- 10 Mar, 2021 23 commits
-
-
Jay Cornwall authored
Similar to arcturus, but ARCH/ACC VGPRs may now be split unevenly. A new field in SQ_WAVE_GPR_ALLOC tracks the boundary between the two sets of VGPRs. Squash below patches: drm/amdkfd: Use preprocessor for IP-specific trap handler code drm/amdkfd: Fix VGPR restore race in gfx8/gfx9 trap handler drm/amdkfd: Remove duplicated code in gfx9 trap handler drm/amdkfd: Separate ARCH/ACC VGPR restore in trap handler drm/amdkfd: Reverse order of ARCH/ACC VGPR restore in trap handler Signed-off-by:
Jay Cornwall <jay.cornwall@amd.com> Signed-off-by:
Alex Deucher <alexander.deucher@amd.com>
-
Kevin Wang authored
add sdma firmware load support for soc model v2: drop some emulator leftovers (Alex) Signed-off-by:
Kevin Wang <kevin1.wang@amd.com> Reviewed-by:
Le Ma <Le.Ma@amd.com> Signed-off-by:
-
Yong Zhao authored
Add initial KFD support. Signed-off-by:
Yong Zhao <Yong.Zhao@amd.com> Signed-off-by:
-
Le Ma authored
Set ip blocks and asic family id Signed-off-by:
Hawking Zhang <Hawking.Zhang@amd.com> Acked-by:
Evan Quan <evan.quan@amd.com> Signed-off-by:
-
Le Ma authored
On aldebaran, mmBIF_SDMA4_DOORBELL_RANGE isn't right next to mmBIF_SDMA3_DOORBELL_RANGE. Signed-off-by:
-
Le Ma authored
Add initial sdma support for aldebaran, and this asic has 5 sdma instances. v2: remove adundant condition check Signed-off-by:
-
Le Ma authored
Add gfx initial support Signed-off-by:
-
Le Ma authored
Set backdoor loading way in current phase v2: change case location to not break other asics Signed-off-by:
-
Le Ma authored
Add gfx memory controller support Signed-off-by:
-
Le Ma authored
v1: dupilcate mmhub_v1_7.c from mmhub_v1_0.c because mmhub register address for aldebaran is different from existing asics (Le) v2: switch to latest mmhub_v9_4_2 register headers (Hawking) v3: squash in init VM_L2_CNTL3 default value for mmhub v1_7 Signed-off-by:
-
Le Ma authored
Initialize aldebaran common ip block Signed-off-by:
-
Le Ma authored
Parses asic configurations stored in gpu_info firmware and make them available for driver to use. Signed-off-by:
-
Le Ma authored
v1: add aldebaran_reg_base_init function to initialize register base for aldebaran (Le) v2: update VCN HWIP and initialize base offset (James) Signed-off-by:
James Zhu <James.Zhu@amd.com> Reviewed-by:
Leo Liu <leo.liu@amd.com> Acked-by:
-
Hawking Zhang authored
v1: re-use arct ip base offset array for aldebaran (Le) v2: create aldebaran ip base offset array for major ip blocks (Hawking) v3: re-use arct VCN ip base offset array for aldebaran (James) v4: correct MP1 ip base offset array (Hawking) v5: update VCN ip base offset array to aldebaran one (Hawking) Signed-off-by:
-
Le Ma authored
Add aldebaran in amdgpu_asic_name array and amdgpu_asic_type enum Signed-off-by:
-
Hawking Zhang authored
v1: Add vcn v2_6_0 register offset and shift masks in header files (Hawking) v2: Clean up vcn v2_6_0 registers (Alex) v3: update registers (Alex) Signed-off-by:
-
Hawking Zhang authored
v1: Add umc v6_7_0 register offset and shift masks in header files (Hawking) v2: Clean up registers (Alex) v3: update registers (Alex) Signed-off-by:
-
Hawking Zhang authored
v1: Add thm v13_0_2 register offset and shift masks in header files (Hawking) v2: Clean up thm v13_0_2 registers (Alex) v3: update registers (Alex) Signed-off-by:
-
Hawking Zhang authored
Add sdma v4_4_0 register offset and shift masks in header files v2: update registers (Alex) Signed-off-by:
-
Hawking Zhang authored
v1: Add smuio v13_0_2 register offset and shift masks in header files (Hawking) v2: Clean up smuio v13_0_2 registers (Alex) v3: update registers (Alex) Signed-off-by:
-
Hawking Zhang authored
v1: Add mp v13_0_2 register offset and shift masks in header files (Hawking) v2: Clean up mp v13_0_2 registers (Alex) v3: update registers (Alex) Signed-off-by:
-
Hawking Zhang authored
v1: Add mmhub v1_7 register offset and shift masks in header files (Hawking) v2: Clean up mmhub v1_7 registers (Alex) v3: Update registers (Alex) Signed-off-by:
-
Hawking Zhang authored
v1: Add gc v9_4_2 register offset and shift masks in header files (Hawking) v2: Clean up gc v9_4_2 registers (Alex) v3: update registers (Alex) Signed-off-by:
-
- 05 Mar, 2021 17 commits
-
-
Holger Hoffstätte authored
After fixing nested FPU contexts caused by 41401ac6 we're still seeing complaints about spurious kernel_fpu_end(). As it turns out this was already fixed for dcn20 in commit f41ed88c ("drm/amdgpu/display: use GFP_ATOMIC in dcn20_validate_bandwidth_internal") but never moved forward to dcn21. Signed-off-by:
Holger Hoffstätte <holger@applied-asynchrony.com> Signed-off-by:
-
Holger Hoffstätte authored
Commit 41401ac6 added FPU wrappers to dcn21_validate_bandwidth(), which was correct. Unfortunately a nested function alredy contained DC_FP_START()/DC_FP_END() calls, which results in nested FPU context enter/exit and complaints by kernel_fpu_begin_mask(). This can be observed e.g. with 5.10.20, which backported 41401ac6 and now emits the following warning on boot: WARNING: CPU: 6 PID: 858 at arch/x86/kernel/fpu/core.c:129 kernel_fpu_begin_mask+0xa5/0xc0 Call Trace: dcn21_calculate_wm+0x47/0xa90 [amdgpu] dcn21_validate_bandwidth_fp+0x15d/0x2b0 [amdgpu] dcn21_validate_bandwidth+0x29/0x40 [amdgpu] dc_validate_global_state+0x3c7/0x4c0 [amdgpu] The warning is emitted due to the additional DC_FP_START/END calls in patch_bounding_box(), which is inlined into dcn21_calculate_wm(), its only caller. Removing the calls brings the code in line with dcn20 and makes the warning disappear. Fixes: 41401ac6 ("drm/amd/display: Add FPU wrappers to dcn21_validate_bandwidth()") Signed-off-by:
-
Gustavo A. R. Silva authored
There is a regular need in the kernel to provide a way to declare having a dynamically sized set of trailing elements in a structure. Kernel code should always use “flexible array members”[1] for these cases. The older style of one-element or zero-length arrays should no longer be used[2]. Refactor the code according to the use of a flexible-array member in struct SISLANDS_SMC_SWSTATE, instead of a one-element array, and use the struct_size() helper to calculate the size for the allocation. Also, this helps with the ongoing efforts to enable -Warray-bounds by fixing the following warnings: drivers/gpu/drm/radeon/si_dpm.c: In function ‘si_convert_power_state_to_smc’: drivers/gpu/drm/radeon/si_dpm.c:2350:20: warning: array subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds] 2350 | smc_state->levels[i].dpm2.MaxPS = (u8)((SISLANDS_DPM2_MAX_PULSE_SKIP * (max_sclk - min_sclk)) / max_sclk); | ~~~~~~~~~~~~~~~~~^~~ drivers/gpu/drm/radeon/si_dpm.c:2351:20: warning: array subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds] 2351 | smc_state->levels[i].dpm2.NearTDPDec = SISLANDS_DPM2_NEAR_TDP_DEC; | ~~~~~~~~~~~~~~~~~^~~ drivers/gpu/drm/radeon/si_dpm.c:2352:20: warning: array subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds] 2352 | smc_state->levels[i].dpm2.AboveSafeInc = SISLANDS_DPM2_ABOVE_SAFE_INC; | ~~~~~~~~~~~~~~~~~^~~ drivers/gpu/drm/radeon/si_dpm.c:2353:20: warning: array subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds] 2353 | smc_state->levels[i].dpm2.BelowSafeInc = SISLANDS_DPM2_BELOW_SAFE_INC; | ~~~~~~~~~~~~~~~~~^~~ drivers/gpu/drm/radeon/si_dpm.c:2354:20: warning: array subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds] 2354 | smc_state->levels[i].dpm2.PwrEfficiencyRatio = cpu_to_be16(pwr_efficiency_ratio); | ~~~~~~~~~~~~~~~~~^~~ drivers/gpu/drm/radeon/si_dpm.c:5105:20: warning: array subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds] 5105 | smc_state->levels[i + 1].aT = cpu_to_be32(a_t); | ~~~~~~~~~~~~~~~~~^~~~~~~ [1] https://en.wikipedia.org/wiki/Flexible_array_member [2] https://www.kernel.org/doc/html/v5.9/process/deprecated.html#zero-length-and-one-element-arrays Link: https://github.com/KSPP/linux/issues/79 Link: https://github.com/KSPP/linux/issues/109Build-tested-by:kernel test robot <lkp@intel.com> Link: https://lore.kernel.org/lkml/603f9a8f.aDLrpMFzzSApzVYQ%25lkp@intel.com/Signed-off-by:
Gustavo A. R. Silva <gustavoars@kernel.org> Signed-off-by:
-
Colin Ian King authored
The continue statement in a for-loop is redudant and can be removed. Clean up the code to address this. Addresses-Coverity: ("Continue as no effect") Fixes: b6f91fc1 ("drm/amdgpu/display: buffer INTERRUPT_LOW_IRQ_CONTEXT interrupt work") Signed-off-by:Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Colin Ian King authored
The variable status is being initialized with a value that is never read and it is being updated later with a new value. The initialization is redundant and can be removed. Addresses-Coverity: ("Unused value") Signed-off-by: -
Jiapeng Chong authored
Fix the following coccicheck warnings: ./drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c:956:52-57: WARNING: conversion to bool not needed here. ./drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c:8311:16-21: WARNING: conversion to bool not needed here. Reported-by:
Abaci Robot <abaci@linux.alibaba.com> Signed-off-by:
Jiapeng Chong <jiapeng.chong@linux.alibaba.com> Signed-off-by:
-
Takashi Iwai authored
There seem devices that don't work with the aux channel backlight control. For allowing such users to test with the other backlight control method, provide a new module option, aux_backlight, to specify enabling or disabling the aux backport support explicitly. As default, the aux support is detected by the hardware capability. v2: make the backlight option generic in case we add future backlight types (Alex) BugLink: https://bugzilla.opensuse.org/show_bug.cgi?id=1180749 BugLink: https://gitlab.freedesktop.org/drm/amd/-/issues/1438Reviewed-by:
Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Signed-off-by:
Takashi Iwai <tiwai@suse.de> Signed-off-by:
-
Alex Deucher authored
Need to fetch it via aux. Reviewed-by:
-
Alex Deucher authored
It just spams the logs. Reviewed-by:
-
Alex Deucher authored
Avoid the extra wrapper function. Reviewed-by:
-
Alex Deucher authored
We set up the parameters, but never called the atom table. Reviewed-by:
Harry Wentland <harry.wentland@amd.com> Signed-off-by:
-
Alex Deucher authored
This has been stable for a while. Reviewed-by:
Christian König <christian.koenig@amd.com> Signed-off-by:
-
Jay Cornwall authored
Trap handler is set per-process per-device and is unrelated to queue management. Move implementation closer to TMA setup code. Signed-off-by:
Felix Kuehling <Felix.Kuehling@amd.com> Signed-off-by:
-
Felix Kuehling authored
Existing, buggy user mode breaks when SRAM ECC is correctly reported as "enabled". To avoid breaking existing user mode, deprecate that bit and leave it as 0. Define a new bit to report the actual SRAM ECC mode that new, correct user mode can use in the future. Fixes: 7ec177bdcfc1 ("drm/amdkfd: fix set kfd node ras properties value") Signed-off-by:Kent Russell <kent.russell@amd.com> Signed-off-by:
-
Emily Deng authored
If have memory leak, maybe it will have issue in ttm_bo_force_list_clean-> ttm_mem_evict_first. Set adev->gart.ptr to null to avoid to call amdgpu_gmc_set_pte_pde to cause ptr issue pointer when calling amdgpu_gart_unbind in amdgpu_bo_fini which is after gart_fini. Signed-off-by:
Emily Deng <Emily.Deng@amd.com> Reviewed-by:
-
Emily Deng authored
When unloading driver after killing some applications, it will hit sdma flush tlb job timeout which is called by ttm_bo_delay_delete. So to avoid the job submit after fence driver fini, call ttm_bo_lock_delayed_workqueue before fence driver fini. And also put drm_sched_fini before waiting fence. Signed-off-by:
-
Jingwen Chen authored
[Why] when try to shutdown guest vm in sriov mode, virt data exchange is not fini. After vram lost, trying to write vram could hang cpu. [How] add fini virt data exchange in ip_suspend Signed-off-by:
Jingwen Chen <Jingwen.Chen2@amd.com> Reviewed-by:
Jack Zhang <Jack.Zhang1@amd.com> Signed-off-by:
-
