- 09 Jan, 2024 8 commits
-
-
Charlene Liu authored
Adjust z8 latency for performance. Reviewed-by:
Muhammad Ahmed <ahmed.ahmed@amd.com> Acked-by:
Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by:
Charlene Liu <charlene.liu@amd.com> Tested-by:
Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by:
Alex Deucher <alexander.deucher@amd.com>
-
Daniel Miess authored
This reverts commit d0f639c5. The previous commit causes failure to light up for 1080p eDP + 8k HDMI panel combo. Reviewed-by:
Daniel Miess <daniel.miess@amd.com> Tested-by:
-
Peichen Huang authored
[WHY] When usb4 bandwidth allocation mode is enabled, driver need to request bandwidth from connection manager. For mst link, the requested bandwidth should be big enough for all remote streams. [HOW] - If mst link, the requested bandwidth should be the sum of all mst streams bandwidth added with dp MTPH overhead. - Allocate/deallcate usb4 bandwidth when setting dpms on/off. - When doing display mode validation, driver also need to consider total bandwidth of all mst streams for mst link. Reviewed-by:
Cruise Hung <cruise.hung@amd.com> Acked-by:
Peichen Huang <peichen.huang@amd.com> Tested-by:
-
Martin Leung authored
This reverts commit 6e4337f6. The original commit causes regression in corner case with HDMI at specific timings. Reverting from staging to get the full suite to retest. Reviewed-by:
Martin Leung <martin.leung@amd.com> Tested-by:
-
Martin Leung authored
This reverts commit 6b2b782a. The original commit causes issues with certain features when DRR is disabled, need to revisit this change later after resolving issues with new DRR policy. Reviewed-by:
-
George Shen authored
[Why] If an OPP is used for a different OPTC without first being disconnected from the previous OPTC, unexpected behaviour can occur. This also applies to phantom pipes, which is what the current logic missed. [How] Disconnect OPPs from OPTC for phantom pipes before disabling OTG master. Also move the disconnection to before the OTG master disable, since the register is double buffered. Reviewed-by:
Dillon Varone <dillon.varone@amd.com> Acked-by:
George Shen <george.shen@amd.com> Tested-by:
-
Martin Tsai authored
[Why] Panels show corruption with high refresh rate timings when ssc is enabled. [How] Read down-spread percentage from lut to adjust dprefclk. Issues come from S0i3 with this commit has been fixed by SMU. Reviewed-by:
Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Acked-by:
Martin Tsai <martin.tsai@amd.com> Tested-by:
-
Felix Kuehling authored
====================================================== WARNING: possible circular locking dependency detected 6.5.0-kfd-fkuehlin #276 Not tainted ------------------------------------------------------ kworker/8:2/2676 is trying to acquire lock: ffff9435aae95c88 ((work_completion)(&svm_bo->eviction_work)){+.+.}-{0:0}, at: __flush_work+0x52/0x550 but task is already holding lock: ffff9435cd8e1720 (&svms->lock){+.+.}-{3:3}, at: svm_range_deferred_list_work+0xe8/0x340 [amdgpu] which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #2 (&svms->lock){+.+.}-{3:3}: __mutex_lock+0x97/0xd30 kfd_ioctl_alloc_memory_of_gpu+0x6d/0x3c0 [amdgpu] kfd_ioctl+0x1b2/0x5d0 [amdgpu] __x64_sys_ioctl+0x86/0xc0 do_syscall_64+0x39/0x80 entry_SYSCALL_64_after_hwframe+0x63/0xcd -> #1 (&mm->mmap_lock){++++}-{3:3}: down_read+0x42/0x160 svm_range_evict_svm_bo_worker+0x8b/0x340 [amdgpu] process_one_work+0x27a/0x540 worker_thread+0x53/0x3e0 kthread+0xeb/0x120 ret_from_fork+0x31/0x50 ret_from_fork_asm+0x11/0x20 -> #0 ((work_completion)(&svm_bo->eviction_work)){+.+.}-{0:0}: __lock_acquire+0x1426/0x2200 lock_acquire+0xc1/0x2b0 __flush_work+0x80/0x550 __cancel_work_timer+0x109/0x190 svm_range_bo_release+0xdc/0x1c0 [amdgpu] svm_range_free+0x175/0x180 [amdgpu] svm_range_deferred_list_work+0x15d/0x340 [amdgpu] process_one_work+0x27a/0x540 worker_thread+0x53/0x3e0 kthread+0xeb/0x120 ret_from_fork+0x31/0x50 ret_from_fork_asm+0x11/0x20 other info that might help us debug this: Chain exists of: (work_completion)(&svm_bo->eviction_work) --> &mm->mmap_lock --> &svms->lock Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&svms->lock); lock(&mm->mmap_lock); lock(&svms->lock); lock((work_completion)(&svm_bo->eviction_work)); I believe this cannot really lead to a deadlock in practice, because svm_range_evict_svm_bo_worker only takes the mmap_read_lock if the BO refcount is non-0. That means it's impossible that svm_range_bo_release is running concurrently. However, there is no good way to annotate this. To avoid the problem, take a BO reference in svm_range_schedule_evict_svm_bo instead of in the worker. That way it's impossible for a BO to get freed while eviction work is pending and the cancel_work_sync call in svm_range_bo_release can be eliminated. v2: Use svm_bo_ref_unless_zero and explained why that's safe. Also removed redundant checks that are already done in amdkfd_fence_enable_signaling. Signed-off-by:Felix Kuehling <felix.kuehling@amd.com> Reviewed-by:
Philip Yang <philip.yang@amd.com> Signed-off-by:
-
- 08 Jan, 2024 1 commit
-
-
https://gitlab.freedesktop.org/agd5f/linuxDave Airlie authored
amd-drm-next-6.8-2024-01-05: amdgpu: - VRR fixes - PSR-SU fixes - SubVP fixes - DCN 3.5 fixes - Documentation updates - DMCUB fixes - DML2 fixes - UMC 12.0 updates - GPUVM fix - Misc code cleanups and whitespace cleanups - DP MST fix - Let KFD sync with GPUVM fences - GFX11 reset fix - SMU 13.0.6 fixes - VSC fix for DP/eDP - Navi12 display fix - RN/CZN system aperture fix - DCN 2.1 bandwidth validation fix - DCN INIT cleanup amdkfd: - SVM fixes - Revert TBA/TMA location change Signed-off-by:
Dave Airlie <airlied@redhat.com> From: Alex Deucher <alexander.deucher@amd.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240105220522.4976-1-alexander.deucher@amd.com
-
- 05 Jan, 2024 29 commits
-
-
Charlene Liu authored
Copy StutterPeriod from DML2 into DML1 StutterPeriod parameter. Tested-by:
Aric Cyr <aric.cyr@amd.com> Acked-by:
-
Melissa Wen authored
IGT `amdgpu/amd_color/crtc-lut-accuracy` fails right at the beginning of the test execution, during atomic check, because DC rejects the bandwidth state for a fb sizing 64x64. The test was previously working with the deprecated dc_commit_state(). Now using dc_validate_with_context() approach, the atomic check needs to perform a full state validation. Therefore, set fast_validation to false in the dc_validate_global_state call for atomic check. Cc: stable@vger.kernel.org Fixes: b8272241 ("drm/amd/display: Drop dc_commit_state in favor of dc_commit_streams") Signed-off-by:
Melissa Wen <mwen@igalia.com> Signed-off-by:
Hamza Mahfooz <hamza.mahfooz@amd.com> Signed-off-by:
-
Alex Deucher authored
These chips needs the same fix. This was previously not seen on then since the AGP aperture expanded the system aperture, but this showed up again when AGP was disabled. Reviewed-and-tested-by:
Jiadong Zhu <Jiadong.Zhu@amd.com> Signed-off-by:
-
Harry Wentland authored
Other environments don't like the unary minus operator on an unsigned value. Signed-off-by:
Harry Wentland <harry.wentland@amd.com> Acked-by:
-
Harry Wentland authored
- Use tabs, not spaces. - Brace and parentheses placement Signed-off-by:
-
Kaibo Ma authored
That commit causes NULL pointer dereferences in dmesgs when running applications using ROCm, including clinfo, blender, and PyTorch, since v6.6.1. Revert it to fix blender again. This reverts commit 96c211f1. Closes: https://github.com/ROCm/ROCm/issues/2596 Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/2991Reviewed-by:
Jay Cornwall <jay.cornwall@amd.com> Signed-off-by:
Kaibo Ma <ent3rm4n@gmail.com> Signed-off-by:
-
Arnd Bergmann authored
gcc prints a warning about a possible array overflow for a couple of callers of dp_decide_lane_settings() after commit 1b56c90018f0 ("Makefile: Enable -Wstringop-overflow globally"): drivers/gpu/drm/amd/amdgpu/../display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c: In function 'dp_perform_fixed_vs_pe_training_sequence_legacy': drivers/gpu/drm/amd/amdgpu/../display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c:426:25: error: 'dp_decide_lane_settings' accessing 4 bytes in a region of size 1 [-Werror=stringop-overflow=] 426 | dp_decide_lane_settings(lt_settings, dpcd_lane_adjust, | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 427 | lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings); | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/gpu/drm/amd/amdgpu/../display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c:426:25: note: referencing argument 4 of type 'union dpcd_training_lane[4]' I'm not entirely sure what caused this, but changing the prototype to expect a pointer instead of an array avoids the warnings. Fixes: 7727e7b6 ("drm/amd/display: Improve robustness of FIXED_VS link training at DP1 rates") Acked-by:Randy Dunlap <rdunlap@infradead.org> Tested-by: Randy Dunlap <rdunlap@infradead.org> # build-tested Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
-
Marcelo Mendes Spessoto Junior authored
Place define macro expression inside () in power_helpers.c file Signed-off-by:
Marcelo Mendes Spessoto Junior <marcelomspessoto@gmail.com> Signed-off-by:
-
Marcelo Mendes Spessoto Junior authored
Place HDCP_EVENT_TRACE(hdcp, event) macro content inside do while loop to avoid if-else issues in hdcp_log.h file v2: fix up build (Alex) Signed-off-by:
-
Marcelo Mendes Spessoto Junior authored
Remove braces for single statement if expressions and change comparison order for hdcp2_execution.c file Signed-off-by:
-
Marcelo Mendes Spessoto Junior authored
Fix identation inside enum and place expressions in define macros inside () for hdcp_psp.h file Signed-off-by:
-
Marcelo Mendes Spessoto Junior authored
Remove braces for single statement if expression for freesync.c file Signed-off-by:
-
Marcelo Mendes Spessoto Junior authored
Fix identation for hdcp_psp.c file Signed-off-by:
-
Marcelo Mendes Spessoto Junior authored
Remove braces from single statement if expression in hdcp1_execution.c file Signed-off-by:
-
Zhipeng Lu authored
The hwmgr->backend, (i.e. data) allocated by kzalloc is not freed in the error-handling paths of smu7_get_evv_voltages and smu7_update_edc_leakage_table. However, it did be freed in the error-handling of phm_initializa_dynamic_state_adjustment_rule_settings, by smu7_hwmgr_backend_fini. So the lack of free in smu7_get_evv_voltages and smu7_update_edc_leakage_table is considered a memleak in this patch. Fixes: 599a7e9f ("drm/amd/powerplay: implement smu7 hwmgr to manager asics with smu ip version 7.") Fixes: 8f0804c6 ("drm/amd/pm: add edc leakage controller setting") Signed-off-by:
Zhipeng Lu <alexious@zju.edu.cn> Signed-off-by:
-
Srinivasan Shanmugam authored
Fix the following about iterator use: drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c:1456 kfd_add_peer_prop() warn: iterator used outside loop: 'iolink3' Cc: Felix Kuehling <Felix.Kuehling@amd.com> Cc: Christian König <christian.koenig@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Signed-off-by:
Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by:
-
Srinivasan Shanmugam authored
Return value of container_of(...) can't be null, so null check is not required for 'fence'. Hence drop its NULL check. Fixes the below: drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c:93 to_amdgpu_amdkfd_fence() warn: can 'fence' even be NULL? Cc: Felix Kuehling <Felix.Kuehling@amd.com> Cc: Christian König <christian.koenig@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Signed-off-by:
-
Srinivasan Shanmugam authored
Before using list_first_entry, make sure to check that list is not empty, if list is empty return -ENODATA. Fixes the below: drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c:1347 kfd_create_indirect_link_prop() warn: can 'gpu_link' even be NULL? drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c:1428 kfd_add_peer_prop() warn: can 'iolink1' even be NULL? drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_topology.c:1433 kfd_add_peer_prop() warn: can 'iolink2' even be NULL? Fixes: 0f28cca8 ("drm/amdkfd: Extend KFD device topology to surface peer-to-peer links") Cc: Felix Kuehling <Felix.Kuehling@amd.com> Cc: Christian König <christian.koenig@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Signed-off-by:
Lijo Lazar <lijo.lazar@amd.com> Reviewed-by:
-
Srinivasan Shanmugam authored
Fixes the below: drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c:1404 amdgpu_ucode_request() warn: '*fw' from request_firmware() not released on lines: 1404. Cc: Mario Limonciello <mario.limonciello@amd.com> Cc: Lijo Lazar <lijo.lazar@amd.com> Cc: Christian König <christian.koenig@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Signed-off-by:
Christian König <christian.koenig@amd.com> Signed-off-by:
-
Srinivasan Shanmugam authored
drm/amdgpu: Fix variable 'mca_funcs' dereferenced before NULL check in 'amdgpu_mca_smu_get_mca_entry()' Fixes the below: drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c:377 amdgpu_mca_smu_get_mca_entry() warn: variable dereferenced before check 'mca_funcs' (see line 368) 357 int amdgpu_mca_smu_get_mca_entry(struct amdgpu_device *adev, enum amdgpu_mca_error_type type, 358 int idx, struct mca_bank_entry *entry) 359 { 360 const struct amdgpu_mca_smu_funcs *mca_funcs = adev->mca.mca_funcs; 361 int count; 362 363 switch (type) { 364 case AMDGPU_MCA_ERROR_TYPE_UE: 365 count = mca_funcs->max_ue_count; mca_funcs is dereferenced here. 366 break; 367 case AMDGPU_MCA_ERROR_TYPE_CE: 368 count = mca_funcs->max_ce_count; mca_funcs is dereferenced here. 369 break; 370 default: 371 return -EINVAL; 372 } 373 374 if (idx >= count) 375 return -EINVAL; 376 377 if (mca_funcs && mca_funcs->mca_get_mca_entry) ^^^^^^^^^ Checked too late! Cc: Yang Wang <kevinyang.wang@amd.com> Cc: Hawking Zhang <Hawking.Zhang@amd.com> Cc: Christian König <christian.koenig@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Signed-off-by:Yang Wang <kevinyang.wang@amd.com> Signed-off-by:
-
Le Ma authored
This param can help isolating data path issues on new systems in early phase. Signed-off-by:
Le Ma <le.ma@amd.com> Reviewed-by:
Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by:
-
Srinivasan Shanmugam authored
Fixes the below: drivers/gpu/drm/amd/amdgpu/atom.c:398 atom_skip_src_int() warn: ignoring unreachable code. Cc: Christian König <christian.koenig@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Signed-off-by:
-
Marcelo Mendes Spessoto Junior authored
mod_freesync header file has duplicated copyright boilerplate. Drop the duplicate. Signed-off-by:
-
Alex Deucher authored
It's no longer required. Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2318Reviewed-by:
-
Alex Deucher authored
This was included in gpu_info firmware, move it into the driver for consistency with other nv1x parts. Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2318Reviewed-by:
-
Joshua Ashton authored
The check for sending the vsc infopacket to the display was gated behind PSR (Panel Self Refresh) being enabled. The vsc infopacket also contains the colorimetry (specifically the container color gamut) information for the stream on modern DP. PSR is typically only supported on mobile phone eDP displays, thus this was not getting sent for typical desktop monitors or TV screens. This functionality is needed for proper HDR10 functionality on DP as it wants BT2020 RGB/YCbCr for the container color space. Cc: stable@vger.kernel.org Cc: Harry Wentland <harry.wentland@amd.com> Cc: Xaver Hugl <xaver.hugl@gmail.com> Cc: Melissa Wen <mwen@igalia.com> Fixes: 15f9dfd5 ("drm/amd/display: Register Colorspace property for DP and HDMI") Tested-by:
Simon Berz <simon@berz.me> Tested-by:
Xaver Hugl <xaver.hugl@kde.org> Signed-off-by:
Joshua Ashton <joshua@froggi.es> Signed-off-by:
-
Hawking Zhang authored
Not needed any more with firmware fixes Signed-off-by:
Tao Zhou <tao.zhou1@amd.com> Signed-off-by:
-
Asad Kamal authored
Expose sysfs entry mem_busy_percent for GC version 9.4.3 APU system Signed-off-by:
Asad Kamal <asad.kamal@amd.com> Reviewed-by:
-
Dave Airlie authored
Merge tag 'drm-misc-next-fixes-2024-01-04' of git://anongit.freedesktop.org/drm/drm-misc into drm-next One fix for drm/plane to avoid a use-after-free and some additional warnings to prevent more of these occurences, a lock inversion dependency fix and an indentation fix for drm/rockchip, and some doc warning fixes for imagination and gpuvm. Signed-off-by:
-
- 03 Jan, 2024 2 commits
-
-
Srinivasan Shanmugam authored
dbg_flags looks to be defined with incorrect data type; to process multiple debug flag options, and hence defined dbg_flags as u32. Fixes the below: drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_packet_manager_v9.c:117 pm_map_process_aldebaran() warn: maybe use && instead of & Fixes: 0de4ec9a ("drm/amdgpu: prepare map process for multi-process debug devices") Suggested-by:
-
Srinivasan Shanmugam authored
Return invalid error code -EINVAL for invalid block id. Fixes the below: drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c:1183 amdgpu_ras_query_error_status_helper() error: we previously assumed 'info' could be null (see line 1176) Suggested-by:
-
