On 4/20/2022 11:51 AM, Zhou1, Tao wrote:
[AMD Official Use Only]
-Original Message-
From: Lazar, Lijo
Sent: Wednesday, April 20, 2022 12:33 PM
To: Zhou1, Tao ; amd-gfx@lists.freedesktop.org; Zhang,
Hawking ; Yang, Stanley
; Ziya, Mohammad zafar
; Chai, Thomas
Subject: Re: [PATCH
[AMD Official Use Only]
> -Original Message-
> From: Lazar, Lijo
> Sent: Wednesday, April 20, 2022 12:33 PM
> To: Zhou1, Tao ; amd-gfx@lists.freedesktop.org; Zhang,
> Hawking ; Yang, Stanley
> ; Ziya, Mohammad zafar
> ; Chai, Thomas
> Subject: Re: [PATCH 2/3] drm/amdgpu: add RAS poison
Please also keep the naming style consistent by using "amdgpu_ras" prefix
Regards,
Hawking
-Original Message-
From: Zhou1, Tao
Sent: Wednesday, April 20, 2022 14:17
To: Lazar, Lijo ; amd-gfx@lists.freedesktop.org; Zhang,
Hawking ; Yang, Stanley ; Ziya,
Mohammad zafar ; Chai, Thomas
S
[AMD Official Use Only]
> -Original Message-
> From: Lazar, Lijo
> Sent: Wednesday, April 20, 2022 12:10 PM
> To: Zhou1, Tao ; amd-gfx@lists.freedesktop.org; Zhang,
> Hawking ; Yang, Stanley
> ; Ziya, Mohammad zafar
> ; Chai, Thomas
> Subject: Re: [PATCH 1/3] drm/amdgpu: implement RAS
On 4/20/2022 9:23 AM, Tao Zhou wrote:
Add support for general RAS poison consumption handler.
Signed-off-by: Tao Zhou
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 43 -
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h | 1 +
2 files changed, 43 insertions(+), 1 deletion(
On 4/20/2022 9:23 AM, Tao Zhou wrote:
Prepare for the implementation of poison consumption handler.
Signed-off-by: Tao Zhou
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 62 ++---
1 file changed, 35 insertions(+), 27 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgp
On Thu, 2022-04-07 at 10:59 +0200, Christian König wrote:
> Rework the internals of the dma_resv object to allow adding more than
> one
> write fence and remember for each fence what purpose it had.
>
> This allows removing the workaround from amdgpu which used a container
> for
> this instead.
>
The fatal error handler is independent from general ras interrupt
handler since there is no related IH ring.
Signed-off-by: Tao Zhou
---
drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c | 15 +--
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 20
drivers/gpu/drm/amd/amdgpu/amd
Add support for general RAS poison consumption handler.
Signed-off-by: Tao Zhou
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 43 -
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h | 1 +
2 files changed, 43 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/am
Prepare for the implementation of poison consumption handler.
Signed-off-by: Tao Zhou
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 62 ++---
1 file changed, 35 insertions(+), 27 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
b/drivers/gpu/drm/amd/amdgpu/
On 4/20/2022 7:32 AM, Prike Liang wrote:
Without MMHUB clock gating being enabled then MMHUB will not disconnect
from DF and will result in DF C-state entry can't be accessed during S2idle
suspend, and eventually s0ix entry will be blocked.
Signed-off-by: Prike Liang
---
drivers/gpu/drm/am
This reverts commit a129dbca92edaeb5a515a0e5b63ada71cf9b981a.
It causes SVM regressions on Vega10 with XNACK-ON. Just revert it
at the moment.
./kfdtest --gtest_filter=KFDSVMRangeTest.MigratePolicyTest
Signed-off-by: Lang Yu
---
drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 4 +---
1 file changed, 1
Without MMHUB clock gating being enabled then MMHUB will not disconnect
from DF and will result in DF C-state entry can't be accessed during S2idle
suspend, and eventually s0ix entry will be blocked.
Signed-off-by: Prike Liang
---
drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c | 9 +
1 file cha
Change SVM range mapping flags or access attributes don't trigger
migration, if range is already mapped on GPUs we should update GPU
mapping, and pass flush_tlb flag to amdgpu vm.
Signed-off-by: Philip Yang
---
drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 41 ++--
1 file change
To avoid unnecessary unmap SVM range from GPUs if range is not mapped on
GPUs when migrating the range. This flag will also be used to flush TLB
when updating the existing mapping on GPUs.
It is protected by prange->migrate_mutex and mmap read lock in MMU
notifier callback.
Signed-off-by: Philip
Hi Paul,
On 4/14/2022 2:52 AM, Paul Menzel wrote:
[Cc: -kernel test robot ]
Dear Alex, dear Richard,
Am 13.04.22 um 15:00 schrieb Alex Deucher:
On Wed, Apr 13, 2022 at 3:43 AM Paul Menzel wrote:
Thank you for sending out v4.
Am 12.04.22 um 23:50 schrieb Richard Gong:
Active State Power M
Hi Nathan,
On 4/13/2022 10:40 AM, Nathan Chancellor wrote:
Hi Richard,
On Tue, Apr 12, 2022 at 04:50:00PM -0500, Richard Gong wrote:
Active State Power Management (ASPM) feature is enabled since kernel 5.14.
There are some AMD GFX cards (such as WX3200 and RX640) that won't work
with ASPM-enab
The Board of Directors election and the vote on the By-laws concluded at
23:59 UTC on 18 April 2022. There are 80 current Members of the X.Org
Foundation, and 52 Members cast votes. This is a 65.0% turn out.
In the election of the Directors to the Board of the X.Org Foundation,
the results were th
On Tue, Apr 19, 2022 at 3:38 PM Alice Wong wrote:
>
> amdgpu_device_ip_hw_init_phase1 and amdgpu_device_ip_hw_init_phase2
> call IP blocks' respective hw_fini when hw_init failed.
I don't think we should call hw_fini() if hw_init() failed. If there
are things that are not properly cleaned up in
[AMD Official Use Only]
Remove firmware.load_type check in amdgpu_ucode_free_bo.
If psp failed to load a firmware, it will set formware.load_type to
AMDGPU_FW_LOAD_DIRECT for bring up purposes. This caused memory leak as
amdpug_ucode_free_bo will skip freeing memory due to the load_type.
BUG:S
[AMD Official Use Only]
amdgpu_device_ip_hw_init_phase1 and amdgpu_device_ip_hw_init_phase2 call IP
blocks' respective hw_fini when hw_init failed.
BUG:SWDEV-332449
Signed-off-by: Alice Wong
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/d
Remove firmware.load_type check in amdgpu_ucode_free_bo.
If psp failed to load a firmware, it will set formware.load_type to
AMDGPU_FW_LOAD_DIRECT for bring up purposes. This caused memory leak
as amdpug_ucode_free_bo will skip freeing memory due to the load_type.
BUG:SWDEV-332449
Signed-off-by:
amdgpu_device_ip_hw_init_phase1 and amdgpu_device_ip_hw_init_phase2
call IP blocks' respective hw_fini when hw_init failed.
BUG:SWDEV-332449
Signed-off-by: Alice Wong
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/a
tree/branch:
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git master
branch HEAD: 634de1db0e9bbeb90d7b01020e59ec3dab4d38a1 Add linux-next specific
files for 20220419
Error/Warning reports:
https://lore.kernel.org/linux-mm/202204081656.6x4pfen4-...@intel.com
https
Am 2022-04-19 um 12:01 schrieb Andrey Grodzovsky:
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -134,6 +134,7 @@ struct amdkfd_process_info {
/* MMU-notifier related fields */
atomic_t evicted_bos;
+atomic_t invalid;
struct delayed_work resto
On 2022-04-19 10:42, Christian König wrote:
Am 19.04.22 um 16:37 schrieb Harry Wentland:
On 2022-04-19 10:19, Christian König wrote:
We just need to reserve the BO here, no need for using ttm_eu.
Signed-off-by: Christian König
Acked-by: Harry Wentland
What about the second patch? Wh
On Sat, Apr 16, 2022 at 11:41 AM Julia Lawall wrote:
>
> From: kernel test robot
>
> Use max to simplify the code.
>
> Generated by: scripts/coccinelle/misc/minmax.cocci
>
> CC: Denis Efremov
> Reported-by: kernel test robot
> Signed-off-by: kernel test robot
> Signed-off-by: Julia Lawall
Th
On Tue, Apr 19, 2022 at 10:55 AM Christian König
wrote:
>
> Am 19.04.22 um 16:49 schrieb Alex Deucher:
> > On Tue, Apr 19, 2022 at 10:19 AM Christian König
> > wrote:
> >> We just need to reserve the BO here, no need for using ttm_eu.
> > Can you include a more detailed description as to why it's
Applied the series with minor fix to capitalize the U in Compute Unit. Thanks!
Alex
On Fri, Apr 15, 2022 at 3:52 PM Tales Lelo da Aparecida
wrote:
>
> Add missing acronyms to the amdgppu glossary.
>
> Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/1939#note_1309737
> Signed-off-by: Tal
Am 19.04.22 um 16:49 schrieb Alex Deucher:
On Tue, Apr 19, 2022 at 10:19 AM Christian König
wrote:
We just need to reserve the BO here, no need for using ttm_eu.
Can you include a more detailed description as to why it's not
necessary?
Well the logic in ttm_eu is for reserving multiple buffe
Applied. Thanks!
On Sat, Apr 16, 2022 at 2:48 PM Tom Rix wrote:
>
> evergreen_default_state and evergreen_default_size are only
> used in evergreen.c. Single file symbols should be static.
> So move their definitions to evergreen_blit_shaders.h
> and change their storage-class-specifier to stat
Applied. Thanks!
On Mon, Apr 18, 2022 at 3:48 PM Tom Rix wrote:
>
> Smatch reports this issue
> virtual_link_hwss.c:32:6: warning: symbol
> 'virtual_setup_stream_attribute' was not declared.
> Should it be static?
>
> virtual_setup_stream_attribute is only used in
> virtual_link_hwss.c, but
On Tue, Apr 19, 2022 at 10:19 AM Christian König
wrote:
>
> We just need to reserve the BO here, no need for using ttm_eu.
Can you include a more detailed description as to why it's not
necessary? Most of this code was copied from radeon originally. Does
radeon need a similar cleanup?
Alex
>
On Tue, Apr 19, 2022 at 10:04 AM Paul Menzel wrote:
>
> Dear Alex,
>
>
> Thank you for the patch.
>
> Am 13.04.22 um 22:15 schrieb Alex Deucher:
> > We normally runtime suspend when there are displays attached if they
> > are in the DPMS off state, however, if something wakes the GPU
> > we send a
Am 19.04.22 um 16:37 schrieb Harry Wentland:
On 2022-04-19 10:19, Christian König wrote:
We just need to reserve the BO here, no need for using ttm_eu.
Signed-off-by: Christian König
Acked-by: Harry Wentland
What about the second patch? Who takes care of amdgpu_vkms.c? You guys
or shou
Applied. Thanks!
On Tue, Apr 19, 2022 at 8:49 AM Keita Suzuki
wrote:
>
> In function si_parse_power_table(), array adev->pm.dpm.ps and its member
> is allocated. If the allocation of each member fails, the array itself
> is freed and returned with an error code. However, the array is later
> fre
On 2022-04-19 10:19, Christian König wrote:
We just need to reserve the BO here, no need for using ttm_eu.
Signed-off-by: Christian König
Acked-by: Harry Wentland
Harry
---
.../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 32 ++-
1 file changed, 17 insertions(+), 15 d
Applied with minor change to drop the prototype in dcn10_hubp.h. Thanks!
Alex
On Fri, Apr 15, 2022 at 2:21 PM Tales Lelo da Aparecida
wrote:
>
> It's a local function, let's make it static.
>
> Signed-off-by: Tales Lelo da Aparecida
> ---
> drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c |
We just need to reserve the BO here, no need for using ttm_eu.
Signed-off-by: Christian König
---
.../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 32 ++-
1 file changed, 17 insertions(+), 15 deletions(-)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
b/drivers/g
We just need to reserve the BO here, no need for using ttm_eu.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c | 32 +---
1 file changed, 17 insertions(+), 15 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c
b/drivers/gpu/drm/am
Dear Alex,
Thank you for the patch.
Am 13.04.22 um 22:15 schrieb Alex Deucher:
We normally runtime suspend when there are displays attached if they
are in the DPMS off state, however, if something wakes the GPU
we send a hotplug event on resume (in case any displays were connected
while the GP
On Tue, Apr 19, 2022 at 7:06 AM Christian König
wrote:
>
> This reverts commit 461fa7b0ac565ef25c1da0ced31005dd437883a7.
>
> We are missing some inter dependencies here so re-introduce the lock
> until we have figured out what's missing. Just drop/retake it while
> adding dependencies.
>
> v2: sti
Ping?
On Wed, Apr 13, 2022 at 4:15 PM Alex Deucher wrote:
>
> We normally runtime suspend when there are displays attached if they
> are in the DPMS off state, however, if something wakes the GPU
> we send a hotplug event on resume (in case any displays were connected
> while the GPU was in suspe
In function si_parse_power_table(), array adev->pm.dpm.ps and its member
is allocated. If the allocation of each member fails, the array itself
is freed and returned with an error code. However, the array is later
freed again in si_dpm_fini() function which is called when the function
returns an er
> -Original Message-
> From: Paul Menzel
> Sent: Tuesday, April 19, 2022 2:54 AM
> To: Yat Sin, David
> Cc: Kuehling, Felix ; amd-
> g...@lists.freedesktop.org
> Subject: Re: [PATCH v2 2/2] drm/amdkfd: CRIU add support for GWS queues
>
>
> Dear David,
>
>
> Thank you for sending out
This reverts commit 461fa7b0ac565ef25c1da0ced31005dd437883a7.
We are missing some inter dependencies here so re-introduce the lock
until we have figured out what's missing. Just drop/retake it while
adding dependencies.
v2: still drop the lock while adding dependencies
Signed-off-by: Christian K
Smatch reports this issue
virtual_link_hwss.c:32:6: warning: symbol
'virtual_setup_stream_attribute' was not declared.
Should it be static?
virtual_setup_stream_attribute is only used in
virtual_link_hwss.c, but the other functions in the
file are declared in the header file and used elsewhere
evergreen_default_state and evergreen_default_size are only
used in evergreen.c. Single file symbols should be static.
So move their definitions to evergreen_blit_shaders.h
and change their storage-class-specifier to static.
Remove unneeded evergreen_blit_shader.c
evergreen_ps/vs definitions wer
48 matches
Mail list logo
