We have added new sysfs pp_power_profile_mode.
The old sysfs (pp_compute_power_profile and pp_gfx_power_profile)
were used internally currently by ROCM tool and KFD to switch mode between gfx
and compute only on VI/CI asics.
We can't extend more modes for other workloads, for example: video,
Thanks for fix.
Regards,
David Zhou
On 2018年02月19日 15:50, Christian König wrote:
Patch is Acked-by: Christian König
Regards,
Christian.
Am 19.02.2018 um 08:25 schrieb Andrey Grodzovsky:
Found time to do some debugging and found error in assigning of
initial
fix:
should do right shift on wb before clearing
cleanups:
1,should memset all wb buffer
2,set max wb number to 128 (total 4KB) is big enough
Change-Id: I43832245c875ce039e7709dc049828e21c50c81f
Signed-off-by: Monk Liu
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h| 2 +-
*** those patches are mainly bug fixing for SR-iov and couple bare-metal case
***
Emily Deng (2):
drm/amdgpu: Remove the memory leak after unload amdgpu driver
drm/amdgpu: Correct sdma_v4 get_wptr
Monk Liu (20):
drm/amdgpu: fix for wb_clear
drm/amdgpu: remove duplicated
if a job is go through scheduler, it is forbidden to
call job_free_resources after ib_schedule, becaust that
would assign wild pointer of fence on the sa_bo->fence
which could lead to weird bug
Change-Id: Iad7ee011c641cb7357569cbce36fdc10f0ed911d
Signed-off-by: Monk Liu
---
Change-Id: I2865a11d1dded214de289787d334ec4a22b5db19
Signed-off-by: Monk Liu
---
drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index
since hotplug_work is initialized under the case of
no dc support
Change-Id: I0d417a5b9f8dfb1863eafc95b6802be9e5c74720
Signed-off-by: Monk Liu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git
issue:
under SR-IOV sometimes the iB test will fail on
gfx ring
fix:
with cond_exec inserted in RB the gfx engine would
skip part packets if RLCV issue PREEMPT on gfx engine
if gfx engine is prior to COND_EXEC packet, this is
okay for regular command from UMD, but for the ib test
since the whole
two reason for this patch:
1) SRIOV doesn't give VF cg/pg feature so the idle_work
is useless
2) VCE's idle work would cause "KMD reload" test failed and
trigger KERNEL PANIC, this is because the idle work is triggered
after VCE ib test and KMD removed right after loaded, which
can hit page fault
1)amdgpu_vce_get_create_msg is only used in ib test so
no reason no to use a static routine for it, and add a
timeout parameter for it.
2)fence handling of MM's ib test part is a little messy
clean it make it easier to read
Change-Id: Ic9bfd9971457600266096e114210e84ce9b4347d
Signed-off-by: Monk
Change-Id: I370966acd0f1925a99dfde888678e6e0fd093b15
Signed-off-by: Monk Liu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
issue:
on bare-metal when doing kmd reload test, there are chance
that kernel hit fatal error afer driver unloaded/reloaded
fix:
the cause is that those "idle work" not really stopped and
if kmd was is unloaded too quick that were chance that
"idle work" run after driver structures already
From: Emily Deng
Need to call function amdgpu_ucode_fini_bo to release ucode bo for
psp firmware load type.
Change-Id: I1c7be8135993e11076c9d46b3cd87615514a9abb
Signed-off-by: Emily Deng
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 +
1 file
and later send req_fini and rel_fini to host for the
finish routine
Change-Id: Ib0347a305ab5f7712d2d76b1a843bb2429acbf3d
Signed-off-by: Monk Liu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 19 ++-
1 file changed, 10 insertions(+), 9 deletions(-)
diff
Change-Id: I35a343b21a007716fc7811781650264339c94273
Signed-off-by: Monk Liu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
two reasons to switch SCRATCH reg method to WB method:
1)Because when doing IB test we don't want to involve KIQ health
status affect, and since SCRATCH register access is go through
KIQ that way GFX IB test would failed due to KIQ fail.
2)acccessing SCRATCH register cost much more time than WB
otherwise there will be DMAR reading error comes out from CP since
GFX is still alive and CPC's WPTR_POLL is still enabled, which would
lead to DMAR read error.
fix:
we can hault CPG after hw_fini, but cannot halt CPC becaues KIQ
stil need to be alive to let RLCV invoke, but its WPTR_POLL could
issue:
sometime GFX/MM ib test hit timeout under SRIOV env, root cause
is that engine doesn't come back soon enough so the current
IB test considered as timed out.
fix:
for SRIOV GFX IB test wait time need to be expanded a lot during
SRIOV runtimei mode since it couldn't really begin before GFX
From: Emily Deng
the original method will change the wptr value in wb.
Change-Id: I984fabca35d9dcf1f5fa8ef7779b2afb7f7d7370
Signed-off-by: Emily Deng
---
drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 10 +-
1 file changed, 5 insertions(+), 5
256MB is too small consider PTE/PDE shadow and TTM
eviction activity
Change-Id: Ifaa30dc730eec36af47fbdeb3cce30de9067b17f
Signed-off-by: Monk Liu
---
drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
issue:
kernel oops or vmc page fault occured during vk_example/vk_cts
test.
root cause:
previously reservation object would kfree the staged when slot
checked available during reserve_shared(), which is incorrect
becasue this way reservation_object->fence will be a wild pointer
referenced by
should call reservation_object_reserve_shared before
amdgpu_bo_fence(), otherwise there are odds kernel
hit BUG in reversation.c
bug:
[12622.076435] [ cut here ]
[12622.076438] kernel BUG at drivers/dma-buf/reservation.c:233!
[12622.078046] invalid opcode: [#1] SMP
To avoid conflicts and confusion, I think better remove two old sysfs.
Well that is not a matter of personal opinion. The rule is simply that
sysfs needs to be stable, otherwise we can't upstream a sysfs interface.
Especially since you already noted that current ROCM tool and KFD depend
on
Adjust dpm optimization settings on smu7 based on
the activity level read from SMU.
in manual dpm mode, auto wattman will be disabled.
Change-Id: I68396f04be76793ece3b07f0ac1b646981ca1e29
Reviewed-by: Alex Deucher
Signed-off-by: Rex Zhu
---
Hi!
> > diff --git a/drivers/soc/imx/gpc.c b/drivers/soc/imx/gpc.c
> > index 53f7275..cfb42f5 100644
> > --- a/drivers/soc/imx/gpc.c
> > +++ b/drivers/soc/imx/gpc.c
> > @@ -348,7 +348,7 @@ static int imx_gpc_old_dt_init(struct device *dev,
> > struct regmap *regmap,
> > if (i == 1) {
Auto wattman feature was used for adjust dpm
optimization settings based on workloads
Change-Id: I6e747beb7786c45f733c3c0f9ef9b48c98f0484c
Reviewed-by: Alex Deucher
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/inc/hwmgr.h | 2 ++
1 file
Change-Id: I4c3e70a5bff3855e0195d77eb12932dc08063ef2
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h| 8 -
drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 180 ---
drivers/gpu/drm/amd/amdgpu/ci_dpm.c| 256
Add new power profile sysfs: pp_power_profile_mode
old sysfs: pp_compute_power_profile/pp_gfx_power_profile
will be removed
Change-Id: Ief35efb81b313115b33c430dfb328ad83ce7975d
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 1 +
1 file changed, 1
based on the engine activity read from smu, driver will
notify smu to adjust the behaviour of natural DPM.
Change-Id: I951390cae82eeacfdaea3917eb570aa57216cbd7
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 65 +-
driver need to monitor the gfx activity and notify
smu to change the dpm policy
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c | 17 +
drivers/gpu/drm/amd/powerplay/inc/hwmgr.h | 3 +++
2 files changed, 20 insertions(+)
diff --git
Add a common hw interface to ctrl autowattman feature and
restore power profile mode.
when autowattman enabled, start the wattman queue when
power state is set. disable this feature when suspend/
enter manual mode.
when autowattman not support, just restore default power
profile mode in auto dpm
Do not need to restore default power profile mode
when force dpm level.
Change-Id: Ib9cabf5efe2bbacd5987598acb79221b4ad4d567
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 5 -
1 file changed, 5 deletions(-)
diff --git
Change-Id: Ifaa7058a4682bafeeef68e78bfcff26012ba71c1
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
AutoWattman changes clocks quite aggressively which can make
framerates jump significantly and lead to a subpar experience
when using Freesync.
Disable AutoWattman (if enabled) when FreeSync is enabled.
Change-Id: I1c8e6b6d0d77b8bb3aa5250d4f5aec81270c27f1
Signed-off-by: Rex Zhu
Change-Id: I7bd4dc5b255ed3edcb28f6c770e7b0199d0662d4
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index
On Thu, Feb 8, 2018 at 11:33 PM, SF Markus Elfring
wrote:
> From: Markus Elfring
> Date: Thu, 8 Feb 2018 22:23:57 +0100
>
> Omit an extra message for a memory allocation failure in this function.
>
> This issue was detected by using
Am 24.02.2018 um 14:55 schrieb Rex Zhu:
Add new power profile sysfs: pp_power_profile_mode
old sysfs: pp_compute_power_profile/pp_gfx_power_profile
will be removed
Well sysfs is supposed to be stable, so removing the old interface won't
be allowed.
Change-Id:
37 matches
Mail list logo