> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Hawking Zhang
> Sent: Thursday, December 01, 2016 4:16 AM
> To: amd-gfx@lists.freedesktop.org
> Cc: Zhang, Hawking
> Subject: [PATCH] drm/amd/powerplay: bypass fan table setup if no fan
>
Hi Rex,
Given how often we diagnose problems by disabling PG/CG I'd have to offer my
NAK to this patch since it prevents me from disabling PG via the command line.
Any reason you want to make it permanently enabled?
Tom
From: amd-gfx
On Thu, Dec 1, 2016 at 3:24 AM, Rex Zhu wrote:
> 1. bit CP_PG_DISABLE was reversed.
Please split this out as a separate patch. Also, with this fixed, we
should be able to enable the AMD_PG_SUPPORT_CP flag for CZ and ST.
> 2. load RLC_SRM_INDEX_CNTL_ADDR/DATA_x pairs
>with
From: Alex Xie
Fixes a rare NULL pointer dereference in amdgpu_ttm_bind.
The issue was found by Nicolai Haehnle.
The patch was tested by Nicolai Haehnle.
Signed-off-by: Alex Xie
Reviewed-by: Chunming Zhou
Reviewed-by: Christian
On Wed, Nov 30, 2016 at 2:01 AM, Xiangliang Yu wrote:
> Vi_structs.h has defined vi_mqd, drop redundant vi_mqd define.
>
> Signed-off-by: Monk Linu
> Signed-off-by: Xiangliang Yu
Patches 1, 2, 5:
Reviewed-by: Alex Deucher
On Thu, Dec 1, 2016 at 1:53 AM, Yu, Xiangliang wrote:
>>
>> > -Original Message-
>> > From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
>> > Of Xiangliang Yu
>> > Sent: Wednesday, November 30, 2016 2:02 AM
>> > To: amd-gfx@lists.freedesktop.org
Hi David,
Some one major review suggestion, don't use file descriptors for
semaphore, as fd's are a limited resource and we don't want to use
them all up.
You create semaphore objects and use them in a single process without
them being fds, then when userspace wants to share with another
process
inside amdgpu and display:
find -name Makefile -o -name Kconfig -o -name "*.c" -o -name "*.h" -o -name
"*.cpp" -o -name "*.hpp" | xargs sed -i 's/DRM_AMD_DAL/DRM_AMD_DC/g'
find -name Makefile -o -name Kconfig -o -name "*.c" -o -name "*.h" -o -name
"*.cpp" -o -name "*.hpp" | xargs sed -i
On 1 December 2016 at 06:11, zhoucm1 wrote:
> Hi Dave,
>
> As the attached, our Vulkan team is verifying it.
>
David, please read through the following documents when designing
ioctls [1] and [im]porting the UABI to libdrm [2].
Thanks
Emil
[1]
On 02/12/16 01:21 AM, Alex Deucher wrote:
> From: Alex Xie
>
> Fixes a rare NULL pointer dereference in amdgpu_ttm_bind.
>
> The issue was found by Nicolai Haehnle.
> The patch was tested by Nicolai Haehnle.
Just as a suggestion for next time, this can be simplified to
1. bit CP_PG_DISABLE was reversed.
2. load RLC_SRM_INDEX_CNTL_ADDR/DATA_x pairs
with valid addr/data.
3. always init gfx pg.
4. delete repeated check for pg mask.
Change-Id: I9fcc8d1f79f5fa1803cb2625aa292188a656ae6b
Signed-off-by: Rex Zhu
---
If vBIOS noFan bit is set, the fan table parameters in thermal controller
will not get initialized. The driver should avoid to use these uninitialized
parameter to do calculation. Otherwise, it may trigger divide 0 error.
Change-Id: I985119837ec818e2101835f79aa4fe6e0e247797
Signed-off-by: Hawking
Hi Dave,
the problem with that approach is that it duplicates the effort done
with android fences but moves some of the logic found in there into the
kernel.
E.g. you could do the same with in/out fences on command submission,
just not share the handle itself with other process before it is
If vBIOS noFan bit is set, the fan table parameters in thermal controller
will not get initialized. The driver should avoid to use these uninitialized
parameter to do calculation. Otherwise, it may trigger divide 0 error.
Change-Id: I76680a5ec5411f59742b65bb70eb7b4a08bda3ef
Signed-off-by: Hawking
SMU buffer is used for power feature, but for virtualization, the
power is controlled by hypervisor. Ignore it.
Signed-off-by: Xiangliang Yu
Signed-off-by: Monk Liu
---
drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c | 13 +++--
1 file
For SR-IOV client, driver shouldn't touch the GFX hw during HW
fini, otherwise, gfx will fail to start after rebooting guest os.
Signed-off-by: shaoyunl
Signed-off-by: Xiangliang Yu
---
drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 4
1 file changed,
For virtualization, FW size need to cut its digest part.
Signed-off-by: Frank Min
Signed-off-by: Monk Liu
Signed-off-by: Xiangliang Yu
---
drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c | 4
1 file changed, 4 insertions(+)
On Fri, Dec 2, 2016 at 1:03 AM, Xiangliang Yu wrote:
> As SR-IOV initialization is difference from bare metal, the series patch
> will change code path of initialization according to HW design.
>
> Changes in V2:
> 1. Remove VM fault patch;
> 2. Split out the third patch
If doesn't enable dpm, the powerplay will not allocate memory for
hw management. So, hw_init_power_state_table function will reference
NULL pointer when resetting.
Signed-off-by: Xiangliang Yu
---
drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 4 ++--
1 file changed, 2
> -Original Message-
> From: Alex Deucher [mailto:alexdeuc...@gmail.com]
> Sent: Friday, December 02, 2016 2:16 PM
> To: Yu, Xiangliang
> Cc: amd-gfx list
> Subject: Re: [V2 0/6] drm/amdgpu: add support SR-IOV initialization
>
> On
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Xiangliang Yu
> Sent: Friday, December 02, 2016 1:04 AM
> To: amd-gfx@lists.freedesktop.org
> Cc: Yu, Xiangliang
> Subject: [V2 6/6] drm/amd/powerplay: Fix potential NULL pointer issue
>
>
New variant.
Signed-off-by: Alex Deucher
Cc: sta...@vger.kernel.org
---
drivers/gpu/drm/radeon/si_dpm.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index c4993452..8b5e697 100644
---
22 matches
Mail list logo