aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2023-12-23drm: Warn when freeing a framebuffer that's still on a listVille Syrjälä1-1/+4
Sprinkle some extra WARNs around so that we might catch premature framebuffer destruction more readily. Signed-off-by: Ville Syrjälä <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected] Acked-by: Javier Martinez Canillas <[email protected]>
2023-12-23drm: Don't unref the same fb many times by mistake due to deadlock handlingVille Syrjälä1-0/+1
If we get a deadlock after the fb lookup in drm_mode_page_flip_ioctl() we proceed to unref the fb and then retry the whole thing from the top. But we forget to reset the fb pointer back to NULL, and so if we then get another error during the retry, before the fb lookup, we proceed the unref the same fb again without having gotten another reference. The end result is that the fb will (eventually) end up being freed while it's still in use. Reset fb to NULL once we've unreffed it to avoid doing it again until we've done another fb lookup. This turned out to be pretty easy to hit on a DG2 when doing async flips (and CONFIG_DEBUG_WW_MUTEX_SLOWPATH=y). The first symptom I saw that drm_closefb() simply got stuck in a busy loop while walking the framebuffer list. Fortunately I was able to convince it to oops instead, and from there it was easier to track down the culprit. Cc: [email protected] Signed-off-by: Ville Syrjälä <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected] Acked-by: Javier Martinez Canillas <[email protected]>
2023-12-22Merge tag 'drm-habanalabs-next-2023-12-19' of ↵Dave Airlie16-184/+333
https://git.kernel.org/pub/scm/linux/kernel/git/ogabbay/linux into drm-next This tag contains habanalabs driver changes for v6.8. The notable changes are: - uAPI changes: - Add sysfs entry to allow users to identify a device minor id with its debugfs path - Add sysfs entry to expose the device's module id as given to us from the f/w - Add signed device information retrieval through the INFO ioctl - New features and improvements: - Update documentation of debugfs paths - Add support for Gaudi2C device (new PCI revision number) - Add pcie reset prepare/done hooks - Firmware related fixes and changes: - Print three instances version numbers of Infineon second stage - Assume hard-reset is done by f/w upon PCIe AXI drain - Bug fixes and code cleanups: - Fix information leak in sec_attest_info() - Avoid overriding existing undefined opcode data in Gaudi2 - Multiple Queue Manager (QMAN) fixes for Gaudi2 - Set hard reset flag if graceful reset is skipped - Remove 'get temperature' debug print - Fix the new Event Queue heartbeat mechanism Signed-off-by: Dave Airlie <[email protected]> From: Oded Gabbay <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/ZYFpihZscr/[email protected]
2023-12-22Merge tag 'drm-misc-next-fixes-2023-12-21' of ↵Dave Airlie6-35/+36
git://anongit.freedesktop.org/drm/drm-misc into drm-next More fixes for the new imagination drier, a DT node refcount fix for the new aux bridge driver and a missing header fix for the LUT management code. Signed-off-by: Dave Airlie <[email protected]> From: Maxime Ripard <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/42dw6ok2g5kz5xljrw7t6lzrgafhwslgw3j4rbaaivluv24vkj@k4smx5r3y2gh
2023-12-22drm/xe: Disable 32bits buildLucas De Marchi1-1/+1
Add a dependency on CONFIG_64BIT since currently the xe driver doesn't build on 32bits. It may be enabled again after all the issues are fixed. Signed-off-by: Lucas De Marchi <[email protected]> Reviewed-by: Rodrigo Vivi <[email protected]> Signed-off-by: Dave Airlie <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2023-12-22Merge tag 'drm-xe-next-2023-12-21-pr1-1' of ↵Dave Airlie352-1/+61425
https://gitlab.freedesktop.org/drm/xe/kernel into drm-next Introduce a new DRM driver for Intel GPUs Xe, is a new driver for Intel GPUs that supports both integrated and discrete platforms. The experimental support starts with Tiger Lake. i915 will continue be the main production driver for the platforms up to Meteor Lake and Alchemist. Then the goal is to make this Intel Xe driver the primary driver for Lunar Lake and newer platforms. It uses most, if not all, of the key drm concepts, in special: TTM, drm-scheduler, drm-exec, drm-gpuvm/gpuva and others. Signed-off-by: Dave Airlie <[email protected]> [airlied: add an extra X86 check, fix a typo, fix drm_exec_init interface change]. From: Rodrigo Vivi <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2023-12-21drm/xe: Remove uninitialized variable from warningLucas De Marchi1-2/+2
"err" is not initialized when failing to create and add the freq0 sysfs file. Remove it from the message. This fixes the following warning with clang: ../drivers/gpu/drm/xe/xe_gt_freq.c:202:30: error: variable 'err' is uninitialized when used here [-Werror,-Wuninitialized] kobject_name(gt->sysfs), err); ^~~ Fixes: bef52b5c7a19 ("drm/xe: Create a xe_gt_freq component for raw management and sysfs") Reviewed-by: Michał Winiarski <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Lucas De Marchi <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Remove ci-only GuC FW definitionsDaniele Ceraolo Spurio1-2/+0
As part of the FW definitions, we declare each blob as required via the MODULE_FIRMWARE() macro. This causes the initramfs update (or equivalent process) to look for the blobs on disk when the kernel is installed; therefore, we need to make sure that all FWs we define are available in linux-firmware. We currently don't plan to push the PVC blob to linux-firmware, while the LNL one will only be pushed once we have machines in CI to test it, so we need to remove them from the list for now. Signed-off-by: Daniele Ceraolo Spurio <[email protected]> Cc: Rodrigo Vivi <[email protected]> Reviewed-by: Rodrigo Vivi <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Fix build without CONFIG_FAULT_INJECTIONRodrigo Vivi2-2/+2
Ideally this header could be included without the CONFIG_FAULT_INJECTION and it would take care itself for the includes it needs. So, let's temporary workaround this by moving this below and including only when CONFIG_FAULT_INJECTION is selected to avoid build breakages. Another solution would be us including the linux/types.h as well, but this creates unnecessary cases. Reference: https://lore.kernel.org/all/[email protected]/ Cc: Himal Prasad Ghimiray <[email protected]> Cc: Oded Gabbay <[email protected]> Cc: Thomas Hellström <[email protected]> Cc: Lucas De Marchi <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]> Reviewed-by: Thomas Hellström <[email protected]>
2023-12-21MAINTAINERS: Updates to Intel DRMRodrigo Vivi1-1/+28
Introduce the Maintainers of the new drm/xe driver for upcoming Intel GPUs. Since it has a shared display with drm/i915, let's also create a dedicated block to group display related files. But without any substantial change to the i915 side. The display patches will continue to flow through i915 from drm-intel-next branches for now. Acked-by: Jani Nikula <[email protected]> Acked-by: Joonas Lahtinen <[email protected]> Acked-by: Tvrtko Ursulin <[email protected]> Acked-by: Lucas De Marchi <[email protected]> Acked-by: Oded Gabbay <[email protected]> Acked-by: Thomas Hellström <[email protected]> Acked-by: Dave Airlie <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Remove reset uevent for nowRodrigo Vivi2-29/+0
This kernel uevent is getting removed for now. It will come back later with a better future proof name. v2: Rebase (Francois Dugast) Cc: Himal Prasad Ghimiray <[email protected]> Cc: Lucas De Marchi <[email protected]> Cc: Francois Dugast <[email protected]> Cc: Aravind Iddamsetty <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]> Reviewed-by: Himal Prasad Ghimiray <[email protected]> Acked-by: Lucas De Marchi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]>
2023-12-21drm/xe/uapi: Move DRM_XE_ACC_GRANULARITY_* where they are usedFrancois Dugast1-14/+8
Bring those defines close to the context where they can be used. Also apply indentation as it is done for other subsets of defines. Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Move CPU_CACHING defines before docFrancois Dugast1-2/+2
Move those defines to align on the rule used elsewhere in the file which was introduced by commit 4f082f2c3a37 ("drm/xe: Move defines before relevant fields"). Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Add examples of user space codeFrancois Dugast1-0/+84
Complete the documentation of some structs by adding functional examples of user space code. Those examples are intentionally kept very simple. Put together, they provide a foundation for a minimal application that executes a job using the Xe driver. v2: Remove use of DRM_XE_VM_BIND_FLAG_ASYNC (Francois Dugast) Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Add block diagram of a deviceFrancois Dugast1-0/+39
In order to make proper use the uAPI, a prerequisite is to understand some key concepts about the discrete GPU devices which are supported by the Xe driver. For example, some structs defined in the uAPI are an abstraction of a hardware component with a specific role. This diagram helps to build a mental representation of a device how it is seen by the Xe driver. As written in the documentation, it does not intend to be a literal representation of an existing device. A lot more information could be added but the intention for the overview is to keep it simple, and go into detail as needed in other sections. v2: Add GT1 inside Tile0 (José Roberto de Souza) Reviewed-by: José Roberto de Souza <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Document the memory_region bitmaskRodrigo Vivi2-5/+37
The uAPI should stay generic in regarding to the bitmask. It is the userspace responsibility to check for the type/class of the memory, without any assumption. Also add comments inside the code to explain how it is actually constructed so we don't accidentally change the assignment of the instance and the masks. No functional change in this patch. It only explains and document the memory_region masks. A further follow-up work with the organization of all memory regions around struct xe_mem_regions is desired, but not part of this patch. Signed-off-by: Rodrigo Vivi <[email protected]> Reviewed-by: Lucas De Marchi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]>
2023-12-21drm/xe/uapi: More uAPI documentation additions and cosmetic updatesRodrigo Vivi1-7/+40
No functional change in this patch. Let's ensure all of our structs are documented and with a certain standard. Also, let's have an overview and list of IOCTLs as the very beginning of the generated HTML doc. v2: Nits (Lucas De Marchi) Signed-off-by: Rodrigo Vivi <[email protected]> Reviewed-by: Lucas De Marchi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]>
2023-12-21drm/xe/uapi: Order sectionsRodrigo Vivi1-122/+130
This patch doesn't modify any text or uapi entries themselves. It only move things up and down aiming a better organization of the uAPI. While fixing the documentation I noticed that query_engine_cs_cycles was in the middle of the memory_region info. Then I noticed more mismatches on the order when compared to the order of the IOCTL and QUERY entries declaration. So this patch aims to bring some order to the uAPI so it gets easier to read and the documentation generated in the end is able to tell a consistent story. Overall order: 1. IOCTL definition 2. Extension definition and helper structs 3. IOCTL's Query structs in the order of the Query's entries. 4. The rest of IOCTL structs in the order of IOCTL declaration. 5. uEvents Signed-off-by: Rodrigo Vivi <[email protected]> Reviewed-by: Lucas De Marchi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]>
2023-12-21drm/xe/uapi: Make constant comments visible in kernel docFrancois Dugast1-116/+155
As there is no direct way to make comments of constants directly visible in the kernel doc, move them to the description of the structure where they can be used. By doing so they appear in the "Description" section of the struct documentation. v2: Remove DRM_XE_UFENCE_WAIT_MASK_* (Francois Dugast) Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Document DRM_XE_DEVICE_QUERY_HWCONFIGFrancois Dugast1-0/+5
Add a documentation on the content and format of when using query type DRM_XE_DEVICE_QUERY_HWCONFIG. The list of keys can be found in IGT under lib/intel_hwconfig_types.h. Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Document drm_xe_query_config keysFrancois Dugast1-0/+20
Provide a description of the keys used the struct drm_xe_query_config info array. Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/637 Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Document use of size in drm_xe_device_queryFrancois Dugast1-3/+9
Document the behavior of the driver for IOCTL DRM_IOCTL_XE_DEVICE_QUERY depending on the size value provided in struct drm_xe_device_query. Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Add missing documentation for struct membersFrancois Dugast1-2/+4
This removes the documentation build warnings below: include/uapi/drm/xe_drm.h:828: warning: Function parameter or \ member 'pad2' not described in 'drm_xe_vm_bind_op' include/uapi/drm/xe_drm.h:875: warning: Function parameter or \ member 'pad2' not described in 'drm_xe_vm_bind' include/uapi/drm/xe_drm.h:1006: warning: Function parameter or \ member 'handle' not described in 'drm_xe_sync' include/uapi/drm/xe_drm.h:1006: warning: Function parameter or \ member 'timeline_value' not described in 'drm_xe_sync' Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Add a comment to each structFrancois Dugast1-3/+40
Add a comment to each struct to complete documentation, ensure all struct appear in the kernel doc, and bind structs to IOCTLs. Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Remove sync bindsMatthew Brost5-88/+20
Remove concept of async vs sync VM bind queues, rather make all binds async. The following bits have dropped from the uAPI: DRM_XE_ENGINE_CLASS_VM_BIND_ASYNC DRM_XE_ENGINE_CLASS_VM_BIND_SYNC DRM_XE_VM_CREATE_FLAG_ASYNC_DEFAULT DRM_XE_VM_BIND_FLAG_ASYNC To implement sync binds the UMD is expected to use the out-fence interface. v2: Send correct version v3: Drop drm_xe_syncs Cc: Rodrigo Vivi <[email protected]> Cc: Thomas Hellström <[email protected]> Cc: Francois Dugast <[email protected]> Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: Thomas Hellström <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Ensure every uapi struct has drm_xe prefixRodrigo Vivi2-10/+10
To ensure consistency and avoid possible later conflicts, let's add drm_xe prefix to xe_user_extension struct. Cc: Francois Dugast <[email protected]> Suggested-by: Lucas De Marchi <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]> Reviewed-by: Matthew Brost <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]>
2023-12-21drm/xe/pmu: Remove PMU from Xe till uapi is finalizedAshutosh Dixit11-798/+2
PMU uapi is likely to change in the future. Till the uapi is finalized, remove PMU from Xe. PMU can be re-added after uapi is finalized. v2: Include xe_drm.h in xe/tests/xe_dma_buf.c (Francois) Signed-off-by: Ashutosh Dixit <[email protected]> Acked-by: Aravind Iddamsetty <[email protected]> Acked-by: Lucas De Marchi <[email protected]> Reviewed-by: Umesh Nerlige Ramappa <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Remove DRM_XE_UFENCE_WAIT_MASK_*Francois Dugast1-5/+7
Those are just possible values for the comparison mask but they are not specific magic values. Let's keep them as examples in the documentation but remove them from the uAPI. Suggested-by: Matthew Brost <[email protected]> Cc: Rodrigo Vivi <[email protected]> Reviewed-by: Matthew Brost <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Remove DRM_IOCTL_XE_EXEC_QUEUE_SET_PROPERTYFrancois Dugast1-1/+0
The exec_queue_set_property feature was removed in a previous commit 0f1d88f27864 ("drm/xe/uapi: Kill exec_queue_set_property") and is no longer usable, struct drm_xe_exec_queue_set_property does not exist anymore, so let's remove this. Reviewed-by: Lucas De Marchi <[email protected]> Acked-by: Rodrigo Vivi <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: Return correct error code for xe_wait_user_fence_ioctlBommu Krishnaiah4-6/+43
Currently xe_wait_user_fence_ioctl is not checking exec_queue state and blocking until timeout, with this patch wakeup the blocking wait if exec_queue reset happen and returning proper error code Signed-off-by: Bommu Krishnaiah <[email protected]> Cc: Oak Zeng <[email protected]> Cc: Kempczynski Zbigniew <[email protected]> Cc: Matthew Brost <[email protected]> Reviewed-by: Rodrigo Vivi <[email protected]> Reviewed-by: Matthew Brost <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/uapi: add exec_queue_id member to drm_xe_wait_user_fence structureBommu Krishnaiah2-75/+7
remove the num_engines/instances members from drm_xe_wait_user_fence structure and add a exec_queue_id member Right now this is only checking if the engine list is sane and nothing else. In the end every operation with this IOCTL is a soft check. So, let's formalize that and only use this IOCTL to wait on the fence. exec_queue_id member will help to user space to get proper error code from kernel while in exec_queue reset Signed-off-by: Bommu Krishnaiah <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]> Acked-by: Matthew Brost <[email protected]> Reviewed-by: Francois Dugast <[email protected]> Acked-by: José Roberto de Souza <[email protected]> Acked-by: Mateusz Naklicki <[email protected]> Signed-off-by: Francois Dugast <[email protected]>
2023-12-21drm/xe: Enable W=1 warnings by defaultLucas De Marchi1-19/+28
Like done in commit 2250c7ead8ad ("drm/i915: enable W=1 warnings by default") for i915, enable W=1 warnings by default in xe. Reviewed-by: Rodrigo Vivi <[email protected]> Acked-by: Jani Nikula <[email protected]> Signed-off-by: Lucas De Marchi <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/display: Fix dummy __i915_inject_probe_error()Lucas De Marchi1-0/+4
When CONFIG_DRM_I915_DEBUG is not set, a dummy __i915_inject_probe_error() is provided on the xe side. Use the same logic as in drivers/gpu/drm/i915/i915_utils.c to ifdef it out. This fixes the build with W=1 and without that config: CC [M] drivers/gpu/drm/xe/display/ext/i915_utils.o ../drivers/gpu/drm/xe/display/ext/i915_utils.c:19:5: error: no previous prototype for ‘__i915_inject_probe_error’ [-Werror=missing-prototypes] 19 | int __i915_inject_probe_error(struct drm_i915_private *i915, int err, | ^~~~~~~~~~~~~~~~~~~~~~~~~ Signed-off-by: Lucas De Marchi <[email protected]> Reviewed-by: Rodrigo Vivi <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/bo: Remove unusued variableLucas De Marchi1-4/+0
bo is not used since all the checks are against tbo. Fix warning: ../drivers/gpu/drm/xe/xe_bo.c: In function ‘xe_evict_flags’: ../drivers/gpu/drm/xe/xe_bo.c:250:23: error: variable ‘bo’ set but not used [-Werror=unused-but-set-variable] 250 | struct xe_bo *bo; Signed-off-by: Lucas De Marchi <[email protected]> Reviewed-by: Rodrigo Vivi <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Return error if drm_buddy_init() failsLucas De Marchi1-0/+2
Fix warning: ../drivers/gpu/drm/xe/xe_ttm_vram_mgr.c: In function ‘__xe_ttm_vram_mgr_init’: ../drivers/gpu/drm/xe/xe_ttm_vram_mgr.c:340:13: error: variable ‘err’ set but not used [-Werror=unused-but-set-variable] 340 | int err; | ^~~ Check for the error return and return it, like done by other drivers. Reviewed-by: Matt Roper <[email protected]> Signed-off-by: Lucas De Marchi <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Drop some unnecessary header includesMatt Roper6-7/+0
Several files were including register headers that they no longer require. Drop the unnecessary includes to reduce build dependencies. Reviewed-by: Lucas De Marchi <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Matt Roper <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Re-sort GT register headerMatt Roper1-38/+38
Keeping the register definitions sorted will make it easy to find existing definitions and prevent accidental introduction of duplicate definitions. v2: - Reorder FUSE3/FUSE4 registers and move GT0_PERF_LIMIT_REASONS / MTL_MEDIA_PERF_LIMIT_REASONS to proper places. (Lucas) Reviewed-by: Lucas De Marchi <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Matt Roper <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Define interrupt vector bits with the interrupt registersMatt Roper3-7/+5
The bit definitions had become a bit orphaned; move them to the same location as the interrupt registers that they're used with. Reviewed-by: Lucas De Marchi <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Matt Roper <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Move GSC HECI base offsets out of register headerMatt Roper2-4/+4
These offsets are only used to setup the auxiliary device BAR information and are never used for driver read/write operations. Move them to the GSC HECI file where they're actually used. Reviewed-by: Lucas De Marchi <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Matt Roper <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Move engine base offsets to engine register headerMatt Roper2-28/+33
These offsets are primarily used as parameters for the engine register definitions, so it makes more sense to define them in the engine header rather than the general register header. Reviewed-by: Lucas De Marchi <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Matt Roper <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Fix whitespace in register definitionsMatt Roper2-5/+5
Our register headers use tabs to align the definition values. Convert a few definitions that were using spaces instead. Reviewed-by: Lucas De Marchi <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Matt Roper <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Move some per-engine register definitions to the engine headerMatt Roper4-28/+30
Although we only work with the RCS instances today, the FF_SLICE_CS_CHICKEN1[1,2] CS_DEBUG_MODE1, CS_CHICKEN1, and FF_THREAD_MODE registers all have instances on both the RCS and CCS engines. Convert these to parameterized macros and move them to the engine register header. Reviewed-by: Lucas De Marchi <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Matt Roper <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe: Drop "_REG" suffix from CSFE_CHICKEN1Matt Roper2-6/+5
We don't use this suffix on any other registers, and it isn't part of the register's official name either, so drop it for consistency. While at it, move the register definition slightly so that it isn't separating RING_CMD_CCTL's definition from its fields. Reviewed-by: Lucas De Marchi <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Matt Roper <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/xe2: Support flat ccsHimal Prasad Ghimiray1-1/+1
Enable flat ccs for XE2_GFX_FEATURES. Cc: Thomas Hellström <[email protected]> Signed-off-by: Himal Prasad Ghimiray <[email protected]> Reviewed-by: Thomas Hellström <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/xe2: Modify xe_bo_test for system memoryHimal Prasad Ghimiray1-17/+27
Modify test to valid ccs clear and copy during evict/restore on igfx. v2: -Vram is associated with tiles not with gt. Use tile based iterator for ccs_test_run_gt. (Matt) Cc: Matt Roper <[email protected]> Cc: Thomas Hellström <[email protected]> Signed-off-by: Himal Prasad Ghimiray <[email protected]> Reviewed-by: Thomas Hellström <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/xe2: Handle flat ccs move for igfx.Himal Prasad Ghimiray5-42/+66
- Clear flat ccs during user bo creation. - copy ccs meta data between flat ccs and bo during eviction and restore. - Add a bool field ccs_cleared in bo, true means ccs region of bo is already cleared. v2: - Rebase. v3: - Maintain order of xe_bo_move_notify for ttm_bo_type_sg. v4: - xe_migrate_copy can be used to copy src to dst bo on igfx too. Add a bool which handles only ccs metadata copy. v5: - on dgfx ccs should be cleared even if the bo is not compression enabled. Cc: Thomas Hellström <[email protected]> Signed-off-by: Himal Prasad Ghimiray <[email protected]> Reviewed-by: Thomas Hellström <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/xe2: Update emit_pte to use compression enabled PAT indexHimal Prasad Ghimiray4-7/+18
For indirect accessed buffer use compression enabled PAT index. v2: - Fix parameter name. v3: - use a relevant define instead of fix number. Cc: Thomas Hellström <[email protected]> Cc: Matthew Auld <[email protected]> Signed-off-by: Himal Prasad Ghimiray <[email protected]> Reviewed-by: Thomas Hellström <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/xe2: Update chunk size for each iteration of ccs copyHimal Prasad Ghimiray1-11/+25
In xe2 platform XY_CTRL_SURF_COPY_BLT can handle ccs copy for max of 1024 main surface pages. v2: - Use better logic to determine chunk size (Matt/Thomas) v3: - use function instead of macro(Thomas) Cc: Matt Roper <[email protected]> Cc: Thomas Hellström <[email protected]> Signed-off-by: Himal Prasad Ghimiray <[email protected]> Reviewed-by: Thomas Hellström <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/xe_migrate: Use NULL 1G PTE mapped at 255GiB VA for ccs clearHimal Prasad Ghimiray1-54/+11
Get rid of the cleared bo, instead use null 1G PTE mapped at 255GiB offset, this can be used for both dgfx and igfx. v2: - Remove xe_migrate::cleared_bo. - Add a comment for NULL mapping.(Thomas) Cc: Thomas Hellström <[email protected]> Signed-off-by: Himal Prasad Ghimiray <[email protected]> Reviewed-by: Thomas Hellström <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
2023-12-21drm/xe/xe2: Updates on XY_CTRL_SURF_COPY_BLTHimal Prasad Ghimiray2-9/+17
- The XY_CTRL_SURF_COPY_BLT instruction operating on ccs data expects size in pages of main memory for which CCS data should be copied. - The bitfield representing copy size in XY_CTRL_SURF_COPY_BLT has shifted one bit higher in the instruction. v2: - Fix the num_pages for ccs size calculation. - Address nits (Thomas) v3: - Use FIELD_PREP and FIELD_FIT instead of shifts and numbers.(Matt) Cc: Matt Roper <[email protected]> Cc: Thomas Hellström <[email protected]> Signed-off-by: Himal Prasad Ghimiray <[email protected]> Reviewed-by: Matt Roper <[email protected]> Reviewed-by: Thomas Hellström <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.