On 2018-05-18 05:57 PM, Oded Gabbay wrote:
>>> +struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void)
>>> +{
>>> + return NULL;
>> I think this will cause an oops in amdgpu_amdkfd_probe because that
>> function doesn't handle kgd2kfd == NULL. You could remove
>> amdgpu_amdkfd_gfx_*_get_
On 2018-05-18 05:57 PM, Oded Gabbay wrote:
> In case CONFIG_HSA_AMD is not chosen, there is no need to compile amdkfd
> files that reside inside amdgpu dirver. In addition, because amdkfd
> depends on x86_64 architecture and amdgpu is not, compiling amdkfd files
> under i386 architecture can cause
On Fri, May 18, 2018 at 11:06 PM, Felix Kuehling wrote:
> Two more comments inline. One cosmetic, one real issue. With that fixed,
> this patch is Reviewed-by: Felix Kuehling
>
> Regards,
> Felix
>
> On 2018-05-18 03:42 PM, Oded Gabbay wrote:
>> In case CONFIG_HSA_AMD is not chosen, there is no
In case CONFIG_HSA_AMD is not chosen, there is no need to compile amdkfd
files that reside inside amdgpu dirver. In addition, because amdkfd
depends on x86_64 architecture and amdgpu is not, compiling amdkfd files
under i386 architecture can cause compiler errors and warnings.
This patch modifies
On 2018-05-18 02:54 PM, Alex Deucher wrote:
> On Thu, May 17, 2018 at 6:20 PM, Samuel Li wrote:
>> Signed-off-by: Samuel Li
>
> Please also add a separate DOC section at the top of this file giving
> a brief overview of how amdgpu bo API works. E.g., how you would
> allocate, free, kmap, etc.
Two more comments inline. One cosmetic, one real issue. With that fixed,
this patch is Reviewed-by: Felix Kuehling
Regards,
Felix
On 2018-05-18 03:42 PM, Oded Gabbay wrote:
> In case CONFIG_HSA_AMD is not chosen, there is no need to compile amdkfd
> files that reside inside amdgpu dirver. In a
On Fri, May 18, 2018 at 2:28 AM, Rex Zhu wrote:
> error: uninitialized symbol ''
>
> Signed-off-by: Rex Zhu
Reviewed-by: Alex Deucher
> ---
> drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c | 24
> +-
> drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 3 ++-
>
On Fri, May 18, 2018 at 3:06 AM, Rex Zhu wrote:
> We should return unlock on the error path
>
> Signed-off-by: Rex Zhu
Reviewed-by: Alex Deucher
> ---
> .../gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c | 31
> +-
> 1 file changed, 19 insertions(+), 12 deletions(-)
>
>
Ping?
On Wed, May 16, 2018 at 4:51 PM, Alex Deucher wrote:
> Signed-off-by: Alex Deucher
> ---
> drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h | 4
> drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h | 4
> 2 files changed, 8 insertions(+)
>
> diff --git a/drivers/gp
In case CONFIG_HSA_AMD is not chosen, there is no need to compile amdkfd
files that reside inside amdgpu dirver. In addition, because amdkfd
depends on x86_64 architecture and amdgpu is not, compiling amdkfd files
under i386 architecture can cause compiler errors and warnings.
This patch modifies
On Fri, May 18, 2018 at 2:22 PM, Samuel Li wrote:
>
>
> On 2018-05-18 04:21 AM, Michel Dänzer wrote:
>> On 2018-05-17 06:55 PM, Alex Deucher wrote:
>>> If the pin domain is set to GTT | VRAM, look at the preferred domains
>>> for the bo and respect that if it's been set explicitly.
>>>
>>> Signed-
On Fri, May 18, 2018 at 4:21 AM, Michel Dänzer wrote:
> On 2018-05-17 06:55 PM, Alex Deucher wrote:
>> If the pin domain is set to GTT | VRAM, look at the preferred domains
>> for the bo and respect that if it's been set explicitly.
>>
>> Signed-off-by: Alex Deucher
>> ---
>> drivers/gpu/drm/amd
On Fri, May 18, 2018 at 5:33 AM, Rex Zhu wrote:
Add a commit message. Something like:
Switch to the VIDEO power profiles when VCN is active which is the
power profile optimized for video playback.
With that fixed:
Reviewed-by: Alex Deucher
Want to take a stab at doing similar patches for UVD a
On Fri, May 18, 2018 at 5:33 AM, Rex Zhu wrote:
> For power saving, default profiling mode was setted
> to power saving mode.
>
> Currently, not support CUSTOM mode and not display
> detailed profiling mode parameters in sysfs.
>
> Signed-off-by: Rex Zhu
Reviewed-by: Alex Deucher
> ---
> driv
On Thu, May 17, 2018 at 6:20 PM, Samuel Li wrote:
> Signed-off-by: Samuel Li
Please also add a separate DOC section at the top of this file giving
a brief overview of how amdgpu bo API works. E.g., how you would
allocate, free, kmap, etc. What pinning means and why you would need
to do it, etc
On Thu, May 17, 2018 at 12:09 AM, Felix Kuehling wrote:
> Hi Oded,
>
> Thanks for working on this! The Makefile changes look good.
>
> Instead of checking and calling function pointers in amdgpu_amdkfd_...
> functions at runtime, couldn't you just define empty stub functions in
> amdgpu_amdkfd.h i
On 2018-05-18 04:21 AM, Michel Dänzer wrote:
> On 2018-05-17 06:55 PM, Alex Deucher wrote:
>> If the pin domain is set to GTT | VRAM, look at the preferred domains
>> for the bo and respect that if it's been set explicitly.
>>
>> Signed-off-by: Alex Deucher
>> ---
>> drivers/gpu/drm/amd/amdgpu/
Fixes memory leak on module unload.
Analogous to mesa commit of the same name.
Signed-off-by: Jan Vesely
---
amdgpu/amdgpu_device.c | 4
1 file changed, 4 insertions(+)
diff --git a/amdgpu/amdgpu_device.c b/amdgpu/amdgpu_device.c
index e23dd3b3..34ac95b8 100644
--- a/amdgpu/amdgpu_device.c
Close the file descriptors under lock as well.
v2: close fds after removing from hash table
Signed-off-by: Jan Vesely
---
amdgpu/amdgpu_device.c | 11 +++
1 file changed, 7 insertions(+), 4 deletions(-)
diff --git a/amdgpu/amdgpu_device.c b/amdgpu/amdgpu_device.c
index 983b19ab..e23dd3b
Analogous to the mesa commit of the same name.
Signed-off-by: Jan Vesely
---
amdgpu/util_hash_table.c | 12
amdgpu/util_hash_table.h | 2 ++
2 files changed, 14 insertions(+)
diff --git a/amdgpu/util_hash_table.c b/amdgpu/util_hash_table.c
index 89a8bf9b..e06d4415 100644
--- a/amdg
On Fri, May 18, 2018 at 12:22 PM, Tom St Denis wrote:
> We're moving to Sphinx style documentation. Conver the comments
> to that format. Also document some new content.
>
Sorry, I should have been more clear. The current comment formatting
is fine. What I'd like is a new separate DOC comment
We're moving to Sphinx style documentation. Conver the comments
to that format. Also document some new content.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 254 +++-
1 file changed, 185 insertions(+), 69 deletions(-)
diff --git a/drive
From: Michel Dänzer
We would store the NULL pointer and continue, which would lead to a
crash down the road.
Bugzilla: https://bugs.freedesktop.org/106293
Signed-off-by: Michel Dänzer
---
src/radeon_dri2.c | 40 +++-
1 file changed, 19 insertions(+), 21 dele
From: Michel Dänzer
Fixes crash in that case.
Bugzilla: https://bugs.freedesktop.org/106293
Signed-off-by: Michel Dänzer
---
src/radeon_glamor.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/src/radeon_glamor.c b/src/radeon_glamor.c
index 7c09abba5..10d513ea9 100644
--- a/src/radeon_g
Den 18. mai 2018 13:18, skrev Håkon Alstadheim:
> TL;DR: How to get some canonical work-around for quirks into official
> module and/or kernel repo ? Running with gpu passed through to virtual
> machine.
>
> -
>
I went ahead and switched to amdgpu driver, and it seems the NULL
dereference of
TL;DR: How to get some canonical work-around for quirks into official
module and/or kernel repo ? Running with gpu passed through to virtual
machine.
-
I'm running this card:
00:06.0 VGA compatible controller: Advanced Micro Devices, Inc.
[AMD/ATI] Curacao PRO [Radeon R7 370 / R9 270/370 OEM
Ah! So we have one RCU grace period caused by the scheduler fence and
another one from the hardware fence.
Thanks for explaining that once more, that was really not obvious from
reading the code.
But this means we could we also fix it by moving the
"dma_fence_put(fence->parent);" from drm_sc
On 2018-05-17 11:43 PM, Leo Li wrote:
> On 2018-05-16 01:09 PM, Michel Dänzer wrote:
>> On 2018-05-03 08:31 PM, sunpeng...@amd.com wrote:
>>> From: "Leo (Sunpeng) Li"
>>>
>>> The dpms_mode flag on the driver-private CRTC was not being set when
>>> it's DPMS state is set to off. This causes some pr
From: Michel Dänzer
Simplifies the latter slightly.
Signed-off-by: Michel Dänzer
---
src/drmmode_display.c | 9 +
1 file changed, 1 insertion(+), 8 deletions(-)
diff --git a/src/drmmode_display.c b/src/drmmode_display.c
index 5d4627810..8a1a201a8 100644
--- a/src/drmmode_display.c
+++
From: Michel Dänzer
Leo pointed out that drmmode_do_crtc_dpms wasn't getting called when
turning off an output with
xrandr --output --off
This meant that the vblank sequence number and timestamp wouldn't be
saved before turning off the CRTC in this case.
Reported-by: Leo (Sunpeng) Li
Signed
On 05/18/2018 05:52 AM, Christian König wrote:
Am 17.05.2018 um 17:34 schrieb Alex Deucher:
On Tue, May 15, 2018 at 10:02 AM, Tom St Denis
wrote:
NFC just comments.
(v2): Updated based on feedback from Alex Deucher.
Signed-off-by: Tom St Denis
Reviewed-by: Alex Deucher
Just one comme
Am 17.05.2018 um 20:25 schrieb Alex Deucher:
On Thu, May 17, 2018 at 2:24 PM, Leo Liu wrote:
On 05/17/2018 02:18 PM, Alex Deucher wrote:
On Thu, May 17, 2018 at 2:12 PM, Leo Liu wrote:
NO_OP register should be writen to 0
Signed-off-by: Leo Liu
We discussed this when we first started wor
Am 17.05.2018 um 20:03 schrieb Alex Deucher:
Take the encode rings into account in the idle work handler.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
b/drivers/gpu/dr
Am 17.05.2018 um 19:38 schrieb Alex Deucher:
This reverts commit 4f7b8507bb4ba19f994e0d72eedd6029961be402.
We don't need separate idle work handles for UVD 7.2. Both instances are
driven by the same clock and power.
Signed-off-by: Alex Deucher
Reviewed-by: Christian König
---
drivers/g
Hi Christian,
When we free an IB fence, we first call one call_rcu in
drm_sched_fence_release_scheduled as the call trace one, then after the call
trace one,
we call the call_rcu second in the amdgpu_fence_release in call trace two, as
below.
The kmem_cache_free(amdgpu_fence_slab, fe
Am 17.05.2018 um 18:55 schrieb Alex Deucher:
If the pin domain is set to GTT | VRAM, look at the preferred domains
for the bo and respect that if it's been set explicitly.
Signed-off-by: Alex Deucher
Reviewed-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 9 +++-
2, 3, 4, 5 are
Reviewed-by: Junwei Zhang
Patch 1:
could you show the reserving VM?
Patch 6:
I could read that code, but not sure the purpose.
Jerry
On 05/17/2018 05:49 PM, Christian König wrote:
Only the moved state needs a separate spin lock protection. All other
states are protected by res
Am 17.05.2018 um 20:41 schrieb Harry Wentland:
On 2018-05-17 11:50 AM, Andrey Grodzovsky wrote:
First of all it's already being called from the display code from
amd_ip_funcs.suspend/resume hooks.
Second of all, the place in amdgpu_device_gpu_recover it's being called is
wrong for GPU stalls s
Am 17.05.2018 um 17:34 schrieb Alex Deucher:
On Tue, May 15, 2018 at 10:02 AM, Tom St Denis wrote:
NFC just comments.
(v2): Updated based on feedback from Alex Deucher.
Signed-off-by: Tom St Denis
Reviewed-by: Alex Deucher
Just one comment "Pin pages of memory pointed to..." better writ
Ok, I'm lost where do we use call_rcu() twice? Cause that sounds
incorrect in the first place.
Christian.
Am 18.05.2018 um 11:41 schrieb Deng, Emily:
Ping..
Best Wishes,
Emily Deng
-Original Message-
From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
Of De
Ping..
Best Wishes,
Emily Deng
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Deng, Emily
> Sent: Friday, May 18, 2018 11:20 AM
> To: Koenig, Christian ; amd-
> g...@lists.freedesktop.org
> Subject: RE: [PATCH] drm/amdgpu: add rcu
For power saving, default profiling mode was setted
to power saving mode.
Currently, not support CUSTOM mode and not display
detailed profiling mode parameters in sysfs.
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 42 +++
1 file changed, 42
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 12
1 file changed, 8 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 3549481..94b221f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amd
On 2018-05-17 06:55 PM, Alex Deucher wrote:
> If the pin domain is set to GTT | VRAM, look at the preferred domains
> for the bo and respect that if it's been set explicitly.
>
> Signed-off-by: Alex Deucher
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 9 +++--
> 1 file changed, 7 ins
On 2018-05-17 11:43 PM, Leo Li wrote:
> On 2018-05-16 01:06 PM, Michel Dänzer wrote:
>> On 2018-05-03 08:31 PM, sunpeng...@amd.com wrote:
>>>
>>> 3. The three color management properties (Degamma LUT, Color
>>> Transform Matrix
>>> (CTM), and Gamma LUT) are hard-coded into the DDX driver, to be
On 2018-05-17 11:44 PM, Leo Li wrote:
> On 2018-05-16 01:10 PM, Michel Dänzer wrote:
>> On 2018-05-03 08:31 PM, sunpeng...@amd.com wrote:
>>> From: "Leo (Sunpeng) Li"
>>>
>>> This will persist color management properties on a CRTC across DPMS
>>> state changes.
>>>
>>> Signed-off-by: Leo (Sunpeng)
On 2018-05-17 11:43 PM, Leo Li wrote:
> On 2018-05-16 01:08 PM, Michel Dänzer wrote:
>> On 2018-05-03 08:31 PM, sunpeng...@amd.com wrote:
>>>
>>> @@ -1458,6 +1586,14 @@ drmmode_crtc_init(ScrnInfoPtr pScrn,
>>> drmmode_ptr drmmode, drmModeResPtr mode_res
>>> drmmode_crtc->ctm->matrix[0] = drmm
We should return unlock on the error path
Signed-off-by: Rex Zhu
---
.../gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c | 31 +-
1 file changed, 19 insertions(+), 12 deletions(-)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
b/drivers/gpu/drm/amd/powerpl
48 matches
Mail list logo
