OSDN Git Service
Felix Kuehling [Sat, 26 Aug 2017 06:10:12 +0000 (02:10 -0400)]
drm/amdkfd: Use PASID manager from KGD
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Felix Kuehling [Sat, 26 Aug 2017 06:00:57 +0000 (02:00 -0400)]
drm/amdkfd: Separate doorbell allocation from PASID
PASID management is moving into KGD. Limiting the PASID range to the
number of doorbell pages is no longer practical.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Felix Kuehling [Sat, 26 Aug 2017 01:30:18 +0000 (21:30 -0400)]
drm/radeon: Add PASID manager for KFD
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Felix Kuehling [Sat, 26 Aug 2017 00:40:26 +0000 (20:40 -0400)]
drm/amdgpu: Add PASID management
Allows assigning a PASID to a VM for identifying VMs involved in page
faults. The global PASID manager is also exported in the KFD
interface so that AMDGPU and KFD can share the PASID space.
PASIDs of different sizes can be requested. On APUs, the PASID size
is deterined by the capabilities of the IOMMU. So KFD must be able
to allocate PASIDs in a smaller range.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Felix Kuehling [Sat, 26 Aug 2017 00:15:04 +0000 (20:15 -0400)]
drm/amdgpu: Fix error handling in amdgpu_vm_init
Make sure vm->root.bo is not left reserved if amdgpu_bo_kmap fails.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 15 Sep 2017 08:30:52 +0000 (16:30 +0800)]
drm/amd/powerplay: Simplify smu7_voting_clients()
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Tue, 12 Sep 2017 05:37:40 +0000 (13:37 +0800)]
drm/amdgpu: add powerplay support for CI asics
currently, for CI asics,
use dpm by default, amdgpu.dpm=-1.
when set amdgpu.dpm=1, enable powplay.
when set amdgpu.dpm=0, disable both dpm and powerplay.
when powerplay is stable on CI asics, ci_dpm will
be removed.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Thu, 14 Sep 2017 13:05:18 +0000 (21:05 +0800)]
drm/amd/powerplay: Add support for CI asics to hwmgr
Add support for CI asics (Bonaire, Hawaii) to
the powerplay hwmgr
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Thu, 14 Sep 2017 13:14:59 +0000 (21:14 +0800)]
drm/amd/powerplay: Add support functions for CI to ppatomctrl.c
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 15 Sep 2017 03:09:20 +0000 (11:09 +0800)]
drm/amd/powerplay: refine dmesg info under powerplay
Use pr_debug to prevent spamming unimportant dmesg.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Thu, 14 Sep 2017 01:10:41 +0000 (09:10 +0800)]
drm/amd/powerplay: fix set target TDP error on tonga/iceland
ConfigurableTDP do not exist from Fiji.
so only use in previous ASIC.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 8 Sep 2017 11:34:33 +0000 (19:34 +0800)]
drm/amd/powerplay: add CI asics support to smumgr (v3)
This ports support for CI asics (Bonaire, Hawaii)
to the powerplay smumgr
v2: warning fix (Alex)
v3: squash in fix for thermal (Tom)
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Tue, 12 Sep 2017 05:18:13 +0000 (13:18 +0800)]
drm/amd/powerplay: fix pcie max lane define error
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 8 Sep 2017 06:31:26 +0000 (14:31 +0800)]
drm/amd/powerplay: fix spelling typo in function name
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 8 Sep 2017 06:05:51 +0000 (14:05 +0800)]
drm/amdgpu: add support for request SI/CI firmware in CGS
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Wed, 6 Sep 2017 10:43:52 +0000 (18:43 +0800)]
drm/amdgpu: unify the interface of amd_pm_funcs
put amd_pm_funcs table in struct powerplay for all
asics.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Wed, 6 Sep 2017 08:08:03 +0000 (16:08 +0800)]
drm/amd/powerplay: use struct amd_pm_funcs in powerplay
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Wed, 6 Sep 2017 07:27:59 +0000 (15:27 +0800)]
drm/amdgpu: rename amdgpu_dpm_funcs to amd_pm_funcs
renamed amdgpu_dpm_funcs and moved to amd_shared.h
so can shared with powerplay.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Mon, 4 Sep 2017 10:22:02 +0000 (18:22 +0800)]
drm/amd/powerplay: delete dead code in powerplay
delete functiontable related codes
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Mon, 4 Sep 2017 10:11:52 +0000 (18:11 +0800)]
drm/amd/powerplay: refine powerplay code for RV
use function points instand of function table.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Mon, 4 Sep 2017 09:51:28 +0000 (17:51 +0800)]
drm/amd/powerplay: refine powerplay code for CZ/ST
1. add function points instand of creat function tables
2. implement stop dpm tasks for CZ/ST
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Mon, 4 Sep 2017 08:25:48 +0000 (16:25 +0800)]
drm/amd/powerplay: refine code for thermal control in powerplay
add function point start_thermal_controller in hwmgr,
delete thermal function table and related functions
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 14 Sep 2017 12:57:26 +0000 (08:57 -0400)]
drm/amd/amdgpu: Support VM environments in amdgpu_ttm_access_memory()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Arnd Bergmann [Fri, 15 Sep 2017 20:06:58 +0000 (22:06 +0200)]
drm/radeon: properly initialize r600_audio_status() data
The structure returned from r600_audio_status() is only partially
initialized, and older gcc versions (4.3 and 4.4) warn about this:
drivers/gpu/drm/radeon/r600_hdmi.c: In function 'r600_audio_status':
drivers/gpu/drm/radeon/r600_hdmi.c:108: error: 'status.id' is used uninitialized in this function
drivers/gpu/drm/radeon/r600_hdmi.c:108: error: 'status.connected' is used uninitialized in this function
drivers/gpu/drm/radeon/r600_hdmi.c:108: error: 'status.offset' is used uninitialized in this function
This is harmless and surprisingly correct in C99, as the caller
only accesses the fields that got initialized, so newer compilers
don't warn about it, but initializing the entire structure feels
like the right thing to do here and avoids the warning.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Deucher [Fri, 15 Sep 2017 21:36:19 +0000 (17:36 -0400)]
drm/amdgpu/psp: declare raven psp firmware
So it gets picked up properly by the kernel.
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Allen Pais [Wed, 13 Sep 2017 08:01:09 +0000 (13:31 +0530)]
drivers:gpu:Use ARRAY_SIZE() for the size calculation of the array.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Wed, 13 Sep 2017 16:35:15 +0000 (12:35 -0400)]
drm/amd/amdgpu: Change vram debugfs to NO_KIQ for VM environments
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Wed, 13 Sep 2017 08:43:09 +0000 (10:43 +0200)]
drm/ttm: fix memory leak while individualizing BOs
We need to free the reservation object before we take the BO
from the delayed delete list.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Monk Liu <monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 12 Sep 2017 08:56:17 +0000 (10:56 +0200)]
drm/amdgpu: fix and cleanup amdgpu_bo_create v2
We adjusted the BO flags for USWC handling, but those never took effect
because the placement was passed in instead of generated inside this
function.
v2: better commit message
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 11 Sep 2017 15:10:26 +0000 (17:10 +0200)]
drm/amd: remove min/max addr handling from cgs
Nobody is actually using this and it causes a bunch of unused and buggy code.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 11 Sep 2017 13:51:30 +0000 (15:51 +0200)]
drm/amdgpu: fix cgs alignment handling
This always allocated on PAGE_SIZE alignment.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Mon, 11 Sep 2017 08:09:17 +0000 (17:09 +0900)]
radeon: Only destroy fbdev framebuffer if it was initialized
Fixes crash when trying to unload the radeon module before the fbdev
framebuffer was initialized, which can happen since the DRM fbdev helper
code supports deferred setup.
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Michel Dänzer [Mon, 11 Sep 2017 08:04:41 +0000 (17:04 +0900)]
amdgpu: Only destroy fbdev framebuffer if it was initialized
Fixes crash when trying to unload the amdgpu module before the fbdev
framebuffer was initialized, which can happen since the DRM fbdev helper
code supports deferred setup.
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 12 Sep 2017 16:29:06 +0000 (12:29 -0400)]
drm/amd/powerplay: lock grbm_gfx index when changing instance
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 12 Sep 2017 14:05:48 +0000 (10:05 -0400)]
drm/amd/powerplay: Tidy up smu7_enable_didt()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 12 Sep 2017 14:01:34 +0000 (10:01 -0400)]
drm/amd/powerplay: Make use of PP_CAP in smu7_powertune.c
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 12 Sep 2017 13:56:45 +0000 (09:56 -0400)]
drm/amd/powerplay: Remove unneeded return from pp_smu7_thermal_fini()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 12 Sep 2017 13:53:25 +0000 (09:53 -0400)]
drm/amd/powerplay: Make use of PP_CAP in smu7_thermal.c
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 12 Sep 2017 13:51:36 +0000 (09:51 -0400)]
drm/amd/powerplay: Tidy up smu7_fan_ctrl_get_fan_speed_rpm()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 12 Sep 2017 13:46:40 +0000 (09:46 -0400)]
drm/amd/powerplay: Tidy up smu7_fan_ctrl_get_fan_speed_info()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 11 Sep 2017 14:54:59 +0000 (16:54 +0200)]
drm/amdgpu: fix amdgpu_vm_handle_moved as well v2
There is no guarantee that the last BO_VA actually needed an update.
Additional to that all command submissions must wait for moved BOs to
be cleared, not just the first one.
v2: Don't overwrite any newer fence.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Himanshu Jha [Mon, 11 Sep 2017 12:37:26 +0000 (18:07 +0530)]
drm/amd/powerplay: remove unnecessary call to memset
call to memset to assign 0 value immediately after allocating
memory with kzalloc is unnecesaary as kzalloc allocates the memory
filled with 0 value.
Semantic patch used to resolve this issue:
@@
expression e,e2; constant c;
statement S;
@@
e = kzalloc(e2, c);
if(e == NULL) S
- memset(e, 0, e2);
Signed-off-by: Himanshu Jha <himanshujha199640@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Deucher [Sat, 2 Sep 2017 06:16:35 +0000 (02:16 -0400)]
drm/amdgpu/gfx9: properly set the hdp flush reg for Raven
Was only being assigned for vega10.
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Fri, 8 Sep 2017 13:45:34 +0000 (09:45 -0400)]
drm/amd/powerplay: Tidy up vega10_thermal_initialize()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 17:08:28 +0000 (13:08 -0400)]
drm/amd/powerplay: Tidy up vega10_thermal_set_temperature_range()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 16:59:32 +0000 (12:59 -0400)]
drm/amd/powerplay: Tidy up vega10_fan_ctrl_set_fan_speed_rpm()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 16:42:38 +0000 (12:42 -0400)]
drm/amd/powerplay: Fix indentation in vega10_fan_ctrl_reset_fan_speed_to_default()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 16:41:51 +0000 (12:41 -0400)]
drm/amd/powerplay: Tidy up vega10_fan_ctrl_set_fan_speed_percent()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 16:26:42 +0000 (12:26 -0400)]
drm/amd/powerplay: Tidy up vega10_fan_ctrl_set_default_mode()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 16:20:48 +0000 (12:20 -0400)]
drm/amd/powerplay: Tidy up vega10_fan_ctrl_set_static_mode()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 14:36:40 +0000 (10:36 -0400)]
drm/amd/powerplay: Tidy up vega10_fan_ctrl_get_fan_speed_rpm()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 11:48:16 +0000 (07:48 -0400)]
drm/amd/powerplay: Port vega10_thermal.c over to PP_CAP
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 11:46:40 +0000 (07:46 -0400)]
drm/amd/powerplay: Port vega10_powertune.c over to PP_CAP
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 11:44:17 +0000 (07:44 -0400)]
drm/amd/powerplay: Port vega10_hwmgr.c over to PP_CAP
Replace and cleanup lengthy phm_cap_enabled() sequences
with PP_CAP.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 7 Sep 2017 11:43:33 +0000 (07:43 -0400)]
drm/amd/powerplay: Add PP_CAP() macro
To replace common lengthy sequence that would create
really long lines all over.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Wed, 6 Sep 2017 12:04:41 +0000 (08:04 -0400)]
drm/amd/powerplay: Port vega10_didt_set_mask() to new macros
Start using new CGS bitmask macros.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Wed, 6 Sep 2017 12:04:10 +0000 (08:04 -0400)]
drm/amd/powerplay: Introduction of bitmask macros for registers
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 5 Sep 2017 16:33:58 +0000 (12:33 -0400)]
drm/amd/powerplay: Simplify vega10_acg_disable()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 5 Sep 2017 16:02:57 +0000 (12:02 -0400)]
drm/amd/powerplay: Simplify vega10_patch_voltage_dependency_tables_with_lookup_table()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Fri, 8 Sep 2017 12:09:41 +0000 (14:09 +0200)]
drm/amdgpu: fix VM sync with always valid BOs v2
All users of a VM must always wait for updates with always
valid BOs to be completed.
v2: remove debugging leftovers, rename struct member
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Roger He <Hongbo.He@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Evan Quan [Mon, 4 Sep 2017 09:48:27 +0000 (17:48 +0800)]
drm/amdgpu: enable raven to load firmwares by psp at default (v2)
- v2: share code with CHIP_VEGA10 case
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Evan Quan [Fri, 8 Sep 2017 05:09:50 +0000 (13:09 +0800)]
drm/amdgpu: stop psp ring on suspend
Otherwise, the ring will fail to create on next resume.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Evan Quan [Fri, 8 Sep 2017 05:04:52 +0000 (13:04 +0800)]
drm/amdgpu: added api for stopping psp ring (v2)
- v2: reuse the ring stop api in ring destory
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Evan Quan [Mon, 4 Sep 2017 09:42:28 +0000 (17:42 +0800)]
drm/amdgpu: fixed raven psp cmd prepare and submit
- fw_size in psp_v10_0_prep_cmd_buf is wrongly set as 0
- fixed the wrong calculation of psp_write_ptr_reg in psp_v10_0_cmd_submit
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 12 Sep 2017 18:29:07 +0000 (14:29 -0400)]
drm/amdgpu: move MMU notifier related defines to amdgpu_mn.h
Just some cleanup.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Fri, 8 Sep 2017 15:19:19 +0000 (17:19 +0200)]
drm/amdgpu: move amdgpu_ttm_tt_* declarations into amdgpu_ttm.h
Just some cleanup.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 5 Sep 2017 15:30:46 +0000 (17:30 +0200)]
drm/amdgpu: keep the MMU lock until the update ends v4
This is quite controversial because it adds another lock which is held during
page table updates, but I don't see much other option.
v2: allow multiple updates to be in flight at the same time
v3: simplify the patch, take the read side only once
v4: correctly fix rebase conflict
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Wed, 6 Sep 2017 14:55:16 +0000 (16:55 +0200)]
drm/amdgpu: rework amdgpu_cs_find_mapping
Use the VM instead of the BO list to find the BO for a virtual address.
This fixes UVD/VCE in physical mode with VM local BOs.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Wed, 6 Sep 2017 14:15:28 +0000 (16:15 +0200)]
drm/amdgpu: move amdgpu_cs_sysvm_access_required into find_mapping
When we need to find the mapping we need sysvm access anyway.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Wed, 6 Sep 2017 14:55:55 +0000 (16:55 +0200)]
drm/ttm: allow mapping BOs while they are still on the swap list
With shared reservation objects it is possible that we want to temporary
kmap an BO while it is still on the swap list.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 12 Sep 2017 18:25:14 +0000 (14:25 -0400)]
drm/amdgpu: stop reserving the BO in the MMU callback v3
Instead take the callback lock during the final parts of CS.
This should solve the last remaining locking order problems with BO reservations.
v2: rebase, make dummy functions static inline
v3: add one more missing inline and comments
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 5 Sep 2017 12:50:24 +0000 (14:50 +0200)]
drm/amdgpu: use a rw_semaphore for MMU notifiers
Allow at least some parallel processing.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 5 Sep 2017 12:36:44 +0000 (14:36 +0200)]
drm/amdgpu: move userptr BOs to CPU domain during CS v2
Instead of moving them in the MMU notifier move them during CS.
v2: still mark pages as accessed/dirty
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com> (v1)
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 5 Sep 2017 12:30:05 +0000 (14:30 +0200)]
drm/amdgpu: stop using BO status for user pages
Instead use a counter to figure out if we need to set new pages or not.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Sun, 3 Sep 2017 13:22:06 +0000 (15:22 +0200)]
drm/amdgpu: move taking mmap_sem into get_user_pages v2
This didn't helped as intended, just simplify the code.
v2: unlock mmap_sem in the error path as well
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Bas Nieuwenhuizen [Thu, 7 Sep 2017 11:23:21 +0000 (13:23 +0200)]
drm/amdgpu: Account for shadow PTs in mapping update IB size.
When amdgpu_vm_frag_ptes calls amdgpu_vm_update_ptes and the pt
has a shadow PT we mirror all the write to the shadow PT too, which
results in twice the commands.
Signed-off-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 4 Sep 2017 18:58:45 +0000 (20:58 +0200)]
drm/ttm: fix ttm_bo_cleanup_refs_or_queue once more
With shared reservation objects __ttm_bo_reserve() can easily fail even on
destroyed BOs. This prevents correct handling when we need to individualize
the reservation object.
Fix this by individualizing the object before even trying to reserve it.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 5 Sep 2017 13:10:50 +0000 (15:10 +0200)]
drm/amdgpu: revert "fix deadlock of reservation between cs and gpu reset v2"
This reverts commit
10e709cb296c98424c03408d23e3addeddcd4088.
The patch doesn't work at all:
1. The CS can still be blocked because of amdgpu_ctx_add_fence().
2. The order of submission isn't correct any more.
3. We could end up using freed up memory because we now drop the
ctx reference to early.
This needs to be fixed cleanly by doing the context handling after the BO
handling, but this is a larger task just avoid the obvious crashes for now.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Monk Liu monk.liu@amd.com
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 5 Sep 2017 11:55:48 +0000 (07:55 -0400)]
drm/amd/powerplay: Fix psm_set_user_performance_state()
We now pass a pointer to a pointer which seems to be
what they meant in the first place. The previous version
was modifying a pointer passed by value.
Fixes bug that was introduced by
commit
332798d40c2e91:drm/amd/powerplay: delete eventmgr layer in poweprlay
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-By: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Sat, 2 Sep 2017 11:21:31 +0000 (13:21 +0200)]
drm/amdgpu: fix userptr put_page handling
Move calling put_page into the unpopulate callback. Otherwise we mess up the pages
reference count when it is unbound multiple times.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Tue, 5 Sep 2017 11:30:59 +0000 (07:30 -0400)]
drm/ttm: Fix configuration error around populate_and_map() functions
Fixed kbuild errors when IOMMU/SWIOTLB are disabled.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Monk Liu [Fri, 11 Aug 2017 09:49:48 +0000 (17:49 +0800)]
drm/amdgpu: fix wait_any_fence
first is incorrect if hit NULL/signaled fence
Signed-off-by: Monk Liu <monk.liu@amd.com>
Reviewed-by: Chunming Zhou <David1.Zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Tue, 29 Aug 2017 09:07:38 +0000 (17:07 +0800)]
drm/amd/powerplay: set uvd/vce/nb/mclk level as UMD P-state required
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Tue, 29 Aug 2017 08:08:56 +0000 (16:08 +0800)]
drm/amd/powerplay: add UMD P-state in powerplay.
This feature is for UMD to run benchmark in a
power state that is as steady as possible. kmd
need to fix the power state as stable as possible.
now, kmd support four level:
profile_standard,peak,min_sclk,min_mclk
move common related code to amd_powerplay.c
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 1 Sep 2017 05:48:11 +0000 (13:48 +0800)]
drm/amd/powerplay: delete eventmgr related files.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rex Zhu [Fri, 1 Sep 2017 05:46:20 +0000 (13:46 +0800)]
drm/amd/powerplay: delete eventmgr layer in poweprlay
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Deucher [Tue, 29 Aug 2017 14:55:05 +0000 (10:55 -0400)]
drm/amd/powerplay: fix sclk setting for profile mode for CZ/ST
Need to select dpm0 to avoid clock fluctuations.
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Thierry Reding [Fri, 1 Sep 2017 14:49:53 +0000 (16:49 +0200)]
drm/radeon: Use correct path to trace include
The header comment in include/trace/define_trace.h specifies that the
TRACE_INCLUDE_PATH needs to be relative to the define_trace.h header
rather than the trace file including it. Most instances get that wrong
and work around it by adding the $(src) directory to the include path.
While this works, it is preferable to refer to the correct path to the
trace file in the first place and avoid any workaround.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Thierry Reding [Fri, 1 Sep 2017 16:54:01 +0000 (12:54 -0400)]
drm/amdgpu: Use correct path to trace include
The header comment in include/trace/define_trace.h specifies that the
TRACE_INCLUDE_PATH needs to be relative to the define_trace.h header
rather than the trace file including it. Most instances get that wrong
and work around it by adding the $(src) directory to the include path.
While this works, it is preferable to refer to the correct path to the
trace file in the first place and avoid any workaround.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Fri, 1 Sep 2017 16:52:38 +0000 (12:52 -0400)]
drm/ttm: Fix trace include path (v2)
Reviewed-by: Thierry Reding <treding@nvidia.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(v2): Drop Makefile change too.
Tom St Denis [Fri, 1 Sep 2017 13:55:04 +0000 (09:55 -0400)]
drm/amd/amdgpu: Cleanup gmc_v9_0_suspend()
Even though fini returns 0 always it could theoretically
fail in the future. Might as well return it instead of 0.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Fri, 1 Sep 2017 13:53:44 +0000 (09:53 -0400)]
drm/amd/amdgpu: Tidy up gmc_v9_0_hw_init()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Fri, 1 Sep 2017 13:52:21 +0000 (09:52 -0400)]
drm/amd/amdgpu: Tidy up gmc_v9_0_gart_enable()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Fri, 1 Sep 2017 13:27:31 +0000 (09:27 -0400)]
drm/amd/amdgpu: Simplify gmc_v9_0_vm_fault_interrupt_state()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Fri, 1 Sep 2017 13:13:04 +0000 (09:13 -0400)]
drm/amd/amdgpu: Support full range of GFX ring names
Right now there's only one but the rest of the code is being
setup to support more so might as well fix this up too.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Fri, 1 Sep 2017 07:22:56 +0000 (09:22 +0200)]
drm/amdgpu: fix placement flags in amdgpu_ttm_bind
Otherwise we lose the NO_EVICT flag and can try to evict pinned BOs.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Fri, 1 Sep 2017 12:46:19 +0000 (14:46 +0200)]
drm/amdgpu: fix moved list handling in the VM
Only move BOs to the moved/relocated list when they aren't already on a list.
This prevents accidential removal from the evicted list.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Evan Quan [Fri, 1 Sep 2017 07:07:41 +0000 (15:07 +0800)]
drm/amdgpu: declare the new firmware files needed by polaris asics
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Flora Cui <Flora.Cui@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Roger He [Wed, 30 Aug 2017 05:01:19 +0000 (13:01 +0800)]
drm/amdgpu: handle all fragment sizes v4
This can improve performance for some cases.
v2 (chk): handle all sizes, simplify the patch quite a bit
v3 (chk): adjust dw estimation as well
v4 (chk): use single loop, make end mask 64bit
Signed-off-by: Roger He <Hongbo.He@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Roger He <Hongbo.He@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 31 Aug 2017 13:48:11 +0000 (09:48 -0400)]
drm/amd/amdgpu: Simplify gfx_v9_0_wait_for_idle()
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>