Am 28.01.19 um 22:52 schrieb Marek Olšák:
From: Marek Olšák
Normal syncobjs signal when an IB finishes. Start syncobjs signal when
an IB starts.
That approach has quite a number of problems (for example you can't
allocate memory at this point).
Better add a flag that we should only sync on
Given a master fd we can then override the priority of the context
in another fd.
Using these overrides was recommended by Christian instead of trying
to submit from a master fd, and I am adding a way to override a
single context instead of the entire process so we can only upgrade
a single Vulkan
Otherwise we interpret the file private data as drm & amdgpu data
while it might not be, possibly allowing one to get memory corruption.
Signed-off-by: Bas Nieuwenhuizen
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 ++
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 16
driver
If we have some disabled HW blocks (say VCN), then the rings are
not initialized. This reuslts in entities that refer to uninitialized
rqs (runqueues?).
In normal usage this does not result in issues because userspace
generally knows to ignore the unsupported blocks, but e.g. setting
the prioritie
Am 29.01.19 um 11:20 schrieb Bas Nieuwenhuizen:
If we have some disabled HW blocks (say VCN), then the rings are
not initialized. This reuslts in entities that refer to uninitialized
rqs (runqueues?).
In normal usage this does not result in issues because userspace
generally knows to ignore the
On Tue, Jan 29, 2019 at 11:33 AM Christian König
wrote:
>
> Am 29.01.19 um 11:20 schrieb Bas Nieuwenhuizen:
> > If we have some disabled HW blocks (say VCN), then the rings are
> > not initialized. This reuslts in entities that refer to uninitialized
> > rqs (runqueues?).
> >
> > In normal usage t
If a non-legacy driver calls these it's valid to assume there is
interrupt support. The flag is really only needed for legacy drivers,
which control IRQ enabling/disabling through the DRM_IOCTL_CONTROL
legacy IOCTL.
Also remove all the flag usage from non-legacy drivers.
v2: Review from Emil:
- i
Am 29.01.19 um 11:41 schrieb Bas Nieuwenhuizen:
> On Tue, Jan 29, 2019 at 11:33 AM Christian König
> wrote:
>> Am 29.01.19 um 11:20 schrieb Bas Nieuwenhuizen:
>>> If we have some disabled HW blocks (say VCN), then the rings are
>>> not initialized. This reuslts in entities that refer to uninitiali
On Tue, Jan 29, 2019, 3:01 AM Christian König <
ckoenig.leichtzumer...@gmail.com wrote:
> Am 28.01.19 um 22:52 schrieb Marek Olšák:
> > From: Marek Olšák
> >
> > Normal syncobjs signal when an IB finishes. Start syncobjs signal when
> > an IB starts.
>
> That approach has quite a number of proble
Testing with the new 5.0.0-rc1 amd-staging-drm-next branch results in
this commit causing the attached lockup on init with my Raven + Polaris
system:
10117450735c7a7c0858095fb46a860e7037cb9a is the first bad commit
commit 10117450735c7a7c0858095fb46a860e7037cb9a
Author: ndesaulni...@google.com
Am 29.01.19 um 14:32 schrieb Marek Olšák:
On Tue, Jan 29, 2019, 3:01 AM Christian König
mailto:ckoenig.leichtzumer...@gmail.com>
wrote:
Am 28.01.19 um 22:52 schrieb Marek Olšák:
> From: Marek Olšák mailto:marek.ol...@amd.com>>
>
> Normal syncobjs signal when an IB finishes. Start syncobjs signa
See:
https://bugs.freedesktop.org/show_bug.cgi?id=109487
From: amd-gfx on behalf of StDenis, Tom
Sent: Tuesday, January 29, 2019 8:44:03 AM
To: Wentland, Harry; amd-gfx mailing list
Cc: Deucher, Alexander
Subject: regression in dcn linked to -msse2
Testing wi
On Fri, Jan 25, 2019 at 10:29 AM Wentland, Harry wrote:
>
> On 2019-01-24 7:52 p.m., ndesaulni...@google.com wrote:
> > arch/x86/Makefile disables SSE and SSE2 for the whole kernel. The
> > AMDGPU drivers modified in this patch re-enable SSE but not SSE2. Turn
> > on SSE2 to support emitting dou
This reverts commit 10117450735c7a7c0858095fb46a860e7037cb9a.
Causes a crash.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=109487
Signed-off-by: Alex Deucher
Cc: sta...@vger.kernel.org # 4.19
---
drivers/gpu/drm/amd/display/dc/calcs/Makefile | 2 +-
drivers/gpu/drm/amd/display/dc/dml/Makef
On Tue, Jan 29, 2019 at 12:38 PM Greg KH wrote:
>
> On Tue, Jan 29, 2019 at 11:57:25AM -0500, Alex Deucher wrote:
> > This reverts commit 10117450735c7a7c0858095fb46a860e7037cb9a.
>
> $ git show 10117450735c7a7c0858095fb46a860e7037cb9a
> fatal: bad object 10117450735c7a7c0858095fb46a860e7037cb9a
On Tue, Jan 29, 2019 at 11:57:25AM -0500, Alex Deucher wrote:
> This reverts commit 10117450735c7a7c0858095fb46a860e7037cb9a.
$ git show 10117450735c7a7c0858095fb46a860e7037cb9a
fatal: bad object 10117450735c7a7c0858095fb46a860e7037cb9a
odd...
___
amd
On 2019-01-29 11:57 a.m., Alex Deucher wrote:
> This reverts commit 10117450735c7a7c0858095fb46a860e7037cb9a.
>
> Causes a crash.
>
> Bug: https://bugs.freedesktop.org/show_bug.cgi?id=109487
> Signed-off-by: Alex Deucher
> Cc: sta...@vger.kernel.org # 4.19
Thanks for finding and reverting.
Rev
On Tue, Jan 29, 2019 at 10:30:31AM -0500, Alex Deucher wrote:
> On Fri, Jan 25, 2019 at 10:29 AM Wentland, Harry
> wrote:
> >
> > On 2019-01-24 7:52 p.m., ndesaulni...@google.com wrote:
> > > arch/x86/Makefile disables SSE and SSE2 for the whole kernel. The
> > > AMDGPU drivers modified in this
On 2019-01-29 1:56 p.m., Guenter Roeck wrote:
> On Tue, Jan 29, 2019 at 10:30:31AM -0500, Alex Deucher wrote:
>> On Fri, Jan 25, 2019 at 10:29 AM Wentland, Harry
>> wrote:
>>>
>>> On 2019-01-24 7:52 p.m., ndesaulni...@google.com wrote:
arch/x86/Makefile disables SSE and SSE2 for the whole ke
Suggestions:
1. revert patch
2. get me the disassembly from gcc of the translation unit in question.
3. land patch that adds clang guards or something different based on 2.
Revert first; ask questions later.
On Tue, Jan 29, 2019 at 11:01 AM Wentland, Harry wrote:
>
> On 2019-01-29 1:56 p.m., Gue
From: Roman Li
[Why]
The earlier change 'Fix 6x4K displays' led to fclk value
idling at higher DPM level.
[How]
Apply the fix only to respective multi-display configuration.
Signed-off-by: Roman Li
---
drivers/gpu/drm/amd/display/dc/dce/dce_clk_mgr.c | 10 +-
1 file changed, 9 inserti
Acked-by: Alex Deucher
From: amd-gfx on behalf of
roman...@amd.com
Sent: Tuesday, January 29, 2019 3:16:53 PM
To: amd-gfx@lists.freedesktop.org; Xu, Feifei; Quan, Evan
Cc: Chan, Carl; Wentland, Harry; Li, Roman
Subject: [PATCH] drm/amd/display: Fix fclk idle st
Some of the grandfathered amd display code does not follow
Linux coding style and emits warnings or errors on checkpatch
No functional changes here - just cleanup
Signed-off-by: David Francis
---
.../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 96 +--
1 file changed, 47 insertio
On Tue, Jan 29, 2019 at 4:03 PM David Francis wrote:
>
> Some of the grandfathered amd display code does not follow
> Linux coding style and emits warnings or errors on checkpatch
>
> No functional changes here - just cleanup
>
> Signed-off-by: David Francis
Acked-by: Alex Deucher
> ---
> ...
On Tue, Jan 29, 2019 at 3:01 AM Christian König <
ckoenig.leichtzumer...@gmail.com> wrote:
> Am 28.01.19 um 22:52 schrieb Marek Olšák:
> > From: Marek Olšák
> >
> > Normal syncobjs signal when an IB finishes. Start syncobjs signal when
> > an IB starts.
>
> That approach has quite a number of pro
On 2019-01-14 5:01 a.m., Evan Quan wrote:
> + features_to_disable =
> + (features_enabled ^ new_ppfeature_masks) & features_enabled;
> + features_to_enable =
> + (features_enabled ^ new_ppfeature_masks) ^ features_to_disable;
This is confusing and unnecessarily com
On 2019-01-21 4:55 a.m., Koenig, Christian wrote:
> Am 21.01.19 um 05:35 schrieb Liu, Monk:
>>
>> > Actually that's not so crazy at all. See the ATC uses the CPU page
>> tables to provide parts of the virtual GPU address space.
>>
>>
>>
>> So aperture 0->hole-start will be translated **not** by G
Some blocks in amdgpu can have 0 rqs.
Job creation already fails with -ENOENT when entity->rq is NULL,
so jobs cannot be pushed. Without a rq there is no scheduler to
pop jobs, and rq selection already does the right thing with a
list of length 0.
So the operations we need to fix are:
- Creatio
Otherwise we interpret the file private data as drm & amdgpu data
while it might not be, possibly allowing one to get memory corruption.
Signed-off-by: Bas Nieuwenhuizen
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 ++
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 16
driver
I don't see another way to figure out if a ring is initialized if
the hardware block might not be initialized.
Entities have been fixed up to handle num_rqs = 0.
Signed-off-by: Bas Nieuwenhuizen
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 11 ---
1 file changed, 8 insertions(+), 3 del
Given a master fd we can then override the priority of the context
in another fd.
Using these overrides was recommended by Christian instead of trying
to submit from a master fd, and I am adding a way to override a
single context instead of the entire process so we can only upgrade
a single Vulkan
sriov's gpu_recover inside xgpu_ai_mailbox_flr_work would cause duplicate
recover in TDR.
TDR's gpu_recover would be triggered by amdgpu_job_timedout,
that could avoid vk-cts failure by unexpected recover.
Change-Id: I840dfc145e4e1be9ece6eac8d9f3501da9b28ebf
Signed-off-by: wentalou
---
drivers/
32 matches
Mail list logo
