Increase the waiting time to 1s from 200ms. This is expected
to fix some mode1 reset failures.
Change-Id: I4c11996628cd40d559c6edea97d092f397122d78
Signed-off-by: Evan Quan
---
drivers/gpu/drm/amd/include/amd_shared.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
This memory allocation flag will be used to indicate BOs containing
sensitive data that should not be leaked to other processes.
Signed-off-by: Felix Kuehling
---
include/uapi/drm/amdgpu_drm.h | 4
1 file changed, 4 insertions(+)
diff --git a/include/uapi/drm/amdgpu_drm.h
Clear VRAM memory containing sensitive data before freeing it. Doing
this in the VRAM manager's put_node callback covers all cases that
release memory, including freeing, moving or evicting BOs. To
minimize the performance impact, use the mman->move fence to delay
future memory allocations rather
Memory used by KFD applications can contain sensitive information that
should not be leaked to other processes. The current approach to prevent
leaks is to clear VRAM at allocation time. This is not effective because
memory can be reused in other ways without being cleared. Synchronously
clearing
Reviewed-by: Evan Quan
> -Original Message-
> From: amd-gfx On Behalf Of Alex
> Deucher
> Sent: Thursday, July 04, 2019 10:58 AM
> To: amd-gfx@lists.freedesktop.org
> Cc: Deucher, Alexander
> Subject: [PATCH 1/2] drm/amdgpu/navi10: add uclk activity sensor
>
> Query the metrics table
I'm seeing some weird intermittent bugs (vm faults, hangs, etc) when
trying to use amdgpu_job_submit_direct. I'm wondering if there is a
possibility of a race condition, when a submit_direct and a GPU
scheduler thread try to submit to the same ring at the same time. I
didn't see any locking to
On Mon, Jul 8, 2019 at 7:50 AM Paul Menzel
wrote:
>
> From: Paul Menzel
> Date: Sun, 7 Jul 2019 14:23:08 +0200
>
> As the clock is already logged, also log the voltage.
>
> Signed-off-by: Paul Menzel
Applied. Thanks.
Alex
> ---
> drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c | 2
Ping?
On Wed, Jul 3, 2019 at 10:58 PM Alex Deucher wrote:
>
> Query the metrics table for the current uclk activity.
>
> Signed-off-by: Alex Deucher
> ---
> drivers/gpu/drm/amd/powerplay/navi10_ppt.c | 16 ++--
> 1 file changed, 14 insertions(+), 2 deletions(-)
>
> diff --git
Reviewed-by: Hawking Zhang
Regards,
Hawking
-Original Message-
From: amd-gfx On Behalf Of Alex Deucher
Sent: 2019年7月9日 2:36
To: amd-gfx@lists.freedesktop.org
Cc: Deucher, Alexander
Subject: [PATCH] drm/amdgpu/psp: add a mutex to protect access to the psp ring
We need to serialize
Series is Reviewed-by: Xiaojie Yuan
BR,
Xiaojie
> On Jul 9, 2019, at 2:48 AM, Alex Deucher wrote:
>
> Use the IP discovery table rather than hardcoding the
> settings in the driver.
>
> Reviewed-by: Xiaojie Yuan
> Signed-off-by: Alex Deucher
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
On Thu, Jul 4, 2019 at 7:52 AM Nathan Chancellor
wrote:
>
> clang warns:
>
> In file included from drivers/gpu/drm/amd/amdgpu/nv.c:53:
> drivers/gpu/drm/amd/amdgpu/../amdgpu/mes_v10_1.h:24:9: warning:
> '__MES_V10_1_H__' is used as a header guard here, followed by #define of
> a different
Use the IP discovery table rather than hardcoding the
settings in the driver.
Reviewed-by: Xiaojie Yuan
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
New parameters added for navi lack documentation.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 13 +
1 file changed, 13 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index
It's only available on navi and newer.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index
We need to serialize access to the psp ring if there are multiple
callers at runtime.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 +
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c| 16 ++--
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h| 1 +
3 files
Acked-by: Slava Abramov
Tested-by: Slava Abramov
From: amd-gfx on behalf of Arnd
Bergmann
Sent: Monday, July 8, 2019 9:52:08 AM
To: Wentland, Harry; Li, Sun peng (Leo); Deucher, Alexander; Koenig, Christian;
Zhou, David(ChunMing); David Airlie; Daniel
Acked-by: Alex Deucher
From: amd-gfx on behalf of Marek Olšák
Sent: Monday, July 8, 2019 1:31 PM
To: amd-gfx mailing list
Subject: Re: [PATCH] drm/amdgpu: don't invalidate caches in RELEASE_MEM, only
do the writeback
ping
On Tue, Jul 2, 2019 at 2:29 PM Marek
ping
On Tue, Jul 2, 2019 at 2:29 PM Marek Olšák wrote:
> From: Marek Olšák
>
> This RELEASE_MEM use has the Release semantic, which means we should write
> back but not invalidate. Invalidations only make sense with the Acquire
> semantic (ACQUIRE_MEM), or when RELEASE_MEM is used to do the
On Mon, Jul 8, 2019 at 4:46 PM Alex Deucher wrote:
>
> On Mon, Jul 8, 2019 at 10:42 AM Arnd Bergmann wrote:
> >
> > Enabling amdgpu but not CONFIG_DRM_AMD_DC leads to a warning:
> >
> > drivers/gpu/drm/amd/amdgpu/nv.c: In function 'nv_set_ip_blocks':
> > drivers/gpu/drm/amd/amdgpu/nv.c:400:3:
Looks like Arnd Bergmann already fixed it with a different patch. That's
fine with me too. Feel free to ignore this.
On 2019-07-08 12:05 p.m., Kuehling, Felix wrote:
> Only define struct amdgpu_mn if HMM_MIRROR is enabled. Add back a
> forward declaration for the case that it isn't.
>
> Fixes:
Only define struct amdgpu_mn if HMM_MIRROR is enabled. Add back a
forward declaration for the case that it isn't.
Fixes: 6f1c3e61d2d8 ("drm/amdgpu: Prepare for hmm_range_register API change")
Signed-off-by: Felix Kuehling
---
drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h | 6 +-
1 file changed, 5
On Mon, Jul 8, 2019 at 4:54 PM Nathan Chancellor
wrote:
>
> Hi Arnd,
>
> On Mon, Jul 08, 2019 at 03:57:06PM +0200, Arnd Bergmann wrote:
> > A couple of calls to smu_get_current_clk_freq() and smu_force_clk_levels()
> > pass constants of the wrong type, leading to warnings with clang-8:
> >
> >
On Mon, Jul 8, 2019 at 10:08 AM Arnd Bergmann wrote:
>
> If smu_get_current_rpm() fails, we can't use the output,
> as that may be uninitialized:
>
> drivers/gpu/drm/amd/amdgpu/../powerplay/vega20_ppt.c:3023:8: error: variable
> 'current_rpm' is used uninitialized whenever '?:' condition is
On Mon, Jul 8, 2019 at 9:53 AM Arnd Bergmann wrote:
>
> Without this header, we get a compiler error in some configurations:
>
> .../dc/dcn20/dcn20_hwseq.c: In function 'dcn20_hwss_wait_for_blank_complete':
> .../dc/dcn20/dcn20_hwseq.c:1493:3: error: implicit declaration of function
> 'udelay'
On 2019-07-05 4:41 p.m., Alex Deucher wrote:
> Need to add appropriate ifdef.
>
Acked-by: Leo Li
> Signed-off-by: Alex Deucher
> ---
> drivers/gpu/drm/amd/amdgpu/nv.c | 4
> 1 file changed, 4 insertions(+)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c
On 7/8/19 11:47 AM, Alex Deucher wrote:
> On Mon, Jul 8, 2019 at 8:56 AM Kazlauskas, Nicholas
> wrote:
>>
>> On 7/8/19 5:15 AM, Michel Dänzer wrote:
>>> On 2019-07-05 10:41 p.m., Alex Deucher wrote:
Need to add appropriate ifdef.
Signed-off-by: Alex Deucher
---
On Mon, Jul 8, 2019 at 9:51 AM Arnd Bergmann wrote:
>
> 'struct hmm_mirror' is not defined without the Kconfig option set,
> so we cannot include it within another struct:
>
> In file included from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:72:
>
On Mon, Jul 8, 2019 at 10:42 AM Arnd Bergmann wrote:
>
> When CONFIG_PERF_EVENTS is disabled, we cannot compile the pmu
> portion of the amdgpu driver:
>
> drivers/gpu/drm/amd/amdgpu/amdgpu_pmu.c:48:38: error: no member named 'hw' in
> 'struct perf_event'
> struct hw_perf_event *hwc =
On Mon, Jul 8, 2019 at 5:02 PM Nathan Chancellor
wrote:
> On Mon, Jul 08, 2019 at 04:07:58PM +0200, Arnd Bergmann wrote:
> > /* if don't has GetDpmClockFreq Message, try get current clock by
> > SmuMetrics_t */
> > - if (smu_msg_get_index(smu, SMU_MSG_GetDpmClockFreq) == 0)
> > +
Applied the series. thanks!
Alex
On Thu, Jul 4, 2019 at 3:26 AM Nathan Chancellor
wrote:
>
> Hi all,
>
> I don't do threaded patches very often so if I have messed something up,
> please forgive me :)
>
> This series fixes all of the clang warnings that I saw added in
> next-20190703. The full
and we might need a follow-up patch to check if (amdgpu_discovery && asic_type
>= CHIP_NAVI10) in amdgpu_device.c.
BR,
Xiaojie
From: amd-gfx on behalf of Yuan,
Xiaojie
Sent: Monday, July 8, 2019 11:45 PM
To: Deucher, Alexander; Alex Deucher;
On Mon, Jul 8, 2019 at 8:56 AM Kazlauskas, Nicholas
wrote:
>
> On 7/8/19 5:15 AM, Michel Dänzer wrote:
> > On 2019-07-05 10:41 p.m., Alex Deucher wrote:
> >> Need to add appropriate ifdef.
> >>
> >> Signed-off-by: Alex Deucher
> >> ---
> >> drivers/gpu/drm/amd/amdgpu/nv.c | 4
> >> 1
Got it. Thanks for the explanation.
Patch is Reviewed-by: Xiaojie Yuan
BR,
Xiaojie
From: Deucher, Alexander
Sent: Monday, July 8, 2019 11:38 PM
To: Yuan, Xiaojie; Alex Deucher; amd-gfx@lists.freedesktop.org
Subject: Re: [PATCH 2/2] drm/amdgpu: enable IP
auto just means do the right thing for each asic. For example, if there is a
new asic program and the discovery tables are not available yet, for that
program, auto means don't use the discovery table. For other programs where
the table is available (e.g., navi10 and 14), then auto means use
Hi Alex,
I see there are many 'auto' module params for amdgpu, could you please explain
about the historical intention of using value '-1'?
and shall we check 'amdgpu_discovery' against '-1' for all the code paths in
amdgpu_device.c and navi10_reg_init.c?
BR,
Xiaojie
On Mon, Jul 08, 2019 at 04:07:58PM +0200, Arnd Bergmann wrote:
> A mistake in the error handling caused an uninitialized
> variable to be used:
>
> drivers/gpu/drm/amd/amdgpu/../powerplay/smu_v11_0.c:1102:10: error: variable
> 'freq' is used uninitialized whenever '?:' condition is false
>
Hi Arnd,
On Mon, Jul 08, 2019 at 03:57:06PM +0200, Arnd Bergmann wrote:
> A couple of calls to smu_get_current_clk_freq() and smu_force_clk_levels()
> pass constants of the wrong type, leading to warnings with clang-8:
>
> drivers/gpu/drm/amd/amdgpu/../powerplay/vega20_ppt.c:995:39: error:
Ping on this series.
Alex
On Wed, Jul 3, 2019 at 11:13 AM Alex Deucher wrote:
>
> Use the IP discovery table rather than hardcoding the
> settings in the driver.
>
> Signed-off-by: Alex Deucher
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 ++-
> 1 file changed, 2 insertions(+), 1
Ping on this series.
Alex
On Wed, Jul 3, 2019 at 10:58 PM Alex Deucher wrote:
>
> This was dropped when the code was refactored. Re-add it
> for navi10.
>
> Signed-off-by: Alex Deucher
> ---
> drivers/gpu/drm/amd/powerplay/navi10_ppt.c | 42 ++
> 1 file changed, 42
On Mon, Jul 8, 2019 at 10:42 AM Arnd Bergmann wrote:
>
> Enabling amdgpu but not CONFIG_DRM_AMD_DC leads to a warning:
>
> drivers/gpu/drm/amd/amdgpu/nv.c: In function 'nv_set_ip_blocks':
> drivers/gpu/drm/amd/amdgpu/nv.c:400:3: error: #warning "Enable
> CONFIG_DRM_AMD_DC for display support on
When CONFIG_PERF_EVENTS is disabled, we cannot compile the pmu
portion of the amdgpu driver:
drivers/gpu/drm/amd/amdgpu/amdgpu_pmu.c:48:38: error: no member named 'hw' in
'struct perf_event'
struct hw_perf_event *hwc = >hw;
~ ^
Enabling amdgpu but not CONFIG_DRM_AMD_DC leads to a warning:
drivers/gpu/drm/amd/amdgpu/nv.c: In function 'nv_set_ip_blocks':
drivers/gpu/drm/amd/amdgpu/nv.c:400:3: error: #warning "Enable
CONFIG_DRM_AMD_DC for display support on navi." [-Werror=cpp]
# warning "Enable CONFIG_DRM_AMD_DC for
On 7/8/19 9:52 AM, Arnd Bergmann wrote:
> On 32-bit architectures, dividing a 64-bit integer in the kernel
> leads to a link error:
>
> ERROR: "__udivdi3" [drivers/gpu/drm/amd/amdgpu/amdgpu.ko] undefined!
> ERROR: "__divdi3" [drivers/gpu/drm/amd/amdgpu/amdgpu.ko] undefined!
>
> Change the two
The navi10_ppt code contains two instances of an incorrect struct initializer:
drivers/gpu/drm/amd/amdgpu/../powerplay/navi10_ppt.c:601:33: error: suggest
braces around initialization of subobject
[-Werror,-Wmissing-braces]
static SmuMetrics_t metrics = {0};
A mistake in the error handling caused an uninitialized
variable to be used:
drivers/gpu/drm/amd/amdgpu/../powerplay/smu_v11_0.c:1102:10: error: variable
'freq' is used uninitialized whenever '?:' condition is false
[-Werror,-Wsometimes-uninitialized]
ret =
If smu_get_current_rpm() fails, we can't use the output,
as that may be uninitialized:
drivers/gpu/drm/amd/amdgpu/../powerplay/vega20_ppt.c:3023:8: error: variable
'current_rpm' is used uninitialized whenever '?:' condition is false
[-Werror,-Wsometimes-uninitialized]
ret =
A couple of calls to smu_get_current_clk_freq() and smu_force_clk_levels()
pass constants of the wrong type, leading to warnings with clang-8:
drivers/gpu/drm/amd/amdgpu/../powerplay/vega20_ppt.c:995:39: error: implicit
conversion from enumeration type 'PPCLK_e' to different enumeration type
Without this header, we get a compiler error in some configurations:
.../dc/dcn20/dcn20_hwseq.c: In function 'dcn20_hwss_wait_for_blank_complete':
.../dc/dcn20/dcn20_hwseq.c:1493:3: error: implicit declaration of function
'udelay' [-Werror=implicit-function-declaration]
Note: the use of udelay
On 32-bit architectures, dividing a 64-bit integer in the kernel
leads to a link error:
ERROR: "__udivdi3" [drivers/gpu/drm/amd/amdgpu/amdgpu.ko] undefined!
ERROR: "__divdi3" [drivers/gpu/drm/amd/amdgpu/amdgpu.ko] undefined!
Change the two recently introduced instances to a multiply+shift
'struct hmm_mirror' is not defined without the Kconfig option set,
so we cannot include it within another struct:
In file included from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:72:
drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu_mn.h:69:20: error: field has
incomplete type 'struct hmm_mirror'
On 7/8/19 5:15 AM, Michel Dänzer wrote:
> On 2019-07-05 10:41 p.m., Alex Deucher wrote:
>> Need to add appropriate ifdef.
>>
>> Signed-off-by: Alex Deucher
>> ---
>> drivers/gpu/drm/amd/amdgpu/nv.c | 4
>> 1 file changed, 4 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c
From: Paul Menzel
Date: Sun, 7 Jul 2019 14:23:08 +0200
As the clock is already logged, also log the voltage.
Signed-off-by: Paul Menzel
---
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
On 2019-07-05 10:41 p.m., Alex Deucher wrote:
> Need to add appropriate ifdef.
>
> Signed-off-by: Alex Deucher
> ---
> drivers/gpu/drm/amd/amdgpu/nv.c | 4
> 1 file changed, 4 insertions(+)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c
> index
We are going to implement per-ip error injection tests. This test
actually only works on UMC.
Signed-off-by: xinhui pan
---
tests/amdgpu/ras_tests.c | 62 +---
1 file changed, 27 insertions(+), 35 deletions(-)
diff --git a/tests/amdgpu/ras_tests.c
From: Alex Deucher
Sent: Saturday, July 6, 2019 5:06 AM
To: Wang, Kevin(Yang)
Cc: amd-gfx@lists.freedesktop.org; Xiao, Jack; kenenth.f...@amd.com; Zhang,
Hawking
Subject: Re: [PATCH] drm/amd/powerplay: add temperature sensor support for
navi10
On Wed, Jul 3,
55 matches
Mail list logo