/xe_configfs_debug_params.h
drivers/gpu/drm/xe/Makefile +1 -1
@@ -161,7 +161,7 @@ xe-$(CONFIG_HWMON) += xe_hwmon.o xe-$(CONFIG_PERF_EVENTS) += xe_pmu.o xe-$(CONFIG_CONFIGFS_FS) += xe_configfs.o -xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o +xe_debug_configfs_obj-$(CONFIG_DRM_XE_DEBUG) := xe_configfs_debug.o xe_configfs_debug_params.o xe-$(CONFIG_CONFIGFS_FS) += $(xe_debug_configfs_obj-y) # graphics virtualization (SR-IOV) support
drivers/gpu/drm/xe/xe_configfs.c +7 -0
@@ -108,6 +108,9 @@ const struct xe_config_device xe_configfs_device_defaults = { .gt_types_allowed = U64_MAX, .guc_log_level = XE_GUC_LOG_LEVEL_UNSET, .guc_log_target = XE_DEFAULT_GUC_LOG_TARGET, + .params = { +XE_CONFIGFS_DEBUG_PARAMS_DEFAULTS + }, }, #endif .sriov = {
@@ -433,6 +436,10 @@ static void dump_custom_dev_config(struct pci_dev *pdev,PRI_CUSTOM_ATTR ("%llx" , debug.gt_types_allowed);PRI_CUSTOM_ATTR ("%d" , debug.guc_log_level);PRI_CUSTOM_ATTR ("%d" , debug.guc_log_target); +#define _XE_PARAM_DUMP (_T, _name, _def, _val) \ + PRI_CUSTOM_ATTR(XE_PARAM_FMT_##_T, debug.params._name); + XE_CONFIGFS_DEBUG_PARAMS_FOR_EACH(_XE_PARAM_DUMP) +#undef _XE_PARAM_DUMP#endif PRI_CUSTOM_ATTR ("%u" , sriov.admin_only_pf);
drivers/gpu/drm/xe/xe_configfs_debug.c +2 -0
@@ -13,6 +13,7 @@#include "abi/guc_log_abi.h" #include "instructions/xe_mi_commands.h" #include "xe_configfs_debug.h" +#include "xe_configfs_debug_params.h" #include "xe_configfs_types.h" #include "xe_gt_types.h" #include "xe_guc_log.h"
@@ -1007,6 +1008,7 @@ static struct configfs_attribute *xe_configfs_debug_attrs[] = { &attr_gt_types_allowed, &attr_guc_log_level, &attr_guc_log_target, +XE_CONFIGFS_DEBUG_PARAMS_FOR_EACH (XE_PARAM_ATTR_PTR ) NULL, };
drivers/gpu/drm/xe/xe_configfs_debug_params.c new file +112 -0
112 lines not on this page · raw message
drivers/gpu/drm/xe/xe_configfs_debug_params.h new file +117 -0
117 lines not on this page · raw message
drivers/gpu/drm/xe/xe_configfs_types.h +2 -0
@@ -9,6 +9,7 @@#include <linux/mutex.h> #include <linux/types.h> +#include "xe_configfs_debug_params.h" #include "xe_hw_engine_types.h" #include "xe_pci_types.h" #include "xe_sriov_types.h"
@@ -39,6 +40,7 @@ struct xe_config_group_device {u64 gt_types_allowed;int guc_log_level;u8 guc_log_target; +struct xe_configfs_debug_params params; } debug;struct {bool admin_only_pf;
Signature · 1 line
2.43.0
PATCH 9/9 drm/xe: Migrate existing debug configfs entries to params infrastructure
In reply to Stuart SummersMove the four existing debug configfs entries whose semantics fit the generic show/store template into the new parameter list:
- enable_psmi (bool) - enable_survivability_mode (bool) - guc_log_level (int, validated against GUC_LOG_LEVEL_MAX,
accepting any negative value as "unset")
- guc_log_target (u8, validated against GUC_LOG_TARGET_MAX)
guc_log_level keeps its module-parameter fallback in the public getter xe_configfs_get_guc_log_level(): a negative value stored in configfs (the default is %XE_GUC_LOG_LEVEL_UNSET, -1) means "use the modparam"; only a value >= 0 overrides it. The validator therefore allows any negative value through so the sentinel can be written back from userspace.
The remaining hand-written entries (ctx_restore_mid_bb, ctx_restore_post_bb, engines_allowed, gt_types_allowed) all have custom string parsing or per-class indexing that does not fit the generic template and stay as-is.
--- drivers/gpu/drm/xe/xe_configfs.c | 8 -- drivers/gpu/drm/xe/xe_configfs_debug.c | 133 +----------------- drivers/gpu/drm/xe/xe_configfs_debug_params.h | 18 ++- drivers/gpu/drm/xe/xe_configfs_types.h | 4 - 4 files changed, 21 insertions(+), 142 deletions(-)
drivers/gpu/drm/xe/xe_configfs.c +0 -8
@@ -102,12 +102,8 @@const struct xe_config_device xe_configfs_device_defaults = { #if IS_ENABLED(CONFIG_DRM_XE_DEBUG) .debug = { - .enable_psmi = false, - .enable_survivability_mode = false, .engines_allowed =U64_MAX , .gt_types_allowed =U64_MAX , - .guc_log_level =XE_GUC_LOG_LEVEL_UNSET , - .guc_log_target =XE_DEFAULT_GUC_LOG_TARGET , .params = {XE_CONFIGFS_DEBUG_PARAMS_DEFAULTS },
@@ -430,12 +426,8 @@ static void dump_custom_dev_config(struct pci_dev *pdev, } while (0)#if IS_ENABLED (CONFIG_DRM_XE_DEBUG ) -PRI_CUSTOM_ATTR ("%d" , debug.enable_psmi); -PRI_CUSTOM_ATTR ("%d" , debug.enable_survivability_mode);PRI_CUSTOM_ATTR ("%llx" , debug.engines_allowed);PRI_CUSTOM_ATTR ("%llx" , debug.gt_types_allowed); -PRI_CUSTOM_ATTR ("%d" , debug.guc_log_level); -PRI_CUSTOM_ATTR ("%d" , debug.guc_log_target);#define _XE_PARAM_DUMP (_T, _name, _def, _val) \ PRI_CUSTOM_ATTR(XE_PARAM_FMT_##_T, debug.params._name);XE_CONFIGFS_DEBUG_PARAMS_FOR_EACH (_XE_PARAM_DUMP)
drivers/gpu/drm/xe/xe_configfs_debug.c +7 -126
189 lines not on this page · raw message
drivers/gpu/drm/xe/xe_configfs_debug_params.h +14 -4
@@ -10,6 +10,10 @@#include <linux/errno.h> #include <linux/types.h> +#include "abi/guc_log_abi.h" +#include "xe_guc_fwif.h" +#include "xe_guc_log.h" +/* * Internal type/format aliases resolved by XE_CONFIGFS_DEBUG_PARAMS_FOR_EACH(). *
@@ -43,12 +47,15 @@* acceptable, a negative errno (typically -EINVAL) otherwise. * * XE_PARAM_VALIDATE_NONE is the no-op validator for parameters whose -* full type range is acceptable. Param-specific validators are defined -* alongside the X-macro list below so that the constraints live next to -* the entries that depend on them. +* full type range is acceptable. */ #define XE_PARAM_VALIDATE_NONE (_v) (0) +#define XE_PARAM_VALIDATE_GUC_LOG_LEVEL (_v) \ + ((((_v) < 0) || ((_v) <= GUC_LOG_LEVEL_MAX)) ? 0 : -EINVAL) +#define XE_PARAM_VALIDATE_GUC_LOG_TARGET (_v) \ + (((_v) <= GUC_LOG_TARGET_MAX) ? 0 : -EINVAL) +/** * XE_CONFIGFS_DEBUG_PARAMS_FOR_EACH - iterate over debug configfs params * @param: function-like macro called as
@@ -73,7 +80,10 @@* at runtime if needed. */ #define XE_CONFIGFS_DEBUG_PARAMS_FOR_EACH (param) \ -/* No parameters yet - add entries above this line. */ + param(bool, enable_psmi, false, XE_PARAM_VALIDATE_NONE) \ + param(bool, enable_survivability_mode, false, XE_PARAM_VALIDATE_NONE) \ + param(int, guc_log_level, XE_GUC_LOG_LEVEL_UNSET, XE_PARAM_VALIDATE_GUC_LOG_LEVEL) \ + param(u8, guc_log_target, XE_DEFAULT_GUC_LOG_TARGET, XE_PARAM_VALIDATE_GUC_LOG_TARGET)/* Generate the struct that backs all debug params in xe_config_device. */ #define _XE_PARAM_MEMBER (_T, _name, _def, _val) XE_PARAM_TYPE_##_T _name;
drivers/gpu/drm/xe/xe_configfs_types.h +0 -4
@@ -34,12 +34,8 @@ struct xe_config_group_device {struct {struct wa_bb ctx_restore_mid_bb[XE_ENGINE_CLASS_MAX ];struct wa_bb ctx_restore_post_bb[XE_ENGINE_CLASS_MAX ]; -bool enable_psmi; -bool enable_survivability_mode;u64 engines_allowed;u64 gt_types_allowed; -int guc_log_level; -u8 guc_log_target;struct xe_configfs_debug_params params; } debug;struct {
Signature · 1 line
2.43.0
Quoted 34 lines · Stuart Summers
Add a new configfs debug group. The intent of this structure is to allow us to separate ABI facing configfs entries from those which are purely for debug purposes. And it allows us more flexibility
in how we define those parameters used for debug.Add a new infrastructure to this debug configfs group that lets us easily define the parameters in a quick list. This is primarily useful for simple, single-type parameters such as enable/disable features or simple values passed. For more complex parameters, we will still need to define these separately.
Pull the GuC target related changes from [1] to fit within this new structure and add a new definition for
guc_log_levelon top of the existing module parameter (to ensure we aren't impacting existing users of the module parameter).Note that the debug parameters here are all to be used "at your own risk". Without having in depth knowledge of how these impact the software and hardware, there could be unforeseen consequences of setting them. As such, they are all wrapped in a
CONFIG_DRM_XE_DEBUGconfigfs option.In terms of the patches here, I'm also making a few organizational changes like naming and sorting. I also know we have a few other module parameters that could migrate here, but I didn't want to overload this series too much, so the focus for now is on the existing configfs entries and demonstrating the new structures with the GuC log level and target parameters.
I used GitHub Copilot with Claude pretty extensively through the process here and attributed as such. Happy to answer any questions around this.
Quoted 8 lines
I also decided to drop John Harrison from the NPK patch. It has been modified quite a bit from the original, but more importantly John is no longer with Intel and that email address isn't available any more. If it makes a difference here, John and I had both separately
implemented this same change at different occasions for debug. The one I used to start that initial series was cherry-picked from his latest variant.
Quick note here that I still need to do some IGT updates here to take the new directory structure into account... I did a little manual testing so far. But I'd really appreciate any feedback you have in the meantime. I just want to make sure there aren't any major complaints with the overall idea here.
Thanks,
Stuart
Quoted 44 lines
Tested on BMG.
[1] https://patchwork.freedesktop.org/series/162087/
Stuart Summers (9):
drm/xe: Renamesurvivability_modetoenable_survivability_modedrm/xe: Sortxe_config_devicefields and defaults alphabetically drm/xe: Split out configfs data structures
drm/xe: Add a new debug focused configfs group drm/xe: Move debug configfs entries toxe_configfs_debug.cdrm/xe/guc: Add configfs support forguc_log_leveldrm/xe/guc: Add support for NPK as a GuC log target drm/xe: Add infrastructure for debug configfs parameters drm/xe: Migrate existing debug configfs entries to params infrastructuredrivers/gpu/drm/xe/Makefile | 2 + drivers/gpu/drm/xe/abi/guc_log_abi.h | 8 + drivers/gpu/drm/xe/xe_configfs.c | 952 ++-------------- --
drivers/gpu/drm/xe/xe_configfs.h | 23 - drivers/gpu/drm/xe/xe_configfs_debug.c | 900
+++++++++++++++++
drivers/gpu/drm/xe/xe_configfs_debug.h | 56 ++drivers/gpu/drm/xe/xe_configfs_debug_params.c| 112 +++drivers/gpu/drm/xe/xe_configfs_debug_params.h| 127 +++ drivers/gpu/drm/xe/xe_configfs_types.h | 73 ++ drivers/gpu/drm/xe/xe_defaults.h | 5 + drivers/gpu/drm/xe/xe_guc.c | 12 +- drivers/gpu/drm/xe/xe_guc_log.c | 3 +- drivers/gpu/drm/xe/xe_hw_engine.c | 1 + drivers/gpu/drm/xe/xe_lrc.c | 1 + drivers/gpu/drm/xe/xe_pci.c | 1 + drivers/gpu/drm/xe/xe_psmi.c | 1 + drivers/gpu/drm/xe/xe_rtp.c | 1 + drivers/gpu/drm/xe/xe_survivability_mode.c | 3 +-18 files changed, 1369 insertions(+), 912 deletions(-)create mode 100644
drivers/gpu/drm/xe/xe_configfs_debug.ccreate mode 100644drivers/gpu/drm/xe/xe_configfs_debug.hcreate mode 100644drivers/gpu/drm/xe/xe_configfs_debug_params.ccreate mode 100644drivers/gpu/drm/xe/xe_configfs_debug_params.hcreate mode 100644drivers/gpu/drm/xe/xe_configfs_types.h
Patchwork
CI✗ CI.checkpatch: warning for Add new debug infrastructure for configfs
In reply to Stuart Summers== Series Details ==
== Series Details ==
Series: Add new debug infrastructure for configfs
URL : https://patchwork.freedesktop.org/series/165879/
State : warning
== Summary ==
+ KERNEL=/kernel
+ git clone https://gitlab.freedesktop.org/drm/maintainer-tools mt Cloning into 'mt'...
warning: redirecting to https://gitlab.freedesktop.org/drm/maintainer-tools.git/
+ git -C mt rev-list -n1 origin/master c8c12e558adaef7a4d125d83b6e1f8824bc13b82
+ cd /kernel
+ git config --global --add safe.directory /kernel
+ git log -n1
commit b22dcaff1c2ae70ee14a72f3a51f0ae7ab2e49e4
Author: Stuart Summers <[email protected]> Date: Mon May 4 04:43:46 2026 +0000
drm/xe: Migrate existing debug configfs entries to params infrastructure
Move the four existing debug configfs entries whose semantics fit the generic show/store template into the new parameter list:
- enable_psmi (bool)
- enable_survivability_mode (bool)
- guc_log_level (int, validated against GUC_LOG_LEVEL_MAX,
accepting any negative value as "unset")
- guc_log_target (u8, validated against GUC_LOG_TARGET_MAX)
guc_log_level keeps its module-parameter fallback in the public getter xe_configfs_get_guc_log_level(): a negative value stored in configfs (the default is %XE_GUC_LOG_LEVEL_UNSET, -1) means "use the modparam"; only a value >= 0 overrides it. The validator therefore allows any negative value through so the sentinel can be written back from userspace.
The remaining hand-written entries (ctx_restore_mid_bb, ctx_restore_post_bb, engines_allowed, gt_types_allowed) all have custom string parsing or per-class indexing that does not fit the generic template and stay as-is.
+ /mt/dim checkpatch 59b0faed1d239bc0f433999671c2d195bd62f9b8 drm-intel a43da1486476 drm/xe: Rename survivability_mode to enable_survivability_mode bcb4520c13a0 drm/xe: Sort xe_config_device fields and defaults alphabetically 1b67b92aeb7d drm/xe: Split out configfs data structures
-:270: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating? #270:
new file mode 100644
total: 0 errors, 1 warnings, 0 checks, 291 lines checked7b8d7a78f666 drm/xe: Add a new debug focused configfs group
-:40: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating? #40:
new file mode 100644
total: 0 errors, 1 warnings, 0 checks, 57 lines checkedf156c79f9fce drm/xe: Move debug configfs entries to xe_configfs_debug.c fab710158ad5 drm/xe/guc: Add configfs support for guc_log_level 8ac835cbc4de drm/xe/guc: Add support for NPK as a GuC log target 0d59bf681606 drm/xe: Add infrastructure for debug configfs parameters -:80: WARNING:MACRO_ARG_UNUSED: Argument '_def' is not used in function-like macro #80: FILE: drivers/gpu/drm/xe/xe_configfs.c:439:
+#define _XE_PARAM_DUMP(_T, _name, _def, _val) \
+ PRI_CUSTOM_ATTR(XE_PARAM_FMT_##_T, debug.params._name);
-:80: WARNING:MACRO_ARG_UNUSED: Argument '_val' is not used in function-like macro #80: FILE: drivers/gpu/drm/xe/xe_configfs.c:439:
+#define _XE_PARAM_DUMP(_T, _name, _def, _val) \
+ PRI_CUSTOM_ATTR(XE_PARAM_FMT_##_T, debug.params._name);
-:80: WARNING:TRAILING_SEMICOLON: macros should not use a trailing semicolon #80: FILE: drivers/gpu/drm/xe/xe_configfs.c:439:
+#define _XE_PARAM_DUMP(_T, _name, _def, _val) \
+ PRI_CUSTOM_ATTR(XE_PARAM_FMT_##_T, debug.params._name);
-:108: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating? #108:
new file mode 100644
-:151: CHECK:MACRO_ARG_REUSE: Macro argument reuse '_name' - possible side-effects? #151: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.c:39:
+#define _XE_PARAM_DEFINE_OPS_NUMERIC(_T, _name, _kstrto, _fmt, _validator) \
+static ssize_t _name##_show(struct config_item *item, char *page) \
+{ \
+ struct xe_config_device *dev = xe_configfs_to_device(item->ci_parent); \
+ return sprintf(page, _fmt "\n", dev->debug.params._name); \
+} \
+static ssize_t _name##_store(struct config_item *item, const char *page, size_t len) \
+{ \
+ struct xe_config_group_device *dev = xe_configfs_to_group_device(item->ci_parent); \
+ XE_PARAM_TYPE_##_T val; \
+ int ret = _kstrto(page, 0, &val); \
+ if (ret) \
+ return ret; \
+ ret = _validator(val); \
+ if (ret) \
+ return ret; \
+ guard(mutex)(&dev->lock); \
+ if (xe_configfs_is_bound(dev)) \
+ return -EBUSY; \
+ dev->config.debug.params._name = val; \
+ return len; \
+}
-:184: CHECK:MACRO_ARG_REUSE: Macro argument reuse '_name' - possible side-effects? #184: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.c:72:
+#define _XE_PARAM_DEFINE_OPS_bool(_name, _validator) \
+static ssize_t _name##_show(struct config_item *item, char *page) \
+{ \
+ struct xe_config_device *dev = xe_configfs_to_device(item->ci_parent); \
+ return sprintf(page, "%d\n", dev->debug.params._name); \
+} \
+static ssize_t _name##_store(struct config_item *item, const char *page, size_t len) \
+{ \
+ struct xe_config_group_device *dev = xe_configfs_to_group_device(item->ci_parent); \
+ bool val; \
+ int ret = kstrtobool(page, &val); \
+ if (ret) \
+ return ret; \
+ ret = _validator(val); \
+ if (ret) \
+ return ret; \
+ guard(mutex)(&dev->lock); \
+ if (xe_configfs_is_bound(dev)) \
+ return -EBUSY; \
+ dev->config.debug.params._name = val; \
+ return len; \
+}
-:207: WARNING:MACRO_ARG_UNUSED: Argument '_def' is not used in function-like macro #207: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.c:95:
+#define _XE_PARAM_DEFINE_OPS(_T, _name, _def, _validator) \
+ _XE_PARAM_DEFINE_OPS_##_T(_name, _validator)
-:215: WARNING:MACRO_ARG_UNUSED: Argument '_T' is not used in function-like macro #215: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.c:103:
+#define _XE_PARAM_DEFINE_ATTR(_T, _name, _def, _val) \
+struct configfs_attribute attr_##_name = { \
+ .ca_name = __stringify(_name), \
+ .ca_owner = THIS_MODULE, \
+ .ca_mode = 0644, \
+ .show = _name##_show, \
+ .store = _name##_store, \
+};
-:215: WARNING:MACRO_ARG_UNUSED: Argument '_def' is not used in function-like macro #215: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.c:103:
+#define _XE_PARAM_DEFINE_ATTR(_T, _name, _def, _val) \
+struct configfs_attribute attr_##_name = { \
+ .ca_name = __stringify(_name), \
+ .ca_owner = THIS_MODULE, \
+ .ca_mode = 0644, \
+ .show = _name##_show, \
+ .store = _name##_store, \
+};
-:215: WARNING:MACRO_ARG_UNUSED: Argument '_val' is not used in function-like macro #215: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.c:103:
+#define _XE_PARAM_DEFINE_ATTR(_T, _name, _def, _val) \
+struct configfs_attribute attr_##_name = { \
+ .ca_name = __stringify(_name), \
+ .ca_owner = THIS_MODULE, \
+ .ca_mode = 0644, \
+ .show = _name##_show, \
+ .store = _name##_store, \
+};
-:215: WARNING:TRAILING_SEMICOLON: macros should not use a trailing semicolon #215: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.c:103:
+#define _XE_PARAM_DEFINE_ATTR(_T, _name, _def, _val) \
+struct configfs_attribute attr_##_name = { \
+ .ca_name = __stringify(_name), \
+ .ca_owner = THIS_MODULE, \
+ .ca_mode = 0644, \
+ .show = _name##_show, \
+ .store = _name##_store, \
+};
-:280: WARNING:MACRO_ARG_UNUSED: Argument '_v' is not used in function-like macro #280: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:50:
+#define XE_PARAM_VALIDATE_NONE(_v) (0)
-:309: WARNING:MACRO_ARG_UNUSED: Argument '_def' is not used in function-like macro #309: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:79:
+#define _XE_PARAM_MEMBER(_T, _name, _def, _val) XE_PARAM_TYPE_##_T _name;
-:309: WARNING:MACRO_ARG_UNUSED: Argument '_val' is not used in function-like macro #309: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:79:
+#define _XE_PARAM_MEMBER(_T, _name, _def, _val) XE_PARAM_TYPE_##_T _name;
-:309: WARNING:TRAILING_SEMICOLON: macros should not use a trailing semicolon #309: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:79:
+#define _XE_PARAM_MEMBER(_T, _name, _def, _val) XE_PARAM_TYPE_##_T _name;
-:323: WARNING:MACRO_ARG_UNUSED: Argument '_T' is not used in function-like macro #323: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:93:
+#define _XE_PARAM_DEFAULT_INIT(_T, _name, _def, _val) ._name = (_def),
-:323: WARNING:MACRO_ARG_UNUSED: Argument '_val' is not used in function-like macro #323: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:93:
+#define _XE_PARAM_DEFAULT_INIT(_T, _name, _def, _val) ._name = (_def),
-:336: WARNING:MACRO_ARG_UNUSED: Argument '_T' is not used in function-like macro #336: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:106:
+#define _XE_PARAM_DECLARE_ATTR(_T, _name, _def, _val) \
+ extern struct configfs_attribute attr_##_name;
-:336: WARNING:MACRO_ARG_UNUSED: Argument '_def' is not used in function-like macro #336: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:106:
+#define _XE_PARAM_DECLARE_ATTR(_T, _name, _def, _val) \
+ extern struct configfs_attribute attr_##_name;
-:336: WARNING:MACRO_ARG_UNUSED: Argument '_val' is not used in function-like macro #336: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:106:
+#define _XE_PARAM_DECLARE_ATTR(_T, _name, _def, _val) \
+ extern struct configfs_attribute attr_##_name;
-:336: WARNING:TRAILING_SEMICOLON: macros should not use a trailing semicolon #336: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:106:
+#define _XE_PARAM_DECLARE_ATTR(_T, _name, _def, _val) \
+ extern struct configfs_attribute attr_##_name;
-:345: ERROR:COMPLEX_MACRO: Macros with complex values should be enclosed in parentheses #345: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:115:
+#define XE_PARAM_ATTR_PTR(_T, _name, _def, _val) &attr_##_name,
BUT SEE:
do {} while (0) advice is over-stated in a few situations:
The more obvious case is macros, like MODULE_PARM_DESC, invoked at file-scope, where C disallows code (it must be in functions). See $exceptions if you have one to add by name.
More troublesome is declarative macros used at top of new scope, like DECLARE_PER_CPU. These might just compile with a do-while-0 wrapper, but would be incorrect. Most of these are handled by detecting struct,union,etc declaration primitives in $exceptions.
Theres also macros called inside an if (block), which "return" an expression. These cannot do-while, and need a ({}) wrapper.
Enjoy this qualification while we work to improve our heuristics.
-:345: WARNING:MACRO_ARG_UNUSED: Argument '_T' is not used in function-like macro #345: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:115:
+#define XE_PARAM_ATTR_PTR(_T, _name, _def, _val) &attr_##_name,
-:345: WARNING:MACRO_ARG_UNUSED: Argument '_def' is not used in function-like macro #345: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:115:
+#define XE_PARAM_ATTR_PTR(_T, _name, _def, _val) &attr_##_name,
-:345: WARNING:MACRO_ARG_UNUSED: Argument '_val' is not used in function-like macro #345: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:115:
+#define XE_PARAM_ATTR_PTR(_T, _name, _def, _val) &attr_##_name,
total: 1 errors, 22 warnings, 2 checks, 284 lines checkedb22dcaff1c2a drm/xe: Migrate existing debug configfs entries to params infrastructure -:288: CHECK:MACRO_ARG_REUSE: Macro argument reuse '_v' - possible side-effects? #288: FILE: drivers/gpu/drm/xe/xe_configfs_debug_params.h:54:
+#define XE_PARAM_VALIDATE_GUC_LOG_LEVEL(_v) \
+ ((((_v) < 0) || ((_v) <= GUC_LOG_LEVEL_MAX)) ? 0 : -EINVAL)
total: 0 errors, 0 warnings, 1 checks, 264 lines checked
Patchwork
CI✓ CI.KUnit: success for Add new debug infrastructure for configfs
In reply to Stuart Summers== Series Details ==
== Series Details ==
Series: Add new debug infrastructure for configfs
URL : https://patchwork.freedesktop.org/series/165879/
State : success
== Summary ==
+ trap cleanup EXIT
+ /kernel/tools/testing/kunit/kunit.py run --kunitconfig /kernel/drivers/gpu/drm/xe/.kunitconfig [05:30:52] Configuring KUnit Kernel ...
Generating .config ...
Populating config with:
$ make ARCH=um O=.kunit olddefconfig [05:30:57] Building KUnit Kernel ... Populating config with: $ make ARCH=um O=.kunit olddefconfig Building with: $ make all compile_commands.json scripts_gdb ARCH=um O=.kunit --jobs=48 [05:31:29] Starting KUnit Kernel (1/1)... [05:31:29] ============================================================ Running tests with: $ .kunit/linux kunit.enable=1 mem=1G console=tty kunit_shutdown=halt [05:31:29] ================== guc_buf (11 subtests) =================== [05:31:29] [PASSED] test_smallest [05:31:29] [PASSED] test_largest [05:31:29] [PASSED] test_granular [05:31:29] [PASSED] test_unique [05:31:29] [PASSED] test_overlap [05:31:29] [PASSED] test_reusable [05:31:29] [PASSED] test_too_big [05:31:29] [PASSED] test_flush [05:31:29] [PASSED] test_lookup [05:31:29] [PASSED] test_data [05:31:29] [PASSED] test_class [05:31:29] ===================== [PASSED] guc_buf ===================== [05:31:29] =================== guc_dbm (7 subtests) =================== [05:31:29] [PASSED] test_empty [05:31:29] [PASSED] test_default [05:31:29] ======================== test_size ======================== [05:31:29] [PASSED] 4 [05:31:29] [PASSED] 8 [05:31:29] [PASSED] 32 [05:31:29] [PASSED] 256 [05:31:29] ==================== [PASSED] test_size ==================== [05:31:29] ======================= test_reuse ======================== [05:31:29] [PASSED] 4 [05:31:29] [PASSED] 8 [05:31:29] [PASSED] 32 [05:31:29] [PASSED] 256 [05:31:29] =================== [PASSED] test_reuse ==================== [05:31:29] =================== test_range_overlap ==================== [05:31:29] [PASSED] 4 [05:31:29] [PASSED] 8 [05:31:29] [PASSED] 32 [05:31:29] [PASSED] 256 [05:31:29] =============== [PASSED] test_range_overlap ================ [05:31:29] =================== test_range_compact ==================== [05:31:29] [PASSED] 4 [05:31:29] [PASSED] 8 [05:31:29] [PASSED] 32 [05:31:29] [PASSED] 256 [05:31:29] =============== [PASSED] test_range_compact ================ [05:31:29] ==================== test_range_spare ===================== [05:31:29] [PASSED] 4 [05:31:29] [PASSED] 8 [05:31:29] [PASSED] 32 [05:31:29] [PASSED] 256 [05:31:29] ================ [PASSED] test_range_spare ================= [05:31:29] ===================== [PASSED] guc_dbm ===================== [05:31:29] =================== guc_idm (6 subtests) =================== [05:31:29] [PASSED] bad_init [05:31:29] [PASSED] no_init [05:31:29] [PASSED] init_fini [05:31:29] [PASSED] check_used [05:31:29] [PASSED] check_quota [05:31:29] [PASSED] check_all [05:31:29] ===================== [PASSED] guc_idm ===================== [05:31:29] ================== no_relay (3 subtests) =================== [05:31:29] [PASSED] xe_drops_guc2pf_if_not_ready [05:31:29] [PASSED] xe_drops_guc2vf_if_not_ready [05:31:29] [PASSED] xe_rejects_send_if_not_ready [05:31:29] ==================== [PASSED] no_relay ===================== [05:31:29] ================== pf_relay (14 subtests) ================== [05:31:29] [PASSED] pf_rejects_guc2pf_too_short [05:31:29] [PASSED] pf_rejects_guc2pf_too_long [05:31:29] [PASSED] pf_rejects_guc2pf_no_payload [05:31:29] [PASSED] pf_fails_no_payload [05:31:29] [PASSED] pf_fails_bad_origin [05:31:29] [PASSED] pf_fails_bad_type [05:31:29] [PASSED] pf_txn_reports_error [05:31:29] [PASSED] pf_txn_sends_pf2guc [05:31:29] [PASSED] pf_sends_pf2guc [05:31:29] [SKIPPED] pf_loopback_nop [05:31:29] [SKIPPED] pf_loopback_echo [05:31:29] [SKIPPED] pf_loopback_fail [05:31:29] [SKIPPED] pf_loopback_busy [05:31:29] [SKIPPED] pf_loopback_retry [05:31:29] ==================== [PASSED] pf_relay ===================== [05:31:29] ================== vf_relay (3 subtests) =================== [05:31:29] [PASSED] vf_rejects_guc2vf_too_short [05:31:29] [PASSED] vf_rejects_guc2vf_too_long [05:31:29] [PASSED] vf_rejects_guc2vf_no_payload [05:31:29] ==================== [PASSED] vf_relay ===================== [05:31:29] ================ pf_gt_config (9 subtests) ================= [05:31:29] [PASSED] fair_contexts_1vf [05:31:29] [PASSED] fair_doorbells_1vf [05:31:29] [PASSED] fair_ggtt_1vf [05:31:29] ====================== fair_vram_1vf ====================== [05:31:29] [PASSED] 3.50 GiB [05:31:29] [PASSED] 11.5 GiB [05:31:29] [PASSED] 15.5 GiB [05:31:29] [PASSED] 31.5 GiB [05:31:29] [PASSED] 63.5 GiB [05:31:29] [PASSED] 1.91 GiB [05:31:29] ================== [PASSED] fair_vram_1vf ================== [05:31:29] ================ fair_vram_1vf_admin_only ================= [05:31:29] [PASSED] 3.50 GiB [05:31:29] [PASSED] 11.5 GiB [05:31:29] [PASSED] 15.5 GiB [05:31:29] [PASSED] 31.5 GiB [05:31:29] [PASSED] 63.5 GiB [05:31:29] [PASSED] 1.91 GiB [05:31:29] ============ [PASSED] fair_vram_1vf_admin_only ============= [05:31:29] ====================== fair_contexts ====================== [05:31:29] [PASSED] 1 VF [05:31:29] [PASSED] 2 VFs [05:31:29] [PASSED] 3 VFs [05:31:29] [PASSED] 4 VFs [05:31:29] [PASSED] 5 VFs [05:31:29] [PASSED] 6 VFs [05:31:29] [PASSED] 7 VFs [05:31:29] [PASSED] 8 VFs [05:31:29] [PASSED] 9 VFs [05:31:29] [PASSED] 10 VFs [05:31:29] [PASSED] 11 VFs [05:31:29] [PASSED] 12 VFs [05:31:29] [PASSED] 13 VFs [05:31:29] [PASSED] 14 VFs [05:31:29] [PASSED] 15 VFs [05:31:29] [PASSED] 16 VFs [05:31:29] [PASSED] 17 VFs [05:31:29] [PASSED] 18 VFs [05:31:29] [PASSED] 19 VFs [05:31:29] [PASSED] 20 VFs [05:31:29] [PASSED] 21 VFs [05:31:29] [PASSED] 22 VFs [05:31:29] [PASSED] 23 VFs [05:31:29] [PASSED] 24 VFs [05:31:29] [PASSED] 25 VFs [05:31:29] [PASSED] 26 VFs [05:31:29] [PASSED] 27 VFs [05:31:29] [PASSED] 28 VFs [05:31:29] [PASSED] 29 VFs [05:31:29] [PASSED] 30 VFs [05:31:29] [PASSED] 31 VFs [05:31:29] [PASSED] 32 VFs [05:31:29] [PASSED] 33 VFs [05:31:29] [PASSED] 34 VFs [05:31:29] [PASSED] 35 VFs [05:31:29] [PASSED] 36 VFs [05:31:29] [PASSED] 37 VFs [05:31:29] [PASSED] 38 VFs [05:31:29] [PASSED] 39 VFs [05:31:29] [PASSED] 40 VFs [05:31:29] [PASSED] 41 VFs [05:31:29] [PASSED] 42 VFs [05:31:29] [PASSED] 43 VFs [05:31:29] [PASSED] 44 VFs [05:31:29] [PASSED] 45 VFs [05:31:29] [PASSED] 46 VFs [05:31:29] [PASSED] 47 VFs [05:31:29] [PASSED] 48 VFs [05:31:29] [PASSED] 49 VFs [05:31:29] [PASSED] 50 VFs [05:31:29] [PASSED] 51 VFs [05:31:29] [PASSED] 52 VFs [05:31:29] [PASSED] 53 VFs [05:31:29] [PASSED] 54 VFs [05:31:29] [PASSED] 55 VFs [05:31:29] [PASSED] 56 VFs [05:31:29] [PASSED] 57 VFs [05:31:29] [PASSED] 58 VFs [05:31:29] [PASSED] 59 VFs [05:31:29] [PASSED] 60 VFs [05:31:29] [PASSED] 61 VFs [05:31:29] [PASSED] 62 VFs [05:31:29] [PASSED] 63 VFs [05:31:29] ================== [PASSED] fair_contexts ================== [05:31:29] ===================== fair_doorbells ====================== [05:31:29] [PASSED] 1 VF [05:31:29] [PASSED] 2 VFs [05:31:29] [PASSED] 3 VFs [05:31:29] [PASSED] 4 VFs [05:31:29] [PASSED] 5 VFs [05:31:29] [PASSED] 6 VFs [05:31:29] [PASSED] 7 VFs [05:31:29] [PASSED] 8 VFs [05:31:29] [PASSED] 9 VFs [05:31:29] [PASSED] 10 VFs [05:31:29] [PASSED] 11 VFs [05:31:29] [PASSED] 12 VFs [05:31:29] [PASSED] 13 VFs [05:31:29] [PASSED] 14 VFs [05:31:29] [PASSED] 15 VFs [05:31:29] [PASSED] 16 VFs [05:31:29] [PASSED] 17 VFs [05:31:29] [PASSED] 18 VFs [05:31:29] [PASSED] 19 VFs [05:31:29] [PASSED] 20 VFs [05:31:29] [PASSED] 21 VFs [05:31:29] [PASSED] 22 VFs [05:31:29] [PASSED] 23 VFs [05:31:29] [PASSED] 24 VFs [05:31:29] [PASSED] 25 VFs [05:31:29] [PASSED] 26 VFs [05:31:29] [PASSED] 27 VFs [05:31:29] [PASSED] 28 VFs [05:31:29] [PASSED] 29 VFs [05:31:29] [PASSED] 30 VFs [05:31:29] [PASSED] 31 VFs [05:31:29] [PASSED] 32 VFs [05:31:29] [PASSED] 33 VFs [05:31:29] [PASSED] 34 VFs [05:31:29] [PASSED] 35 VFs [05:31:29] [PASSED] 36 VFs [05:31:29] [PASSED] 37 VFs [05:31:29] [PASSED] 38 VFs [05:31:29] [PASSED] 39 VFs [05:31:29] [PASSED] 40 VFs [05:31:29] [PASSED] 41 VFs [05:31:29] [PASSED] 42 VFs [05:31:29] [PASSED] 43 VFs [05:31:29] [PASSED] 44 VFs [05:31:29] [PASSED] 45 VFs [05:31:29] [PASSED] 46 VFs [05:31:29] [PASSED] 47 VFs [05:31:29] [PASSED] 48 VFs [05:31:29] [PASSED] 49 VFs [05:31:29] [PASSED] 50 VFs [05:31:29] [PASSED] 51 VFs [05:31:29] [PASSED] 52 VFs [05:31:29] [PASSED] 53 VFs [05:31:29] [PASSED] 54 VFs [05:31:29] [PASSED] 55 VFs [05:31:29] [PASSED] 56 VFs [05:31:29] [PASSED] 57 VFs [05:31:29] [PASSED] 58 VFs [05:31:29] [PASSED] 59 VFs [05:31:29] [PASSED] 60 VFs [05:31:29] [PASSED] 61 VFs [05:31:29] [PASSED] 62 VFs [05:31:29] [PASSED] 63 VFs [05:31:29] ================= [PASSED] fair_doorbells ================== [05:31:29] ======================== fair_ggtt ======================== [05:31:29] [PASSED] 1 VF [05:31:29] [PASSED] 2 VFs [05:31:29] [PASSED] 3 VFs [05:31:29] [PASSED] 4 VFs [05:31:29] [PASSED] 5 VFs [05:31:29] [PASSED] 6 VFs [05:31:29] [PASSED] 7 VFs [05:31:29] [PASSED] 8 VFs [05:31:29] [PASSED] 9 VFs [05:31:29] [PASSED] 10 VFs [05:31:29] [PASSED] 11 VFs [05:31:29] [PASSED] 12 VFs [05:31:29] [PASSED] 13 VFs [05:31:29] [PASSED] 14 VFs [05:31:29] [PASSED] 15 VFs [05:31:29] [PASSED] 16 VFs [05:31:29] [PASSED] 17 VFs [05:31:29] [PASSED] 18 VFs [05:31:29] [PASSED] 19 VFs [05:31:29] [PASSED] 20 VFs [05:31:29] [PASSED] 21 VFs [05:31:29] [PASSED] 22 VFs [05:31:29] [PASSED] 23 VFs [05:31:29] [PASSED] 24 VFs [05:31:29] [PASSED] 25 VFs [05:31:29] [PASSED] 26 VFs [05:31:29] [PASSED] 27 VFs [05:31:29] [PASSED] 28 VFs [05:31:29] [PASSED] 29 VFs [05:31:29] [PASSED] 30 VFs [05:31:29] [PASSED] 31 VFs [05:31:29] [PASSED] 32 VFs [05:31:29] [PASSED] 33 VFs [05:31:29] [PASSED] 34 VFs [05:31:29] [PASSED] 35 VFs [05:31:29] [PASSED] 36 VFs [05:31:29] [PASSED] 37 VFs [05:31:29] [PASSED] 38 VFs [05:31:29] [PASSED] 39 VFs [05:31:29] [PASSED] 40 VFs [05:31:29] [PASSED] 41 VFs [05:31:29] [PASSED] 42 VFs [05:31:29] [PASSED] 43 VFs [05:31:29] [PASSED] 44 VFs [05:31:29] [PASSED] 45 VFs [05:31:29] [PASSED] 46 VFs [05:31:29] [PASSED] 47 VFs [05:31:29] [PASSED] 48 VFs [05:31:29] [PASSED] 49 VFs [05:31:29] [PASSED] 50 VFs [05:31:29] [PASSED] 51 VFs [05:31:29] [PASSED] 52 VFs [05:31:29] [PASSED] 53 VFs [05:31:29] [PASSED] 54 VFs [05:31:29] [PASSED] 55 VFs [05:31:29] [PASSED] 56 VFs [05:31:29] [PASSED] 57 VFs [05:31:29] [PASSED] 58 VFs [05:31:29] [PASSED] 59 VFs [05:31:29] [PASSED] 60 VFs [05:31:29] [PASSED] 61 VFs [05:31:29] [PASSED] 62 VFs [05:31:29] [PASSED] 63 VFs [05:31:29] ==================== [PASSED] fair_ggtt ==================== [05:31:29] ======================== fair_vram ======================== [05:31:29] [PASSED] 1 VF [05:31:29] [PASSED] 2 VFs [05:31:29] [PASSED] 3 VFs [05:31:29] [PASSED] 4 VFs [05:31:29] [PASSED] 5 VFs [05:31:29] [PASSED] 6 VFs [05:31:29] [PASSED] 7 VFs [05:31:29] [PASSED] 8 VFs [05:31:29] [PASSED] 9 VFs [05:31:29] [PASSED] 10 VFs [05:31:29] [PASSED] 11 VFs [05:31:29] [PASSED] 12 VFs [05:31:29] [PASSED] 13 VFs [05:31:29] [PASSED] 14 VFs [05:31:29] [PASSED] 15 VFs [05:31:29] [PASSED] 16 VFs [05:31:29] [PASSED] 17 VFs [05:31:29] [PASSED] 18 VFs [05:31:29] [PASSED] 19 VFs [05:31:29] [PASSED] 20 VFs [05:31:29] [PASSED] 21 VFs [05:31:29] [PASSED] 22 VFs [05:31:29] [PASSED] 23 VFs [05:31:29] [PASSED] 24 VFs [05:31:29] [PASSED] 25 VFs [05:31:29] [PASSED] 26 VFs [05:31:29] [PASSED] 27 VFs [05:31:29] [PASSED] 28 VFs [05:31:29] [PASSED] 29 VFs [05:31:29] [PASSED] 30 VFs [05:31:29] [PASSED] 31 VFs [05:31:29] [PASSED] 32 VFs [05:31:29] [PASSED] 33 VFs [05:31:29] [PASSED] 34 VFs [05:31:29] [PASSED] 35 VFs [05:31:29] [PASSED] 36 VFs [05:31:29] [PASSED] 37 VFs [05:31:29] [PASSED] 38 VFs [05:31:29] [PASSED] 39 VFs [05:31:29] [PASSED] 40 VFs [05:31:29] [PASSED] 41 VFs [05:31:29] [PASSED] 42 VFs [05:31:29] [PASSED] 43 VFs [05:31:29] [PASSED] 44 VFs [05:31:29] [PASSED] 45 VFs [05:31:29] [PASSED] 46 VFs [05:31:29] [PASSED] 47 VFs [05:31:29] [PASSED] 48 VFs [05:31:29] [PASSED] 49 VFs [05:31:29] [PASSED] 50 VFs [05:31:29] [PASSED] 51 VFs [05:31:29] [PASSED] 52 VFs [05:31:29] [PASSED] 53 VFs [05:31:29] [PASSED] 54 VFs [05:31:29] [PASSED] 55 VFs [05:31:29] [PASSED] 56 VFs [05:31:29] [PASSED] 57 VFs [05:31:29] [PASSED] 58 VFs [05:31:29] [PASSED] 59 VFs [05:31:29] [PASSED] 60 VFs [05:31:29] [PASSED] 61 VFs [05:31:29] [PASSED] 62 VFs [05:31:29] [PASSED] 63 VFs [05:31:29] ==================== [PASSED] fair_vram ==================== [05:31:29] ================== [PASSED] pf_gt_config =================== [05:31:29] ===================== lmtt (1 subtest) ===================== [05:31:29] ======================== test_ops ========================= [05:31:29] [PASSED] 2-level [05:31:29] [PASSED] multi-level [05:31:29] ==================== [PASSED] test_ops ===================== [05:31:29] ====================== [PASSED] lmtt ======================= [05:31:29] ================= pf_service (11 subtests) ================= [05:31:29] [PASSED] pf_negotiate_any [05:31:29] [PASSED] pf_negotiate_base_match [05:31:29] [PASSED] pf_negotiate_base_newer [05:31:29] [PASSED] pf_negotiate_base_next [05:31:29] [SKIPPED] pf_negotiate_base_older [05:31:29] [PASSED] pf_negotiate_base_prev [05:31:29] [PASSED] pf_negotiate_latest_match [05:31:29] [PASSED] pf_negotiate_latest_newer [05:31:29] [PASSED] pf_negotiate_latest_next [05:31:29] [SKIPPED] pf_negotiate_latest_older [05:31:29] [SKIPPED] pf_negotiate_latest_prev [05:31:29] =================== [PASSED] pf_service ==================== [05:31:29] ================= xe_guc_g2g (2 subtests) ================== [05:31:29] ============== xe_live_guc_g2g_kunit_default ============== [05:31:29] ========= [SKIPPED] xe_live_guc_g2g_kunit_default ========== [05:31:29] ============== xe_live_guc_g2g_kunit_allmem =============== [05:31:29] ========== [SKIPPED] xe_live_guc_g2g_kunit_allmem ========== [05:31:29] =================== [SKIPPED] xe_guc_g2g =================== [05:31:29] =================== xe_mocs (2 subtests) =================== [05:31:29] ================ xe_live_mocs_kernel_kunit ================ [05:31:29] =========== [SKIPPED] xe_live_mocs_kernel_kunit ============ [05:31:29] ================ xe_live_mocs_reset_kunit ================= [05:31:29] ============ [SKIPPED] xe_live_mocs_reset_kunit ============ [05:31:29] ==================== [SKIPPED] xe_mocs ===================== [05:31:29] ================= xe_migrate (2 subtests) ================== [05:31:29] ================= xe_migrate_sanity_kunit ================= [05:31:29] ============ [SKIPPED] xe_migrate_sanity_kunit ============= [05:31:29] ================== xe_validate_ccs_kunit ================== [05:31:29] ============= [SKIPPED] xe_validate_ccs_kunit ============== [05:31:29] =================== [SKIPPED] xe_migrate =================== [05:31:29] ================== xe_dma_buf (1 subtest) ================== [05:31:29] ==================== xe_dma_buf_kunit ===================== [05:31:29] ================ [SKIPPED] xe_dma_buf_kunit ================ [05:31:29] =================== [SKIPPED] xe_dma_buf =================== [05:31:29] ================= xe_bo_shrink (1 subtest) ================= [05:31:29] =================== xe_bo_shrink_kunit ==================== [05:31:29] =============== [SKIPPED] xe_bo_shrink_kunit =============== [05:31:29] ================== [SKIPPED] xe_bo_shrink ================== [05:31:29] ==================== xe_bo (2 subtests) ==================== [05:31:29] ================== xe_ccs_migrate_kunit =================== [05:31:29] ============== [SKIPPED] xe_ccs_migrate_kunit ============== [05:31:29] ==================== xe_bo_evict_kunit ==================== [05:31:29] =============== [SKIPPED] xe_bo_evict_kunit ================ [05:31:29] ===================== [SKIPPED] xe_bo ====================== [05:31:29] ==================== args (13 subtests) ==================== [05:31:29] [PASSED] count_args_test [05:31:29] [PASSED] call_args_example [05:31:29] [PASSED] call_args_test [05:31:29] [PASSED] drop_first_arg_example [05:31:29] [PASSED] drop_first_arg_test [05:31:29] [PASSED] first_arg_example [05:31:29] [PASSED] first_arg_test [05:31:29] [PASSED] last_arg_example [05:31:29] [PASSED] last_arg_test [05:31:29] [PASSED] pick_arg_example [05:31:29] [PASSED] if_args_example [05:31:29] [PASSED] if_args_test [05:31:29] [PASSED] sep_comma_example [05:31:29] ====================== [PASSED] args ======================= [05:31:29] =================== xe_pci (3 subtests) ==================== [05:31:29] ==================== check_graphics_ip ==================== [05:31:29] [PASSED] 12.00 Xe_LP [05:31:29] [PASSED] 12.10 Xe_LP+ [05:31:29] [PASSED] 12.55 Xe_HPG [05:31:29] [PASSED] 12.60 Xe_HPC [05:31:29] [PASSED] 12.70 Xe_LPG [05:31:29] [PASSED] 12.71 Xe_LPG [05:31:29] [PASSED] 12.74 Xe_LPG+ [05:31:29] [PASSED] 20.01 Xe2_HPG [05:31:29] [PASSED] 20.02 Xe2_HPG [05:31:29] [PASSED] 20.04 Xe2_LPG [05:31:29] [PASSED] 30.00 Xe3_LPG [05:31:29] [PASSED] 30.01 Xe3_LPG [05:31:29] [PASSED] 30.03 Xe3_LPG [05:31:29] [PASSED] 30.04 Xe3_LPG [05:31:29] [PASSED] 30.05 Xe3_LPG [05:31:29] [PASSED] 35.10 Xe3p_LPG [05:31:29] [PASSED] 35.11 Xe3p_XPC [05:31:29] ================ [PASSED] check_graphics_ip ================ [05:31:29] ===================== check_media_ip ====================== [05:31:29] [PASSED] 12.00 Xe_M [05:31:29] [PASSED] 12.55 Xe_HPM [05:31:29] [PASSED] 13.00 Xe_LPM+ [05:31:29] [PASSED] 13.01 Xe2_HPM [05:31:29] [PASSED] 20.00 Xe2_LPM [05:31:29] [PASSED] 30.00 Xe3_LPM [05:31:29] [PASSED] 30.02 Xe3_LPM [05:31:29] [PASSED] 35.00 Xe3p_LPM [05:31:29] [PASSED] 35.03 Xe3p_HPM [05:31:29] ================= [PASSED] check_media_ip ================== [05:31:29] =================== check_platform_desc =================== [05:31:29] [PASSED] 0x9A60 (TIGERLAKE) [05:31:29] [PASSED] 0x9A68 (TIGERLAKE) [05:31:29] [PASSED] 0x9A70 (TIGERLAKE) [05:31:29] [PASSED] 0x9A40 (TIGERLAKE) [05:31:29] [PASSED] 0x9A49 (TIGERLAKE) [05:31:29] [PASSED] 0x9A59 (TIGERLAKE) [05:31:29] [PASSED] 0x9A78 (TIGERLAKE) [05:31:29] [PASSED] 0x9AC0 (TIGERLAKE) [05:31:29] [PASSED] 0x9AC9 (TIGERLAKE) [05:31:29] [PASSED] 0x9AD9 (TIGERLAKE) [05:31:29] [PASSED] 0x9AF8 (TIGERLAKE) [05:31:29] [PASSED] 0x4C80 (ROCKETLAKE) [05:31:29] [PASSED] 0x4C8A (ROCKETLAKE) [05:31:29] [PASSED] 0x4C8B (ROCKETLAKE) [05:31:29] [PASSED] 0x4C8C (ROCKETLAKE) [05:31:29] [PASSED] 0x4C90 (ROCKETLAKE) [05:31:29] [PASSED] 0x4C9A (ROCKETLAKE) [05:31:29] [PASSED] 0x4680 (ALDERLAKE_S) [05:31:29] [PASSED] 0x4682 (ALDERLAKE_S) [05:31:29] [PASSED] 0x4688 (ALDERLAKE_S) [05:31:29] [PASSED] 0x468A (ALDERLAKE_S) [05:31:29] [PASSED] 0x468B (ALDERLAKE_S) [05:31:29] [PASSED] 0x4690 (ALDERLAKE_S) [05:31:29] [PASSED] 0x4692 (ALDERLAKE_S) [05:31:29] [PASSED] 0x4693 (ALDERLAKE_S) [05:31:29] [PASSED] 0x46A0 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46A1 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46A2 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46A3 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46A6 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46A8 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46AA (ALDERLAKE_P) [05:31:29] [PASSED] 0x462A (ALDERLAKE_P) [05:31:29] [PASSED] 0x4626 (ALDERLAKE_P) [05:31:29] [PASSED] 0x4628 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46B0 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46B1 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46B2 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46B3 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46C0 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46C1 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46C2 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46C3 (ALDERLAKE_P) [05:31:29] [PASSED] 0x46D0 (ALDERLAKE_N) [05:31:29] [PASSED] 0x46D1 (ALDERLAKE_N) [05:31:29] [PASSED] 0x46D2 (ALDERLAKE_N) [05:31:29] [PASSED] 0x46D3 (ALDERLAKE_N) [05:31:29] [PASSED] 0x46D4 (ALDERLAKE_N) [05:31:29] [PASSED] 0xA721 (ALDERLAKE_P) [05:31:29] [PASSED] 0xA7A1 (ALDERLAKE_P) [05:31:29] [PASSED] 0xA7A9 (ALDERLAKE_P) [05:31:29] [PASSED] 0xA7AC (ALDERLAKE_P) [05:31:29] [PASSED] 0xA7AD (ALDERLAKE_P) [05:31:29] [PASSED] 0xA720 (ALDERLAKE_P) [05:31:29] [PASSED] 0xA7A0 (ALDERLAKE_P) [05:31:29] [PASSED] 0xA7A8 (ALDERLAKE_P) [05:31:29] [PASSED] 0xA7AA (ALDERLAKE_P) [05:31:29] [PASSED] 0xA7AB (ALDERLAKE_P) [05:31:29] [PASSED] 0xA780 (ALDERLAKE_S) [05:31:29] [PASSED] 0xA781 (ALDERLAKE_S) [05:31:29] [PASSED] 0xA782 (ALDERLAKE_S) [05:31:29] [PASSED] 0xA783 (ALDERLAKE_S) [05:31:29] [PASSED] 0xA788 (ALDERLAKE_S) [05:31:29] [PASSED] 0xA789 (ALDERLAKE_S) [05:31:29] [PASSED] 0xA78A (ALDERLAKE_S) [05:31:29] [PASSED] 0xA78B (ALDERLAKE_S) [05:31:29] [PASSED] 0x4905 (DG1) [05:31:29] [PASSED] 0x4906 (DG1) [05:31:29] [PASSED] 0x4907 (DG1) [05:31:29] [PASSED] 0x4908 (DG1) [05:31:29] [PASSED] 0x4909 (DG1) [05:31:29] [PASSED] 0x56C0 (DG2) [05:31:29] [PASSED] 0x56C2 (DG2) [05:31:29] [PASSED] 0x56C1 (DG2) [05:31:29] [PASSED] 0x7D51 (METEORLAKE) [05:31:29] [PASSED] 0x7DD1 (METEORLAKE) [05:31:29] [PASSED] 0x7D41 (METEORLAKE) [05:31:29] [PASSED] 0x7D67 (METEORLAKE) [05:31:29] [PASSED] 0xB640 (METEORLAKE) [05:31:29] [PASSED] 0x56A0 (DG2) [05:31:29] [PASSED] 0x56A1 (DG2) [05:31:29] [PASSED] 0x56A2 (DG2) [05:31:29] [PASSED] 0x56BE (DG2) [05:31:29] [PASSED] 0x56BF (DG2) [05:31:29] [PASSED] 0x5690 (DG2) [05:31:29] [PASSED] 0x5691 (DG2) [05:31:29] [PASSED] 0x5692 (DG2) [05:31:29] [PASSED] 0x56A5 (DG2) [05:31:29] [PASSED] 0x56A6 (DG2) [05:31:29] [PASSED] 0x56B0 (DG2) [05:31:29] [PASSED] 0x56B1 (DG2) [05:31:29] [PASSED] 0x56BA (DG2) [05:31:29] [PASSED] 0x56BB (DG2) [05:31:29] [PASSED] 0x56BC (DG2) [05:31:29] [PASSED] 0x56BD (DG2) [05:31:29] [PASSED] 0x5693 (DG2) [05:31:29] [PASSED] 0x5694 (DG2) [05:31:29] [PASSED] 0x5695 (DG2) [05:31:29] [PASSED] 0x56A3 (DG2) [05:31:29] [PASSED] 0x56A4 (DG2) [05:31:29] [PASSED] 0x56B2 (DG2) [05:31:29] [PASSED] 0x56B3 (DG2) [05:31:29] [PASSED] 0x5696 (DG2) [05:31:29] [PASSED] 0x5697 (DG2) [05:31:29] [PASSED] 0xB69 (PVC) [05:31:29] [PASSED] 0xB6E (PVC) [05:31:29] [PASSED] 0xBD4 (PVC) [05:31:29] [PASSED] 0xBD5 (PVC) [05:31:29] [PASSED] 0xBD6 (PVC) [05:31:29] [PASSED] 0xBD7 (PVC) [05:31:29] [PASSED] 0xBD8 (PVC) [05:31:29] [PASSED] 0xBD9 (PVC) [05:31:29] [PASSED] 0xBDA (PVC) [05:31:29] [PASSED] 0xBDB (PVC) [05:31:29] [PASSED] 0xBE0 (PVC) [05:31:29] [PASSED] 0xBE1 (PVC) [05:31:29] [PASSED] 0xBE5 (PVC) [05:31:29] [PASSED] 0x7D40 (METEORLAKE) [05:31:29] [PASSED] 0x7D45 (METEORLAKE) [05:31:29] [PASSED] 0x7D55 (METEORLAKE) [05:31:29] [PASSED] 0x7D60 (METEORLAKE) [05:31:29] [PASSED] 0x7DD5 (METEORLAKE) [05:31:29] [PASSED] 0x6420 (LUNARLAKE) [05:31:29] [PASSED] 0x64A0 (LUNARLAKE) [05:31:29] [PASSED] 0x64B0 (LUNARLAKE) [05:31:29] [PASSED] 0xE202 (BATTLEMAGE) [05:31:29] [PASSED] 0xE209 (BATTLEMAGE) [05:31:29] [PASSED] 0xE20B (BATTLEMAGE) [05:31:29] [PASSED] 0xE20C (BATTLEMAGE) [05:31:29] [PASSED] 0xE20D (BATTLEMAGE) [05:31:29] [PASSED] 0xE210 (BATTLEMAGE) [05:31:29] [PASSED] 0xE211 (BATTLEMAGE) [05:31:29] [PASSED] 0xE212 (BATTLEMAGE) [05:31:29] [PASSED] 0xE216 (BATTLEMAGE) [05:31:29] [PASSED] 0xE220 (BATTLEMAGE) [05:31:29] [PASSED] 0xE221 (BATTLEMAGE) [05:31:29] [PASSED] 0xE222 (BATTLEMAGE) [05:31:29] [PASSED] 0xE223 (BATTLEMAGE) [05:31:29] [PASSED] 0xB080 (PANTHERLAKE) [05:31:29] [PASSED] 0xB081 (PANTHERLAKE) [05:31:29] [PASSED] 0xB082 (PANTHERLAKE) [05:31:29] [PASSED] 0xB083 (PANTHERLAKE) [05:31:29] [PASSED] 0xB084 (PANTHERLAKE) [05:31:29] [PASSED] 0xB085 (PANTHERLAKE) [05:31:29] [PASSED] 0xB086 (PANTHERLAKE) [05:31:29] [PASSED] 0xB087 (PANTHERLAKE) [05:31:29] [PASSED] 0xB08F (PANTHERLAKE) [05:31:29] [PASSED] 0xB090 (PANTHERLAKE) [05:31:29] [PASSED] 0xB0A0 (PANTHERLAKE) [05:31:29] [PASSED] 0xB0B0 (PANTHERLAKE) [05:31:29] [PASSED] 0xFD80 (PANTHERLAKE) [05:31:29] [PASSED] 0xFD81 (PANTHERLAKE) [05:31:29] [PASSED] 0xD740 (NOVALAKE_S) [05:31:29] [PASSED] 0xD741 (NOVALAKE_S) [05:31:29] [PASSED] 0xD742 (NOVALAKE_S) [05:31:29] [PASSED] 0xD743 (NOVALAKE_S) [05:31:29] [PASSED] 0xD744 (NOVALAKE_S) [05:31:29] [PASSED] 0xD745 (NOVALAKE_S) [05:31:29] [PASSED] 0x674C (CRESCENTISLAND) [05:31:29] [PASSED] 0xD750 (NOVALAKE_P) [05:31:29] [PASSED] 0xD751 (NOVALAKE_P) [05:31:29] [PASSED] 0xD752 (NOVALAKE_P) [05:31:29] [PASSED] 0xD753 (NOVALAKE_P) [05:31:29] [PASSED] 0xD754 (NOVALAKE_P) [05:31:29] [PASSED] 0xD755 (NOVALAKE_P) [05:31:29] [PASSED] 0xD756 (NOVALAKE_P) [05:31:29] [PASSED] 0xD757 (NOVALAKE_P) [05:31:29] [PASSED] 0xD75F (NOVALAKE_P) [05:31:29] =============== [PASSED] check_platform_desc =============== [05:31:29] ===================== [PASSED] xe_pci ====================== [05:31:29] =================== xe_rtp (2 subtests) ==================== [05:31:29] =============== xe_rtp_process_to_sr_tests ================ [05:31:29] [PASSED] coalesce-same-reg [05:31:29] [PASSED] no-match-no-add [05:31:29] [PASSED] match-or [05:31:29] [PASSED] match-or-xfail [05:31:29] [PASSED] no-match-no-add-multiple-rules [05:31:29] [PASSED] two-regs-two-entries [05:31:29] [PASSED] clr-one-set-other [05:31:29] [PASSED] set-field [05:31:29] [PASSED] conflict-duplicate [05:31:29] [PASSED] conflict-not-disjoint [05:31:29] [PASSED] conflict-reg-type [05:31:29] =========== [PASSED] xe_rtp_process_to_sr_tests ============ [05:31:29] ================== xe_rtp_process_tests =================== [05:31:29] [PASSED] active1 [05:31:29] [PASSED] active2 [05:31:29] [PASSED] active-inactive [05:31:29] [PASSED] inactive-active [05:31:29] [PASSED] inactive-1st_or_active-inactive [05:31:29] [PASSED] inactive-2nd_or_active-inactive [05:31:29] [PASSED] inactive-last_or_active-inactive [05:31:29] [PASSED] inactive-no_or_active-inactive [05:31:29] ============== [PASSED] xe_rtp_process_tests =============== [05:31:29] ===================== [PASSED] xe_rtp ====================== [05:31:29] ==================== xe_wa (1 subtest) ===================== [05:31:29] ======================== xe_wa_gt ========================= [05:31:29] [PASSED] TIGERLAKE B0 [05:31:29] [PASSED] DG1 A0 [05:31:29] [PASSED] DG1 B0 [05:31:29] [PASSED] ALDERLAKE_S A0 [05:31:29] [PASSED] ALDERLAKE_S B0 [05:31:29] [PASSED] ALDERLAKE_S C0 [05:31:29] [PASSED] ALDERLAKE_S D0 [05:31:29] [PASSED] ALDERLAKE_P A0 [05:31:29] [PASSED] ALDERLAKE_P B0 [05:31:29] [PASSED] ALDERLAKE_P C0 [05:31:29] [PASSED] ALDERLAKE_S RPLS D0 [05:31:29] [PASSED] ALDERLAKE_P RPLU E0 [05:31:29] [PASSED] DG2 G10 C0 [05:31:29] [PASSED] DG2 G11 B1 [05:31:29] [PASSED] DG2 G12 A1 [05:31:29] [PASSED] METEORLAKE 12.70(Xe_LPG) A0 13.00(Xe_LPM+) A0 [05:31:29] [PASSED] METEORLAKE 12.71(Xe_LPG) A0 13.00(Xe_LPM+) A0 [05:31:29] [PASSED] METEORLAKE 12.74(Xe_LPG+) A0 13.00(Xe_LPM+) A0 [05:31:29] [PASSED] LUNARLAKE 20.04(Xe2_LPG) A0 20.00(Xe2_LPM) A0 [05:31:29] [PASSED] LUNARLAKE 20.04(Xe2_LPG) B0 20.00(Xe2_LPM) A0 [05:31:29] [PASSED] BATTLEMAGE 20.01(Xe2_HPG) A0 13.01(Xe2_HPM) A1 [05:31:29] [PASSED] PANTHERLAKE 30.00(Xe3_LPG) A0 30.00(Xe3_LPM) A0 [05:31:29] ==================== [PASSED] xe_wa_gt ===================== [05:31:29] ====================== [PASSED] xe_wa ====================== [05:31:29] ============================================================ [05:31:29] Testing complete. Ran 597 tests: passed: 579, skipped: 18 [05:31:30] Elapsed time: 37.624s total, 4.773s configuring, 32.185s building, 0.628s running
+ /kernel/tools/testing/kunit/kunit.py run --kunitconfig /kernel/drivers/gpu/drm/tests/.kunitconfig [05:31:30] Configuring KUnit Kernel ...
Regenerating .config ...
Populating config with:
$ make ARCH=um O=.kunit olddefconfig [05:31:31] Building KUnit Kernel ... Populating config with: $ make ARCH=um O=.kunit olddefconfig Building with: $ make all compile_commands.json scripts_gdb ARCH=um O=.kunit --jobs=48 [05:31:56] Starting KUnit Kernel (1/1)... [05:31:56] ============================================================ Running tests with: $ .kunit/linux kunit.enable=1 mem=1G console=tty kunit_shutdown=halt [05:31:56] ============ drm_test_pick_cmdline (2 subtests) ============ [05:31:56] [PASSED] drm_test_pick_cmdline_res_1920_1080_60 [05:31:56] =============== drm_test_pick_cmdline_named =============== [05:31:56] [PASSED] NTSC [05:31:56] [PASSED] NTSC-J [05:31:56] [PASSED] PAL [05:31:56] [PASSED] PAL-M [05:31:56] =========== [PASSED] drm_test_pick_cmdline_named =========== [05:31:56] ============== [PASSED] drm_test_pick_cmdline ============== [05:31:56] == drm_test_atomic_get_connector_for_encoder (1 subtest) === [05:31:56] [PASSED] drm_test_drm_atomic_get_connector_for_encoder [05:31:56] ==== [PASSED] drm_test_atomic_get_connector_for_encoder ==== [05:31:56] =========== drm_validate_clone_mode (2 subtests) =========== [05:31:56] ============== drm_test_check_in_clone_mode =============== [05:31:56] [PASSED] in_clone_mode [05:31:56] [PASSED] not_in_clone_mode [05:31:56] ========== [PASSED] drm_test_check_in_clone_mode =========== [05:31:56] =============== drm_test_check_valid_clones =============== [05:31:56] [PASSED] not_in_clone_mode [05:31:56] [PASSED] valid_clone [05:31:56] [PASSED] invalid_clone [05:31:56] =========== [PASSED] drm_test_check_valid_clones =========== [05:31:56] ============= [PASSED] drm_validate_clone_mode ============= [05:31:56] ============= drm_validate_modeset (1 subtest) ============= [05:31:56] [PASSED] drm_test_check_connector_changed_modeset [05:31:56] ============== [PASSED] drm_validate_modeset =============== [05:31:56] ====== drm_test_bridge_get_current_state (2 subtests) ====== [05:31:56] [PASSED] drm_test_drm_bridge_get_current_state_atomic [05:31:56] [PASSED] drm_test_drm_bridge_get_current_state_legacy [05:31:56] ======== [PASSED] drm_test_bridge_get_current_state ======== [05:31:56] ====== drm_test_bridge_helper_reset_crtc (3 subtests) ====== [05:31:56] [PASSED] drm_test_drm_bridge_helper_reset_crtc_atomic [05:31:56] [PASSED] drm_test_drm_bridge_helper_reset_crtc_atomic_disabled [05:31:56] [PASSED] drm_test_drm_bridge_helper_reset_crtc_legacy [05:31:56] ======== [PASSED] drm_test_bridge_helper_reset_crtc ======== [05:31:56] ============== drm_bridge_alloc (2 subtests) =============== [05:31:56] [PASSED] drm_test_drm_bridge_alloc_basic [05:31:56] [PASSED] drm_test_drm_bridge_alloc_get_put [05:31:56] ================ [PASSED] drm_bridge_alloc ================= [05:31:56] ============= drm_cmdline_parser (40 subtests) ============= [05:31:56] [PASSED] drm_test_cmdline_force_d_only [05:31:56] [PASSED] drm_test_cmdline_force_D_only_dvi [05:31:56] [PASSED] drm_test_cmdline_force_D_only_hdmi [05:31:56] [PASSED] drm_test_cmdline_force_D_only_not_digital [05:31:56] [PASSED] drm_test_cmdline_force_e_only [05:31:56] [PASSED] drm_test_cmdline_res [05:31:56] [PASSED] drm_test_cmdline_res_vesa [05:31:56] [PASSED] drm_test_cmdline_res_vesa_rblank [05:31:56] [PASSED] drm_test_cmdline_res_rblank [05:31:56] [PASSED] drm_test_cmdline_res_bpp [05:31:56] [PASSED] drm_test_cmdline_res_refresh [05:31:56] [PASSED] drm_test_cmdline_res_bpp_refresh [05:31:56] [PASSED] drm_test_cmdline_res_bpp_refresh_interlaced [05:31:56] [PASSED] drm_test_cmdline_res_bpp_refresh_margins [05:31:56] [PASSED] drm_test_cmdline_res_bpp_refresh_force_off [05:31:56] [PASSED] drm_test_cmdline_res_bpp_refresh_force_on [05:31:56] [PASSED] drm_test_cmdline_res_bpp_refresh_force_on_analog [05:31:56] [PASSED] drm_test_cmdline_res_bpp_refresh_force_on_digital [05:31:56] [PASSED] drm_test_cmdline_res_bpp_refresh_interlaced_margins_force_on [05:31:56] [PASSED] drm_test_cmdline_res_margins_force_on [05:31:56] [PASSED] drm_test_cmdline_res_vesa_margins [05:31:56] [PASSED] drm_test_cmdline_name [05:31:56] [PASSED] drm_test_cmdline_name_bpp [05:31:56] [PASSED] drm_test_cmdline_name_option [05:31:56] [PASSED] drm_test_cmdline_name_bpp_option [05:31:56] [PASSED] drm_test_cmdline_rotate_0 [05:31:56] [PASSED] drm_test_cmdline_rotate_90 [05:31:56] [PASSED] drm_test_cmdline_rotate_180 [05:31:56] [PASSED] drm_test_cmdline_rotate_270 [05:31:56] [PASSED] drm_test_cmdline_hmirror [05:31:56] [PASSED] drm_test_cmdline_vmirror [05:31:56] [PASSED] drm_test_cmdline_margin_options [05:31:56] [PASSED] drm_test_cmdline_multiple_options [05:31:56] [PASSED] drm_test_cmdline_bpp_extra_and_option [05:31:56] [PASSED] drm_test_cmdline_extra_and_option [05:31:56] [PASSED] drm_test_cmdline_freestanding_options [05:31:56] [PASSED] drm_test_cmdline_freestanding_force_e_and_options [05:31:56] [PASSED] drm_test_cmdline_panel_orientation [05:31:56] ================ drm_test_cmdline_invalid ================= [05:31:56] [PASSED] margin_only [05:31:56] [PASSED] interlace_only [05:31:56] [PASSED] res_missing_x [05:31:56] [PASSED] res_missing_y [05:31:56] [PASSED] res_bad_y [05:31:56] [PASSED] res_missing_y_bpp [05:31:56] [PASSED] res_bad_bpp [05:31:56] [PASSED] res_bad_refresh [05:31:56] [PASSED] res_bpp_refresh_force_on_off [05:31:56] [PASSED] res_invalid_mode [05:31:56] [PASSED] res_bpp_wrong_place_mode [05:31:56] [PASSED] name_bpp_refresh [05:31:56] [PASSED] name_refresh [05:31:56] [PASSED] name_refresh_wrong_mode [05:31:56] [PASSED] name_refresh_invalid_mode [05:31:56] [PASSED] rotate_multiple [05:31:56] [PASSED] rotate_invalid_val [05:31:56] [PASSED] rotate_truncated [05:31:56] [PASSED] invalid_option [05:31:56] [PASSED] invalid_tv_option [05:31:56] [PASSED] truncated_tv_option [05:31:56] ============ [PASSED] drm_test_cmdline_invalid ============= [05:31:56] =============== drm_test_cmdline_tv_options =============== [05:31:56] [PASSED] NTSC [05:31:56] [PASSED] NTSC_443 [05:31:56] [PASSED] NTSC_J [05:31:56] [PASSED] PAL [05:31:56] [PASSED] PAL_M [05:31:56] [PASSED] PAL_N [05:31:56] [PASSED] SECAM [05:31:56] [PASSED] MONO_525 [05:31:56] [PASSED] MONO_625 [05:31:56] =========== [PASSED] drm_test_cmdline_tv_options =========== [05:31:56] =============== [PASSED] drm_cmdline_parser ================ [05:31:56] ========== drmm_connector_hdmi_init (20 subtests) ========== [05:31:56] [PASSED] drm_test_connector_hdmi_init_valid [05:31:56] [PASSED] drm_test_connector_hdmi_init_bpc_8 [05:31:56] [PASSED] drm_test_connector_hdmi_init_bpc_10 [05:31:56] [PASSED] drm_test_connector_hdmi_init_bpc_12 [05:31:56] [PASSED] drm_test_connector_hdmi_init_bpc_invalid [05:31:56] [PASSED] drm_test_connector_hdmi_init_bpc_null [05:31:56] [PASSED] drm_test_connector_hdmi_init_formats_empty [05:31:56] [PASSED] drm_test_connector_hdmi_init_formats_no_rgb [05:31:56] === drm_test_connector_hdmi_init_formats_yuv420_allowed === [05:31:56] [PASSED] supported_formats=0x9 yuv420_allowed=1 [05:31:56] [PASSED] supported_formats=0x9 yuv420_allowed=0 [05:31:56] [PASSED] supported_formats=0x5 yuv420_allowed=1 [05:31:56] [PASSED] supported_formats=0x5 yuv420_allowed=0 [05:31:56] === [PASSED] drm_test_connector_hdmi_init_formats_yuv420_allowed === [05:31:56] [PASSED] drm_test_connector_hdmi_init_null_ddc [05:31:56] [PASSED] drm_test_connector_hdmi_init_null_product [05:31:56] [PASSED] drm_test_connector_hdmi_init_null_vendor [05:31:56] [PASSED] drm_test_connector_hdmi_init_product_length_exact [05:31:56] [PASSED] drm_test_connector_hdmi_init_product_length_too_long [05:31:56] [PASSED] drm_test_connector_hdmi_init_product_valid [05:31:56] [PASSED] drm_test_connector_hdmi_init_vendor_length_exact [05:31:56] [PASSED] drm_test_connector_hdmi_init_vendor_length_too_long [05:31:56] [PASSED] drm_test_connector_hdmi_init_vendor_valid [05:31:56] ========= drm_test_connector_hdmi_init_type_valid ========= [05:31:56] [PASSED] HDMI-A [05:31:56] [PASSED] HDMI-B [05:31:56] ===== [PASSED] drm_test_connector_hdmi_init_type_valid ===== [05:31:56] ======== drm_test_connector_hdmi_init_type_invalid ======== [05:31:56] [PASSED] Unknown [05:31:56] [PASSED] VGA [05:31:56] [PASSED] DVI-I [05:31:56] [PASSED] DVI-D [05:31:56] [PASSED] DVI-A [05:31:56] [PASSED] Composite [05:31:56] [PASSED] SVIDEO [05:31:56] [PASSED] LVDS [05:31:56] [PASSED] Component [05:31:56] [PASSED] DIN [05:31:56] [PASSED] DP [05:31:56] [PASSED] TV [05:31:56] [PASSED] eDP [05:31:56] [PASSED] Virtual [05:31:56] [PASSED] DSI [05:31:56] [PASSED] DPI [05:31:56] [PASSED] Writeback [05:31:56] [PASSED] SPI [05:31:56] [PASSED] USB [05:31:56] ==== [PASSED] drm_test_connector_hdmi_init_type_invalid ==== [05:31:56] ============ [PASSED] drmm_connector_hdmi_init ============= [05:31:56] ============= drmm_connector_init (3 subtests) ============= [05:31:56] [PASSED] drm_test_drmm_connector_init [05:31:56] [PASSED] drm_test_drmm_connector_init_null_ddc [05:31:56] ========= drm_test_drmm_connector_init_type_valid ========= [05:31:56] [PASSED] Unknown [05:31:56] [PASSED] VGA [05:31:56] [PASSED] DVI-I [05:31:56] [PASSED] DVI-D [05:31:56] [PASSED] DVI-A [05:31:56] [PASSED] Composite [05:31:56] [PASSED] SVIDEO [05:31:56] [PASSED] LVDS [05:31:56] [PASSED] Component [05:31:56] [PASSED] DIN [05:31:56] [PASSED] DP [05:31:56] [PASSED] HDMI-A [05:31:56] [PASSED] HDMI-B [05:31:56] [PASSED] TV [05:31:56] [PASSED] eDP [05:31:56] [PASSED] Virtual [05:31:56] [PASSED] DSI [05:31:56] [PASSED] DPI [05:31:56] [PASSED] Writeback [05:31:56] [PASSED] SPI [05:31:56] [PASSED] USB [05:31:56] ===== [PASSED] drm_test_drmm_connector_init_type_valid ===== [05:31:56] =============== [PASSED] drmm_connector_init =============== [05:31:56] ========= drm_connector_dynamic_init (6 subtests) ========== [05:31:56] [PASSED] drm_test_drm_connector_dynamic_init [05:31:56] [PASSED] drm_test_drm_connector_dynamic_init_null_ddc [05:31:56] [PASSED] drm_test_drm_connector_dynamic_init_not_added [05:31:56] [PASSED] drm_test_drm_connector_dynamic_init_properties [05:31:56] ===== drm_test_drm_connector_dynamic_init_type_valid ====== [05:31:56] [PASSED] Unknown [05:31:56] [PASSED] VGA [05:31:56] [PASSED] DVI-I [05:31:56] [PASSED] DVI-D [05:31:56] [PASSED] DVI-A [05:31:56] [PASSED] Composite [05:31:56] [PASSED] SVIDEO [05:31:56] [PASSED] LVDS [05:31:56] [PASSED] Component [05:31:56] [PASSED] DIN [05:31:56] [PASSED] DP [05:31:56] [PASSED] HDMI-A [05:31:56] [PASSED] HDMI-B [05:31:56] [PASSED] TV [05:31:56] [PASSED] eDP [05:31:56] [PASSED] Virtual [05:31:56] [PASSED] DSI [05:31:56] [PASSED] DPI [05:31:56] [PASSED] Writeback [05:31:56] [PASSED] SPI [05:31:56] [PASSED] USB [05:31:56] = [PASSED] drm_test_drm_connector_dynamic_init_type_valid == [05:31:56] ======== drm_test_drm_connector_dynamic_init_name ========= [05:31:56] [PASSED] Unknown [05:31:56] [PASSED] VGA [05:31:56] [PASSED] DVI-I [05:31:56] [PASSED] DVI-D [05:31:56] [PASSED] DVI-A [05:31:56] [PASSED] Composite [05:31:56] [PASSED] SVIDEO [05:31:56] [PASSED] LVDS [05:31:56] [PASSED] Component [05:31:56] [PASSED] DIN [05:31:56] [PASSED] DP [05:31:56] [PASSED] HDMI-A [05:31:56] [PASSED] HDMI-B [05:31:56] [PASSED] TV [05:31:56] [PASSED] eDP [05:31:56] [PASSED] Virtual [05:31:56] [PASSED] DSI [05:31:56] [PASSED] DPI [05:31:56] [PASSED] Writeback [05:31:56] [PASSED] SPI [05:31:56] [PASSED] USB [05:31:56] ==== [PASSED] drm_test_drm_connector_dynamic_init_name ===== [05:31:56] =========== [PASSED] drm_connector_dynamic_init ============ [05:31:56] ==== drm_connector_dynamic_register_early (4 subtests) ===== [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_early_on_list [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_early_defer [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_early_no_init [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_early_no_mode_object [05:31:56] ====== [PASSED] drm_connector_dynamic_register_early ======= [05:31:56] ======= drm_connector_dynamic_register (7 subtests) ======== [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_on_list [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_no_defer [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_no_init [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_mode_object [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_sysfs [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_sysfs_name [05:31:56] [PASSED] drm_test_drm_connector_dynamic_register_debugfs [05:31:56] ========= [PASSED] drm_connector_dynamic_register ========== [05:31:56] = drm_connector_attach_broadcast_rgb_property (2 subtests) = [05:31:56] [PASSED] drm_test_drm_connector_attach_broadcast_rgb_property [05:31:56] [PASSED] drm_test_drm_connector_attach_broadcast_rgb_property_hdmi_connector [05:31:56] === [PASSED] drm_connector_attach_broadcast_rgb_property === [05:31:56] ========== drm_get_tv_mode_from_name (2 subtests) ========== [05:31:56] ========== drm_test_get_tv_mode_from_name_valid =========== [05:31:56] [PASSED] NTSC [05:31:56] [PASSED] NTSC-443 [05:31:56] [PASSED] NTSC-J [05:31:56] [PASSED] PAL [05:31:56] [PASSED] PAL-M [05:31:56] [PASSED] PAL-N [05:31:56] [PASSED] SECAM [05:31:56] [PASSED] Mono [05:31:56] ====== [PASSED] drm_test_get_tv_mode_from_name_valid ======= [05:31:56] [PASSED] drm_test_get_tv_mode_from_name_truncated [05:31:56] ============ [PASSED] drm_get_tv_mode_from_name ============ [05:31:56] = drm_test_connector_hdmi_compute_mode_clock (12 subtests) = [05:31:56] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb [05:31:56] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_10bpc [05:31:56] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_10bpc_vic_1 [05:31:56] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_12bpc [05:31:56] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_12bpc_vic_1 [05:31:56] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_double [05:31:56] = drm_test_connector_hdmi_compute_mode_clock_yuv420_valid = [05:31:56] [PASSED] VIC 96 [05:31:56] [PASSED] VIC 97 [05:31:56] [PASSED] VIC 101 [05:31:56] [PASSED] VIC 102 [05:31:56] [PASSED] VIC 106 [05:31:56] [PASSED] VIC 107 [05:31:56] === [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv420_valid === [05:31:56] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv420_10_bpc [05:31:56] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv420_12_bpc [05:31:56] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv422_8_bpc [05:31:56] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv422_10_bpc [05:31:56] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv422_12_bpc [05:31:56] === [PASSED] drm_test_connector_hdmi_compute_mode_clock ==== [05:31:56] == drm_hdmi_connector_get_broadcast_rgb_name (2 subtests) == [05:31:56] === drm_test_drm_hdmi_connector_get_broadcast_rgb_name ==== [05:31:56] [PASSED] Automatic [05:31:56] [PASSED] Full [05:31:56] [PASSED] Limited 16:235 [05:31:56] === [PASSED] drm_test_drm_hdmi_connector_get_broadcast_rgb_name === [05:31:56] [PASSED] drm_test_drm_hdmi_connector_get_broadcast_rgb_name_invalid [05:31:56] ==== [PASSED] drm_hdmi_connector_get_broadcast_rgb_name ==== [05:31:56] == drm_hdmi_connector_get_output_format_name (2 subtests) == [05:31:56] === drm_test_drm_hdmi_connector_get_output_format_name ==== [05:31:56] [PASSED] RGB [05:31:56] [PASSED] YUV 4:2:0 [05:31:56] [PASSED] YUV 4:2:2 [05:31:56] [PASSED] YUV 4:4:4 [05:31:56] === [PASSED] drm_test_drm_hdmi_connector_get_output_format_name === [05:31:56] [PASSED] drm_test_drm_hdmi_connector_get_output_format_name_invalid [05:31:56] ==== [PASSED] drm_hdmi_connector_get_output_format_name ==== [05:31:56] ============= drm_damage_helper (21 subtests) ============== [05:31:56] [PASSED] drm_test_damage_iter_no_damage [05:31:56] [PASSED] drm_test_damage_iter_no_damage_fractional_src [05:31:56] [PASSED] drm_test_damage_iter_no_damage_src_moved [05:31:56] [PASSED] drm_test_damage_iter_no_damage_fractional_src_moved [05:31:56] [PASSED] drm_test_damage_iter_no_damage_not_visible [05:31:56] [PASSED] drm_test_damage_iter_no_damage_no_crtc [05:31:56] [PASSED] drm_test_damage_iter_no_damage_no_fb [05:31:56] [PASSED] drm_test_damage_iter_simple_damage [05:31:56] [PASSED] drm_test_damage_iter_single_damage [05:31:56] [PASSED] drm_test_damage_iter_single_damage_intersect_src [05:31:56] [PASSED] drm_test_damage_iter_single_damage_outside_src [05:31:56] [PASSED] drm_test_damage_iter_single_damage_fractional_src [05:31:56] [PASSED] drm_test_damage_iter_single_damage_intersect_fractional_src [05:31:56] [PASSED] drm_test_damage_iter_single_damage_outside_fractional_src [05:31:56] [PASSED] drm_test_damage_iter_single_damage_src_moved [05:31:56] [PASSED] drm_test_damage_iter_single_damage_fractional_src_moved [05:31:56] [PASSED] drm_test_damage_iter_damage [05:31:56] [PASSED] drm_test_damage_iter_damage_one_intersect [05:31:56] [PASSED] drm_test_damage_iter_damage_one_outside [05:31:56] [PASSED] drm_test_damage_iter_damage_src_moved [05:31:56] [PASSED] drm_test_damage_iter_damage_not_visible [05:31:56] ================ [PASSED] drm_damage_helper ================ [05:31:56] ============== drm_dp_mst_helper (3 subtests) ============== [05:31:56] ============== drm_test_dp_mst_calc_pbn_mode ============== [05:31:56] [PASSED] Clock 154000 BPP 30 DSC disabled [05:31:56] [PASSED] Clock 234000 BPP 30 DSC disabled [05:31:56] [PASSED] Clock 297000 BPP 24 DSC disabled [05:31:56] [PASSED] Clock 332880 BPP 24 DSC enabled [05:31:56] [PASSED] Clock 324540 BPP 24 DSC enabled [05:31:56] ========== [PASSED] drm_test_dp_mst_calc_pbn_mode ========== [05:31:56] ============== drm_test_dp_mst_calc_pbn_div =============== [05:31:56] [PASSED] Link rate 2000000 lane count 4 [05:31:56] [PASSED] Link rate 2000000 lane count 2 [05:31:56] [PASSED] Link rate 2000000 lane count 1 [05:31:56] [PASSED] Link rate 1350000 lane count 4 [05:31:56] [PASSED] Link rate 1350000 lane count 2 [05:31:56] [PASSED] Link rate 1350000 lane count 1 [05:31:56] [PASSED] Link rate 1000000 lane count 4 [05:31:56] [PASSED] Link rate 1000000 lane count 2 [05:31:56] [PASSED] Link rate 1000000 lane count 1 [05:31:56] [PASSED] Link rate 810000 lane count 4 [05:31:56] [PASSED] Link rate 810000 lane count 2 [05:31:56] [PASSED] Link rate 810000 lane count 1 [05:31:56] [PASSED] Link rate 540000 lane count 4 [05:31:56] [PASSED] Link rate 540000 lane count 2 [05:31:56] [PASSED] Link rate 540000 lane count 1 [05:31:56] [PASSED] Link rate 270000 lane count 4 [05:31:56] [PASSED] Link rate 270000 lane count 2 [05:31:56] [PASSED] Link rate 270000 lane count 1 [05:31:56] [PASSED] Link rate 162000 lane count 4 [05:31:56] [PASSED] Link rate 162000 lane count 2 [05:31:56] [PASSED] Link rate 162000 lane count 1 [05:31:56] ========== [PASSED] drm_test_dp_mst_calc_pbn_div =========== [05:31:56] ========= drm_test_dp_mst_sideband_msg_req_decode ========= [05:31:56] [PASSED] DP_ENUM_PATH_RESOURCES with port number [05:31:56] [PASSED] DP_POWER_UP_PHY with port number [05:31:56] [PASSED] DP_POWER_DOWN_PHY with port number [05:31:56] [PASSED] DP_ALLOCATE_PAYLOAD with SDP stream sinks [05:31:56] [PASSED] DP_ALLOCATE_PAYLOAD with port number [05:31:56] [PASSED] DP_ALLOCATE_PAYLOAD with VCPI [05:31:56] [PASSED] DP_ALLOCATE_PAYLOAD with PBN [05:31:56] [PASSED] DP_QUERY_PAYLOAD with port number [05:31:56] [PASSED] DP_QUERY_PAYLOAD with VCPI [05:31:56] [PASSED] DP_REMOTE_DPCD_READ with port number [05:31:56] [PASSED] DP_REMOTE_DPCD_READ with DPCD address [05:31:56] [PASSED] DP_REMOTE_DPCD_READ with max number of bytes [05:31:56] [PASSED] DP_REMOTE_DPCD_WRITE with port number [05:31:56] [PASSED] DP_REMOTE_DPCD_WRITE with DPCD address [05:31:56] [PASSED] DP_REMOTE_DPCD_WRITE with data array [05:31:56] [PASSED] DP_REMOTE_I2C_READ with port number [05:31:56] [PASSED] DP_REMOTE_I2C_READ with I2C device ID [05:31:56] [PASSED] DP_REMOTE_I2C_READ with transactions array [05:31:56] [PASSED] DP_REMOTE_I2C_WRITE with port number [05:31:56] [PASSED] DP_REMOTE_I2C_WRITE with I2C device ID [05:31:56] [PASSED] DP_REMOTE_I2C_WRITE with data array [05:31:56] [PASSED] DP_QUERY_STREAM_ENC_STATUS with stream ID [05:31:56] [PASSED] DP_QUERY_STREAM_ENC_STATUS with client ID [05:31:56] [PASSED] DP_QUERY_STREAM_ENC_STATUS with stream event [05:31:56] [PASSED] DP_QUERY_STREAM_ENC_STATUS with valid stream event [05:31:56] [PASSED] DP_QUERY_STREAM_ENC_STATUS with stream behavior [05:31:56] [PASSED] DP_QUERY_STREAM_ENC_STATUS with a valid stream behavior [05:31:56] ===== [PASSED] drm_test_dp_mst_sideband_msg_req_decode ===== [05:31:56] ================ [PASSED] drm_dp_mst_helper ================ [05:31:56] ================== drm_exec (7 subtests) =================== [05:31:56] [PASSED] sanitycheck [05:31:56] [PASSED] test_lock [05:31:56] [PASSED] test_lock_unlock [05:31:56] [PASSED] test_duplicates [05:31:56] [PASSED] test_prepare [05:31:56] [PASSED] test_prepare_array [05:31:56] [PASSED] test_multiple_loops [05:31:56] ==================== [PASSED] drm_exec ===================== [05:31:56] =========== drm_format_helper_test (17 subtests) =========== [05:31:56] ============== drm_test_fb_xrgb8888_to_gray8 ============== [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ========== [PASSED] drm_test_fb_xrgb8888_to_gray8 ========== [05:31:56] ============= drm_test_fb_xrgb8888_to_rgb332 ============== [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ========= [PASSED] drm_test_fb_xrgb8888_to_rgb332 ========== [05:31:56] ============= drm_test_fb_xrgb8888_to_rgb565 ============== [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ========= [PASSED] drm_test_fb_xrgb8888_to_rgb565 ========== [05:31:56] ============ drm_test_fb_xrgb8888_to_xrgb1555 ============= [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ======== [PASSED] drm_test_fb_xrgb8888_to_xrgb1555 ========= [05:31:56] ============ drm_test_fb_xrgb8888_to_argb1555 ============= [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ======== [PASSED] drm_test_fb_xrgb8888_to_argb1555 ========= [05:31:56] ============ drm_test_fb_xrgb8888_to_rgba5551 ============= [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ======== [PASSED] drm_test_fb_xrgb8888_to_rgba5551 ========= [05:31:56] ============= drm_test_fb_xrgb8888_to_rgb888 ============== [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ========= [PASSED] drm_test_fb_xrgb8888_to_rgb888 ========== [05:31:56] ============= drm_test_fb_xrgb8888_to_bgr888 ============== [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ========= [PASSED] drm_test_fb_xrgb8888_to_bgr888 ========== [05:31:56] ============ drm_test_fb_xrgb8888_to_argb8888 ============= [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ======== [PASSED] drm_test_fb_xrgb8888_to_argb8888 ========= [05:31:56] =========== drm_test_fb_xrgb8888_to_xrgb2101010 =========== [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ======= [PASSED] drm_test_fb_xrgb8888_to_xrgb2101010 ======= [05:31:56] =========== drm_test_fb_xrgb8888_to_argb2101010 =========== [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ======= [PASSED] drm_test_fb_xrgb8888_to_argb2101010 ======= [05:31:56] ============== drm_test_fb_xrgb8888_to_mono =============== [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ========== [PASSED] drm_test_fb_xrgb8888_to_mono =========== [05:31:56] ==================== drm_test_fb_swab ===================== [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ================ [PASSED] drm_test_fb_swab ================= [05:31:56] ============ drm_test_fb_xrgb8888_to_xbgr8888 ============= [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ======== [PASSED] drm_test_fb_xrgb8888_to_xbgr8888 ========= [05:31:56] ============ drm_test_fb_xrgb8888_to_abgr8888 ============= [05:31:56] [PASSED] single_pixel_source_buffer [05:31:56] [PASSED] single_pixel_clip_rectangle [05:31:56] [PASSED] well_known_colors [05:31:56] [PASSED] destination_pitch [05:31:56] ======== [PASSED] drm_test_fb_xrgb8888_to_abgr8888 ========= [05:31:56] ================= drm_test_fb_clip_offset ================= [05:31:56] [PASSED] pass through [05:31:56] [PASSED] horizontal offset [05:31:56] [PASSED] vertical offset [05:31:56] [PASSED] horizontal and vertical offset [05:31:56] [PASSED] horizontal offset (custom pitch) [05:31:56] [PASSED] vertical offset (custom pitch) [05:31:56] [PASSED] horizontal and vertical offset (custom pitch) [05:31:56] ============= [PASSED] drm_test_fb_clip_offset ============= [05:31:56] =================== drm_test_fb_memcpy ==================== [05:31:56] [PASSED] single_pixel_source_buffer: XR24 little-endian (0x34325258) [05:31:56] [PASSED] single_pixel_source_buffer: XRA8 little-endian (0x38415258) [05:31:56] [PASSED] single_pixel_source_buffer: YU24 little-endian (0x34325559) [05:31:56] [PASSED] single_pixel_clip_rectangle: XB24 little-endian (0x34324258) [05:31:56] [PASSED] single_pixel_clip_rectangle: XRA8 little-endian (0x38415258) [05:31:56] [PASSED] single_pixel_clip_rectangle: YU24 little-endian (0x34325559) [05:31:56] [PASSED] well_known_colors: XB24 little-endian (0x34324258) [05:31:56] [PASSED] well_known_colors: XRA8 little-endian (0x38415258) [05:31:56] [PASSED] well_known_colors: YU24 little-endian (0x34325559) [05:31:56] [PASSED] destination_pitch: XB24 little-endian (0x34324258) [05:31:56] [PASSED] destination_pitch: XRA8 little-endian (0x38415258) [05:31:56] [PASSED] destination_pitch: YU24 little-endian (0x34325559) [05:31:56] =============== [PASSED] drm_test_fb_memcpy ================ [05:31:56] ============= [PASSED] drm_format_helper_test ============== [05:31:56] ================= drm_format (18 subtests) ================= [05:31:56] [PASSED] drm_test_format_block_width_invalid [05:31:56] [PASSED] drm_test_format_block_width_one_plane [05:31:56] [PASSED] drm_test_format_block_width_two_plane [05:31:56] [PASSED] drm_test_format_block_width_three_plane [05:31:56] [PASSED] drm_test_format_block_width_tiled [05:31:56] [PASSED] drm_test_format_block_height_invalid [05:31:56] [PASSED] drm_test_format_block_height_one_plane [05:31:56] [PASSED] drm_test_format_block_height_two_plane [05:31:56] [PASSED] drm_test_format_block_height_three_plane [05:31:56] [PASSED] drm_test_format_block_height_tiled [05:31:56] [PASSED] drm_test_format_min_pitch_invalid [05:31:56] [PASSED] drm_test_format_min_pitch_one_plane_8bpp [05:31:56] [PASSED] drm_test_format_min_pitch_one_plane_16bpp [05:31:56] [PASSED] drm_test_format_min_pitch_one_plane_24bpp [05:31:56] [PASSED] drm_test_format_min_pitch_one_plane_32bpp [05:31:56] [PASSED] drm_test_format_min_pitch_two_plane [05:31:56] [PASSED] drm_test_format_min_pitch_three_plane_8bpp [05:31:56] [PASSED] drm_test_format_min_pitch_tiled [05:31:56] =================== [PASSED] drm_format ==================== [05:31:56] ============== drm_framebuffer (10 subtests) =============== [05:31:56] ========== drm_test_framebuffer_check_src_coords ========== [05:31:56] [PASSED] Success: source fits into fb [05:31:56] [PASSED] Fail: overflowing fb with x-axis coordinate [05:31:56] [PASSED] Fail: overflowing fb with y-axis coordinate [05:31:56] [PASSED] Fail: overflowing fb with source width [05:31:56] [PASSED] Fail: overflowing fb with source height [05:31:56] ====== [PASSED] drm_test_framebuffer_check_src_coords ====== [05:31:56] [PASSED] drm_test_framebuffer_cleanup [05:31:56] =============== drm_test_framebuffer_create =============== [05:31:56] [PASSED] ABGR8888 normal sizes [05:31:56] [PASSED] ABGR8888 max sizes [05:31:56] [PASSED] ABGR8888 pitch greater than min required [05:31:56] [PASSED] ABGR8888 pitch less than min required [05:31:56] [PASSED] ABGR8888 Invalid width [05:31:56] [PASSED] ABGR8888 Invalid buffer handle [05:31:56] [PASSED] No pixel format [05:31:56] [PASSED] ABGR8888 Width 0 [05:31:56] [PASSED] ABGR8888 Height 0 [05:31:56] [PASSED] ABGR8888 Out of bound height * pitch combination [05:31:56] [PASSED] ABGR8888 Large buffer offset [05:31:56] [PASSED] ABGR8888 Buffer offset for inexistent plane [05:31:56] [PASSED] ABGR8888 Invalid flag [05:31:56] [PASSED] ABGR8888 Set DRM_MODE_FB_MODIFIERS without modifiers [05:31:56] [PASSED] ABGR8888 Valid buffer modifier [05:31:56] [PASSED] ABGR8888 Invalid buffer modifier(DRM_FORMAT_MOD_SAMSUNG_64_32_TILE) [05:31:56] [PASSED] ABGR8888 Extra pitches without DRM_MODE_FB_MODIFIERS [05:31:56] [PASSED] ABGR8888 Extra pitches with DRM_MODE_FB_MODIFIERS [05:31:56] [PASSED] NV12 Normal sizes [05:31:56] [PASSED] NV12 Max sizes [05:31:56] [PASSED] NV12 Invalid pitch [05:31:56] [PASSED] NV12 Invalid modifier/missing DRM_MODE_FB_MODIFIERS flag [05:31:56] [PASSED] NV12 different modifier per-plane [05:31:56] [PASSED] NV12 with DRM_FORMAT_MOD_SAMSUNG_64_32_TILE [05:31:56] [PASSED] NV12 Valid modifiers without DRM_MODE_FB_MODIFIERS [05:31:56] [PASSED] NV12 Modifier for inexistent plane [05:31:56] [PASSED] NV12 Handle for inexistent plane [05:31:56] [PASSED] NV12 Handle for inexistent plane without DRM_MODE_FB_MODIFIERS [05:31:56] [PASSED] YVU420 DRM_MODE_FB_MODIFIERS set without modifier [05:31:56] [PASSED] YVU420 Normal sizes [05:31:56] [PASSED] YVU420 Max sizes [05:31:56] [PASSED] YVU420 Invalid pitch [05:31:56] [PASSED] YVU420 Different pitches [05:31:56] [PASSED] YVU420 Different buffer offsets/pitches [05:31:56] [PASSED] YVU420 Modifier set just for plane 0, without DRM_MODE_FB_MODIFIERS [05:31:56] [PASSED] YVU420 Modifier set just for planes 0, 1, without DRM_MODE_FB_MODIFIERS [05:31:56] [PASSED] YVU420 Modifier set just for plane 0, 1, with DRM_MODE_FB_MODIFIERS [05:31:56] [PASSED] YVU420 Valid modifier [05:31:56] [PASSED] YVU420 Different modifiers per plane [05:31:56] [PASSED] YVU420 Modifier for inexistent plane [05:31:56] [PASSED] YUV420_10BIT Invalid modifier(DRM_FORMAT_MOD_LINEAR) [05:31:56] [PASSED] X0L2 Normal sizes [05:31:56] [PASSED] X0L2 Max sizes [05:31:56] [PASSED] X0L2 Invalid pitch [05:31:56] [PASSED] X0L2 Pitch greater than minimum required [05:31:56] [PASSED] X0L2 Handle for inexistent plane [05:31:56] [PASSED] X0L2 Offset for inexistent plane, without DRM_MODE_FB_MODIFIERS set [05:31:56] [PASSED] X0L2 Modifier without DRM_MODE_FB_MODIFIERS set [05:31:56] [PASSED] X0L2 Valid modifier [05:31:56] [PASSED] X0L2 Modifier for inexistent plane [05:31:56] =========== [PASSED] drm_test_framebuffer_create =========== [05:31:56] [PASSED] drm_test_framebuffer_free [05:31:56] [PASSED] drm_test_framebuffer_init [05:31:56] [PASSED] drm_test_framebuffer_init_bad_format [05:31:56] [PASSED] drm_test_framebuffer_init_dev_mismatch [05:31:56] [PASSED] drm_test_framebuffer_lookup [05:31:56] [PASSED] drm_test_framebuffer_lookup_inexistent [05:31:56] [PASSED] drm_test_framebuffer_modifiers_not_supported [05:31:56] ================= [PASSED] drm_framebuffer ================= [05:31:56] ================ drm_gem_shmem (8 subtests) ================ [05:31:56] [PASSED] drm_gem_shmem_test_obj_create [05:31:56] [PASSED] drm_gem_shmem_test_obj_create_private [05:31:56] [PASSED] drm_gem_shmem_test_pin_pages [05:31:56] [PASSED] drm_gem_shmem_test_vmap [05:31:56] [PASSED] drm_gem_shmem_test_get_sg_table [05:31:56] [PASSED] drm_gem_shmem_test_get_pages_sgt [05:31:56] [PASSED] drm_gem_shmem_test_madvise [05:31:56] [PASSED] drm_gem_shmem_test_purge [05:31:56] ================== [PASSED] drm_gem_shmem ================== [05:31:56] === drm_atomic_helper_connector_hdmi_check (27 subtests) === [05:31:56] [PASSED] drm_test_check_broadcast_rgb_auto_cea_mode [05:31:56] [PASSED] drm_test_check_broadcast_rgb_auto_cea_mode_vic_1 [05:31:56] [PASSED] drm_test_check_broadcast_rgb_full_cea_mode [05:31:56] [PASSED] drm_test_check_broadcast_rgb_full_cea_mode_vic_1 [05:31:56] [PASSED] drm_test_check_broadcast_rgb_limited_cea_mode [05:31:56] [PASSED] drm_test_check_broadcast_rgb_limited_cea_mode_vic_1 [05:31:56] ====== drm_test_check_broadcast_rgb_cea_mode_yuv420 ======= [05:31:56] [PASSED] Automatic [05:31:56] [PASSED] Full [05:31:56] [PASSED] Limited 16:235 [05:31:56] == [PASSED] drm_test_check_broadcast_rgb_cea_mode_yuv420 === [05:31:56] [PASSED] drm_test_check_broadcast_rgb_crtc_mode_changed [05:31:56] [PASSED] drm_test_check_broadcast_rgb_crtc_mode_not_changed [05:31:56] [PASSED] drm_test_check_disable_connector [05:31:56] [PASSED] drm_test_check_hdmi_funcs_reject_rate [05:31:56] [PASSED] drm_test_check_max_tmds_rate_bpc_fallback_rgb [05:31:56] [PASSED] drm_test_check_max_tmds_rate_bpc_fallback_yuv420 [05:31:56] [PASSED] drm_test_check_max_tmds_rate_bpc_fallback_ignore_yuv422 [05:31:56] [PASSED] drm_test_check_max_tmds_rate_bpc_fallback_ignore_yuv420 [05:31:56] [PASSED] drm_test_check_driver_unsupported_fallback_yuv420 [05:31:56] [PASSED] drm_test_check_output_bpc_crtc_mode_changed [05:31:56] [PASSED] drm_test_check_output_bpc_crtc_mode_not_changed [05:31:56] [PASSED] drm_test_check_output_bpc_dvi [05:31:56] [PASSED] drm_test_check_output_bpc_format_vic_1 [05:31:56] [PASSED] drm_test_check_output_bpc_format_display_8bpc_only [05:31:56] [PASSED] drm_test_check_output_bpc_format_display_rgb_only [05:31:56] [PASSED] drm_test_check_output_bpc_format_driver_8bpc_only [05:31:56] [PASSED] drm_test_check_output_bpc_format_driver_rgb_only [05:31:56] [PASSED] drm_test_check_tmds_char_rate_rgb_8bpc [05:31:56] [PASSED] drm_test_check_tmds_char_rate_rgb_10bpc [05:31:56] [PASSED] drm_test_check_tmds_char_rate_rgb_12bpc [05:31:56] ===== [PASSED] drm_atomic_helper_connector_hdmi_check ====== [05:31:56] === drm_atomic_helper_connector_hdmi_reset (6 subtests) ==== [05:31:56] [PASSED] drm_test_check_broadcast_rgb_value [05:31:56] [PASSED] drm_test_check_bpc_8_value [05:31:56] [PASSED] drm_test_check_bpc_10_value [05:31:56] [PASSED] drm_test_check_bpc_12_value [05:31:56] [PASSED] drm_test_check_format_value [05:31:56] [PASSED] drm_test_check_tmds_char_value [05:31:56] ===== [PASSED] drm_atomic_helper_connector_hdmi_reset ====== [05:31:56] = drm_atomic_helper_connector_hdmi_mode_valid (4 subtests) = [05:31:56] [PASSED] drm_test_check_mode_valid [05:31:56] [PASSED] drm_test_check_mode_valid_reject [05:31:56] [PASSED] drm_test_check_mode_valid_reject_rate [05:31:56] [PASSED] drm_test_check_mode_valid_reject_max_clock [05:31:56] === [PASSED] drm_atomic_helper_connector_hdmi_mode_valid === [05:31:56] = drm_atomic_helper_connector_hdmi_infoframes (5 subtests) = [05:31:56] [PASSED] drm_test_check_infoframes [05:31:56] [PASSED] drm_test_check_reject_avi_infoframe [05:31:56] [PASSED] drm_test_check_reject_hdr_infoframe_bpc_8 [05:31:56] [PASSED] drm_test_check_reject_hdr_infoframe_bpc_10 [05:31:56] [PASSED] drm_test_check_reject_audio_infoframe [05:31:56] === [PASSED] drm_atomic_helper_connector_hdmi_infoframes === [05:31:56] ================= drm_managed (2 subtests) ================= [05:31:56] [PASSED] drm_test_managed_release_action [05:31:56] [PASSED] drm_test_managed_run_action [05:31:56] =================== [PASSED] drm_managed =================== [05:31:56] =================== drm_mm (6 subtests) ==================== [05:31:56] [PASSED] drm_test_mm_init [05:31:56] [PASSED] drm_test_mm_debug [05:31:56] [PASSED] drm_test_mm_align32 [05:31:56] [PASSED] drm_test_mm_align64 [05:31:56] [PASSED] drm_test_mm_lowest [05:31:56] [PASSED] drm_test_mm_highest [05:31:56] ===================== [PASSED] drm_mm ====================== [05:31:56] ============= drm_modes_analog_tv (5 subtests) ============= [05:31:56] [PASSED] drm_test_modes_analog_tv_mono_576i [05:31:56] [PASSED] drm_test_modes_analog_tv_ntsc_480i [05:31:56] [PASSED] drm_test_modes_analog_tv_ntsc_480i_inlined [05:31:56] [PASSED] drm_test_modes_analog_tv_pal_576i [05:31:56] [PASSED] drm_test_modes_analog_tv_pal_576i_inlined [05:31:56] =============== [PASSED] drm_modes_analog_tv =============== [05:31:56] ============== drm_plane_helper (2 subtests) =============== [05:31:56] =============== drm_test_check_plane_state ================ [05:31:56] [PASSED] clipping_simple [05:31:56] [PASSED] clipping_rotate_reflect [05:31:56] [PASSED] positioning_simple [05:31:56] [PASSED] upscaling [05:31:56] [PASSED] downscaling [05:31:56] [PASSED] rounding1 [05:31:56] [PASSED] rounding2 [05:31:56] [PASSED] rounding3 [05:31:56] [PASSED] rounding4 [05:31:56] =========== [PASSED] drm_test_check_plane_state ============ [05:31:56] =========== drm_test_check_invalid_plane_state ============ [05:31:56] [PASSED] positioning_invalid [05:31:56] [PASSED] upscaling_invalid [05:31:56] [PASSED] downscaling_invalid [05:31:56] ======= [PASSED] drm_test_check_invalid_plane_state ======== [05:31:56] ================ [PASSED] drm_plane_helper ================= [05:31:56] ====== drm_connector_helper_tv_get_modes (1 subtest) ======= [05:31:56] ====== drm_test_connector_helper_tv_get_modes_check ======= [05:31:56] [PASSED] None [05:31:56] [PASSED] PAL [05:31:56] [PASSED] NTSC [05:31:56] [PASSED] Both, NTSC Default [05:31:56] [PASSED] Both, PAL Default [05:31:56] [PASSED] Both, NTSC Default, with PAL on command-line [05:31:56] [PASSED] Both, PAL Default, with NTSC on command-line [05:31:56] == [PASSED] drm_test_connector_helper_tv_get_modes_check === [05:31:56] ======== [PASSED] drm_connector_helper_tv_get_modes ======== [05:31:56] ================== drm_rect (9 subtests) =================== [05:31:56] [PASSED] drm_test_rect_clip_scaled_div_by_zero [05:31:56] [PASSED] drm_test_rect_clip_scaled_not_clipped [05:31:56] [PASSED] drm_test_rect_clip_scaled_clipped [05:31:56] [PASSED] drm_test_rect_clip_scaled_signed_vs_unsigned [05:31:56] ================= drm_test_rect_intersect ================= [05:31:56] [PASSED] top-left x bottom-right: 2x2+1+1 x 2x2+0+0 [05:31:56] [PASSED] top-right x bottom-left: 2x2+0+0 x 2x2+1-1 [05:31:56] [PASSED] bottom-left x top-right: 2x2+1-1 x 2x2+0+0 [05:31:56] [PASSED] bottom-right x top-left: 2x2+0+0 x 2x2+1+1 [05:31:56] [PASSED] right x left: 2x1+0+0 x 3x1+1+0 [05:31:56] [PASSED] left x right: 3x1+1+0 x 2x1+0+0 [05:31:56] [PASSED] up x bottom: 1x2+0+0 x 1x3+0-1 [05:31:56] [PASSED] bottom x up: 1x3+0-1 x 1x2+0+0 [05:31:56] [PASSED] touching corner: 1x1+0+0 x 2x2+1+1 [05:31:56] [PASSED] touching side: 1x1+0+0 x 1x1+1+0 [05:31:56] [PASSED] equal rects: 2x2+0+0 x 2x2+0+0 [05:31:56] [PASSED] inside another: 2x2+0+0 x 1x1+1+1 [05:31:56] [PASSED] far away: 1x1+0+0 x 1x1+3+6 [05:31:56] [PASSED] points intersecting: 0x0+5+10 x 0x0+5+10 [05:31:56] [PASSED] points not intersecting: 0x0+0+0 x 0x0+5+10 [05:31:56] ============= [PASSED] drm_test_rect_intersect ============= [05:31:56] ================ drm_test_rect_calc_hscale ================ [05:31:56] [PASSED] normal use [05:31:56] [PASSED] out of max range [05:31:56] [PASSED] out of min range [05:31:56] [PASSED] zero dst [05:31:56] [PASSED] negative src [05:31:56] [PASSED] negative dst [05:31:56] ============ [PASSED] drm_test_rect_calc_hscale ============ [05:31:56] ================ drm_test_rect_calc_vscale ================ [05:31:56] [PASSED] normal use [05:31:56] [PASSED] out of max range [05:31:56] [PASSED] out of min range [05:31:56] [PASSED] zero dst [05:31:56] [PASSED] negative src [05:31:56] [PASSED] negative dst [05:31:56] ============ [PASSED] drm_test_rect_calc_vscale ============ [05:31:56] ================== drm_test_rect_rotate =================== [05:31:56] [PASSED] reflect-x [05:31:56] [PASSED] reflect-y [05:31:56] [PASSED] rotate-0 [05:31:56] [PASSED] rotate-90 [05:31:56] [PASSED] rotate-180 [05:31:56] [PASSED] rotate-270 [05:31:56] ============== [PASSED] drm_test_rect_rotate =============== [05:31:56] ================ drm_test_rect_rotate_inv ================= [05:31:56] [PASSED] reflect-x [05:31:56] [PASSED] reflect-y [05:31:56] [PASSED] rotate-0 [05:31:56] [PASSED] rotate-90 [05:31:56] [PASSED] rotate-180 [05:31:56] [PASSED] rotate-270 [05:31:56] ============ [PASSED] drm_test_rect_rotate_inv ============= [05:31:56] ==================== [PASSED] drm_rect ===================== [05:31:56] ============ drm_sysfb_modeset_test (1 subtest) ============ [05:31:56] ============ drm_test_sysfb_build_fourcc_list ============= [05:31:56] [PASSED] no native formats [05:31:56] [PASSED] XRGB8888 as native format [05:31:56] [PASSED] remove duplicates [05:31:56] [PASSED] convert alpha formats [05:31:56] [PASSED] random formats [05:31:56] ======== [PASSED] drm_test_sysfb_build_fourcc_list ========= [05:31:56] ============= [PASSED] drm_sysfb_modeset_test ============== [05:31:56] ================== drm_fixp (2 subtests) =================== [05:31:56] [PASSED] drm_test_int2fixp [05:31:56] [PASSED] drm_test_sm2fixp [05:31:56] ==================== [PASSED] drm_fixp ===================== [05:31:56] ============================================================ [05:31:56] Testing complete. Ran 621 tests: passed: 621 [05:31:56] Elapsed time: 26.369s total, 1.745s configuring, 24.493s building, 0.130s running
+ /kernel/tools/testing/kunit/kunit.py run --kunitconfig /kernel/drivers/gpu/drm/ttm/tests/.kunitconfig [05:31:56] Configuring KUnit Kernel ...
Regenerating .config ...
Populating config with:
$ make ARCH=um O=.kunit olddefconfig [05:31:58] Building KUnit Kernel ... Populating config with: $ make ARCH=um O=.kunit olddefconfig Building with: $ make all compile_commands.json scripts_gdb ARCH=um O=.kunit --jobs=48 [05:32:07] Starting KUnit Kernel (1/1)... [05:32:07] ============================================================ Running tests with: $ .kunit/linux kunit.enable=1 mem=1G console=tty kunit_shutdown=halt [05:32:08] ================= ttm_device (5 subtests) ================== [05:32:08] [PASSED] ttm_device_init_basic [05:32:08] [PASSED] ttm_device_init_multiple [05:32:08] [PASSED] ttm_device_fini_basic [05:32:08] [PASSED] ttm_device_init_no_vma_man [05:32:08] ================== ttm_device_init_pools ================== [05:32:08] [PASSED] No DMA allocations, no DMA32 required [05:32:08] [PASSED] DMA allocations, DMA32 required [05:32:08] [PASSED] No DMA allocations, DMA32 required [05:32:08] [PASSED] DMA allocations, no DMA32 required [05:32:08] ============== [PASSED] ttm_device_init_pools ============== [05:32:08] =================== [PASSED] ttm_device ==================== [05:32:08] ================== ttm_pool (8 subtests) =================== [05:32:08] ================== ttm_pool_alloc_basic =================== [05:32:08] [PASSED] One page [05:32:08] [PASSED] More than one page [05:32:08] [PASSED] Above the allocation limit [05:32:08] [PASSED] One page, with coherent DMA mappings enabled [05:32:08] [PASSED] Above the allocation limit, with coherent DMA mappings enabled [05:32:08] ============== [PASSED] ttm_pool_alloc_basic =============== [05:32:08] ============== ttm_pool_alloc_basic_dma_addr ============== [05:32:08] [PASSED] One page [05:32:08] [PASSED] More than one page [05:32:08] [PASSED] Above the allocation limit [05:32:08] [PASSED] One page, with coherent DMA mappings enabled [05:32:08] [PASSED] Above the allocation limit, with coherent DMA mappings enabled [05:32:08] ========== [PASSED] ttm_pool_alloc_basic_dma_addr ========== [05:32:08] [PASSED] ttm_pool_alloc_order_caching_match [05:32:08] [PASSED] ttm_pool_alloc_caching_mismatch [05:32:08] [PASSED] ttm_pool_alloc_order_mismatch [05:32:08] [PASSED] ttm_pool_free_dma_alloc [05:32:08] [PASSED] ttm_pool_free_no_dma_alloc [05:32:08] [PASSED] ttm_pool_fini_basic [05:32:08] ==================== [PASSED] ttm_pool ===================== [05:32:08] ================ ttm_resource (8 subtests) ================= [05:32:08] ================= ttm_resource_init_basic ================= [05:32:08] [PASSED] Init resource in TTM_PL_SYSTEM [05:32:08] [PASSED] Init resource in TTM_PL_VRAM [05:32:08] [PASSED] Init resource in a private placement [05:32:08] [PASSED] Init resource in TTM_PL_SYSTEM, set placement flags [05:32:08] ============= [PASSED] ttm_resource_init_basic ============= [05:32:08] [PASSED] ttm_resource_init_pinned [05:32:08] [PASSED] ttm_resource_fini_basic [05:32:08] [PASSED] ttm_resource_manager_init_basic [05:32:08] [PASSED] ttm_resource_manager_usage_basic [05:32:08] [PASSED] ttm_resource_manager_set_used_basic [05:32:08] [PASSED] ttm_sys_man_alloc_basic [05:32:08] [PASSED] ttm_sys_man_free_basic [05:32:08] ================== [PASSED] ttm_resource =================== [05:32:08] =================== ttm_tt (15 subtests) =================== [05:32:08] ==================== ttm_tt_init_basic ==================== [05:32:08] [PASSED] Page-aligned size [05:32:08] [PASSED] Extra pages requested [05:32:08] ================ [PASSED] ttm_tt_init_basic ================ [05:32:08] [PASSED] ttm_tt_init_misaligned [05:32:08] [PASSED] ttm_tt_fini_basic [05:32:08] [PASSED] ttm_tt_fini_sg [05:32:08] [PASSED] ttm_tt_fini_shmem [05:32:08] [PASSED] ttm_tt_create_basic [05:32:08] [PASSED] ttm_tt_create_invalid_bo_type [05:32:08] [PASSED] ttm_tt_create_ttm_exists [05:32:08] [PASSED] ttm_tt_create_failed [05:32:08] [PASSED] ttm_tt_destroy_basic [05:32:08] [PASSED] ttm_tt_populate_null_ttm [05:32:08] [PASSED] ttm_tt_populate_populated_ttm [05:32:08] [PASSED] ttm_tt_unpopulate_basic [05:32:08] [PASSED] ttm_tt_unpopulate_empty_ttm [05:32:08] [PASSED] ttm_tt_swapin_basic [05:32:08] ===================== [PASSED] ttm_tt ====================== [05:32:08] =================== ttm_bo (14 subtests) =================== [05:32:08] =========== ttm_bo_reserve_optimistic_no_ticket =========== [05:32:08] [PASSED] Cannot be interrupted and sleeps [05:32:08] [PASSED] Cannot be interrupted, locks straight away [05:32:08] [PASSED] Can be interrupted, sleeps [05:32:08] ======= [PASSED] ttm_bo_reserve_optimistic_no_ticket ======= [05:32:08] [PASSED] ttm_bo_reserve_locked_no_sleep [05:32:08] [PASSED] ttm_bo_reserve_no_wait_ticket [05:32:08] [PASSED] ttm_bo_reserve_double_resv [05:32:08] [PASSED] ttm_bo_reserve_interrupted [05:32:08] [PASSED] ttm_bo_reserve_deadlock [05:32:08] [PASSED] ttm_bo_unreserve_basic [05:32:08] [PASSED] ttm_bo_unreserve_pinned [05:32:08] [PASSED] ttm_bo_unreserve_bulk [05:32:08] [PASSED] ttm_bo_fini_basic [05:32:08] [PASSED] ttm_bo_fini_shared_resv [05:32:08] [PASSED] ttm_bo_pin_basic [05:32:08] [PASSED] ttm_bo_pin_unpin_resource [05:32:08] [PASSED] ttm_bo_multiple_pin_one_unpin [05:32:08] ===================== [PASSED] ttm_bo ====================== [05:32:08] ============== ttm_bo_validate (22 subtests) =============== [05:32:08] ============== ttm_bo_init_reserved_sys_man =============== [05:32:08] [PASSED] Buffer object for userspace [05:32:08] [PASSED] Kernel buffer object [05:32:08] [PASSED] Shared buffer object [05:32:08] ========== [PASSED] ttm_bo_init_reserved_sys_man =========== [05:32:08] ============== ttm_bo_init_reserved_mock_man ============== [05:32:08] [PASSED] Buffer object for userspace [05:32:08] [PASSED] Kernel buffer object [05:32:08] [PASSED] Shared buffer object [05:32:08] ========== [PASSED] ttm_bo_init_reserved_mock_man ========== [05:32:08] [PASSED] ttm_bo_init_reserved_resv [05:32:08] ================== ttm_bo_validate_basic ================== [05:32:08] [PASSED] Buffer object for userspace [05:32:08] [PASSED] Kernel buffer object [05:32:08] [PASSED] Shared buffer object [05:32:08] ============== [PASSED] ttm_bo_validate_basic ============== [05:32:08] [PASSED] ttm_bo_validate_invalid_placement [05:32:08] ============= ttm_bo_validate_same_placement ============== [05:32:08] [PASSED] System manager [05:32:08] [PASSED] VRAM manager [05:32:08] ========= [PASSED] ttm_bo_validate_same_placement ========== [05:32:08] [PASSED] ttm_bo_validate_failed_alloc [05:32:08] [PASSED] ttm_bo_validate_pinned [05:32:08] [PASSED] ttm_bo_validate_busy_placement [05:32:08] ================ ttm_bo_validate_multihop ================= [05:32:08] [PASSED] Buffer object for userspace [05:32:08] [PASSED] Kernel buffer object [05:32:08] [PASSED] Shared buffer object [05:32:08] ============ [PASSED] ttm_bo_validate_multihop ============= [05:32:08] ========== ttm_bo_validate_no_placement_signaled ========== [05:32:08] [PASSED] Buffer object in system domain, no page vector [05:32:08] [PASSED] Buffer object in system domain with an existing page vector [05:32:08] ====== [PASSED] ttm_bo_validate_no_placement_signaled ====== [05:32:08] ======== ttm_bo_validate_no_placement_not_signaled ======== [05:32:08] [PASSED] Buffer object for userspace [05:32:08] [PASSED] Kernel buffer object [05:32:08] [PASSED] Shared buffer object [05:32:08] ==== [PASSED] ttm_bo_validate_no_placement_not_signaled ==== [05:32:08] [PASSED] ttm_bo_validate_move_fence_signaled [05:32:08] ========= ttm_bo_validate_move_fence_not_signaled ========= [05:32:08] [PASSED] Waits for GPU [05:32:08] [PASSED] Tries to lock straight away [05:32:08] ===== [PASSED] ttm_bo_validate_move_fence_not_signaled ===== [05:32:08] [PASSED] ttm_bo_validate_swapout [05:32:08] [PASSED] ttm_bo_validate_happy_evict [05:32:08] [PASSED] ttm_bo_validate_all_pinned_evict [05:32:08] [PASSED] ttm_bo_validate_allowed_only_evict [05:32:08] [PASSED] ttm_bo_validate_deleted_evict [05:32:08] [PASSED] ttm_bo_validate_busy_domain_evict [05:32:08] [PASSED] ttm_bo_validate_evict_gutting [05:32:08] [PASSED] ttm_bo_validate_recrusive_evict [05:32:08] ================= [PASSED] ttm_bo_validate ================= [05:32:08] ============================================================ [05:32:08] Testing complete. Ran 102 tests: passed: 102 [05:32:08] Elapsed time: 11.634s total, 1.742s configuring, 9.677s building, 0.173s running
+ cleanup
++ stat -c %u:%g /kernel
+ chown -R 1003:1003 /kernel
Patchwork
CI✓ Xe.CI.BAT: success for Add new debug infrastructure for configfs
In reply to Stuart Summers== Series Details ==
== Series Details ==
Series: Add new debug infrastructure for configfs
URL : https://patchwork.freedesktop.org/series/165879/
State : success
== Summary ==
CI Bug Log - changes from xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c_BAT -> xe-pw-165879v1_BAT
====================================================
Summary
-------
**SUCCESS**
No regressions found.
Participating hosts (13 -> 13)
------------------------------
No changes in participating hosts
Changes
-------
No changes found
Build changes
-------------
* Linux: xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c -> xe-pw-165879v1
IGT_8881: aa5853ef5b379b1e4558218c21ef4caeae112184 @ https://gitlab.freedesktop.org/drm/igt-gpu-tools.git
xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c: 9756f0821b775d39e62f2524aa3f7421f7a9e76c
xe-pw-165879v1: 165879v1
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/index.html
Patchwork
CI✗ Xe.CI.FULL: failure for Add new debug infrastructure for configfs
In reply to Stuart Summers== Series Details ==
== Series Details ==
Series: Add new debug infrastructure for configfs
URL : https://patchwork.freedesktop.org/series/165879/
State : failure
== Summary ==
CI Bug Log - changes from xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c_FULL -> xe-pw-165879v1_FULL
====================================================
Summary
-------
**FAILURE**
Serious unknown changes coming with xe-pw-165879v1_FULL absolutely need to be verified manually.
If you think the reported changes have nothing to do with the changes
introduced in xe-pw-165879v1_FULL, please notify your bug team ([email protected]) to allow them to document this new failure mode, which will reduce false positives in CI.
Participating hosts (2 -> 2)
------------------------------
No changes in participating hosts
Possible new issues
-------------------
Here are the unknown changes that may have been introduced in xe-pw-165879v1_FULL:
### IGT changes ###
#### Possible regressions ####
* igt@xe_configfs@ctx-restore-post-bb:
- shard-lnl: [PASS][1] -> [FAIL][2] +2 other tests fail
[1]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-lnl-3/igt@[email protected]
[2]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-1/igt@[email protected]
* igt@xe_configfs@engines-allowed:
- shard-bmg: [PASS][3] -> [FAIL][4] +2 other tests fail
[3]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-7/igt@[email protected]
[4]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-6/igt@[email protected]
- shard-lnl: NOTRUN -> [FAIL][5]
[5]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@xe_configfs@gt-types-allowed:
- shard-bmg: NOTRUN -> [FAIL][6] +1 other test fail
[6]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@xe_exec_reset@multi-queue-gt-reset:
- shard-bmg: NOTRUN -> [SKIP][7] +3 other tests skip
[7]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
#### Warnings ####
* igt@xe_module_load@load:
- shard-bmg: ([PASS][8], [PASS][9], [PASS][10], [PASS][11], [PASS][12], [PASS][13], [PASS][14], [PASS][15], [PASS][16], [PASS][17], [PASS][18], [PASS][19], [PASS][20], [PASS][21], [PASS][22], [PASS][23], [PASS][24], [PASS][25], [PASS][26], [PASS][27], [PASS][28], [ABORT][29]) -> ([ABORT][30], [PASS][31], [SKIP][32], [PASS][33], [PASS][34], [PASS][35], [PASS][36], [PASS][37], [PASS][38], [PASS][39], [PASS][40], [PASS][41], [PASS][42], [ABORT][43], [PASS][44], [PASS][45], [PASS][46], [PASS][47], [PASS][48], [PASS][49], [PASS][50], [PASS][51], [PASS][52]) ([Intel XE#2457] / [Intel XE#7405])
[8]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-7/igt@[email protected]
[9]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-2/igt@[email protected]
[10]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-4/igt@[email protected]
[11]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-10/igt@[email protected]
[12]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-10/igt@[email protected]
[13]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-10/igt@[email protected]
[14]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-2/igt@[email protected]
[15]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-8/igt@[email protected]
[16]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-8/igt@[email protected]
[17]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-4/igt@[email protected]
[18]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-9/igt@[email protected]
[19]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-9/igt@[email protected]
[20]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-6/igt@[email protected]
[21]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-5/igt@[email protected]
[22]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-5/igt@[email protected]
[23]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-7/igt@[email protected]
[24]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-7/igt@[email protected]
[25]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-6/igt@[email protected]
[26]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-6/igt@[email protected]
[27]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-1/igt@[email protected]
[28]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-1/igt@[email protected]
[29]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-3/igt@[email protected]
[30]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-3/igt@[email protected]
[31]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-2/igt@[email protected]
[32]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
[33]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-6/igt@[email protected]
[34]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-6/igt@[email protected]
[35]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
[36]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
[37]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
[38]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-10/igt@[email protected]
[39]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
[40]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-10/igt@[email protected]
[41]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
[42]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
[43]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-3/igt@[email protected]
[44]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
[45]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-1/igt@[email protected]
[46]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-1/igt@[email protected]
[47]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
[48]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
[49]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
[50]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-4/igt@[email protected]
[51]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-4/igt@[email protected]
[52]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-4/igt@[email protected]
Known issues
------------
Here are the changes found in xe-pw-165879v1_FULL that come from known issues:
### IGT changes ###
#### Issues hit ####
* igt@kms_big_fb@4-tiled-32bpp-rotate-90:
- shard-bmg: NOTRUN -> [SKIP][53] ([Intel XE#2327]) +3 other tests skip
[53]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_big_fb@4-tiled-64bpp-rotate-270:
- shard-lnl: NOTRUN -> [SKIP][54] ([Intel XE#1407])
[54]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@kms_big_fb@linear-max-hw-stride-32bpp-rotate-0-hflip:
- shard-bmg: NOTRUN -> [SKIP][55] ([Intel XE#7059] / [Intel XE#7085]) +1 other test skip
[55]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@kms_big_fb@y-tiled-16bpp-rotate-0:
- shard-bmg: NOTRUN -> [SKIP][56] ([Intel XE#1124]) +9 other tests skip
[56]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@kms_big_fb@y-tiled-max-hw-stride-64bpp-rotate-180-async-flip:
- shard-lnl: NOTRUN -> [SKIP][57] ([Intel XE#1124])
[57]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-2/igt@kms_big_fb@y-tiled-max-hw-stride-64bpp-rotate-180-async-flip.html
* igt@kms_bw@connected-linear-tiling-2-displays-target-1920x1080p:
- shard-bmg: NOTRUN -> [SKIP][58] ([Intel XE#7679]) +2 other tests skip
[58]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_bw@linear-tiling-4-displays-target-1920x1080p:
- shard-bmg: NOTRUN -> [SKIP][59] ([Intel XE#367])
[59]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_ccs@crc-primary-suspend-4-tiled-mtl-rc-ccs:
- shard-bmg: NOTRUN -> [SKIP][60] ([Intel XE#3432]) +1 other test skip
[60]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_ccs@random-ccs-data-4-tiled-dg2-rc-ccs:
- shard-bmg: NOTRUN -> [SKIP][61] ([Intel XE#2887]) +14 other tests skip
[61]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_cdclk@mode-transition-all-outputs:
- shard-bmg: NOTRUN -> [SKIP][62] ([Intel XE#2724] / [Intel XE#7449])
[62]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@kms_chamelium_color@ctm-0-25:
- shard-bmg: NOTRUN -> [SKIP][63] ([Intel XE#2325] / [Intel XE#7358])
[63]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_chamelium_edid@dp-edid-change-during-hibernate:
- shard-bmg: NOTRUN -> [SKIP][64] ([Intel XE#2252]) +7 other tests skip
[64]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@kms_content_protection@content-type-change:
- shard-bmg: NOTRUN -> [SKIP][65] ([Intel XE#7642]) +1 other test skip
[65]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@kms_content_protection@dp-mst-lic-type-0:
- shard-bmg: NOTRUN -> [SKIP][66] ([Intel XE#2390] / [Intel XE#6974])
[66]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_content_protection@uevent:
- shard-bmg: NOTRUN -> [FAIL][67] ([Intel XE#6707] / [Intel XE#7439]) +1 other test fail
[67]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@kms_cursor_crc@cursor-random-512x170:
- shard-bmg: NOTRUN -> [SKIP][68] ([Intel XE#2321] / [Intel XE#7355]) +1 other test skip
[68]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_cursor_crc@cursor-sliding-256x85:
- shard-bmg: NOTRUN -> [SKIP][69] ([Intel XE#2320]) +4 other tests skip
[69]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@kms_cursor_legacy@cursora-vs-flipa-atomic-transitions-varying-size:
- shard-bmg: [PASS][70] -> [DMESG-WARN][71] ([Intel XE#5354])
[70]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-9/igt@kms_cursor_legacy@cursora-vs-flipa-atomic-transitions-varying-size.html
[71]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-10/igt@kms_cursor_legacy@cursora-vs-flipa-atomic-transitions-varying-size.html
* igt@kms_cursor_legacy@cursorb-vs-flipb-atomic-transitions-varying-size:
- shard-lnl: NOTRUN -> [SKIP][72] ([Intel XE#309] / [Intel XE#7343])
[72]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@kms_cursor_legacy@cursorb-vs-flipb-atomic-transitions-varying-size.html
* igt@kms_dp_link_training@uhbr-mst:
- shard-bmg: NOTRUN -> [SKIP][73] ([Intel XE#4354] / [Intel XE#7386])
[73]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@kms_dsc@dsc-fractional-bpp-with-bpc:
- shard-bmg: NOTRUN -> [SKIP][74] ([Intel XE#2244])
[74]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_fbc_dirty_rect@fbc-dirty-rectangle-different-formats:
- shard-bmg: NOTRUN -> [SKIP][75] ([Intel XE#4422] / [Intel XE#7442])
[75]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@kms_fbc_dirty_rect@fbc-dirty-rectangle-out-visible-area:
- shard-lnl: NOTRUN -> [SKIP][76] ([Intel XE#4422] / [Intel XE#7442])
[76]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@kms_feature_discovery@display-3x:
- shard-bmg: NOTRUN -> [SKIP][77] ([Intel XE#2373] / [Intel XE#7448])
[77]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_feature_discovery@dp-mst:
- shard-bmg: NOTRUN -> [SKIP][78] ([Intel XE#2375])
[78]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@kms_flip@flip-vs-expired-vblank@a-edp1:
- shard-lnl: [PASS][79] -> [FAIL][80] ([Intel XE#301]) +1 other test fail
[79]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-lnl-1/igt@kms_flip@[email protected]
[80]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-7/igt@kms_flip@[email protected]
* igt@kms_flip_scaled_crc@flip-64bpp-yftile-to-16bpp-yftile-downscaling:
- shard-bmg: NOTRUN -> [SKIP][81] ([Intel XE#7178] / [Intel XE#7351]) +3 other tests skip
[81]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@kms_flip_scaled_crc@flip-64bpp-yftile-to-16bpp-yftile-downscaling.html
* igt@kms_frontbuffer_tracking@drrs-2p-primscrn-pri-indfb-draw-mmap-wc:
- shard-bmg: NOTRUN -> [SKIP][82] ([Intel XE#2311]) +26 other tests skip
[82]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@kms_frontbuffer_tracking@drrs-2p-primscrn-pri-indfb-draw-mmap-wc.html
* igt@kms_frontbuffer_tracking@drrs-abgr161616f-draw-mmap-wc:
- shard-bmg: NOTRUN -> [SKIP][83] ([Intel XE#7061] / [Intel XE#7356]) +6 other tests skip
[83]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@kms_frontbuffer_tracking@fbc-2p-primscrn-pri-shrfb-draw-blt:
- shard-lnl: NOTRUN -> [SKIP][84] ([Intel XE#5812] / [Intel XE#656]) +1 other test skip
[84]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@kms_frontbuffer_tracking@fbc-2p-scndscrn-spr-indfb-onoff:
- shard-bmg: NOTRUN -> [SKIP][85] ([Intel XE#4141]) +12 other tests skip
[85]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_frontbuffer_tracking@fbcdrrs-1p-primscrn-spr-indfb-draw-render:
- shard-lnl: NOTRUN -> [SKIP][86] ([Intel XE#6312] / [Intel XE#651])
[86]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@kms_frontbuffer_tracking@fbcdrrs-1p-primscrn-spr-indfb-draw-render.html
* igt@kms_frontbuffer_tracking@psr-2p-primscrn-indfb-plflip-blt:
- shard-bmg: NOTRUN -> [SKIP][87] ([Intel XE#2313]) +30 other tests skip
[87]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_hdr@invalid-hdr:
- shard-bmg: NOTRUN -> [SKIP][88] ([Intel XE#1503])
[88]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@kms_pipe_stress@stress-xrgb8888-ytiled:
- shard-bmg: NOTRUN -> [SKIP][89] ([Intel XE#4329] / [Intel XE#6912] / [Intel XE#7375])
[89]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_plane@pixel-format-4-tiled-dg2-rc-ccs-cc-modifier-source-clamping:
- shard-bmg: NOTRUN -> [SKIP][90] ([Intel XE#7283]) +2 other tests skip
[90]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@kms_plane@pixel-format-4-tiled-dg2-rc-ccs-cc-modifier-source-clamping.html
* igt@kms_plane_multiple@2x-tiling-y:
- shard-bmg: NOTRUN -> [SKIP][91] ([Intel XE#5021] / [Intel XE#7377])
[91]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_pm_backlight@fade:
- shard-bmg: NOTRUN -> [SKIP][92] ([Intel XE#7376] / [Intel XE#870]) +1 other test skip
[92]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_pm_rpm@dpms-lpsp:
- shard-bmg: NOTRUN -> [SKIP][93] ([Intel XE#1439] / [Intel XE#3141] / [Intel XE#7383] / [Intel XE#836]) +1 other test skip
[93]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_psr2_sf@pr-primary-plane-update-sf-dmg-area:
- shard-bmg: NOTRUN -> [SKIP][94] ([Intel XE#1489]) +4 other tests skip
[94]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_psr2_su@frontbuffer-xrgb8888:
- shard-bmg: NOTRUN -> [SKIP][95] ([Intel XE#2387] / [Intel XE#7429]) +1 other test skip
[95]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@kms_psr@fbc-pr-sprite-plane-move:
- shard-lnl: NOTRUN -> [SKIP][96] ([Intel XE#1406])
[96]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@kms_psr@psr-primary-page-flip:
- shard-bmg: NOTRUN -> [SKIP][97] ([Intel XE#2234] / [Intel XE#2850]) +12 other tests skip
[97]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_psr_stress_test@flip-primary-invalidate-overlay:
- shard-bmg: NOTRUN -> [SKIP][98] ([Intel XE#7795])
[98]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@kms_rotation_crc@bad-pixel-format:
- shard-bmg: NOTRUN -> [SKIP][99] ([Intel XE#3904] / [Intel XE#7342]) +2 other tests skip
[99]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_rotation_crc@primary-y-tiled-reflect-x-0:
- shard-bmg: NOTRUN -> [SKIP][100] ([Intel XE#2330] / [Intel XE#5813])
[100]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@kms_scaling_modes@scaling-mode-center:
- shard-bmg: NOTRUN -> [SKIP][101] ([Intel XE#2413])
[101]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@kms_sharpness_filter@filter-strength:
- shard-bmg: NOTRUN -> [SKIP][102] ([Intel XE#6503])
[102]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_tiled_display@basic-test-pattern-with-chamelium:
- shard-bmg: NOTRUN -> [SKIP][103] ([Intel XE#2426] / [Intel XE#5848])
[103]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@kms_vrr@flipline:
- shard-bmg: NOTRUN -> [SKIP][104] ([Intel XE#1499]) +1 other test skip
[104]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@xe_compute@ccs-mode-compute-kernel:
- shard-bmg: NOTRUN -> [SKIP][105] ([Intel XE#6599])
[105]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@xe_eudebug@vm-bind-clear:
- shard-bmg: NOTRUN -> [SKIP][106] ([Intel XE#7636]) +12 other tests skip
[106]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@xe_eudebug_online@writes-caching-sram-bb-vram-target-vram:
- shard-lnl: NOTRUN -> [SKIP][107] ([Intel XE#7636])
[107]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@xe_evict@evict-beng-mixed-many-threads-small:
- shard-bmg: NOTRUN -> [INCOMPLETE][108] ([Intel XE#6321])
[108]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@xe_evict@evict-small-multi-queue-cm:
- shard-bmg: NOTRUN -> [SKIP][109] ([Intel XE#7140]) +1 other test skip
[109]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@xe_exec_basic@multigpu-once-null-rebind:
- shard-bmg: NOTRUN -> [SKIP][110] ([Intel XE#2322] / [Intel XE#7372]) +9 other tests skip
[110]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@xe_exec_fault_mode@many-execqueues-multi-queue-prefetch:
- shard-lnl: NOTRUN -> [SKIP][111] ([Intel XE#7136])
[111]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@xe_exec_fault_mode@twice-multi-queue-imm:
- shard-bmg: NOTRUN -> [SKIP][112] ([Intel XE#7136]) +10 other tests skip
[112]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@xe_exec_multi_queue@many-execs-preempt-mode-close-fd-smem:
- shard-lnl: NOTRUN -> [SKIP][113] ([Intel XE#6874]) +1 other test skip
[113]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-2/igt@[email protected]
* igt@xe_exec_multi_queue@one-queue-preempt-mode-fault-dyn-priority-smem:
- shard-bmg: NOTRUN -> [SKIP][114] ([Intel XE#6874]) +29 other tests skip
[114]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@xe_exec_multi_queue@one-queue-preempt-mode-fault-dyn-priority-smem.html
* igt@xe_exec_threads@threads-multi-queue-mixed-fd-userptr:
- shard-bmg: NOTRUN -> [SKIP][115] ([Intel XE#7138]) +8 other tests skip
[115]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@xe_exec_threads@threads-multi-queue-userptr-invalidate-race:
- shard-lnl: NOTRUN -> [SKIP][116] ([Intel XE#7138])
[116]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@xe_live_ktest@xe_eudebug:
- shard-bmg: NOTRUN -> [SKIP][117] ([Intel XE#2833])
[117]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@xe_live_ktest@xe_eudebug.html
* igt@xe_multigpu_svm@mgpu-coherency-fail-prefetch:
- shard-bmg: NOTRUN -> [SKIP][118] ([Intel XE#6964]) +1 other test skip
[118]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@xe_multigpu_svm@mgpu-pagefault-prefetch:
- shard-lnl: NOTRUN -> [SKIP][119] ([Intel XE#6964])
[119]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@xe_page_reclaim@binds-null-vma:
- shard-bmg: NOTRUN -> [SKIP][120] ([Intel XE#7793]) +2 other tests skip
[120]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@xe_page_reclaim@random:
- shard-lnl: NOTRUN -> [SKIP][121] ([Intel XE#7793])
[121]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@xe_pat@pat-sw-hw-compare:
- shard-bmg: NOTRUN -> [FAIL][122] ([Intel XE#7695])
[122]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@xe_pat@xa-app-transient-media-on:
- shard-bmg: NOTRUN -> [SKIP][123] ([Intel XE#7590])
[123]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
* igt@xe_pm@d3hot-i2c:
- shard-bmg: NOTRUN -> [SKIP][124] ([Intel XE#5742] / [Intel XE#7328] / [Intel XE#7400])
[124]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-5/igt@[email protected]
* igt@xe_pm@s3-multiple-execs:
- shard-lnl: NOTRUN -> [SKIP][125] ([Intel XE#584] / [Intel XE#7369])
[125]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@[email protected]
* igt@xe_pm@s4-d3cold-basic-exec:
- shard-bmg: NOTRUN -> [SKIP][126] ([Intel XE#2284] / [Intel XE#7370]) +1 other test skip
[126]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@xe_prefetch_fault@prefetch-fault:
- shard-bmg: NOTRUN -> [SKIP][127] ([Intel XE#7599])
[127]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
* igt@xe_pxp@pxp-stale-bo-exec-post-suspend:
- shard-bmg: NOTRUN -> [SKIP][128] ([Intel XE#4733] / [Intel XE#7417]) +1 other test skip
[128]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
* igt@xe_query@multigpu-query-hwconfig:
- shard-bmg: NOTRUN -> [SKIP][129] ([Intel XE#944]) +1 other test skip
[129]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-7/igt@[email protected]
#### Possible fixes ####
* igt@xe_evict@evict-mixed-many-threads-small:
- shard-bmg: [INCOMPLETE][130] ([Intel XE#6321]) -> [PASS][131]
[130]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-10/igt@[email protected]
[131]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-9/igt@[email protected]
#### Warnings ####
* igt@kms_plane_scaling@planes-upscale-factor-0-25-downscale-factor-0-75:
- shard-lnl: [ABORT][132] -> [SKIP][133] ([Intel XE#2763] / [Intel XE#6886]) +1 other test skip
[132]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-lnl-6/igt@kms_plane_scaling@planes-upscale-factor-0-25-downscale-factor-0-75.html
[133]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-lnl-5/igt@kms_plane_scaling@planes-upscale-factor-0-25-downscale-factor-0-75.html
* igt@kms_tiled_display@basic-test-pattern:
- shard-bmg: [FAIL][134] ([Intel XE#1729] / [Intel XE#7424]) -> [SKIP][135] ([Intel XE#2426] / [Intel XE#5848])
[134]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c/shard-bmg-5/igt@[email protected]
[135]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/shard-bmg-8/igt@[email protected]
[Intel XE#1124]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1124
[Intel XE#1406]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1406
[Intel XE#1407]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1407
[Intel XE#1439]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1439
[Intel XE#1489]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1489
[Intel XE#1499]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1499
[Intel XE#1503]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1503
[Intel XE#1729]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1729
[Intel XE#2234]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2234
[Intel XE#2244]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2244
[Intel XE#2252]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2252
[Intel XE#2284]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2284
[Intel XE#2311]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2311
[Intel XE#2313]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2313
[Intel XE#2320]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2320
[Intel XE#2321]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2321
[Intel XE#2322]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2322
[Intel XE#2325]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2325
[Intel XE#2327]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2327
[Intel XE#2330]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2330
[Intel XE#2373]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2373
[Intel XE#2375]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2375
[Intel XE#2387]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2387
[Intel XE#2390]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2390
[Intel XE#2413]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2413
[Intel XE#2426]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2426
[Intel XE#2457]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2457
[Intel XE#2724]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2724
[Intel XE#2763]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2763
[Intel XE#2833]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2833
[Intel XE#2850]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2850
[Intel XE#2887]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2887
[Intel XE#301]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/301
[Intel XE#309]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/309
[Intel XE#3141]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/3141
[Intel XE#3432]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/3432
[Intel XE#367]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/367
[Intel XE#3904]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/3904
[Intel XE#4141]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/4141
[Intel XE#4329]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/4329
[Intel XE#4354]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/4354
[Intel XE#4422]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/4422
[Intel XE#4733]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/4733
[Intel XE#5021]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/5021
[Intel XE#5354]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/5354
[Intel XE#5742]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/5742
[Intel XE#5812]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/5812
[Intel XE#5813]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/5813
[Intel XE#584]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/584
[Intel XE#5848]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/5848
[Intel XE#6312]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6312
[Intel XE#6321]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6321
[Intel XE#6503]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6503
[Intel XE#651]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/651
[Intel XE#656]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/656
[Intel XE#6599]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6599
[Intel XE#6707]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6707
[Intel XE#6874]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6874
[Intel XE#6886]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6886
[Intel XE#6912]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6912
[Intel XE#6964]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6964
[Intel XE#6974]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6974
[Intel XE#7059]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7059
[Intel XE#7061]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7061
[Intel XE#7085]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7085
[Intel XE#7136]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7136
[Intel XE#7138]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7138
[Intel XE#7140]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7140
[Intel XE#7178]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7178
[Intel XE#7283]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7283
[Intel XE#7328]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7328
[Intel XE#7342]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7342
[Intel XE#7343]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7343
[Intel XE#7351]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7351
[Intel XE#7355]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7355
[Intel XE#7356]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7356
[Intel XE#7358]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7358
[Intel XE#7369]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7369
[Intel XE#7370]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7370
[Intel XE#7372]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7372
[Intel XE#7375]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7375
[Intel XE#7376]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7376
[Intel XE#7377]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7377
[Intel XE#7383]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7383
[Intel XE#7386]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7386
[Intel XE#7400]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7400
[Intel XE#7405]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7405
[Intel XE#7417]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7417
[Intel XE#7424]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7424
[Intel XE#7429]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7429
[Intel XE#7439]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7439
[Intel XE#7442]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7442
[Intel XE#7448]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7448
[Intel XE#7449]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7449
[Intel XE#7590]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7590
[Intel XE#7599]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7599
[Intel XE#7636]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7636
[Intel XE#7642]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7642
[Intel XE#7679]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7679
[Intel XE#7695]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7695
[Intel XE#7793]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7793
[Intel XE#7795]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7795
[Intel XE#836]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/836
[Intel XE#870]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/870
[Intel XE#944]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/944
Build changes
-------------
* Linux: xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c -> xe-pw-165879v1
IGT_8881: aa5853ef5b379b1e4558218c21ef4caeae112184 @ https://gitlab.freedesktop.org/drm/igt-gpu-tools.git
xe-4969-9756f0821b775d39e62f2524aa3f7421f7a9e76c: 9756f0821b775d39e62f2524aa3f7421f7a9e76c
xe-pw-165879v1: 165879v1
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-165879v1/index.html