On 10/04/17 06:42 PM, Emil Velikov wrote:
> On 10 April 2017 at 09:37, Michel Dänzer wrote:
>> From: Daniel Stone
>>
>> DRM_EVENT_CONTEXT_VERSION is the latest context version supported by
>> whatever version of libdrm is present. We were blindly
Don't change bare-metal logic, so please use amdgpu_sriov_vf() to differentiate
-4 or -4 for the ID_MAXIMUM
With that addressed, this patch is :
Reviewed-by: Monk Liu
-邮件原件-
发件人: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] 代表 Huang Rui
发送时间: Monday, April
Reviewed-by: Xiangliang Yu
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Trigger Huang
> Sent: Monday, April 10, 2017 2:46 PM
> To: amd-gfx@lists.freedesktop.org
> Cc: Huang, Ray ; Huang,
On 11/04/17 08:21 AM, Felix Kuehling wrote:
> Provide convenient compile time and boot time options for selecting
> CIK ASIC support in either or both drivers.
>
> v2: git add missing files
>
> Signed-off-by: Felix Kuehling
[...]
> diff --git
Reviewed-by: Ken Wang
发件人: amd-gfx 代表 Huang Rui
发送时间: 2017年4月11日 10:58:21
收件人: amd-gfx@lists.freedesktop.org; Deucher, Alexander
抄送: Huang, Ray
主题: [PATCH] drm/amdgpu: fix to add
On Mon, Apr 10, 2017 at 08:09:17PM +0800, Grazvydas Ignotas wrote:
> On Mon, Apr 10, 2017 at 12:37 PM, Huang Rui wrote:
>
> ASIC_INIT_COMPLETE bit must be cleared during S3 resuming phase,
> because VBIOS will check the bit to decide if execute ASIC_Init
> posting
From: Dave Airlie
This object can be used to implement the Vulkan semaphores.
The object behaviour differs from fence, in that you can
replace the underlying fence, and you cannot merge semaphores.
Signed-off-by: Dave Airlie
---
From: Dave Airlie
This just adds two helper interfaces to bridge the gap from
drivers to sync_file for the semaphore objects.
These will be used by the amdgpu driver.
Signed-off-by: Dave Airlie
---
drivers/gpu/drm/drm_syncobj.c | 77
This set of sync object patches should address most of the issues
raised in review.
The major changes:
Race on destroy should be gone,
Documentation fixups
amdgpu internal use p instead of adev fixups
My plans are to write some igt tests this week, and try
to get some more review on what the API
From: Dave Airlie
These are just alloc and fdget interfaces needed by the drm sync
objects code.
Reviewed-by: Christian König
Reviewed-by: Daniel Vetter
Signed-off-by: Dave Airlie
---
From: Dave Airlie
This creates a new command submission chunk for amdgpu
to add wait and signal sync objects around the submission.
Sync objects are managed via the drm syncobj ioctls.
The command submission interface is enhanced with two new
chunks, one for semaphore
From: Dave Airlie
This allows us to create sync files with different semantics,
and clearly define the interoperation between them it also
provides flags to allow for tweaks on those semantics.
This provides a validation interface for drivers that accept
types from userspace
From: Dave Airlie
This patch allows the underlying fence in a sync_file to be changed
or set to NULL. This isn't currently required but for Vulkan
semaphores we need to be able to swap and reset the fence.
In order to faciliate this, it uses rcu to protect the fence,
along
From: Dave Airlie
Sync objects are new toplevel drm object, that have the same
semantics as sync_file objects, but without requiring an fd
to be consumed to support them.
This patch just enables the DRM interface to create these
objects, it doesn't actually provide any
This patch fixes the case when buffer funcs is empty and bo evict is
executing. It must double check buffer funcs, otherwise, a NULL
pointer dereference kernel panic will be encountered.
BUG: unable to handle kernel NULL pointer dereference at 01a4
IP: []
On 10/04/17 07:24 PM, Monk Liu wrote:
>
> struct drm_amdgpu_wait_cs_in {
> - /** Command submission handle */
> + /* Command submission handle
Any particular reason for dropping the /** comment opening?
> + * handle equals 0 means none to wait for
This line and the following
On 11/04/17 08:29 AM, Felix Kuehling wrote:
> I tested this with Hawaii on the KFD branch and
> DRM_CIK_BOTH_DEFAULT_AMDGPU. Both modules get loaded, but radeon doesn't
> initialize the device. Amdgpu works with kfdtest.
Did radeon print the "CIK support disabled by module param" message?
--
On 11/04/17 03:13 AM, Felix Kuehling wrote:
> On 17-04-09 09:28 PM, Michel Dänzer wrote:
>> On 10/04/17 09:56 AM, Michel Dänzer wrote:
>>> On 08/04/17 05:15 AM, Felix Kuehling wrote:
Advertise CIK PCI IDs only when they are not supported by amdgpu.
Use the CONFIG_DRM_AMDGPU_CIK to check
I tested this with Hawaii on the KFD branch and
DRM_CIK_BOTH_DEFAULT_AMDGPU. Both modules get loaded, but radeon doesn't
initialize the device. Amdgpu works with kfdtest.
To keep the driver code clean I only use straight-forward per-driver
options in #ifdefs. The CONFIG_DRM_CIK_... options that
Provide convenient compile time and boot time options for selecting
CIK ASIC support in either or both drivers.
v2: git add missing files
Signed-off-by: Felix Kuehling
---
drivers/gpu/drm/Kconfig | 51 ++
drivers/gpu/drm/amd/amdgpu/Kconfig
Provide convenient compile time and boot time options for selecting
CIK ASIC support in either or both drivers.
Signed-off-by: Felix Kuehling
---
drivers/gpu/drm/amd/amdgpu/Kconfig | 10 ++-
drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 +
On 17-04-08 04:50 AM, Nicolai Hähnle wrote:
> On 07.04.2017 22:15, Felix Kuehling wrote:
>> Change the wording of the CONFIG_DRM_AMDGPU_CIK option to indicate
>> that it's no longer experimental.
>>
>> Signed-off-by: Felix Kuehling
>> ---
>>
Even if we disable clockgating, we still need to make sure the
cp/rlc interrupts are enabled for powergating which might still
be enabled.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 2 ++
1 file changed, 2 insertions(+)
diff --git
Even if we disable clockgating, we still need to make sure the
cp/rlc interrupts are enabled for powergating which might still
be enabled.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c | 11 +++
1 file changed, 7 insertions(+), 4
On 17-04-10 02:50 PM, Alex Deucher wrote:
> On Mon, Apr 10, 2017 at 2:13 PM, Felix Kuehling
> wrote:
>>
>> Are there any other ideas or requests, before I go spend more time on this?
>
> In the short term, it would probably be good to just have a tri-state option:
> 1.
Signed-off-by: Alex Deucher
---
include/uapi/drm/amdgpu_drm.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index 92262d8..95260e5 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++
fixed in: "drm/amdgpu:fix race condition"
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
b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
On Mon, Apr 10, 2017 at 2:13 PM, Felix Kuehling wrote:
> On 17-04-09 09:28 PM, Michel Dänzer wrote:
>> On 10/04/17 09:56 AM, Michel Dänzer wrote:
>>> On 08/04/17 05:15 AM, Felix Kuehling wrote:
Advertise CIK PCI IDs only when they are not supported by amdgpu.
Use
On 17-04-09 09:28 PM, Michel Dänzer wrote:
> On 10/04/17 09:56 AM, Michel Dänzer wrote:
>> On 08/04/17 05:15 AM, Felix Kuehling wrote:
>>> Advertise CIK PCI IDs only when they are not supported by amdgpu.
>>> Use the CONFIG_DRM_AMDGPU_CIK to check so that a single option in
>>> the kernel config
Patch was Tested-by and Reviewed-by: Rex Zhu
Best Regards
Rex
From: amd-gfx on behalf of Alex Deucher
Sent: Tuesday, April 11, 2017 1:02:02 AM
To: amd-gfx@lists.freedesktop.org
Patch was Tested-by and Reviewed-by: Rex Zhu
Best Regards
Rex
From: amd-gfx on behalf of Alex Deucher
Sent: Tuesday, April 11, 2017 12:51:01 AM
To:
Am 10.04.2017 um 18:19 schrieb Tom St Denis:
This series ports a variety of vega10 drivers over
to the new SOC15 macros and introduces a pair of
OFFSET ready macros to help with MC related code.
Patch #1 is Reviewed-by: Christian König
Patches #2-#10 are Acked-by:
It's global, not queue specific, so move it out of the
kiq register init function.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 26 --
1 file changed, 12 insertions(+), 14 deletions(-)
diff --git
Add missing chips to the doorbell range setup. These
were missed in the KIQ code. Fixes power and performance
regressions with KIQ. Spotted by Rex.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 7 +--
1 file changed, 5 insertions(+), 2
This series ports a variety of vega10 drivers over
to the new SOC15 macros and introduces a pair of
OFFSET ready macros to help with MC related code.
___
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/psp_v3_1.c | 34 +-
1 file changed, 17 insertions(+), 17 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
index
No functional changes.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c | 201
1 file changed, 100 insertions(+), 101 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
Note this commit also fixes a few typos in the commented
out portions which were missing closing parenthesis on the
original coding style.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/vce_v4_0.c | 176 +-
1 file changed, 88
Swap a couple of read/set/write patterns for WREG32_FIELD15.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c | 12
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c
Change read/set/write patterns into WREG32_FIELD15 calls.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c | 11 +++
1 file changed, 3 insertions(+), 8 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
Allows reading/writing via SOC15 macros with offset for
various register banks.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/soc15_common.h | 14 ++
1 file changed, 14 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15_common.h
No functional changes in this patch. Simply re-writing
read/writes into new macros.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c | 182 +++
1 file changed, 91 insertions(+), 91 deletions(-)
diff --git
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c | 44 ++
1 file changed, 18 insertions(+), 26 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c
b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c
index
As well swap a read/set/write pattern over to WREG32_FIELD15.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 11 ---
1 file changed, 4 insertions(+), 7 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
This patch also fixes some bugs in commented out code where
parenthesis were missing on various SOC15 macros.
Signed-off-by: Tom St Denis
---
drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 206 +-
1 file changed, 103 insertions(+), 103 deletions(-)
On Mon, Apr 10, 2017 at 8:02 AM, Daniel Vetter wrote:
> On Mon, Apr 10, 2017 at 12:12:01PM +0200, Gerd Hoffmann wrote:
>> Ok, this is really a kickstart for a discussion. While working on
>> graphics support for virtual machines on ppc64 (which exists in both
>> little and big
On Mon, Apr 10, 2017 at 10:45 AM, Ilia Mirkin wrote:
> On Mon, Apr 10, 2017 at 10:17 AM, Gerd Hoffmann wrote:
>> Hi,
>>
>>> which software have you used as representative of "reality"?
>>
>> ppc64 (big endian) virtual machine, running with qemu stdvga &
On Mon, Apr 10, 2017 at 11:09 AM, Pekka Paalanen wrote:
> On Mon, 10 Apr 2017 16:17:27 +0200
> Gerd Hoffmann wrote:
>
>> Hi,
>>
>> > which software have you used as representative of "reality"?
>>
>> ppc64 (big endian) virtual machine, running with qemu
On Mon, 10 Apr 2017 16:17:27 +0200
Gerd Hoffmann wrote:
> Hi,
>
> > which software have you used as representative of "reality"?
>
> ppc64 (big endian) virtual machine, running with qemu stdvga & bochs-drm
> driver. Xorg with modesetting driver uses DRM_FORMAT_XRGB
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Huang Rui
> Sent: Monday, April 10, 2017 5:38 AM
> To: amd-gfx@lists.freedesktop.org; Deucher, Alexander
> Cc: Huang, Ray
> Subject: [PATCH 3/3] drm/amdgpu: fix to clear ASIC INIT COMPLETE
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Michel Dänzer
> Sent: Monday, April 10, 2017 4:38 AM
> To: amd-gfx@lists.freedesktop.org
> Subject: [PATCH xf86-video-ati] Set correct DRM event context version
>
> From: Daniel Stone
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Michel Dänzer
> Sent: Monday, April 10, 2017 4:38 AM
> To: amd-gfx@lists.freedesktop.org
> Subject: [PATCH xf86-video-amdgpu] Set correct DRM event context version
>
> From: Daniel Stone
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Rex Zhu
> Sent: Monday, April 10, 2017 4:08 AM
> To: amd-gfx@lists.freedesktop.org
> Cc: Zhu, Rex
> Subject: [PATCH] drm/amd/powerplay: set sensitivity of fan reaction to
> temepature
Hi,
> which software have you used as representative of "reality"?
ppc64 (big endian) virtual machine, running with qemu stdvga & bochs-drm
driver. Xorg with modesetting driver uses DRM_FORMAT_XRGB (one and
only format supported by bochs-drm), and we have to interpret that in
bigendian
On Mon, Apr 10, 2017 at 12:37 PM, Huang Rui wrote:
> ASIC_INIT_COMPLETE bit must be cleared during S3 resuming phase,
> because VBIOS will check the bit to decide if execute ASIC_Init
> posting via kernel driver.
>
> Signed-off-by: Huang Rui
> ---
>
On Mon, Apr 10, 2017 at 12:12:01PM +0200, Gerd Hoffmann wrote:
> Ok, this is really a kickstart for a discussion. While working on
> graphics support for virtual machines on ppc64 (which exists in both
> little and big endian variants) I've figured the comments in the header
> file don't match
Am 10.04.2017 um 12:24 schrieb Monk Liu:
sequence is protected by spinlock so don't access sequence
in paramter seq when invoking this function.
~0 means to get the latest sequence number and 0 means none to
get.
Change-Id: Ib7a03f3cf5594deeb4ad333cc59b47a6bddfd1ad
Signed-off-by: Monk Liu
sequence is protected by spinlock so don't access sequence
in paramter seq when invoking this function.
~0 means to get the latest sequence number and 0 means none to
get.
Change-Id: Ib7a03f3cf5594deeb4ad333cc59b47a6bddfd1ad
Signed-off-by: Monk Liu
---
Am 10.04.2017 um 12:16 schrieb Liu, Monk:
Is that so ?
Seq is already type of "uint64_t", GCC still compiles ~0 as 0x, but not
0x ?
You mean that it sign extend the value? Good point, that could be as well.
I'll change to use ~0ull anyway,
Yeah, that should make
Is that so ?
Seq is already type of "uint64_t", GCC still compiles ~0 as 0x, but not
0x ?
I'll change to use ~0ull anyway,
BR Monk
-邮件原件-
发件人: Christian König [mailto:deathsim...@vodafone.de]
发送时间: Monday, April 10, 2017 4:10 PM
收件人: Liu, Monk
Ok, this is really a kickstart for a discussion. While working on
graphics support for virtual machines on ppc64 (which exists in both
little and big endian variants) I've figured the comments in the header
file don't match reality. They are not considered little endian (as
suggested by the
On 10 April 2017 at 09:37, Michel Dänzer wrote:
> From: Daniel Stone
>
> DRM_EVENT_CONTEXT_VERSION is the latest context version supported by
> whatever version of libdrm is present. We were blindly asserting we
> supported whatever version that may be,
Fence buf needs to be used on suspend/resume phase.
Signed-off-by: Huang Rui
---
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
ASIC_INIT_COMPLETE bit must be cleared during S3 resuming phase,
because VBIOS will check the bit to decide if execute ASIC_Init
posting via kernel driver.
Signed-off-by: Huang Rui
---
drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c | 5 -
Signed-off-by: Huang Rui
---
drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 9da5b0b..f73e80c 100644
---
From: Daniel Stone
DRM_EVENT_CONTEXT_VERSION is the latest context version supported by
whatever version of libdrm is present. We were blindly asserting we
supported whatever version that may be, even if we actually didn't.
With libdrm 2.4.78, setting a higher context
From: Daniel Stone
DRM_EVENT_CONTEXT_VERSION is the latest context version supported by
whatever version of libdrm is present. We were blindly asserting we
supported whatever version that may be, even if we actually didn't.
With libdrm 2.4.78, setting a higher context
Change-Id: Ia7b9b3e619368d1fd3009be1352f5a587485d245
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c
Am 10.04.2017 um 09:39 schrieb Monk Liu:
sequence is protected by spinlock so don't access sequence
in paramter seq when invoking this function.
~0 means to get the latest sequence number and 0 means none to
get.
Well you are still not using ~0ull as I said. Using ~0 here will cause
sequence is protected by spinlock so don't access sequence
in paramter seq when invoking this function.
~0 means to get the latest sequence number and 0 means none to
get.
Change-Id: Ib7a03f3cf5594deeb4ad333cc59b47a6bddfd1ad
Signed-off-by: Monk Liu
---
> Am 09.04.2017 um 14:31 schrieb Xiangliang Yu:
> > For SRIOV, multi vce engine will hang when encoding. Add VMHUB flush
> > to workaround it, will continue to find the root cause later.
>
> NAK, that won't work and can cause hangs as well.
>
> The hang you experience could be a known issue with
>That is unnecessary the "if (seq + amdgpu_sched_jobs < cring->sequence) {"
>will catch this case.
But what if user land pass in seq 0 on purpose , but the latest cring->sequence
value is only 3 ?
Without this "if (seq == 0)" the logic seems goes to a wrong direction
BR Monk
-邮件原件-
Oh yeah~~ thanks
-邮件原件-
发件人: Christian König [mailto:deathsim...@vodafone.de]
发送时间: Monday, April 10, 2017 3:22 PM
收件人: Liu, Monk ; amd-gfx@lists.freedesktop.org
主题: Re: [PATCH] drm/amdgpu:fix race condition
Am 10.04.2017 um 09:18 schrieb Monk Liu:
> sequence is
Am 10.04.2017 um 09:18 schrieb Monk Liu:
sequence is protected by spinlock so don't access sequence
in paramter seq when invoking this function.
~0 means to get the latest sequence number and 0 means none to
get.
Change-Id: Ib7a03f3cf5594deeb4ad333cc59b47a6bddfd1ad
Signed-off-by: Monk Liu
sequence is protected by spinlock so don't access sequence
in paramter seq when invoking this function.
~0 means to get the latest sequence number and 0 means none to
get.
Change-Id: Ib7a03f3cf5594deeb4ad333cc59b47a6bddfd1ad
Signed-off-by: Monk Liu
---
Mhm, did you run into an actual issue or was that just from reading the
code?
The caller should have reserved a shared slot before calling
amdgpu_vm_update_directories() and I just double checked that and we
seem to do this everywhere.
Regards,
Christian.
Am 10.04.2017 um 08:54 schrieb
In Tonga's virtualization environment, for firmware UCODE_ID_STORAGE,
there is no actual firmware data, but we still need alloc a BO and
tell the BO's mc address to HW, or world switch will hang on VFs.
Signed-off-by: Trigger Huang
---
Signed-off-by: Junwei Zhang
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 157ae50..55af909 100644
---
On Mon, Apr 10, 2017 at 02:45:49PM +0800, Trigger Huang wrote:
> In Tonga's virtualization environment, for firmware UCODE_ID_STORAGE,
> there is no actual firmware data, but we still need alloc a BO and
> tell the BO's mc address to HW, or world switch will hang on VFs.
>
> Signed-off-by:
Am 09.04.2017 um 14:31 schrieb Xiangliang Yu:
For SRIOV, multi vce engine will hang when encoding. Add VMHUB
flush to workaround it, will continue to find the root cause later.
NAK, that won't work and can cause hangs as well.
The hang you experience could be a known issue with old libdrm, do
80 matches
Mail list logo