leandrof/linux
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
768,455 Commits 2 Branches 0 Tags
c9b51a017065bf514ab5a380f2fb3b0dada5bc68
Commit Graph

4194 Commits

Author SHA1 Message Date
Alex Deucher
273a14cd15 drm/amdgpu/gmc9: add vega12 support (v2) 
Same as vega10.

v2: squash in golden regs fix from Feifei

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
 2018-03-21 14:36:46 -05:00
Alex Deucher
dca7b4015c drm/amdgpu: add vega12 to dc support check 
DC is used for modesetting on vega12.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
 2018-03-21 14:24:51 -05:00
Alex Deucher
f79f3fc141 drm/amdgpu/virtual_dce: add vega12 support 
Add virtual dce support for vega12.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
 2018-03-21 14:24:35 -05:00
Alex Deucher
9aa52bc487 drm/amdgpu: specify vega12 vce firmware 
Declare firmware and add support for the file.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
 2018-03-21 14:24:30 -05:00
Alex Deucher
2327e6261a drm/amdgpu: specify vega12 uvd firmware 
Declare firmware and add support for the file.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
 2018-03-21 14:24:24 -05:00
Feifei Xu
6b9c6e1b15 drm/amdgpu: add vega12 ucode loading method 
Same as vega10.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
 2018-03-21 14:24:18 -05:00
Evan Quan
eddfa8dfad drm/amdgpu: initilize vega12 psp firmwares 
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Acked-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-21 14:24:12 -05:00
Alex Deucher
ff13dc671d drm/amdgpu/psp: initial vega12 support 
Same as vega10 for now.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
 2018-03-21 14:24:07 -05:00
Hawking Zhang
ad0ad9aa20 drm/amdgpu: vega12 to smu firmware 
Add the cgs interface to query the smu firmware for vega12
and declare the firmware.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
 2018-03-21 14:24:00 -05:00
Alex Deucher
e48a3cd9cb drm/amdgpu: set asic family and ip blocks for vega12 
soc15 just like vega10 and raven.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
 2018-03-21 14:23:55 -05:00
Alex Deucher
3f76dcedb3 drm/amdgpu: add gpu_info firmware for vega12 
Stores gpu configuration details.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
 2018-03-21 14:23:49 -05:00
Feifei Xu
8fab806ad1 drm/amdgpu: add vega12 to asic_type enum 
Add vega12 to amd_asic_type enum and amdgpu_asic_name[].

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
 2018-03-21 14:23:39 -05:00
Sean Paul
1c7095d283 Merge airlied/drm-next into drm-misc-next 
Refresh -misc-next

Signed-off-by: Sean Paul <seanpaul@chromium.org>
 2018-03-21 09:40:55 -04:00
Rex Zhu
eb4900aa4c drm/amdgpu: Fix kernel NULL pointer dereference in dpm functions 
caused by
'commit 83e3c4615872 ("drm/amdgpu: Remove wrapper layer of smu ip functions")'

BUG: unable to handle kernel NULL pointer dereference at 00000000000005d8
[  313.241459] IP: ci_dpm_read_sensor+0x37/0xf0 [amdgpu]

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>
 2018-03-20 23:44:50 -05:00
Rex Zhu
81ce8bea03 drm/amdgpu: Fix kernel NULL pointer dereference when amdgpu fini 
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>
 2018-03-20 23:44:21 -05:00
Mikita Lipski
e5b03032e3 drm/amdgpu - Disable all irqs before disabling all CRTCs 
By moving amdgpu_irq_disable_all earlier in the sequence
fixes an issue with disabling pflip interrupts:

*ERROR* dal_irq_service_dummy_ack: called for non-implemented irq source

Earlier patch fixed a memory corruption and revealed irq
warnings.This way it seems to be there no obvious issues
with unloading the module.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-20 23:44:12 -05:00
Mikita Lipski
088e7c1617 drm/amdgpu: Disable irq on device before destroying it 
Disable irq on devices before destroying them. That prevents
use-after-free memory access when unloading the driver.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-20 23:44:03 -05:00
Mikita Lipski
ff97cba8c1 drm/amdgpu: Use atomic function to disable crtcs with dc enabled 
This change fixes the deadlock when unloading the driver with displays
connected.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-20 23:43:42 -05:00
Tom St Denis
864917a3b8 drm/amd/amdgpu: fix offset into page with amdgpu_iomem debugfs file 
The offset inside the page wasn't included in the copy call meaning
the start of the page was being read/written instead.

Reported-by: Jay Cornwall <Jay.Cornwall@amd.com>
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>
 2018-03-20 23:43:34 -05:00
Hawking Zhang
bd08a8d9e8 drm/amdgpu/nbio6: Correct PCIE_INDEX/DATA pair used for smn register accessing 
PCIE_INDEX2/DATA2 pair will be used for smn register accessing since from vega.
PCIE_INDEX/DATA pair should be reserved for smu

Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-20 23:43:25 -05:00
Alex Deucher
e3ecdffac9 drm/amdgpu: add documentation for amdgpu_device.c 
Add kernel doc for the functions in amdgpu_device.c

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-20 23:43:05 -05:00
Alex Deucher
3ac952b10d drm/amdgpu: add VCN to firmware query interface 
Need to be able to query the VCN firmware version from
userspace to determine supported features, etc.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Acked-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-20 23:42:49 -05:00
Rex Zhu
2538090cb6 drm/amd/pp: Delete get_xclk function in powerplay (v2) 
use asic's callback function get_xclk in amdgpu

v2: squash in removal of leftover debug info
(drm/amd/pp: Delete debug info in smu7_hwmgr.c) (Rex)

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-19 13:38:31 -05:00
Rex Zhu
699f47951e drm/amdgpu: Delete dead code when early init 
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-19 13:34:46 -05:00
Rex Zhu
59156faf81 drm/amd/pp: Remove the cgs wrapper for notify smu version on APU 
Refine commit f49e9bac19 ("drm/amd/pp: Get and save Rv smu version")

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>
 2018-03-16 16:16:49 -05:00
Felix Kuehling
ede0dd86f4 drm/amdgpu: Add kfd2kgd interface to acquire an existing VM 
This allows acquiring an existing VM from a render node FD to use it
for a compute process.

Such VMs get destroyed when the original file descriptor is released.
Added a callback from amdgpu_vm_fini to handle KFD VM destruction
correctly in this case.

v2:
* Removed vm->vm_context check in amdgpu_amdkfd_gpuvm_destroy_cb,
  check vm->process_info earlier instead

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
 2018-03-15 17:27:43 -04:00
Felix Kuehling
b236fa1d33 drm/amdgpu: Add helper to turn an existing VM into a compute VM 
v2: Removed updating and checking of vm->vm_context
v3: Enable amdgpu_vm_clear_bo in amdgpu_vm_make_compute

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
 2018-03-15 17:27:42 -04:00
Felix Kuehling
3486625bbf drm/amdgpu: Fix initial validation of PD BO for KFD VMs 
Make sure the PD BO is valid and attach the eviction fence during VM
creation. This ensures that the pd_phys_address is actually valid
and an eviction that would invalidate it triggers a KFD process
eviction like it should.

v2: Use uninterruptible waiting in initial PD validation

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
 2018-03-15 17:27:41 -04:00
Felix Kuehling
5b21d3e5fd drm/amdgpu: Move KFD-specific fields into struct amdgpu_vm 
Remove struct amdkfd_vm and move the fields into struct amdgpu_vm.
This will allow turning a VM created by a DRM render node into a
KFD VM.

v2: Removed vm_context field

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
 2018-03-15 17:27:40 -04:00
Andrey Grodzovsky
64350f1b7f drm/amdgpu: Improve documentation of bo_ptr in amdgpu_bo_create_kernel 
and amdgpu_bo_create_reserved.

Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@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>
 2018-03-15 14:34:15 -05:00
Mikita Lipski
d1bb88e078 drm/amdgpu: Use atomic function to disable crtcs with dc enabled 
This change fixes the deadlock when unloading the driver with displays
connected.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-15 10:46:37 -05:00
Rex Zhu
2dac5936e5 drm/amdgpu: Call amdgpu_ucode_fini_bo in amd_powerplay.c 
make it symmetric with amdgpu_ucode_init_bo in amd_powerplay.c

refine the "commit b22558bb4ff8fc9fe925222f90297d7a03a5fb20"

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-15 09:57:58 -05:00
Rex Zhu
b905090d2b drm/amdgpu: Remove wrapper layer of smu ip functions 
1. delete amdgpu_powerplay.c used for wrapping smu ip functions
2. delete struct pp_instance,
3. make struct hwmgr as the smu hw handle.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-15 09:57:50 -05:00
Rex Zhu
5b2a3d2c15 drm/amdgpu: Don't compared ip_block_type with ip_block_index 
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-15 09:57:41 -05:00
Rex Zhu
5771632723 drm/amdgpu: Plus NULL function pointer check 
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-15 09:57:32 -05:00
Oded Gabbay
7420f482ea drm/amdkfd: add missing include of mm.h 
This patch fixes kernel build in ARCH=frv

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
 2018-03-15 10:08:35 +02:00
Feifei Xu
2f51d6e8e4 drm/amdgpu/sdma4: Remove unused header file from sdma_v4_0.c 
Remove mmhub header files inclusion which not used.

Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 16:01:19 -05:00
Alex Deucher
fcb7d51571 drm/amdgpu/sdma4: use num_instances for clock/powergating config 
Rather then relying on the asic type for the second instance.
Makes it more consistent with the rest of the code.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
 2018-03-14 16:01:18 -05:00
Michel Dänzer
daf8809626 drm/amdgpu/dce: Don't turn off DP sink when disconnected 
Turning off the sink in this case causes various issues, because
userspace expects it to stay on until it turns it off explicitly.

Instead, turn the sink off and back on when a display is connected
again. This dance seems necessary for link training to work correctly.

Bugzilla: https://bugs.freedesktop.org/105308
Cc: stable@vger.kernel.org
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>
 2018-03-14 16:01:17 -05:00
Alex Deucher
5494d8640f drm/amdgpu: move getting pcie info to common code 
No need to replicate it in several places.

Reviewed-by: Rex Zhu <rezhu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 16:01:16 -05:00
Alex Deucher
3cdfe70056 drm/amdgpu/soc15: always load the psp module 
Regardless of whether the user has selected psp fw loading or
not.  It's still needed for GPU reset among other things.
There are already guards in place to avoid setting up the full
psp if PSP fw loading is not enabled.

Reviewed-by: Rex Zhu <rezhu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 16:01:16 -05:00
Alex Deucher
454bbbf915 drm/amdgpu: use adev->firmware to determine whether to load the PSP module 
The per device firmware load method is limited to what makes sense for
that asic rather than whatever arbitrary value may have been set by the
user.

Reviewed-by: Rex Zhu <rezhu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 16:01:16 -05:00
Alex Deucher
19aede7791 drm/amdgpu: move firmware loading type setup to common code 
No need to replicate it in several places.

Reviewed-by: Rex Zhu <rezhu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 16:01:15 -05:00
Alex Deucher
a9f3636019 drm/amdgpu/psp: add a few more fw load type checks 
We already checked and returned early in most of the IP
functions, fill in the rest as well.

Reviewed-by: Rex Zhu <rezhu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 16:01:15 -05:00
Michel Dänzer
7d617264eb drm/amdgpu/dce: Don't turn off DP sink when disconnected 
Turning off the sink in this case causes various issues, because
userspace expects it to stay on until it turns it off explicitly.

Instead, turn the sink off and back on when a display is connected
again. This dance seems necessary for link training to work correctly.

Bugzilla: https://bugs.freedesktop.org/105308
Cc: stable@vger.kernel.org
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>
 2018-03-14 15:40:00 -05:00
Alex Deucher
b5e3241316 drm/amdgpu: save/restore backlight level in legacy dce code 
Save/restore the backlight level scratch register in S3/S4 so the
backlight level comes back at the previously requested level.

Bug: https://bugzilla.kernel.org/show_bug.cgi?id=199047
Fixes: 4ec6ecf48c (drm/amdgpu: drop scratch regs save and restore from S3/S4 handling)
Acked-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 15:39:51 -05:00
Christian König
342038d924 drm/amdgpu: fix prime teardown order 
We unmapped imported DMA-bufs when the GEM handle was dropped, not when the
hardware was done with the buffere.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
CC: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 15:39:35 -05:00
Oak Zeng
3760f76cbe drm/amdgpu: Move IH clientid defs to separate file 
This is preparation for sharing client ID definitions
between amdgpu and amdkfd

Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 15:16:35 -05:00
Harry Wentland
d179664a2a drm/amdgpu: Remove some unused elements from amdgpu_connector struct 
They were used by amdgpu_dm at some point but since it has its own
amdgpu_dm_connector now these aren't needed anymore.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 15:08:40 -05:00
Alex Deucher
a59b3c80fd drm/amdgpu: save/restore backlight level in legacy dce code 
Save/restore the backlight level scratch register in S3/S4 so the
backlight level comes back at the previously requested level.

Bug: https://bugzilla.kernel.org/show_bug.cgi?id=199047
Fixes: 4ec6ecf48c (drm/amdgpu: drop scratch regs save and restore from S3/S4 handling)
Acked-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
 2018-03-14 15:08:40 -05:00