- Oct 30, 2024
-
-
This reverts commit 55858fa7. '55858fa7 ("drm/xe/xe_guc_ads: save/restore OA registers and allowlist regs")' was not properly reviewed and also causes dmesg asserts in CI. Revert it. Closes: drm/xe/kernel#3295 Fixes: 55858fa7 ("drm/xe/xe_guc_ads: save/restore OA registers and allowlist regs") Signed-off-by:
Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by:
Jonathan Cavitt <jonathan.cavitt@intel.com> Signed-off-by:
Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241029200147.1476513-1-ashutosh.dixit@intel.com
-
- Oct 29, 2024
-
-
John Harrison authoredThe guc_info debugfs file is meant to be a quick view of the current software state of the GuC interface. Including the full CTB contents makes the file as a whole much less human readable and is not partiular useful in the general case. So don't pollute the info dump with the full buffers. Instead, move those into a separate debugfs entry that can be read when that information is actually required. Also, improve the human readability by adding a few extra blank lines to delimt the sections. v2: Hide the internal capture/print params from external callers that don't need to know (review feedback from Matthew Brost). Signed-off-by:
John Harrison <John.C.Harrison@Intel.com> Reviewed-by:
Matthew Brost <matthew.brost@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241024002554.1983101-3-John.C.Harrison@Intel.com
-
John Harrison authored
The extra bits are not hugely useful because the GuC log only uses 32bit time stamps. But they exist so might as well provide them. Signed-off-by:
-
- Oct 28, 2024
-
-
Several OA registers and allowlist registers were missing from the save/restore list for GuC and could be lost during an engine reset. Add them to the list. v2: - Fix commit message (Umesh) - Add missing closes (Ashutosh) v3: - Add missing fixes (Ashutosh) Closes: #2249 Fixes: dd08ebf6 ("drm/xe: Introduce a new DRM driver for Intel GPUs") Suggested-by:
Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com> Suggested-by:
-
- Oct 23, 2024
-
-
Ashutosh Dixit authored
Whereas all properties can be specified during OA stream open, when the OA stream is reconfigured only the config_id and syncs can be specified. v2: Use separate function table for reconfig case (Jonathan) Change bool function args to enum (Matt B) v3: s/xe_oa_set_property_funcs/xe_oa_set_property_funcs_open/ (Jonathan) Reviewed-by: -
Ashutosh Dixit authored
In addition to stream open, add xe_sync support to the OA config ioctl, where it is even more useful. This allows e.g. Mesa to replay a workload repeatedly on the GPU, each time with a different OA configuration, while precisely controlling (at batch buffer granularity) the workload segment for which a particular OA configuration is active, without introducing stalls in the userspace pipeline. v2: Emit OA config even when config id is same as previous, to ensure consistent sync behavior (Jose) Reviewed-by: -
Ashutosh Dixit authored
No code changes, only code movement so that functions used during stream open can be reused for the stream reconfiguration ioctl (DRM_XE_OBSERVATION_IOCTL_CONFIG). Reviewed-by:
-
Ashutosh Dixit authored
Introduce 'struct xe_oa_fence' which includes the dma_fence used to signal output fences in the xe_sync array. The fences are signaled asynchronously. When there are no output fences to signal, the OA configuration wait is synchronously re-introduced into the ioctl. v2: Don't wait in the work, use callback + delayed work (Matt B) Use a single, not a per-fence spinlock (Matt Brost) v3: Move ofence alloc before job submission (Matt) Assert, don't fail, from dma_fence_add_callback (Matt) Additional dma_fence_get for dma_fence_wait (Matt) Change dma_fence_wait to non-interruptible (Matt) v4: Introduce last_fence to prevent uaf if stream is closed with pending OA config jobs v5: Remove oa_fence_lock, move spinlock back into xe_oa_fence to prevent uaf Suggested-by: -
Ashutosh Dixit authored
Add input fence dependencies which will make OA configuration wait till these dependencies are met (till input fences signal). v2: Change add_deps arg to xe_oa_submit_bb from bool to enum (Matt Brost) Reviewed-by:
-
Ashutosh Dixit authored
Now that we have laid the groundwork, introduce OA sync properties in the uapi and parse the input xe_sync array as is done elsewhere in the driver. Also add DRM_XE_OA_CAPS_SYNCS bit in OA capabilities for userspace. v2: Fix and document DRM_XE_SYNC_TYPE_USER_FENCE for OA (Matt B) Add DRM_XE_OA_CAPS_SYNCS bit to OA capabilities (Jose) Acked-by:José Roberto de Souza <jose.souza@intel.com> Reviewed-by:
-
Ashutosh Dixit authored
When we introduce xe_syncs, we don't wait for internal OA programming batches to complete. That is, xe_syncs are signaled asynchronously. In anticipation for this, separate out batch submission from waiting for completion of those batches. v2: Change return type of xe_oa_submit_bb to "struct dma_fence *" (Matt B) v3: Retain init "int err = 0;" in xe_oa_submit_bb (Jose) Reviewed-by:
-
Matthew Brost authoredGT ordered work queue can be used to free memory via resets and fence signaling thus we should allow this work queue to run during reclaim. Mark with GT ordered work queue with WQ_MEM_RECLAIM appropriately. Signed-off-by:
Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com> Reviewed-by:
Badal Nilawar <badal.nilawar@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241021175705.1584521-5-matthew.brost@intel.com
-
Matthew Brost authored
G2H work queue can be used to free memory thus we should allow this work queue to run during reclaim. Mark with G2H work queue with WQ_MEM_RECLAIM appropriately. Signed-off-by:
-
Matthew Brost authored
GGTT work queue is used to free memory thus we should allow this work queue to run during reclaim. Mark with GGTT work queue with WQ_MEM_RECLAIM appropriately. Signed-off-by:
-
Matthew Brost authored
Take ref to job's fence in arm rather than run job. This ref is owned by the drm scheduler so it makes sense to take the ref before handing over the job to the scheduler. Also removes an atomic from the run job path. Suggested-by:
Matthew Auld <matthew.auld@intel.com> Signed-off-by:
-
Nirmoy Das authored
In case of parallel submissions multiple GuC id will point to the same exec queue and on GT reset such exec queues will get restarted multiple times which is not desirable. v2: don't use exec_queue_enabled() which could race, do the same for xe_guc_submit_stop (Matt B) Link: drm/xe/kernel#2295 Cc: Jonathan Cavitt <jonathan.cavitt@intel.com> Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com> Cc: Matthew Auld <matthew.auld@intel.com> Cc: Matthew Brost <matthew.brost@intel.com> Cc: Tejas Upadhyay <tejas.upadhyay@intel.com> Reviewed-by:Nirmoy Das <nirmoy.das@intel.com>
-
- Oct 22, 2024
-
-
mwajdecz authored
While we can already view individual VF LMEM provisioning using lmem_quota debugfs attribute, we want to have unified way to show summary across all VFs, like we do for GGTT or GuC doorbells/IDs. Signed-off-by:
Michal Wajdeczko <michal.wajdeczko@intel.com> Reviewed-by:
Marcin Bernatowicz <marcin.bernatowicz@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241021201506.1771-1-michal.wajdeczko@intel.com
-
Zhanjun Dong authored
GuC based register capture is not supported by VF, thus prevent it running on VF. Signed-off-by:
Zhanjun Dong <zhanjun.dong@intel.com> Reviewed-by:
Rodrigo Vivi <rodrigo.vivi@intel.com>
-
- Oct 21, 2024
-
-
The lite restore is a performance improvement feature which avoids unnecessary context switch (flush, save and restore) if the incoming context has a ContextID matching that of the outgoing context. The scheduling is done by the GuC firmware, so on the driver side it's just a matter of setting corresponding GUC_CTL_FEATURE flag. This is supposed to be enabled by default, thus the flag is set unconditionally. Signed-off-by:
Fei Yang <fei.yang@intel.com> Reviewed-by:
-
- Oct 20, 2024
-
-
Matthew Brost authored
Good practice to use __counted_by in kernel coding for flexible arrays. Signed-off-by:
-
- Oct 18, 2024
-
-
Nirmoy Das authored
This shouldn't happen but seen this while debugging ufence timeout issue time to time so log it to isolate this particular case. v2: s/XE_WARN_ON/drm_dbg(Maarten) Link: drm/xe/kernel#1630 Cc: Francois Dugast <francois.dugast@intel.com> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Matthew Auld <matthew.auld@intel.com> Cc: Matthew Brost <matthew.brost@intel.com> Reviewed-by:
Francois Dugast <francois.dugast@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241016082304.66009-3-nirmoy.das@intel.com Signed-off-by:
-
Nirmoy Das authored
access_ok() only checks for addr overflow so also try to read the addr to catch invalid addr sent from userspace. Link: #1630 Cc: Francois Dugast <francois.dugast@intel.com> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Matthew Auld <matthew.auld@intel.com> Cc: Matthew Brost <matthew.brost@intel.com> Reviewed-by:
-
Shuicheng Lin authored
In some cases, when the driver attempts to read an MMIO register, the hardware may return 0xFFFFFFFF. The current force wake path code treats this as a valid response, as it only checks the BIT. However, 0xFFFFFFFF should be considered an invalid value, indicating a potential issue. To address this, we should add a log entry to highlight this condition and return failure. The force wake failure log level is changed from notice to err to match the failure return value. v2 (Matt Brost): - set ret value (-EIO) to kick the error to upper layers v3 (Rodrigo): - add commit message for the log level promotion from notice to err v4: - update reviewed info Suggested-by:
Alex Zuo <alex.zuo@intel.com> Signed-off-by:
Shuicheng Lin <shuicheng.lin@intel.com> Cc: Matthew Brost <matthew.brost@intel.com> Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Reviewed-by:
-
- Oct 17, 2024
-
-
As part of this WA, GuC will hold a forcewake for certain MMIO accesses outside the GT/media domains. Cc: Matt Roper <matthew.d.roper@intel.com> Signed-off-by:
Vinay Belgaumkar <vinay.belgaumkar@intel.com> Reviewed-by:
-
In case if g2h worker doesn't get opportunity to within specified timeout delay then flush the g2h worker explicitly. v2: - Describe change in the comment and add TODO (Matt B/John H) - Add xe_gt_warn on fence done after G2H flush (John H) v3: - Updated the comment with root cause - Clean up xe_gt_warn message (John H) Closes: #1620 Closes: #2902 Signed-off-by:
-
Himal Ghimiray authored
There is no need to return an error from xe_force_wake_put(), as a failure implicitly indicates that the domain failed to sleep. v3 - Move kernel-doc to this patch (Badal) v5 - change parameter to unsigned int in xe_force_wake_put() v6 - Remove unneccsary wrapping (Michal) - Remove non required header (Michal) - Mention timeout(Michal) v8 - Fix kernel-doc Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Cc: Badal Nilawar <badal.nilawar@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Nirmoy Das <nirmoy.das@intel.com> Reviewed-by:
-
Himal Ghimiray authored
Add __must_check attribute for xe_force_wake_get(). Cc: Badal Nilawar <badal.nilawar@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Nirmoy Das <nirmoy.das@intel.com> Reviewed-by:
-
Himal Ghimiray authored
A failure in xe_force_wake_get() no longer increments the domain's refcount. Therefore, if xe_force_wake_get() fails during forcewake debugfs open, return an error. This ensures there are no valid file descriptors to close via forcewake debugfs, preventing refcount mismanagement. v3 - return xe_wakeref_t instead of int in xe_force_wake_get() v5 - return unsigned int from xe_force_wake_get() v6 - Use helper xe_force_wake_ref_has_domain() to determine the status of the call. Cc: Badal Nilawar <badal.nilawar@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
xe_force_wake_get() now returns the reference count-incremented domain mask. If it fails for individual domains, the return value will always be 0. However, for XE_FORCEWAKE_ALL, it may return a non-zero value even in the event of failure. Update the return handling of xe_force_wake_get() to reflect this behavior, and ensure that the return value is passed as input to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() - xe_force_wake_put() error doesn't need to be escalated/considered as probing error. It internally WARNS on domain ack failure. v5 - return unsigned int from xe_force_wake_get() v7 - Fix commit message Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
With xe_force_wake_get() now returning the refcount-incremented domain mask, a non-zero return value in the case of XE_FORCEWAKE_ALL does not necessarily indicate success. Use xe_force_wake_ref_has_domain() to determine the status of the call. Modify the return handling of xe_force_wake_get() accordingly and pass the return value to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() - xe_force_wake_put() error doesn't need to be checked. It internally WARNS on domain ack failure. v5 - return unsigned int from xe_force_wake_get() v6 - Use helper Use xe_force_wake_ref_has_domain() Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
With xe_force_wake_get() now returning the refcount-incremented domain mask, a non-zero return value in the case of XE_FORCEWAKE_ALL does not necessarily indicate success. Use xe_force_wake_ref_has_domain() to determine the status of the call. Modify the return handling of xe_force_wake_get() accordingly and pass the return value to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() - xe_force_wake_put() error doesn't need to be checked. It internally WARNS on domain ack failure. v5 - return unsigned int from xe_force_wake_get() v6 - use helper xe_force_wake_ref_has_domain() Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
xe_force_wake_get() now returns the reference count-incremented domain mask. If it fails for individual domains, the return value will always be 0. However, for XE_FORCEWAKE_ALL, it may return a non-zero value even in the event of failure. Update the return handling of xe_force_wake_get() to reflect this behavior, and ensure that the return value is passed as input to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() v5 - return unsigned int from xe_force_wake_get() - remove redundant warns v7 - Fix commit message Cc: Matthew Brost <matthew.brost@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
xe_force_wake_get() now returns the reference count-incremented domain mask. If it fails for individual domains, the return value will always be 0. However, for XE_FORCEWAKE_ALL, it may return a non-zero value even in the event of failure. Update the return handling of xe_force_wake_get() to reflect this behavior, and ensure that the return value is passed as input to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() - xe_force_wake_put() error doesn't need to be checked. It internally WARNS on domain ack failure. - don't use xe_assert() to report HW errors (Michal) v5 - return unsigned int from xe_force_wake_get() - remove redundant warns v7 - Fix commit message - Remove redundant header Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
With xe_force_wake_get() now returning the refcount-incremented domain mask, a non-zero return value in the case of XE_FORCEWAKE_ALL does not necessarily indicate success. use xe_force_wake_ref_has_domain () to determine the status of the call. Modify the return handling of xe_force_wake_get() accordingly and pass the return value to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() - xe_force_wake_put() error doesn't need to be checked. It internally WARNS on domain ack failure. v5 - return unsigned int from xe_force_wake_get() v6 - Use helper xe_force_wake_ref_has_domain() Cc: Ashutosh Dixit <ashutosh.dixit@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
xe_force_wake_get() now returns the reference count-incremented domain mask. If it fails for individual domains, the return value will always be 0. However, for XE_FORCEWAKE_ALL, it may return a non-zero value even in the event of failure. Update the return handling of xe_force_wake_get() to reflect this behavior, and ensure that the return value is passed as input to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() v5 - return unsigned int from xe_force_wake_get() v7 - Fix commit message Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
xe_force_wake_get() now returns the reference count-incremented domain mask. If it fails for individual domains, the return value will always be 0. However, for XE_FORCEWAKE_ALL, it may return a non-zero value even in the event of failure. Use helper xe_force_wake_ref_has_domain to verify all domains are initialized or not. Update the return handling of xe_force_wake_get() to reflect this behavior, and ensure that the return value is passed as input to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() - xe_force_wake_put() error doesn't need to be checked. It internally WARNS on domain ack failure. v5 - return unsigned int from xe_force_wake_get() - Remove redundant xe_gt_WARN_ON v6 - use helper xe_force_wake_ref_has_domain() v7 - Fix commit message v9 - Rebase Cc: Matthew Brost <matthew.brost@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
With xe_force_wake_get() now returning the refcount-incremented domain mask, a non-zero return value in the case of XE_FORCEWAKE_ALL does not necessarily indicate success. Use xe_force_wake_ref_has_domain() determine the status of the call. Modify the return handling of xe_force_wake_get() accordingly and pass the return value to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() - xe_force_wake_put() error doesn't need to be checked. It internally WARNS on domain ack failure. v5 - return unsigned int for xe_force_wake_get() v6 - use helper xe_force_wake_ref_has_domain() Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
xe_force_wake_get() now returns the reference count-incremented domain mask. If it fails for individual domains, the return value will always be 0. However, for XE_FORCEWAKE_ALL, it may return a non-zero value even in the event of failure. Use helper xe_force_wake_ref_has_domain to verify all domains are initialized or not. Update the return handling of xe_force_wake_get() to reflect this behavior, and ensure that the return value is passed as input to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() - xe_force_wake_put() error doesn't need to be checked. It internally WARNS on domain ack failure. v5 - return unsigned int from xe_force_wake_get() v6 - use xe_force_wake_ref_has_domain() v7 - Fix commit message Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
xe_force_wake_get() now returns the reference count-incremented domain mask. If it fails for individual domains, the return value will always be 0. However, for XE_FORCEWAKE_ALL, it may return a non-zero value even in the event of failure. Update the return handling of xe_force_wake_get() to reflect this behavior, and ensure that the return value is passed as input to xe_force_wake_put(). v3 - return xe_wakeref_t instead of int in xe_force_wake_get() - don't use xe_assert() to report HW errors (Michal) v5 - return unsigned int from xe_force_wake_get() - Remove redundant warn v7 - Fix commit message Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
Himal Ghimiray authored
With xe_force_wake_get() now returning the refcount-incremented domain mask, a return value of 0 indicates failure for single domains. Change assert condition to incorporate this change in return and pass the return value to xe_force_wake_put() v3 - return xe_wakeref_t instead of int in xe_force_wake_get() v5 - return unsigned int for xe_force_wake_get() Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
