- 18 Dec, 2020 3 commits
-
-
Chris Wilson authored
When inserting a VMA, we restrict the placement to the low 4G unless the caller opts into using the full range. This was done to allow usersapce the opportunity to transition slowly from a 32b address space, and to avoid breaking inherent 32b assumptions of some commands. However, for insert we limited ourselves to 4G-4K, but on verification we allowed the full 4G. This causes some attempts to bind a new buffer to sporadically fail with -ENOSPC, but at other times be bound successfully. commit 48ea1e32 ("drm/i915/gen9: Set PIN_ZONE_4G end to 4GB - 1 page") suggests that there is a genuine problem with stateless addressing that cannot utilize the last page in 4G and so we purposefully excluded it. This means that the quick pin pass may cause us to utilize a buggy placement. Reported-by:
CQ Tang <cq.tang@intel.com> Testcase: igt/gem_exec_params/larger-than-life-batch Fixes: 48ea1e32 ("drm/i915/gen9: Set PIN_ZONE_4G end to 4GB - 1 page") Signed-off-by:
Chris Wilson <chris@chris-wilson.co.uk> Cc: CQ Tang <cq.tang@intel.com> Reviewed-by:
Matthew Auld <matthew.auld@intel.com> Cc: <stable@vger.kernel.org> # v4.5+ Link: https://patchwork.freedesktop.org/patch/msgid/20201216092951.7124-1-chris@chris-wilson.co.uk (cherry picked from commit 5f22cc0b) Signed-off-by:
Jani Nikula <jani.nikula@intel.com>
-
Lionel Landwerlin authored
CI shows this workaround is also needed on Gen11. Signed-off-by:
Lionel Landwerlin <lionel.g.landwerlin@intel.com> Fixes: 059a0beb ("drm/i915/perf: workaround register corruption in OATAILPTR") Reviewed-by:
Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20201126105155.540350-1-lionel.g.landwerlin@intel.com (cherry picked from commit fa5d598b) Signed-off-by:
-
Chris Wilson authored
We now use ilk_hpd_irq_setup for all GMCH platforms that do not have hotplug. These are early gen3 and gen2 devices that now explode on boot as they try to access non-existent registers. Fixes: 794d61a1 ("drm/i915: re-order if/else ladder for hpd_irq_setup") Signed-off-by:
-
- 16 Dec, 2020 5 commits
-
-
Daniel Vetter authored
Merge tag 'amd-drm-fixes-5.11-2020-12-16' of git://people.freedesktop.org/~agd5f/linux into drm-next amd-drm-fixes-5.11-2020-12-16: amdgpu: - Fix a eDP regression for DCE asics - SMU fixes for sienna cichlid - Misc W=1 fixes - SDMA 5.2 reset fix - Suspend/resume fix - Misc display fixes - Misc runtime PM fixes and cleanups - Dimgrey Cavefish fixes - printk cleanup - Documentation warning fixes amdkfd: - Error logging fix - Fix pipe offset calculation radeon: - printk cleanup Signed-off-by:
Daniel Vetter <daniel.vetter@ffwll.ch> From: Alex Deucher <alexdeucher@gmail.com> Link: https://patchwork.freedesktop.org/patch/msgid/20201216192421.18627-1-alexander.deucher@amd.com
-
Alex Deucher authored
Add documentation for crc window. Fixes: c920888c ("drm/amd/display: Expose new CRC window property") Signed-off-by:
Alex Deucher <alexander.deucher@amd.com> Cc: Wayne Lin <Wayne.Lin@amd.com> Reported-by:
Stephen Rothwell <sfr@canb.auug.org.au> Reviewed-by:
Wayne Lin <Wayne.Lin@amd.com>
-
Tao Zhou authored
This makes it easier to debug what block is causing the fault, same as sienna_cichlid. Signed-off-by:
Tao Zhou <tao.zhou1@amd.com> Reviewed-by:
Jiansong Chen <Jiansong.Chen@amd.com> Signed-off-by:
-
Tao Zhou authored
Use mode1 reset for dimgrey_cavefish by default. Signed-off-by:
-
Rodrigo Siqueira authored
Some old ASICs might not implement/require get_dig_frontend helper; in this scenario, we can have a NULL pointer exception when we try to call it inside vbios disable operation. For example, this situation might happen when using Polaris12 with an eDP panel. This commit avoids this situation by adding a specific get_dig_frontend implementation for DCEx. Cc: Alex Deucher <alexander.deucher@amd.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Harry Wentland <Harry.Wentland@amd.com> Cc: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com> Cc: Chiawen Huang <chiawen.huang@amd.com> Reported-and-tested-by:
Borislav Petkov <bp@suse.de> Acked-by:
Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Signed-off-by:
-
- 15 Dec, 2020 32 commits
-
-
Tom Rix authored
See Documentation/core-api/printk-formats.rst. h should no longer be used in the format specifier for printk. Reviewed-by:
Christian König <christian.koenig@amd.com> Signed-off-by:
Tom Rix <trix@redhat.com> Signed-off-by:
-
Tom Rix authored
See Documentation/core-api/printk-formats.rst. h should no longer be used in the format specifier for printk. Reviewed-by:
-
Colin Ian King authored
There is a spelling mistake in a comment in the Kconfig. Fix it. Signed-off-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Alex Deucher authored
We need to move the check under the non-headless case, otherwise we always reserve the VGA save size. Fixes: 157fe68d ("drm/amdgpu: fix size calculation with stolen vga memory") Reviewed-by:
Guchun Chen <guchun.chen@amd.com> Signed-off-by:
-
Jiange Zhao authored
In Virtualization case, when one VF is sending too many FLR requests, hypervisor would stop responding to this VF's request for a long period of time. This is called event guard. During this period of cooling time, guest driver should wait instead of doing other things. After this period of time, guest driver would resume reset process and return to normal. Currently, guest driver would wait 12 seconds and return fail if it doesn't get response from host. Solution: extend this waiting time in guest driver and poll response periodically. Poll happens every 6 seconds and it will last for 60 seconds. v2: change the max repetition times from number to macro. Signed-off-by:
Jiange Zhao <Jiange.Zhao@amd.com> Acked-by:
Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by:
-
Yifan Zhang authored
it could also be insufficient vram that makes amdgpu_amdkfd_reserve_mem_limit fail. Signed-off-by:
Yifan Zhang <yifan1.zhang@amd.com> Reviewed-by:
Felix Kuehling <Felix.Kuehling@amd.com> Signed-off-by:
-
Souptick Joarder authored
Kernel test robot throws below warning -> drivers/gpu/drm/amd/amdgpu/../display/dc/dcn21/dcn21_dccg.c:46:6: warning: no previous prototype for 'dccg21_update_dpp_dto' [-Wmissing-prototypes] Adding prototype for dccg21_update_dpp_dto(). Reported-by:
kernel test robot <lkp@intel.com> Signed-off-by:
Souptick Joarder <jrdr.linux@gmail.com> Signed-off-by:
-
Alex Deucher authored
So we know when it's enabled and what method we are using. Acked-by:
Evan Quan <evan.quan@amd.com> Signed-off-by:
-
Alex Deucher authored
Simplify the logic in the runtime resume handling for atpx Acked-by:
-
Alex Deucher authored
The platform knows it's doing d3cold. Acked-by:
-
Alex Deucher authored
Enable runtime pm on non HG/PX BOCO capable boards. Acked-by:
-
Alex Deucher authored
Change it to check if the device has ACPI power resources. Acked-by:
-
Alex Deucher authored
Check if the device has ACPI power resources so we can enable runtime pm if so. Acked-by:
-
Alex Deucher authored
In preparation for systems that support d3cold on dGPUs independent of PX/HG. No functional change intended. Acked-by:
-
Souptick Joarder authored
Kernel test robot throws below warning -> drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:5349:5: warning: no previous prototype for 'amdgpu_dm_crtc_atomic_set_property' [-Wmissing-prototypes] drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:5349:5: warning: no previous prototype for function 'amdgpu_dm_crtc_atomic_set_property' [-Wmissing-prototypes] drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:5373:5: warning: no previous prototype for 'amdgpu_dm_crtc_atomic_get_property' [-Wmissing-prototypes] drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:5373:5: warning: no previous prototype for function 'amdgpu_dm_crtc_atomic_get_property' [-Wmissing-prototypes] As these functions are only used inside amdgpu_dm.c, these can be made static. Reported-by:
-
Aric Cyr authored
Signed-off-by:
Aric Cyr <aric.cyr@amd.com> Acked-by:
Qingqing Zhuo <qingqing.zhuo@amd.com> Signed-off-by:
-
Anthony Koo authored
- Add new aux_channel_type - Changed port_index to instance in dmub_cmd_dp_aux_control_data - Change aux_return_code_type to sync up with driver - param for ramping abm based on backlight level Signed-off-by:
Anthony Koo <Anthony.Koo@amd.com> Acked-by:
-
Aric Cyr authored
[Why] Typo in MPCC visual confirmation. [How] Fix to correct values. Signed-off-by:
Jun Lei <Jun.Lei@amd.com> Acked-by:
-
Felipe authored
[Why] The OGAM LUT precision was accumulating too much error in the higher end. [How] Instead of calculating all points of the LUT in relation to the previous ones, perform a full calculation in one of the intermediate segments to stop error propagation. Signed-off-by:
Felipe Clark <Felipe.Clark@amd.com> Reviewed-by:
Krunoslav Kovac <Krunoslav.Kovac@amd.com> Acked-by:
-
Aric Cyr authored
[Why] FP2 is not double buffered and must wait for VACTIVE before programming. [How] Only update when there is a full update we should change FP2 to avoid delay every flip. Signed-off-by:
-
Jake Wang authored
[Why] For certain timings, Renoir may underflow due to sr exit latency being too slow. [How] Updated wm table for renoir. Signed-off-by:
Jake Wang <haonan.wang2@amd.com> Reviewed-by:
Yongqiang Sun <yongqiang.sun@amd.com> Acked-by:
-
Martin Leung authored
[Why] race condition of programming FP2 wrt pipe locking and vactive state can cause underflow/black screen [How] Enforce the FP2 is only programmed during vactive, and unlock pipe soon afterwards. Signed-off-by:
Martin Leung <martin.leung@amd.com> Reviewed-by:
-
Leo (Hanghong) Ma authored
[Why] We are missing the DP info frame update on dcn30, and this will lead to DP SDPs not being sent; [How] Add the DP info frame update for dcn30; Signed-off-by:
Leo (Hanghong) Ma <hanghong.ma@amd.com> Reviewed-by:
Roman Li <Roman.Li@amd.com> Acked-by:
-
Eric Bernstein authored
[Why] dcn30_link_encoder_validate_output_with_stream was a static function. [How] remove the static define and include it in the header. Signed-off-by:
Eric Bernstein <eric.bernstein@amd.com> Reviewed-by:
Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com> Acked-by:
-
Wayne Lin authored
[Why] Find out when we try to disable CRC calculation, crc generation is still enabled. Main reason is that dc_stream_configure_crc() will never get called when the source is AMDGPU_DM_PIPE_CRC_SOURCE_NONE. [How] Add checking condition that when source is AMDGPU_DM_PIPE_CRC_SOURCE_NONE, we should also call dc_stream_configure_crc() to disable crc calculation. Also, clean up crc window when disable crc calculation. Signed-off-by:
Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com> Acked-by:
-
Aric Cyr authored
[Why] Many VR headsets require a HSYNC width of 4, but DCN has default minimum of 8. [How] Change the arbitrary minimum HSYNC width to 4 to match DCN20. Signed-off-by:
Charlene Liu <Charlene.Liu@amd.com> Reviewed-by:
-
Max Tseng authored
[Why] some DP_SEC register defs and masks are missing. [How] add the missing defs and masks. Signed-off-by:
Max Tseng <chuan-yu.tseng@amd.com> Reviewed-by:
-
Victor Lu authored
[Why] There is a warning that triggers when pstate takes too long. Pstate can take up to ~200us on Linux without hanging but it is currently set to 40us. [How] Change the timeout for the warning to be 180us on Linux. Signed-off-by:
Victor Lu <victorchengchi.lu@amd.com> Reviewed-by:
-
Stanley.Yang authored
skip load smu and sdma fw on sriov due to sos, ta and asd fw have been skipped for SIENNA_CICHLID. V2: move asic check into smu11 Signed-off-by:Stanley.Yang <Stanley.Yang@amd.com> Reviewed-by:
Kevin Wang <kevin1.wang@amd.com> Signed-off-by:
-
Flora Cui authored
CONFIG_DRM_AMD_DC_DCN3_0 is retired. drop it Signed-off-by:
Flora Cui <flora.cui@amd.com> Reviewed-by:
-
Likun Gao authored
S0ix only makes sense on APUs since they are part of the platform, so only when the ASIC is APU should set amdgpu_acpi_is_s0ix_supported flag to deal with the related situation. Signed-off-by:
Likun Gao <Likun.Gao@amd.com> Reviewed-by:
-
Xiaomeng Hou authored
Without doing the soft reset, register mmSDMA0_GFX_RB_WPTR's value could not be reset to 0 when sdma block resumes. That would cause the ring buffer's read and write pointers not equal and ring test fail. So add the soft reset step. Signed-off-by:
Xiaomeng Hou <Xiaomeng.Hou@amd.com> Reviewed-by:
Huang Rui <ray.huang@amd.com> Acked-by:
-
