On 04/12/2018 12:54 AM, Alex Deucher wrote:
On Thu, Apr 12, 2018 at 12:48 AM, Shawn Starr wrote:
When trying an early 4.17-rc0 RPM from Fedora,
I get:
[2.794243] [drm:construct [amdgpu]] *ERROR* construct: Invalid Connector
ObjectID from Adapter Service for
On Thu, Apr 12, 2018 at 12:48 AM, Shawn Starr wrote:
> When trying an early 4.17-rc0 RPM from Fedora,
>
> I get:
>
> [2.794243] [drm:construct [amdgpu]] *ERROR* construct: Invalid Connector
> ObjectID from Adapter Service for connector index:1! type 0 expected 3
>
>
When trying an early 4.17-rc0 RPM from Fedora,
I get:
[2.794243] [drm:construct [amdgpu]] *ERROR* construct: Invalid
Connector ObjectID from Adapter Service for connector index:1! type 0
expected 3
We've had issues with connector types on this laptop before and found
some workarounds,
On Thu, Apr 12, 2018 at 12:08 AM, Andrey Grodzovsky
wrote:
> Reserved VRAM is used to avoid overriding pre OS FB.
> Once our display stack takes over we don't need the reserved
> VRAM anymore.
>
> v2:
> Remove comment, we know actually why we need to reserve the stolen
On 04/12/2018 12:16 AM, Alex Deucher wrote:
On Thu, Apr 12, 2018 at 12:08 AM, Andrey Grodzovsky
wrote:
From: Alex Deucher
Steal 9 MB for vga emulation and fb if vga is enabled, otherwise,
steal enough to cover the current display size as
On Thu, Apr 12, 2018 at 12:08 AM, Andrey Grodzovsky
wrote:
> From: Alex Deucher
>
> Steal 9 MB for vga emulation and fb if vga is enabled, otherwise,
> steal enough to cover the current display size as set by the vbios.
>
> If no memory is used
Reserved VRAM is used to avoid overriding pre OS FB.
Once our display stack takes over we don't need the reserved
VRAM anymore.
v2:
Remove comment, we know actually why we need to reserve the stolen VRAM.
Fix return type for amdgpu_ttm_late_init.
v3:
Return 0 in amdgpu_bo_late_init, rebase on
From: Alex Deucher
Steal 9 MB for vga emulation and fb if vga is enabled, otherwise,
steal enough to cover the current display size as set by the vbios.
If no memory is used (e.g., secondary or headless card), skip
stolen memory reserve.
Signed-off-by: Alex Deucher
Reviewed-by: Rex Zhu
Best Regards
Rex
From: amd-gfx on behalf of Alex Deucher
Sent: Thursday, April 12, 2018 7:11 AM
To: amd-gfx@lists.freedesktop.org
Cc: Deucher, Alexander
Reviewed-by: Rex Zhu
Best Regards
Rex
From: amd-gfx on behalf of Alex Deucher
Sent: Thursday, April 12, 2018 6:59 AM
To: amd-gfx@lists.freedesktop.org
Cc: Deucher, Alexander
This is useful for drivers (which will probably be all of them soon)
which need to track state that is exclusive to the topology, and not a
specific connector on said topology. This includes things such as the
link rate and lane count that are shared by all of the connectors on the
topology.
Next version of https://patchwork.freedesktop.org/series/41576/
Only changes are removing duplicate SoBs that git send-email annoyingly
added. Sorry about that :(
Lyude Paul (10):
drm/atomic: Print debug message on atomic check failure
drm/i915: Move DP modeset retry work into intel_dp
> From: Michel Dänzer [mailto:mic...@daenzer.net]
> Sent: Wednesday, April 11, 2018 05:50
> On 2018-04-11 08:57 AM, Nicolai Hähnle wrote:
> > On 10.04.2018 23:45, Cyr, Aric wrote:
> >> How does it work fine today given that all kernel seems to know is
> >> 'current' or 'current+1' vsyncs.
> >>
It doesn't actually upload any firmware is just
checks the version. The actual upload happens in
the gmc modules.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git
Fiji uses a different register than other smu7 asics, but
we already have this info in the base driver so just
use that.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 8 ++--
1 file changed, 2 insertions(+), 6 deletions(-)
This is useful for drivers (which will probably be all of them soon)
which need to track state that is exclusive to the topology, and not a
specific connector on said topology. This includes things such as the
link rate and lane count that are shared by all of the connectors on the
topology.
Next version of https://patchwork.freedesktop.org/series/41576/
All changes in this patch series are just to make checkpatch a little
happier, no functional changes.
Lyude Paul (10):
drm/atomic: Print debug message on atomic check failure
drm/i915: Move DP modeset retry work into intel_dp
Code is expecing to observe the same number of buffers returned from
dma_map_sg() function compared to sg_alloc_table_from_pages(). This
doesn't hold true universally especially for systems with IOMMU.
The hardware actually needs the dma_address on a page by page basis.
Joining multiple
On 2018-04-11 04:39 AM, Michel Dänzer wrote:
On 2018-04-10 08:02 PM, Leo Li wrote:
On 2018-04-09 11:03 AM, Michel Dänzer wrote:
On 2018-03-26 10:00 PM, sunpeng...@amd.com wrote:
From: "Leo (Sunpeng) Li"
In cases where CRTC properties are updated without going through
On 2018-04-11 04:39 AM, Michel Dänzer wrote:
On 2018-04-10 08:02 PM, Leo Li wrote:
On 2018-04-09 11:03 AM, Michel Dänzer wrote:
On 2018-03-26 10:00 PM, sunpeng...@amd.com wrote:
From: "Leo (Sunpeng) Li"
This change adds a few functions in preparation of enabling CRTC
On 2018-04-11 04:48 AM, Michel Dänzer wrote:
On 2018-04-10 08:02 PM, Leo Li wrote:
On 2018-04-09 11:03 AM, Michel Dänzer wrote:
On 2018-03-26 10:00 PM, sunpeng...@amd.com wrote:
From: "Leo (Sunpeng) Li"
The functions insert into the output resource creation, and
On 2018-04-11 03:47 PM, Shaoyun Liu wrote:
> Even reset failed, kfd post reset need to be called to make lock balance on
> kfd side
>
> Change-Id: I8b6ef29d7527915611be0b96a9cd039bc75bb0a9
> Signed-off-by: Shaoyun Liu
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7
On Wed, Apr 11, 2018 at 3:47 PM, Shaoyun Liu wrote:
> Even reset failed, kfd post reset need to be called to make lock balance on
> kfd side
>
> Change-Id: I8b6ef29d7527915611be0b96a9cd039bc75bb0a9
> Signed-off-by: Shaoyun Liu
Reviewed-by: Alex Deucher
Even reset failed, kfd post reset need to be called to make lock balance on
kfd side
Change-Id: I8b6ef29d7527915611be0b96a9cd039bc75bb0a9
Signed-off-by: Shaoyun Liu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 +++
1 file changed, 3 insertions(+), 4 deletions(-)
This is useful for drivers (which will probably be all of them soon)
which need to track state that is exclusive to the topology, and not a
specific connector on said topology. This includes things such as the
link rate and lane count that are shared by all of the connectors on the
topology.
Latest version of PW series 39642, hopefully this should also actually
come up on intel-gfx and go through CI.
No changes other than rebasing to the current drm-intel-next-queued
Lyude Paul (10):
drm/atomic: Print debug message on atomic check failure
drm/i915: Move DP modeset retry work
Am 11.04.2018 um 19:11 schrieb Robin Murphy:
Now that drm_prime_sg_to_page_addr_arrays() understands the case where
dma_map_sg() has coalesced segments and returns 0 < count < nents, we
can relax the check to only consider genuine failure.
That pattern is repeated in pretty much all drivers
Am 11.04.2018 um 19:11 schrieb Robin Murphy:
For dma_map_sg(), DMA API implementations are free to merge consecutive
segments into a single DMA mapping if conditions are suitable, thus the
resulting DMA addresses may be packed into fewer entries than
ttm->sg->nents implies.
On 11/04/18 18:11, Robin Murphy wrote:
For dma_map_sg(), DMA API implementations are free to merge consecutive
segments into a single DMA mapping if conditions are suitable, thus the
resulting DMA addresses may be packed into fewer entries than
ttm->sg->nents implies.
I am OK with this change.
Regards,
Eric
On 2018-04-11 01:25 PM, Tom St Denis wrote:
On 04/11/2018 01:23 PM, Alex Deucher wrote:
On Wed, Apr 11, 2018 at 2:31 AM, Rex Zhu wrote:
This struct can't be used for all asics.
Currently smu only calculate average gpu power in real
This patch change the power registers reading from average to maximum.
If SMU team verifies it, I am OK with it.
Regards,
Eric
On 2018-04-11 01:21 PM, Alex Deucher wrote:
On Wed, Apr 11, 2018 at 2:31 AM, Rex Zhu wrote:
pkgpwr is the average gpu power of 100ms. it is
If it is verified by SMU team, I am OK with it.
Regards,
Eric
On 2018-04-11 01:19 PM, Alex Deucher wrote:
On Wed, Apr 11, 2018 at 2:31 AM, Rex Zhu wrote:
we don't have limit of [50ms, 4sec] sampling period.
smu calculate average gpu power in real time.
we can read average
For dma_map_sg(), DMA API implementations are free to merge consecutive
segments into a single DMA mapping if conditions are suitable, thus the
resulting DMA addresses may be packed into fewer entries than
ttm->sg->nents implies.
drm_prime_sg_to_page_addr_arrays() does not account for this,
Now that drm_prime_sg_to_page_addr_arrays() understands the case where
dma_map_sg() has coalesced segments and returns 0 < count < nents, we
can relax the check to only consider genuine failure.
Signed-off-by: Robin Murphy
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 2 +-
On Wed, Apr 11, 2018 at 6:20 AM, Rex Zhu wrote:
> smu firmware do not update response register immediately under
> some delay tasks. so we read out the original value.
>
> so clear the register first.
>
> Signed-off-by: Rex Zhu
Acked-by: Alex Deucher
On Wed, Apr 11, 2018 at 8:00 AM, Rex Zhu wrote:
> Signed-off-by: Rex Zhu
Is this really worth it? It just saves a few cycles, but I'm worried
if we ever extend or rewrite the function we may miss later updates.
At the very least, I'd add a comment saying
On 04/11/2018 01:23 PM, Alex Deucher wrote:
On Wed, Apr 11, 2018 at 2:31 AM, Rex Zhu wrote:
This struct can't be used for all asics.
Currently smu only calculate average gpu power in real time.
for vddc/vddci/max power,
we can read them per 1ms through some registers. but
On Wed, Apr 11, 2018 at 2:31 AM, Rex Zhu wrote:
> This struct can't be used for all asics.
>
> Currently smu only calculate average gpu power in real time.
>
> for vddc/vddci/max power,
> we can read them per 1ms through some registers. but
>
> 1. the type of return values is not
On Wed, Apr 11, 2018 at 2:31 AM, Rex Zhu wrote:
> pkgpwr is the average gpu power of 100ms. it is calculated by
> firmware in real time.
>
> 1. we can send smu message PPSMC_MSG_GetCurrPkgPwr to read currentpkgpwr
> directly.
>
> 2. On Fiji/tonga/bonaire/hawwii, without
On Wed, Apr 11, 2018 at 2:31 AM, Rex Zhu wrote:
> we don't have limit of [50ms, 4sec] sampling period.
> smu calculate average gpu power in real time.
> we can read average gpu power through smu message or
> read special register.
>
> This reverts commit
On 2018-04-10 06:26 PM, Cyr, Aric wrote:> > My guess is they prefer to
“do nothing” and let driver/HW manage it,
> otherwise you exempt all existing games from supporting adaptive sync
> without a rewrite or update.
Nobody is saying adaptive sync should only work with explicit target
presentation
On 11/04/18 15:33, Sinan Kaya wrote:
On 4/11/2018 8:03 AM, Robin Murphy wrote:
On 10/04/18 21:59, Sinan Kaya wrote:
Code is expecing to observe the same number of buffers returned
from dma_map_sg() function compared to
sg_alloc_table_from_pages(). This doesn't hold true universally
especially
On 2018-04-10 09:02 PM, Laura Abbott wrote:
> There's an ongoing effort to remove VLAs[1] from the kernel to eventually
> turn on -Wvla. Switch to a constant value that covers all hardware.
>
> [1] https://lkml.org/lkml/2018/3/7/621
>
> Signed-off-by: Laura Abbott
> ---
> v2:
On 04/11/2018 08:17 AM, Christian König wrote:
Am 11.04.2018 um 14:03 schrieb Tom St Denis:
On 04/11/2018 07:58 AM, Christian König wrote:
- if (size & 3 || *pos & 3)
+ if (size & 3 || size > (4 * AMDGPU_DEBUGFS_MAX_SGPR_READ))
I think checking the position alignment here is
On Wed, Apr 11, 2018 at 2:08 AM, Rex Zhu wrote:
> 1. fix set vce clocks failed on Cz/St
>which lead 1s delay when boot up.
> 2. remove the workaround in vce_v3_0.c
>
> Reviewed-by: Alex Deucher
> Reviewed-by: Huang Rui
>
On 10/04/18 21:59, Sinan Kaya wrote:
Code is expecing to observe the same number of buffers returned from
dma_map_sg() function compared to sg_alloc_table_from_pages(). This
doesn't hold true universally especially for systems with IOMMU.
So why not fix said code? It's clearly not a real
On Wed, Apr 11, 2018 at 02:09:29PM +0300, Heikki Krogerus wrote:
> On Wed, Apr 11, 2018 at 08:28:44AM +, andrey.ara...@nixaid.com wrote:
> > Thank you for the insights, Heikki!
> >
> > Please find the acpi.dump.tgz file is a attached.
> >
> > I do not have the USBC* and INT3515* devices,
>
Pipeline removal of the BOs backing store when the placement is given
during validation.
Signed-off-by: Christian König
---
drivers/gpu/drm/ttm/ttm_bo.c | 12
1 file changed, 12 insertions(+)
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c
Move the parameters into a structure to make it simpler to extend it in
follow up patches.
This also adds the importer private as parameter so that we can directly
work with a completely filled in attachment structure.
Signed-off-by: Christian König
---
Instead of relying on the DRM functions just implement our own export
functions. This adds support for taking care of unpinned DMA-buf.
v2: fix unintended recursion, remove debugging leftovers
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h
Instead of relying on the DRM functions just implement our own import
functions. This adds support for taking care of unpinned DMA-buf.
v2: enable for all exporters, not just amdgpu, fix invalidation
handling, lock reservation object while setting callback
v3: change to new dma_buf attach
Each importer can now provide an invalidate_mappings callback.
This allows the exporter to provide the mappings without the need to pin
the backing store.
v2: don't try to invalidate mappings when the callback is NULL,
lock the reservation obj while using the attachments,
add helper to
Just note if a BO was imported/exported.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 11 +++
1 file changed, 11 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 68b9e0b..31d271f 100644
Am 11.04.2018 um 14:03 schrieb Tom St Denis:
On 04/11/2018 07:58 AM, Christian König wrote:
- if (size & 3 || *pos & 3)
+ if (size & 3 || size > (4 * AMDGPU_DEBUGFS_MAX_SGPR_READ))
I think checking the position alignment here is still necessary,
cause we can't read from not dw
On 04/11/2018 07:58 AM, Christian König wrote:
- if (size & 3 || *pos & 3)
+ if (size & 3 || size > (4 * AMDGPU_DEBUGFS_MAX_SGPR_READ))
I think checking the position alignment here is still necessary, cause
we can't read from not dw boundaries don't we?
The index is a dword
- if (size & 3 || *pos & 3)
+ if (size & 3 || size > (4 * AMDGPU_DEBUGFS_MAX_SGPR_READ))
I think checking the position alignment here is still necessary, cause
we can't read from not dw boundaries don't we?
Christian.
Am 11.04.2018 um 13:55 schrieb Tom St Denis:
Ping?
On
Ping?
On 04/09/2018 08:16 AM, Tom St Denis wrote:
We don't need to check the alignment of the offset and there was
potential a buffer overflow as well.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 8 ++--
1 file changed, 6
smu firmware do not update response register immediately under
some delay tasks. so we read out the original value.
so clear the register first.
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c | 4 +---
On 2018-04-10 07:25 PM, Cyr, Aric wrote:
>> From: Michel Dänzer [mailto:mic...@daenzer.net]
>> On 2018-04-10 07:13 PM, Cyr, Aric wrote:
From: Michel Dänzer [mailto:mic...@daenzer.net]
On 2018-04-10 06:26 PM, Cyr, Aric wrote:
> From: Koenig, Christian Sent: Tuesday, April 10, 2018
On 2018-04-11 08:57 AM, Nicolai Hähnle wrote:
> On 10.04.2018 23:45, Cyr, Aric wrote:
> For video games we have a similar situation where a frame is
> rendered
> for a certain world time and in the ideal case we would actually
> display the frame at this world time.
Am 11.04.2018 um 03:02 schrieb Laura Abbott:
There's an ongoing effort to remove VLAs[1] from the kernel to eventually
turn on -Wvla. Switch to a constant value that covers all hardware.
[1] https://lkml.org/lkml/2018/3/7/621
Signed-off-by: Laura Abbott
It would be nicer
Am 11.04.2018 um 04:38 schrieb zhoucm1:
On 2018年04月10日 21:42, Christian König wrote:
When GEM needs to fallback to GTT for VRAM BOs we still want the
preferred domain to be untouched so that the BO has a cance to move back
to VRAM in the future.
Signed-off-by: Christian König
Am 11.04.2018 um 08:26 schrieb Huang Rui:
On Tue, Apr 10, 2018 at 04:59:55PM -0400, Sinan Kaya wrote:
Code is expecing to observe the same number of buffers returned from
dma_map_sg() function compared to sg_alloc_table_from_pages(). This
doesn't hold true universally especially for systems
On 2018-04-10 08:02 PM, Leo Li wrote:
> On 2018-04-09 11:03 AM, Michel Dänzer wrote:
>> On 2018-03-26 10:00 PM, sunpeng...@amd.com wrote:
>>> From: "Leo (Sunpeng) Li"
>>>
>>> The functions insert into the output resource creation, and property
>>> change functions. CRTC
On 2018-04-10 08:02 PM, Leo Li wrote:
> On 2018-04-09 11:03 AM, Michel Dänzer wrote:
>> On 2018-03-26 10:00 PM, sunpeng...@amd.com wrote:
>>> From: "Leo (Sunpeng) Li"
>>>
>>> This change adds a few functions in preparation of enabling CRTC color
>>> managment via the randr
On 2018-04-10 08:02 PM, Leo Li wrote:
> On 2018-04-09 11:03 AM, Michel Dänzer wrote:
>> On 2018-03-26 10:00 PM, sunpeng...@amd.com wrote:
>>> From: "Leo (Sunpeng) Li"
>>>
>>> In cases where CRTC properties are updated without going through
>>> RRChangeOutputProperty, we don't
Code is expecing to observe the same number of buffers returned from
dma_map_sg() function compared to sg_alloc_table_from_pages(). This
doesn't hold true universally especially for systems with IOMMU.
IOMMU driver tries to combine buffers into a single DMA address as much
as it can. The right
On 10.04.2018 23:45, Cyr, Aric wrote:
For video games we have a similar situation where a frame is rendered
for a certain world time and in the ideal case we would actually
display the frame at this world time.
That seems like it would be a poorly written game that flips like
that, unless they
This struct can't be used for all asics.
Currently smu only calculate average gpu power in real time.
for vddc/vddci/max power,
we can read them per 1ms through some registers. but
1. the type of return values is not unified.
For Vi, return type is uint. For vega, return type is float.
2.
pkgpwr is the average gpu power of 100ms. it is calculated by
firmware in real time.
1. we can send smu message PPSMC_MSG_GetCurrPkgPwr to read currentpkgpwr
directly.
2. On Fiji/tonga/bonaire/hawwii, without PPSMC_MSG_GetCurrPkgPwr support.
Send PPSMC_MSG_PmStatusLogStart/Sample to let smu
we don't have limit of [50ms, 4sec] sampling period.
smu calculate average gpu power in real time.
we can read average gpu power through smu message or
read special register.
This reverts commit 462d8dcc9fec0d89f1ff6a1f93f1d4f670878c71.
Signed-off-by: Rex Zhu
---
1. fix set vce clocks failed on Cz/St
which lead 1s delay when boot up.
2. remove the workaround in vce_v3_0.c
Reviewed-by: Alex Deucher
Reviewed-by: Huang Rui
Acked-by: Christian König
Acked-by: Shirish S
fix the issue set uvd clock failed on CZ/ST
which lead 1s delay when boot up.
Reviewed-by: Alex Deucher
Reviewed-by: Huang Rui
Acked-by: Christian König
Acked-by: Shirish S
Signed-off-by: Rex Zhu
On Tue, Apr 10, 2018 at 04:59:55PM -0400, Sinan Kaya wrote:
> Code is expecing to observe the same number of buffers returned from
> dma_map_sg() function compared to sg_alloc_table_from_pages(). This
> doesn't hold true universally especially for systems with IOMMU.
>
> IOMMU driver tries to
75 matches
Mail list logo