Age | Commit message (Collapse) | Author | Files | Lines | |
---|---|---|---|---|---|
2017-12-19 | drm/amdgpu: use defines for mmBIF_IOV_FUNC_IDENTIFIER fields | Alex Deucher | 1 | -4/+3 | |
Rather than magic numbers. Reviewed-by: Xiangliang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-19 | drm/amdgpu: only check mmBIF_IOV_FUNC_IDENTIFIER on tonga/fiji | Alex Deucher | 1 | -8/+13 | |
We only support SR-IOV on tonga/fiji. Don't check this register on other VI parts. Fixes: 048765ad5af7c89 (amdgpu: fix asic initialization for virtualized environments (v2)) Reviewed-by: Xiangliang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]> Cc: [email protected] | |||||
2017-12-18 | Revert "drm/amd/amdgpu: set gtt size according to system memory size only" | Andrey Grodzovsky | 1 | -3/+5 | |
This reverts commit ba851eed895c76be0eb4260bdbeb7e26f9ccfaa2. With that change piglit max size tests (running with -t max.*size) are causing OOM and hard hang on my CZ with 1GB RAM. Signed-off-by: Andrey Grodzovsky <[email protected]> Acked-by: Alex Deucher <[email protected]> Reviewed-by: Christian König <[email protected]> Reviewed-by: Roger He <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: implement 2+1 PD support for Raven v3 | Christian König | 5 | -35/+89 | |
Instead of falling back to 2 level and very limited address space use 2+1 PD support and 128TB + 512GB of virtual address space. v2: cleanup defines, rebase on top of level enum v3: fix inverted check in hardware setup Signed-off-by: Christian König <[email protected]> Reviewed-and-Tested-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: restore uvd fence seq in uvd resume | Jim Qu | 1 | -0/+2 | |
otherwise, uvd block will be never powered up in ring begin_use() callback. uvd ring test will be fail in resume in rumtime pm. Signed-off-by: Jim Qu <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: always cancel uvd idle handler in uvd suspend | Jim Qu | 1 | -2/+2 | |
Signed-off-by: Jim Qu <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: rename amdgpu_get_pcie_info | Alex Deucher | 5 | -5/+5 | |
add device to the name for consistency. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: move amdgpu_need_backup to amdgpu_object.c | Alex Deucher | 3 | -9/+8 | |
It's the only place it's used. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: rename amdgpu_gpu_recover | Alex Deucher | 7 | -8/+10 | |
add device to the name for consistency. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: move dummy page functions to amdgpu_gart.c | Alex Deucher | 3 | -50/+47 | |
It's the only place they are used. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: rename amdgpu_need_post | Alex Deucher | 3 | -6/+6 | |
add device to the name for consistency. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: rename ip block helper functions | Alex Deucher | 14 | -246/+252 | |
add device to the name for consistency. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: move fw_reserve functions to amdgpu_ttm.c | Alex Deucher | 3 | -100/+97 | |
It's the only place they are used. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: rename amdgpu_*_location functions | Alex Deucher | 6 | -14/+18 | |
add device to the name for consistency. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: move amdgpu_doorbell_get_kfd_info to amdgpu_amdkfd.c | Alex Deucher | 3 | -38/+34 | |
It's the only place it's used. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: rename amdgpu_pci_config_reset | Alex Deucher | 4 | -4/+4 | |
add device for consistency with other functions in this file. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: rename amdgpu_program_register_sequence | Alex Deucher | 14 | -311/+311 | |
add device for consistency with other functions in this file. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: rename amdgpu_wb_* functions | Alex Deucher | 10 | -46/+46 | |
add device for consistency. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: move debugfs functions to their own file | Alex Deucher | 6 | -785/+838 | |
amdgpu_device.c was getting pretty cluttered. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: rename amdgpu_suspend to amdgpu_device_ip_suspend | Alex Deucher | 3 | -6/+6 | |
for consistency with the other functions in that file. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: use consistent naming for static funcs in amdgpu_device.c | Alex Deucher | 1 | -86/+90 | |
Prefix the functions with device or device_ip for functions which deal with ip blocks for consistency. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-18 | drm/amdgpu: move atom functions from amdgpu_device.c | Alex Deucher | 3 | -235/+236 | |
and move them to amdgpu_atombios.c for consistency. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-15 | drm/amdgpu: setup the shared and private apertures on gfx9 | Alex Deucher | 2 | -5/+20 | |
Same as previous asics. This was not yet set for gfx9. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-15 | drm/amdgpu: Simplify amdgpu_lockup_timeout usage. | Andrey Grodzovsky | 5 | -18/+11 | |
With introduction of amdgpu_gpu_recovery we don't need any more to rely on amdgpu_lockup_timeout == 0 for disabling GPU reset. Signed-off-by: Andrey Grodzovsky <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-15 | drm/amdgpu: Add gpu_recovery parameter | Andrey Grodzovsky | 8 | -7/+19 | |
Add new parameter to control GPU recovery procedure. v2: Add auto logic where reset is disabled for bare metal and enabled for SR-IOV. Allow forced reset from debugfs. Signed-off-by: Andrey Grodzovsky <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-15 | drm/ttm: use an ttm operation ctx for ttm_bo_move_xxx | Roger He | 1 | -4/+3 | |
include ttm_bo_move_memcpy and ttm_bo_move_ttm Signed-off-by: Roger He <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-15 | drm/amdgpu: set allow_reserved_eviction and resv when bo allocation and cs | Roger He | 2 | -2/+12 | |
enable eviction of other per VM BOs during allocation and allows reaping of deleted BOs during CS. Reviewed-by: Chunming Zhou <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Roger He <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-14 | drm/amdgpu: add enumerate for PDB/PTB v3 | Chunming Zhou | 2 | -19/+61 | |
v2: remove SUBPTB member v3: remove last_level, use AMDGPU_VM_PTB directly instead. Signed-off-by: Chunming Zhou <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-13 | drm/amdgpu: fix huge page setting for ATS case | Chunming Zhou | 1 | -1/+1 | |
Signed-off-by: Chunming Zhou <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-13 | drm/amdgpu: drop amdgpu_atombios_scratch_regs_save/restore | Alex Deucher | 2 | -24/+0 | |
No longer used. Reviewed-by: Harry Wentland <[email protected]> Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-13 | drm/amdgpu: drop scratch regs save and restore from GPU reset handling | Alex Deucher | 1 | -2/+0 | |
The expectation is that the base driver doesn't mess with these. Some components interact with these directly so let the components handle these directly. Reviewed-by: Harry Wentland <[email protected]> Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-13 | drm/amdgpu: drop scratch regs save and restore from S3/S4 handling | Alex Deucher | 1 | -2/+0 | |
The expectation is that the base driver doesn't mess with these. Some components interact with these directly so let the components handle these directly. Reviewed-by: Harry Wentland <[email protected]> Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-13 | drm/amdgpu: remove some old gc 9.x registers | Alex Deucher | 1 | -4/+4 | |
Leftover from bring up. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-13 | drm/amdgpu: drop soc15_init_golden_registers | Alex Deucher | 1 | -31/+0 | |
The golden register arrays were empty so the function was effectively useless. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-13 | drm/amdgpu: drop the bios scratch reg callbacks from nbio | Alex Deucher | 3 | -31/+0 | |
They are not used any longer. We get the scratch register locations from the vbios directly now. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-13 | drm/amdgpu: convert nbio to use callbacks (v2) | Alex Deucher | 11 | -183/+209 | |
Cleans up and consolidates all of the per-asic logic. v2: squash in "drm/amdgpu: fix NULL err for sriov detect" (Chunming) Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: make function names consistent in nbio files | Alex Deucher | 2 | -16/+16 | |
All functions should have nbio_v* prefix. Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: correct vce fw data and stack size | Frank Min | 1 | -1/+1 | |
this fix the VCE world switch hang issue Signed-off-by: Frank Min <[email protected]> Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: fix MAP_QUEUES paramter | Monk Liu | 1 | -1/+1 | |
Should be 0. Signed-off-by: Monk Liu <[email protected]> Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: no need with INT for fence polling | Monk Liu | 1 | -1/+1 | |
We are polling so no need for INT. Signed-off-by: Monk Liu <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: no need to evict VRAM in device_fini | Monk Liu | 1 | -2/+1 | |
this VRAM evict is not needed and also cost 2seconds to finish because the IRQ is software side disabled before it. Signed-off-by: Monk Liu <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: allow get_vm_pde to change flags as well | Christian König | 11 | -38/+54 | |
And also provide the level for which we need a PDE. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: batch PDE updates again | Christian König | 1 | -109/+94 | |
Now instead of one submission for each PDE batch them together over all PDs who need an update. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: remove keeping the addr of the VM PDs | Christian König | 2 | -10/+5 | |
No more double house keeping. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: remove last_entry_used from the VM code | Christian König | 2 | -24/+29 | |
Not needed any more. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: avoid the modulo in amdgpu_vm_get_entry | Christian König | 1 | -3/+3 | |
We can do this with a simple mask as well. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: use polling mem to set SDMA3 wptr for VF | Pixel Ding | 2 | -8/+20 | |
On Tonga VF, there're 2 sources updating wptr registers for sdma3: 1) polling mem and 2) doorbell. When doorbell and polling mem are both enabled on sdma3, there will be collision hit in occasion between those two sources when ucode and h/w are doing the updating on wptr register in parallel. Issue doesn't happen on CP GFX/Compute since CP drops all doorbell writes when VF is inactive. So enable polling mem and don't use doorbell for SDMA3. Signed-off-by: Pixel Ding <[email protected]> Reviewed-by: Monk Liu <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: update one PDE at a time v2 | Christian König | 1 | -46/+34 | |
Horrible inefficient, but avoids problems when the root PD size becomes to big. v2: remove incr as well. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: stop joining PDEs | Christian König | 1 | -34/+7 | |
That doesn't hit any more most of the time anyway. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]> | |||||
2017-12-12 | drm/amdgpu: add amdgpu_evict_vram debugfs file | Christian König | 1 | -1/+12 | |
Torture test for MM and VM support, can be used to evict all VRAM while the system is under load. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]> |