This breaks the "size" out-parameter.
Hi Felix,
I just tested your patch. It works fine on my test set with the following
little fix.
Regards,
Ma Jun
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index 7ea3ec1e9e75..7d6fbfbfeb79 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topolo
Register irq handler.
Signed-off-by: Tao Zhou
Reviewed-by: Hawking Zhang
---
drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c
b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c
index b067fb63010e..403d054cf51b 1006
Register related irq handler.
Signed-off-by: Tao Zhou
Reviewed-by: Hawking Zhang
---
drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c | 18 ++
1 file changed, 18 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c
b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c
index aa5323c9f481
Add interrupt source id macros.
Signed-off-by: Tao Zhou
Reviewed-by: Hawking Zhang
---
drivers/gpu/drm/amd/include/ivsrcid/vcn/irqsrcs_vcn_4_0.h | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/gpu/drm/amd/include/ivsrcid/vcn/irqsrcs_vcn_4_0.h
b/drivers/gpu/drm/amd/include/ivsrci
Initialize VCN RAS structure and add RAS status query function.
Signed-off-by: Tao Zhou
Reviewed-by: Hawking Zhang
---
drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c | 60 +++
drivers/gpu/drm/amd/amdgpu/vcn_v4_0.h | 6 +++
2 files changed, 66 insertions(+)
diff --git a/drivers/
Make the code reusable.
Signed-off-by: Tao Zhou
Reviewed-by: Hawking Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.c | 17 +
drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h | 1 +
drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c | 13 +
3 files changed, 19 insertions(+), 12 del
Initialize JPEG RAS structure and add error query interface.
Signed-off-by: Tao Zhou
Reviewed-by: Hawking Zhang
---
drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c | 63 ++
drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.h | 7 +++
2 files changed, 70 insertions(+)
diff --git a/drivers/gp
So the code can be reused.
Signed-off-by: Tao Zhou
Reviewed-by: Hawking Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 17 +
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h | 1 +
drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c | 13 +
3 files changed, 19 insertions(+), 12 del
Set support flag for VCN/JPEG 4.0.
Signed-off-by: Tao Zhou
Reviewed-by: Hawking Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 693bce
On Wed, Nov 16, 2022 at 03:49:18PM -0500, Felix Kuehling wrote:
> Am 2022-11-16 um 03:04 schrieb Ma Jun:
> > Release the topology_lock in error case
> >
> > Signed-off-by: Ma Jun
> > Reported-by: Dan Carpenter
> Dan, did you change your email address, is this one correct?
>
Yep.
I'm still aro
Hi Dave, Daniel,
Fixes for 6.1.
The following changes since commit 094226ad94f471a9f19e8f8e7140a09c2625abaa:
Linux 6.1-rc5 (2022-11-13 13:12:55 -0800)
are available in the Git repository at:
https://gitlab.freedesktop.org/agd5f/linux.git
tags/amd-drm-fixes-6.1-2022-11-16
for you to fetch
On Wed, Nov 16, 2022 at 8:25 PM Luben Tuikov wrote:
>
> Don't assume FRU MCU memory locations for the FRU data fields, or their sizes,
> instead reading and interpret the IPMI data, as stipulated in the IPMI spec
> version 1.0 rev 1.2.
>
> Extract the Product Name, Product Part/Model Number, and t
[AMD Official Use Only - General]
Hi Michel,
I didn't reproduce the hang yet but find a race condition related with fence
signaling time. I updated the patch series based on kernel 5.18.
Thanks,
Jiadong
-Original Message-
From: Michel Dänzer
Sent: Tuesday, November 15, 2022 1:15 AM
To
Hi,
On 10/14/22 11:46, Christian König wrote:
> +/* Remove the entity from the scheduler and kill all pending jobs */
> +static void drm_sched_entity_kill(struct drm_sched_entity *entity)
> +{
> + struct drm_sched_job *job;
> + struct dma_fence *prev;
> +
> + if (!entity->rq)
> +
Thanks for the fix, feel free to add my:
Tested-by: Guilherme G. Piccoli
From: "Jiadong.Zhu"
Trigger Mid-Command Buffer Preemption according to the priority of the software
rings and the hw fence signalling condition.
The muxer saves the locations of the indirect buffer frames from the software
ring together with the fence sequence number in its fifo queue, and pops
From: "Jiadong.Zhu"
1. Modify the unmap_queue package on gfx9. Add trailing fence to track the
preemption done.
2. Modify emit_ce_meta emit_de_meta functions for the resumed ibs.
v2: Restyle code not to use ternary operator.
v3: Modify code format.
v4: Enable Mid-Command Buffer Preemption for
From: "Jiadong.Zhu"
The software ring is created to support priority context while there is only
one hardware queue for gfx.
Every software ring has its fence driver and could be used as an ordinary ring
for the GPU scheduler.
Multiple software rings are bound to a real ring with the ring muxer.
From: "Jiadong.Zhu"
Set ring functions with software ring callbacks on gfx9.
The software ring could be tested by debugfs_test_ib case.
v2: Set sw_ring 2 to enable software ring by default.
v3: Remove the parameter for software ring enablement.
v4: Use amdgpu_ring_init/fini for software rings.
[Why]
[ 754.862560] refcount_t: underflow; use-after-free.
[ 754.862898] Call Trace:
[ 754.862903]
[ 754.862913] amdgpu_job_free_cb+0xc2/0xe1 [amdgpu]
[ 754.863543] drm_sched_main.cold+0x34/0x39 [amd_sched]
[How]
The fw_fence may be not init, check whether dma_fenc
Don't assume FRU MCU memory locations for the FRU data fields, or their sizes,
instead reading and interpret the IPMI data, as stipulated in the IPMI spec
version 1.0 rev 1.2.
Extract the Product Name, Product Part/Model Number, and the Product Serial
Number by interpreting the IPMI data.
Check t
Add support for RAS table at I2C EEPROM address of 0x4, since on some
ASICs it is not at 0, but at 0x4.
Cc: Alex Deucher
Cc: Kent Russell
Signed-off-by: Luben Tuikov
Tested-by: Kent Russell
Reviewed-by: Kent Russell
Reviewed-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras
Bug fix reading the product name, product number, and serial number
on newer boards, especially new Aldebaran boards.
This edition fixes reading the actual serial number rather than the
product version, in patch 3. Also rebase to latest
amd-staging-drm-next.
Clarify changes in commits 1-3.
Add t
Allow non-standard EEPROM I2C address of 0x58, where the Device Type
Identifier is 1011b, where we form 1011000b = 0x58 I2C address, as on some
ASICs the FRU data lives there.
Cc: Alex Deucher
Cc: Kent Russell
Signed-off-by: Luben Tuikov
Tested-by: Kent Russell
Reviewed-by: Kent Russell
Revie
Set the new correct default FRU MCU I2C address for newer ASICs, so that we
can correctly read the Product Name, Product Part/Model Number and Serial
Number.
On newer ASICs, the FRU MCU was moved to I2C address 0x58.
Cc: Alex Deucher
Cc: Kent Russell
Signed-off-by: Luben Tuikov
Tested-by: Kent
On 2022-11-16 15:58, Alex Deucher wrote:
> On Wed, Nov 16, 2022 at 2:49 PM Luben Tuikov wrote:
>>
>> Read and interpret IPMI data to get the product name, product model, and
>> product serial number.
>
> Patches 1,2,4 are:
> Reviewed-by: Alex Deucher
> for patch 3:
> It's not clear from the comm
Not sure how much more to say, than what the commit description says now:
Read and interpret IPMI data to get the product name, product model, and
product serial number.
It reads IPMI data, and interpets it, to get the prodcut name, product model,
and product serial number.
What
It is actually reading the IPMI data as it should. Previously to this, it was
just assuming where the data would be, it's size, and so on. In talking
to some engineers internally, we concluded that it shouldn't do that, and it
should instead follow the IPMI spec to read the data--in the same way
th
On Wed, 2022-11-16 at 04:39 +, Lin, Wayne wrote:
> [Public]
>
> All the patch set looks good to me. Feel free to add:
> Reviewed-by: Wayne Lin
>
> Again, thank you Lyude for helping on this!!!
No problem! I was the one who introduced the bug anyway :P, I'm just glad we
were able to fix this
On 11/16/22 11:58, Alex Deucher wrote:
> Move the new callback outside of the guard.
>
> Fixes: dc55b106ad47 ("drm/amd/display: Disable phantom OTG after enable for
> plane disable")
> CC: Alvin Lee
> CC: Alan Liu
> Signed-off-by: Alex Deucher
Reviewed-by: Harry Wentland
Harry
> ---
> dri
[AMD Official Use Only - General]
You can add my
Reviewed-by: Kent Russell
as well. And I have no issue with a little elaboration on the commit. We
definitely changed a few things here.
Kent
> -Original Message-
> From: Alex Deucher
> Sent: Wednesday, November 16, 2022 3:59 PM
> To
Am 2022-11-16 um 11:54 schrieb Ramesh Errabolu:
Allow user to know number of compute units (CU) that are in use at any
given moment. Enable access to the method kgd_gfx_v9_get_cu_occupancy
that computes CU occupancy.
Signed-off-by: Ramesh Errabolu
Reviewed-by: Felix Kuehling
---
drivers
On Wed, Nov 16, 2022 at 12:03 AM Dave Airlie wrote:
>
> arm32 build fails
>
> /home/airlied/devel/kernel/dim/src/drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc.c:
> In function ‘disable_dangling_plane’:
> /home/airlied/devel/kernel/dim/src/drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc.c:1134:
Am 2022-11-16 um 11:40 schrieb Alex Deucher:
All of the IP specific versions are the same now, so
we can just use a common function.
Signed-off-by: Alex Deucher
The series is
Acked-by: Felix Kuehling
---
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 26 +++
drivers/gp
On Wed, Nov 16, 2022 at 11:54 AM Ramesh Errabolu
wrote:
>
> Allow user to know number of compute units (CU) that are in use at any
> given moment. Enable access to the method kgd_gfx_v9_get_cu_occupancy
> that computes CU occupancy.
>
> Signed-off-by: Ramesh Errabolu
Acked-by: Alex Deucher
> -
On Wed, Nov 16, 2022 at 2:49 PM Luben Tuikov wrote:
>
> Read and interpret IPMI data to get the product name, product model, and
> product serial number.
Patches 1,2,4 are:
Reviewed-by: Alex Deucher
for patch 3:
It's not clear from the commit message what this change is doing. Is
this just a re
Am 2022-11-16 um 03:04 schrieb Ma Jun:
Release the topology_lock in error case
Signed-off-by: Ma Jun
Reported-by: Dan Carpenter
Dan, did you change your email address, is this one correct?
Ma Jun, thanks for looking into this. Some of this problem predates your
patch that was flagged by Dan
tree/branch:
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git master
branch HEAD: 15f3bff12cf6a888ec2ad39652828c60e6836b3d Add linux-next specific
files for 20221116
Error/Warning reports:
https://lore.kernel.org/linux-mm/202210261404.b6ulzg7h-...@intel.com
https
Mostly fix reading the product name, product number, and serial number
for newer Aldebaran boards.
Luben Tuikov (4):
drm/amdgpu: Allow non-standard EEPROM I2C address
drm/amdgpu: Set new default I2C FRU EEPROM address
drm/amdgpu: Read IPMI data for product information
drm/amdgpu: Add suppo
Add support for RAS table at I2C EEPROM address of 0x4, since on some
ASICs it is not at 0, but at 0x4.
Cc: Alex Deucher
Cc: Kent Russell
Signed-off-by: Luben Tuikov
Tested-by: Kent Russell
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c | 7 ++-
1 file changed, 6 insertions(+)
Set the new default I2C FRU EEPROM address for newer ASICs.
Cc: Alex Deucher
Cc: Kent Russell
Signed-off-by: Luben Tuikov
Tested-by: Kent Russell
---
.../gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c| 36 +--
1 file changed, 25 insertions(+), 11 deletions(-)
diff --git a/drivers
Read and interpret IPMI data to get the product name, product model, and
product serial number.
Cc: Alex Deucher
Cc: Kent Russell
Signed-off-by: Luben Tuikov
Tested-by: Kent Russell
---
.../gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c| 183 --
1 file changed, 85 insertions(+), 9
Allow non-standard EEPROM I2C address of 0x58, where the Device Type
Identifier is 1011b, where we form 1011000b = 0x58 I2C address, as on some
ASICs the FRU lives there.
Cc: Alex Deucher
Cc: Kent Russell
Signed-off-by: Luben Tuikov
Tested-by: Kent Russell
---
drivers/gpu/drm/amd/amdgpu/amdgp
Move the new callback outside of the guard.
Fixes: dc55b106ad47 ("drm/amd/display: Disable phantom OTG after enable for
plane disable")
CC: Alvin Lee
CC: Alan Liu
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h | 2 +-
1 file changed, 1 insertion(+), 1
Allow user to know number of compute units (CU) that are in use at any
given moment. Enable access to the method kgd_gfx_v9_get_cu_occupancy
that computes CU occupancy.
Signed-off-by: Ramesh Errabolu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_aldebaran.c | 1 +
1 file changed, 1 insertion(+)
All of the IP specific versions are the same now, so
we can just use a common function.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 26 +++
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h | 2 --
drivers/gpu/drm/amd/amdgpu/psp_v10_0.c | 27 -
This matches what we do for psp 3.1 and makes ring_init
common for all PSP versions.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/psp_v12_0.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
b/drivers/gpu/drm/amd/amdgpu
We can reuse the same buffers on resume.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 56 +
1 file changed, 29 insertions(+), 27 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
inde
I was thinking something like this would be more straightforward.
Alex
On Wed, Nov 16, 2022 at 11:01 AM Christian König
wrote:
>
> That the PSP code tries to free the memory during suspend is quite
> broken and leads to problems during resume.
>
> Signed-off-by: Christian König
> ---
> drivers
When buffers are freed during suspend there is no guarantee that
they can be re-allocated during resume.
The PSP subsystem seems to be quite buggy regarding this, so add
a WARN_ON() to point out those bugs.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 2 ++
1
That the PSP code tries to free the memory during suspend is quite
broken and leads to problems during resume.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 30 ++---
1 file changed, 12 insertions(+), 18 deletions(-)
diff --git a/drivers/gpu/dr
Some kernel buffers can only be allocated after asking the firmware how
large they should be. But since this happens on every resume don't pin
them multiple times.
Also bail out with an error if the requested size should ever change.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu
Am 16.11.22 um 09:50 schrieb Somalapuram Amaranath:
remove page shift operations as ttm_resource moved
from num_pages to size_t size in bytes.
Signed-off-by: Somalapuram Amaranath
---
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 4 +---
drivers/gpu/drm/ttm/ttm_range_manager.c| 2 +-
Am 16.11.22 um 10:20 schrieb Arunpravin Paneer Selvam:
Hi Amar,
On 11/16/2022 2:20 PM, Somalapuram Amaranath wrote:
ttm_resource allocate size in bytes i.e less than page size.
Signed-off-by: Somalapuram Amaranath
---
drivers/gpu/drm/drm_gem.c | 2 +-
1 file changed, 1 insertion(+), 1 dele
Am 16.11.22 um 06:47 schrieb Arunpravin Paneer Selvam:
This patch fixes the VRAM BO eviction issue during resume when
playing the steam game cuphead.
During psp resume, it requests a VRAM buffer of size 10240 KiB for
the trusted memory region, as part of this memory allocation we are
trying to e
Hi Amar,
On 11/16/2022 2:20 PM, Somalapuram Amaranath wrote:
ttm_resource allocate size in bytes i.e less than page size.
Signed-off-by: Somalapuram Amaranath
---
drivers/gpu/drm/drm_gem.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/drm_gem.c b/driver
ttm_resource allocate size in bytes i.e less than page size.
Signed-off-by: Somalapuram Amaranath
---
drivers/gpu/drm/drm_gem.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
index b8db675e7fb5..a346e3b7f9a8 100644
--- a/
remove page shift operations as ttm_resource moved
from num_pages to size_t size in bytes.
Signed-off-by: Somalapuram Amaranath
---
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 4 +---
drivers/gpu/drm/ttm/ttm_range_manager.c| 2 +-
2 files changed, 2 insertions(+), 4 deletions(-)
diff --git
Release the topology_lock in error case
Signed-off-by: Ma Jun
Reported-by: Dan Carpenter
---
drivers/gpu/drm/amd/amdkfd/kfd_topology.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index ef9c6fdfb88d
60 matches
Mail list logo
