There is a typo so this unreserves "sobj" instead of "dobj".
Fixes: bb812f1ea87d ("drm/amdgpu: allocate gart memory when it's required (v3)")
Signed-off-by: Dan Carpenter
---
From static analysis. Not tested.
drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c | 2 +-
1 file changed, 1 insertion(+),
Hi Andrey,
Currently, I am busying with another issue, maybe will try next week.
Best wishes
Emily Deng
>-Original Message-
>From: Grodzovsky, Andrey
>Sent: Friday, November 15, 2019 6:14 AM
>To: Koenig, Christian ; Deng, Emily
>; amd-gfx@lists.freedesktop.org
>Subject: Re: [PATCH
DMA's atomic behavir is unlike GFX,If the comparing data is not
equal to destination data,
For GFX, loop again till gfx timeout(system hang).
For DMA, loop again till timer expired and then send interrupt.
So testcase can't use interrupt mechanism.
We take another way to verify. When the comparing
This patch is to test the command submission with secure context.
Signed-off-by: Aaron Liu
---
tests/amdgpu/security_tests.c | 9 +
1 file changed, 9 insertions(+)
diff --git a/tests/amdgpu/security_tests.c b/tests/amdgpu/security_tests.c
index a3807ee..0bb4ad2 100644
--- a/tests/amdgpu
This patch add test to submit a sdma command with secure context.
Signed-off-by: Aaron Liu
---
tests/amdgpu/security_tests.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/tests/amdgpu/security_tests.c b/tests/amdgpu/security_tests.c
index 0bb4ad2..eb239f4 100644
--- a/tests/amd
From: Huang Rui
This patch is to add add device handle as input param for exec_cs_helper
and write_linear_helper.
Because they are needed in security tests.
v2: fix typo that basic tests should be un-secure.
v3: refine the function implementation.
v4: remove amdgpu_cs_ctx_create3 calling.
Sign
Secure buffer is only able to be read with trusted ip block.
So we need use GFX ip to read it back instead of CPU.
Steps:
1. use write_data packet to write 0xdeadbeaf to secure buffer,
2. use atmoic_mem packet and ATOMIC_CMPSWAP_RTN_32 opcode to compare
the cmp_data(0xdeadbeaf) to the written data
From: Huang Rui
This patch expand write linear helper for security to submit the command
with secure context.
v2: refine the function implementation.
v3: remove amdgpu_cs_ctx_create3.
Signed-off-by: Huang Rui
Signed-off-by: Aaron Liu
Reviewed-by: Alex Deucher
---
tests/amdgpu/amdgpu_test.h
This patch expands secure param for amdgpu_test_exec_cs_helper_raw.
The flag is transfered to kernel with cs.
Signed-off-by: Aaron Liu
---
tests/amdgpu/basic_tests.c | 8 +---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/tests/amdgpu/basic_tests.c b/tests/amdgpu/basic_tests.
In kernel, cs->in.flags is used for TMZ. Hence libdrm should transfer
the flag to kernel.
Signed-off-by: Aaron Liu
---
amdgpu/amdgpu.h| 4 +++-
amdgpu/amdgpu_cs.c | 4
2 files changed, 7 insertions(+), 1 deletion(-)
diff --git a/amdgpu/amdgpu.h b/amdgpu/amdgpu.h
index f45f9f7..aee3f3d
From: Huang Rui
To align the kernel uapi change from Alex:
"Add a flag to the GEM_CREATE ioctl to create encrypted buffers. Buffers with
this flag set will be created with the TMZ bit set in the PTEs or engines
accessing them. This is required in order to properly access the data from the
engine
These series of patches are to initial TMZ support for libdrm and amdgpu_test.
They are based on amd-staging-hybrid-master20190125 branch. So far, gfx/sdma
TMZ tests are passed on Raven.
Please review them.
Thanks,
Aaron
Aaron Liu (6):
amdgpu: using IB flags for TMZ or non-TMZ
tests/amd
From: Huang Rui
This patch is to add a new test suite to store security tests.
In Raven+ asics, it will support TMZ (trust memory zone), and it is
page-based protection feature.
v2: remove tests/amdgpu/Makefile.am and update to
tests/amdgpu/meson.build
Signed-off-by: Huang Rui
Signed-off-by: A
From: Huang Rui
This patch is to add secure buffer allocation test for system memory.
Signed-off-by: Huang Rui
Reviewed-by: Alex Deucher
---
tests/amdgpu/security_tests.c | 9 +
1 file changed, 9 insertions(+)
diff --git a/tests/amdgpu/security_tests.c b/tests/amdgpu/security_tests.c
From: Huang Rui
This patch is to add secure buffer allocation test for invisible VRAM.
Signed-off-by: Huang Rui
Reviewed-by: Alex Deucher
---
tests/amdgpu/security_tests.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/tests/amdgpu/security_tests.c b/tests/amdgpu/security_tes
Sorry, I confused reserve_bad_pages with add_bad_pages, you're right.
If vram allocated info could be reserved after gpu reset, it seems your patch
is unnecessary.
If there is a risk that the info is lost, I think [data->0, data->count) pages
instead of only [data->last_reserved, data->count) pa
Thanks Alex for the quick reply. The kernel in question is from Fedora
that somehow failed to install the firmware in the initrd. The firmware
is present as highlighted on the bugzilla
https://bugzilla.redhat.com/show_bug.cgi?id=1772313#c4
Awaiting for the Fedora Kernel team to investigate the
Hi Harry,
As suggested I created the issue here
https://gitlab.freedesktop.org/drm/amd/issues/2 with a picture of the
problem attached.
Please take a look, thx!
Julien
On Mon, Nov 11, 2019 at 1:11 PM Harry Wentland wrote:
> On 2019-10-08 2:15 p.m., Julien Isorce wrote:
> > Hi Harry,
> >
> > I
Attached.
Emily - can you give it a try ?
Andrey
On 11/14/19 3:12 AM, Christian König wrote:
What about instead of peeking at the job to actually remove it from
ring_mirror_list right there,
Also an interesting idea. We would need to protect the mirror list
with a lock again, but that should
Hi Dave, Daniel,
Just one fix for 5.4.
The following changes since commit 31f4f5b495a62c9a8b15b1c3581acd5efeb9af8c:
Linux 5.4-rc7 (2019-11-10 16:17:15 -0800)
are available in the Git repository at:
git://people.freedesktop.org/~agd5f/linux tags/drm-fixes-5.4-2019-11-14
for you to fetch ch
Yes
-Original Message-
From: Wentland, Harry
Sent: November 14, 2019 4:22 PM
To: Cornij, Nikola ; Lipski, Mikita
; amd-gfx@lists.freedesktop.org
Cc: Wentland, Harry ; Deucher, Alexander
; manasi.d.nav...@intel.com; David Francis
Subject: Re: [PATCH 1/2] drm/dsc: Update drm_dsc to ref
On 2019-11-14 4:20 p.m., Cornij, Nikola wrote:
> This looks good, too.
>
Can we treat this as your reviewed-by for both changes?
Usually people mark these as reviewed with a
Reviewed-by: Harry Wentland
Thanks,
Harry
> -Original Message-
> From: mikita.lip...@amd.com
> Sent: November
This looks good, too.
-Original Message-
From: mikita.lip...@amd.com
Sent: November 13, 2019 2:07 PM
To: amd-gfx@lists.freedesktop.org
Cc: Wentland, Harry ; Deucher, Alexander
; Cornij, Nikola ;
manasi.d.nav...@intel.com; Lipski, Mikita ; David
Francis
Subject: [PATCH 1/2] drm/dsc: U
Looks good to me
-Original Message-
From: mikita.lip...@amd.com
Sent: November 13, 2019 2:07 PM
To: amd-gfx@lists.freedesktop.org
Cc: Wentland, Harry ; Deucher, Alexander
; Cornij, Nikola ;
manasi.d.nav...@intel.com; Lipski, Mikita
Subject: [PATCH 2/2] drm/dsc: Use 32-bit integers for
On Thu, Nov 14, 2019 at 11:41 AM Harry Wentland wrote:
>
> On 2019-11-14 7:36 a.m., zhengbin wrote:
> > zhengbin (4):
> > drm/amd/display: remove set but not used variable 'old_plane_crtc'
> > drm/amd/display: remove set but not used variable 'bp' in
> > bios_parser2.c
> > drm/amd/displa
When gfxoff is enabled, accessing gfx registers via MMIO
can lead to a hang.
Bug: https://bugzilla.kernel.org/show_bug.cgi?id=205497
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/am
This reverts commit 5e49d6f654c569c2de920babbaf5cf7c4c4a353f.
Drop this workaround in favor of a better one.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/nv.c| 27 ++
drivers/gpu/drm/amd/amdgpu/soc15.c | 31 --
2 files change
On 11/13/19 10:09 PM, Zhou1, Tao wrote:
Two questions:
1. "we lose all reservation during ASIC reset"
Are you sure of it? I remember the content of vram may be lost after reset but
the allocated status could be reserved.
Yea, now that I am thinking of it I think i might have confused it with
On 2019-11-14 7:36 a.m., zhengbin wrote:
> zhengbin (4):
> drm/amd/display: remove set but not used variable 'old_plane_crtc'
> drm/amd/display: remove set but not used variable 'bp' in
> bios_parser2.c
> drm/amd/display: remove set but not used variable 'bp' in
> bios_parser.c
> dr
On 2019-11-14 11:03 a.m., Alex Deucher wrote:
On Tue, Nov 12, 2019 at 1:04 PM Leo Liu wrote:
By using its own enabling function
Signed-off-by: Leo Liu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 8
drivers/gpu/drm/amd/amdgpu/amdgpu_pm.h | 1 +
drivers/gpu/drm/amd/amdgpu/jpeg_
On Tue, Nov 12, 2019 at 1:04 PM Leo Liu wrote:
>
> By using its own enabling function
>
> Signed-off-by: Leo Liu
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 8
> drivers/gpu/drm/amd/amdgpu/amdgpu_pm.h | 1 +
> drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c | 10 +-
> 3 files chan
ok
Andrey
On 11/14/19 3:12 AM, Christian König wrote:
What about instead of peeking at the job to actually remove it from
ring_mirror_list right there,
Also an interesting idea. We would need to protect the mirror list
with a lock again, but that should be the lesser evil.
Maybe prototype th
Acked-by: Alex Deucher
From: amd-gfx on behalf of Xiaojie Yuan
Sent: Thursday, November 14, 2019 4:13 AM
To: amd-gfx@lists.freedesktop.org
Cc: Long, Gang ; Xiao, Jack ; Yuan,
Xiaojie ; Zhang, Hawking
Subject: [PATCH] drm/amdgpu/gfx10: explicitly wait for cp i
zhengbin (4):
drm/amd/display: remove set but not used variable 'old_plane_crtc'
drm/amd/display: remove set but not used variable 'bp' in
bios_parser2.c
drm/amd/display: remove set but not used variable 'bp' in
bios_parser.c
drm/amd/display: remove set but not used variable 'min_co
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/gpu/drm/amd/display/modules/color/color_gamma.c: In function
build_freesync_hdr:
drivers/gpu/drm/amd/display/modules/color/color_gamma.c:830:20: warning:
variable min_content set but not used [-Wunused-but-set-variable]
It is not used since
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c: In function
bios_get_board_layout_info:
drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c:1826:22: warning: variable
bp set but not used [-Wunused-but-set-variable]
It is introduced by commit 1ee
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c: In function
dm_determine_update_type_for_commit:
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c:6516:36: warning: variable
old_plane_crtc set but not used [-Wunused-but-set-variable]
It is intro
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/gpu/drm/amd/display/dc/bios/bios_parser.c: In function
bios_get_board_layout_info:
drivers/gpu/drm/amd/display/dc/bios/bios_parser.c:2743:22: warning: variable bp
set but not used [-Wunused-but-set-variable]
It is introduced by commit 1eeed
On Thu, Nov 14, 2019 at 1:34 AM Luya Tshimbalanga
wrote:
>
> Hello team,
>
> I filed a bug on the kernel with a broken Raven Ridge firmware resulting
> a blank screen thus unable to boot to either graphical and text mode.
> Could someone investigate the issue on:
>
> https://bugzilla.kernel.org/sh
Acked-by: Alex Deucher
From: amd-gfx on behalf of Evan Quan
Sent: Thursday, November 14, 2019 2:33 AM
To: amd-gfx@lists.freedesktop.org
Cc: Quan, Evan
Subject: [PATCH] drm/amd/powerplay: issue no PPSMC_MSG_GetCurrPkgPwr on
unsupported ASICs
Otherwise, the er
Am 14.11.19 um 11:17 schrieb Changfeng.Zhu:
From: changzhu
MMHUB may lose GPUVM invalidate acknowledge state across power-gating off
cycle when it does invalidation req/ack work.
So we must acquire/release one of the vm_invalidate_eng*_sem around the
invalidation req/ack.
Besides, vm_invalida
From: changzhu
mmhub power gate structure is changed in renoir compared with raven. It
goes through smu_dpm_set_power_gate other than
adev->powerplay.pp_funcs->set_powergating_by_smu in renoir.
So we can realize mmhub power gate in
smu_dpm_set_power_gate ->
smu_powergate_mmhub ->
powergate_mmhub
From: changzhu
MMHUB may lose GPUVM invalidate acknowledge state across power-gating off
cycle when it does invalidation req/ack work.
So we must acquire/release one of the vm_invalidate_eng*_sem around the
invalidation req/ack.
Besides, vm_invalidate_eng*_sem will be read-only after acquire it
Reviewed-by: Kevin Wang
From: amd-gfx on behalf of Evan Quan
Sent: Thursday, November 14, 2019 5:07 PM
To: amd-gfx@lists.freedesktop.org
Cc: Quan, Evan
Subject: [PATCH] drm/amd/powerplay: correct fine grained dpm force level setting
For fine grained dpm, the
50us is not enough to wait for cp ready after gpu reset on some navi asics.
Signed-off-by: Xiaojie Yuan
Suggested-by: Jack Xiao
---
drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 12 +++-
1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
b/d
For fine grained dpm, there is only two levels supported. However
to reflect correctly the current clock frequency, there is an
intermediate level faked. Thus on forcing level setting, we
need to treat level 2 correctly as level 1.
Change-Id: I32f936636f27eb8d8d9002bedd701f2bb0d3060a
Signed-off-by
Adding zeis...@freebsd.org for FreeBSD. I'll try to see if I can ping
some NetBSD/OpenBSD folks too.
On Tuesday, November 12, 2019 4:03 PM, Daniel Vetter
wrote:
> Hi all,
>
> Dave and me chatted about this last week on irc. Essentially we have:
>
> $ git grep SPDX.*GPL -- ':(glob)drivers/gpu/dr
What about instead of peeking at the job to actually remove it from
ring_mirror_list right there,
Also an interesting idea. We would need to protect the mirror list with
a lock again, but that should be the lesser evil.
Maybe prototype that and see if it works or not.
Regards,
Christian.
Am 1
48 matches
Mail list logo
