Hi Gary,
not sure what driver re-initialize feature you are talking about, but
the last time I tried to re-initialize the driver it deadlocks in the
modeset code because of some DC problem.
It's probably a good idea to fix that first, but in general please
explain further what are you workin
From: pding
This lock is used during register accessing in SRIOV guest.
The register accessing could happen both in irq enabled and
irq disabled cases. Always use irq-safe lock.
Signed-off-by: pding
---
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 10 ++
1 file changed, 6 insertions(+),
Hi Christian,
The patch is for driver re- initialize feature, not for driver exit or rmmod.
When the driver initialize failed at some point, the re- initialize feature
will do some little clean and then try to initialize driver again, then it will
re-register some registered debugfs , so it
From: pding
This lock is used during register accessing in SRIOV guest
since KIQ uses general ring submission (amdgpu_ring_commit).
The register accessing could happen both in irq enabled and
irq disabled cases. Always use irq-safe lock.
Signed-off-by: pding
---
drivers/gpu/drm/amd/amdgpu/amdg
Do you have any work actually going into multiple pipes? My understanding
is that opencl will only use one queue at a time (but I'm not really
certain about that).
What you can also check is if the app works correctly when it executed on
pipe0, and if it hangs on pipe 1+. I removed all the locatio
From: Shirish S
This patch fixes static checker warning of
"warn: cast after binop" introduced by
4d3e00dad80a: "drm/amd/display : add high part address calculation for underlay"
Signed-off-by: Shirish S
---
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 ++--
1 file changed, 2 insertio
Then snychronization should have no problem, it maybe relate to multipipe hw
setting issue.
Regards,
David Zhou
From: Andres Rodriguez
Sent: Tuesday, November 7, 2017 2:00:57 AM
To: Zhou, David(ChunMing); amd-gfx list
Cc: Deucher, Alexander
Subject: Re: [PATCH
From: pding
It introduces 900ms latency in exclusive mode which causes failure
of driver loading. Host can resize the BAR before guest staring,
so the resizing is not necessary here.
Signed-off-by: pding
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4
1 file changed, 4 insertions(+)
We must remove the fence callback.
Change-Id: I5d58a3a43b82fefd6c211c4128b0c9187c191e7f
Signed-off-by: Chunming Zhou
---
drivers/gpu/drm/amd/scheduler/gpu_scheduler.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c
b/drivers/gpu/drm/amd/s
Otherwise there could be page protection.
Change-Id: I1f6c81002fb2ba21c17cdc14fdde86579b28374e
Signed-off-by: Chunming Zhou
---
drivers/gpu/drm/amd/scheduler/gpu_scheduler.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c
b/drivers/gpu/drm
It doesn’t work without that write. Meanwhile, this functionality doesn’t
change BAR0 and BAR2 when operating on VF, just as you said it’s not necessary
for VF.
So I think we can skip the whole stuff. Please help to review the patch coming
later.
—
Sincerely Yours,
Pixel
On 06/11/2017,
Signed-off-by: Ernst Sjöstrand
---
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 9 +
1 file changed, 9 insertions(+)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index e6bfa9f30900..2301589e4cc3 100644
--- a/dr
More "warn: inconsistent indenting" fixes from smatch.
Signed-off-by: Ernst Sjöstrand
---
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 10 +-
.../gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c | 2 +-
2 files changed, 6 insertions(+), 6 deletions(-)
diff
From smatch:
error: we previously assumed X could be null
Signed-off-by: Ernst Sjöstrand
---
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 15 +++
1 file changed, 11 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
b/drivers/gpu/drm
I saw the previous smatch fixes and thought I would try to fix a few.
Since the other patches were applied to agd5f/amd-staging-drm-next
I continued there.
___
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listi
> -Original Message-
> From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Felix Kuehling
> Sent: Monday, November 06, 2017 2:52 PM
> To: amd-gfx@lists.freedesktop.org; oded.gab...@gmail.com
> Cc: Kuehling, Felix
> Subject: [PATCH 2/2] drm/amdkfd: Use order_base_2 to
On Sat, Nov 4, 2017 at 8:21 AM, Nicolas Iooss
wrote:
> On Sun, Sep 3, 2017 at 2:00 PM, Nicolas Iooss
> wrote:
>>
>> Function vega10_apply_state_adjust_rules() only initializes
>> stable_pstate_sclk_dpm_percentage when
>> data->registry_data.stable_pstate_sclk_dpm_percentage is not between 1
>> an
On Mon, Nov 6, 2017 at 6:59 AM, Christian König
wrote:
> Am 06.11.2017 um 12:44 schrieb Dan Carpenter:
>>
>> We assign "v_init = asic_blank_start;" a few lines earlier so there is
>> no need to do it again inside the if statements. Also "v_init" is
>> unsigned so it can't be less than zero.
>>
>>
On Mon, Nov 6, 2017 at 6:51 AM, Christian König
wrote:
> Am 06.11.2017 um 12:43 schrieb Dan Carpenter:
>>
>> backlight_device_register() never returns NULL, it returns error
>> pointers on error so the check here is wrong.
>>
>> Signed-off-by: Dan Carpenter
>
>
> Acked-by: Christian König
Appli
On Mon, Nov 6, 2017 at 3:44 AM, Christian König
wrote:
> Am 06.11.2017 um 08:07 schrieb Dan Carpenter:
>>
>> Static analysis tools get annoyed that we don't indent this if
>> statement. Actually, the if statement isn't required because kfree()
>> can handle NULL pointers just fine. The DCE110STR
Don't use sizeof(uint32_t) or similar types for hardware or firmware
DWORD size. The hardware and firmware don't care about Linux types.
Signed-off-by: Felix Kuehling
---
drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c | 14 +-
drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c| 2 +
Replace (ffs(size) - 1) with order_base_2(size) as a more straight
forward way to get log2 of buffer sizes.
Signed-off-by: Felix Kuehling
---
drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c | 6 +++---
drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c | 6 +++---
2 files changed, 6 insertions(+)
Am 06.11.2017 um 20:17 schrieb Dave Airlie:
From: Dave Airlie
smatch reported:
drivers/gpu/drm/amd/amdgpu/../display/dc/bios/dce80/command_table_helper_dce80.c:351:71:
warning: non-ANSI function declaration of function
'dal_cmd_tbl_helper_dce80_get_table'
drivers/gpu/drm/amd/amdgpu/../display
From: Dave Airlie
This fixes all the current smatch:
warn: inconsistent indenting
Signed-off-by: Dave Airlie
---
drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c | 2 +-
.../gpu/drm/amd/display/dc/bios/command_table2.c | 18 ++---
drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
From: Dave Airlie
Reported-by smatch:
drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc.c:966
dc_commit_planes_to_stream() error: potential null dereference 'flip_addr'.
(kcalloc returns null)
drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc.c:968
dc_commit_planes_to_stream() error: potential
From: Dave Airlie
smatch reported:
drivers/gpu/drm/amd/amdgpu/../display/dc/bios/dce80/command_table_helper_dce80.c:351:71:
warning: non-ANSI function declaration of function
'dal_cmd_tbl_helper_dce80_get_table'
drivers/gpu/drm/amd/amdgpu/../display/dc/bios/dce110/command_table_helper_dce110.c:
Am 06.11.2017 um 19:39 schrieb Tom St Denis:
On 06/11/17 01:34 PM, Christian König wrote:
Am 06.11.2017 um 19:28 schrieb Tom St Denis:
On 06/11/17 05:01 AM, Christian König wrote:
Am 04.11.2017 um 18:15 schrieb Tom St Denis:
Signed-off-by: Tom St Denis
Still not perfect, but good enough fo
On 06/11/17 01:34 PM, Christian König wrote:
Am 06.11.2017 um 19:28 schrieb Tom St Denis:
On 06/11/17 05:01 AM, Christian König wrote:
Am 04.11.2017 um 18:15 schrieb Tom St Denis:
Signed-off-by: Tom St Denis
Still not perfect, but good enough for now. Patch is Tested-by:
Christian König .
Am 06.11.2017 um 19:28 schrieb Tom St Denis:
On 06/11/17 05:01 AM, Christian König wrote:
Am 04.11.2017 um 18:15 schrieb Tom St Denis:
Signed-off-by: Tom St Denis
Still not perfect, but good enough for now. Patch is Tested-by:
Christian König .
I think you need to rework the VM walking a
On 06/11/17 05:01 AM, Christian König wrote:
Am 04.11.2017 um 18:15 schrieb Tom St Denis:
Signed-off-by: Tom St Denis
Still not perfect, but good enough for now. Patch is Tested-by:
Christian König .
I think you need to rework the VM walking a bit, cause we need to
support the T bit as we
Additionally, we should also just fix this ratelimit macro anyway
since it's intended purpose is not to print anything when debugging isn't
enabled. What do you think Alex?
Well, at least I think that the described behavior of the macro is a bug
which should be fixed.
But I think that is indepe
Sorry my mail client seems to have blown up. My reply got cut off,
here is the full version:
On 2017-11-06 01:49 AM, Chunming Zhou wrote:
> Hi Andres,
>
Hi David,
> With your this patch, OCLperf hung.
Is this on all ASICs or just a specific one?
>
> Could you explain more?
>
> If I am correctl
On Mon, Nov 6, 2017 at 12:20 PM, Randy Dunlap wrote:
> On 11/05/2017 11:30 PM, Stephen Rothwell wrote:
>> Hi all,
>>
>> Changes since 20171103:
>>
>
> on i386, when CONFIG_MODULES is not set:
>
> CC drivers/gpu/drm/amd/amdgpu/amdgpu_virt.o
> In file included from ../arch/x86/include/asm/ato
On 11/05/2017 11:30 PM, Stephen Rothwell wrote:
> Hi all,
>
> Changes since 20171103:
>
on i386, when CONFIG_MODULES is not set:
CC drivers/gpu/drm/amd/amdgpu/amdgpu_virt.o
In file included from ../arch/x86/include/asm/atomic.h:5:0,
from ../include/linux/atomic.h:5,
On Fri, Nov 03, 2017 at 04:35:45PM -0400, Alex Deucher wrote:
> Minimum time required between power On of codec and read
> of RT5645_VENDOR_ID2 is 400msec. We should wait and attempt
> before erroring out.
So the description says we have to wait 400ms before attempting a
read...
> BUG=b:66978383
Quoting Christian König (2017-10-30 14:59:04)
> From: Christian König
>
> The amdgpu issue to also need signaled fences in the reservation objects
> should
> be fixed by now.
>
> Optimize the handling by replacing a signaled fence when adding a new
> shared one.
>
> Signed-off-by: Christian Kö
Quoting Christian König (2017-10-30 14:59:03)
> From: Christian König
>
> The amdgpu issue to also need signaled fences in the reservation objects
> should be fixed by now.
>
> Optimize the list by keeping only the not signaled yet fences around.
>
> v2: temporary put the signaled fences at the
The main reason I added this was because the radeon driver's hotplugging paths
for DP do a ton of unnessecary probing, and because the driver usually also
checks all connectors every time there's a hotplug (there isn't much of a good
reason for this, it's just an old driver) it's not at all difficu
On Fri, Nov 03, 2017 at 04:35:43PM -0400, Alex Deucher wrote:
> Signed-off-by: Vijendar Mukunda
> Signed-off-by: Akshu Agrawal
> Reviewed-on: https://chromium-review.googlesource.com/659699
> Commit-Ready: Akshu Agrawal
> Tested-by: Akshu Agrawal
> Reviewed-by: Jason Clinton
> Reviewed-on: ht
[+Oded]
Hi Pixel,
Thanks, this looks a lot cleaner. The series is Reviewed-by: Felix
Kuehling . Oded, are you OK with this as well?
Regards,
Felix
On 2017-11-06 03:18 AM, Pixel Ding wrote:
> Hi Felix,
>
> Please review.
>
> [PATCH 1/2] drm/amdkfd: initialise kfd inside amdgpu_device_init
> A
On 06/11/17 09:57 AM, Vladimir Klebanov wrote:
> On Sun, Nov 5, 2017 at 11:53 PM, Bridgman, John wrote:
>
>> The only problems I see in the xorg log are the following lines at the end,
>> but not sure if they are serious or not.
>>
>> [ 205.542] (WW) RADEON(0): flip queue failed: Invalid argum
Hi John,
On Sun, Nov 5, 2017 at 11:53 PM, Bridgman, John wrote:
> For clarity, are you saying that when you go back to whatever distro you had
> installed on the machine previously it is still not booting correctly ?
Indeed. Even worse, nothing boots correctly as the machine does not
get to/co
Hi all,
commit 92c177b7947d9c889ea7b024871445015ea74221
Author: Lyude
Date: Wed Feb 22 16:34:53 2017 -0500
drm/radeon/dp_auxch: Ratelimit aux transfer debug messages
does more harm than good in my opinion. Since this commit, I see
several occurrences of the following message in my kernel l
Should fix the indentation now when a PDE entry is actually a PTE.
Signed-off-by: Tom St Denis
---
src/lib/read_vram.c | 28 ++--
1 file changed, 14 insertions(+), 14 deletions(-)
diff --git a/src/lib/read_vram.c b/src/lib/read_vram.c
index 51823d71021e..89d55ff1bef6 100
Am 06.11.2017 um 12:44 schrieb Dan Carpenter:
We assign "v_init = asic_blank_start;" a few lines earlier so there is
no need to do it again inside the if statements. Also "v_init" is
unsigned so it can't be less than zero.
Signed-off-by: Dan Carpenter
Acked-by: Christian König
diff --git
Am 06.11.2017 um 12:43 schrieb Dan Carpenter:
backlight_device_register() never returns NULL, it returns error
pointers on error so the check here is wrong.
Signed-off-by: Dan Carpenter
Acked-by: Christian König
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
b/drivers/gpu
We assign "v_init = asic_blank_start;" a few lines earlier so there is
no need to do it again inside the if statements. Also "v_init" is
unsigned so it can't be less than zero.
Signed-off-by: Dan Carpenter
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_timing_generator.c
b/drivers/gpu
backlight_device_register() never returns NULL, it returns error
pointers on error so the check here is wrong.
Signed-off-by: Dan Carpenter
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 33a15a1d818c..75f9383f5b9b 100644
Linux-next was offline for the last month and the AMD drm driver went
through major changes. Anyway, I'm a bit overwhelmed by the number of
warnings and I'm not going to be able to go through them all so I'm just
sending them to you unfiltered.
Part of the problem is that I'm not running the rele
remove debugfs file in amdgpu_device_finish
Signed-off-by: Gary Sun
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h|1 +
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 18 ++
2 files changed, 19 insertions(+), 0 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
Am 04.11.2017 um 18:15 schrieb Tom St Denis:
Signed-off-by: Tom St Denis
Still not perfect, but good enough for now. Patch is Tested-by:
Christian König .
I think you need to rework the VM walking a bit, cause we need to
support the T bit as well in the future and your code make a few
ass
What benefits will the larger VRAM bar bring in,
Well the obvious one that the CPU can access all of VRAM. There are some
games/workloads which rely quite a bit on this and it simplifies MM
largely when you don't need to shuffle buffers around for CPU access.
or are there new features relying
Hello Shirish S,
The patch 4d3e00dad80a: "drm/amd/display : add high part address
calculation for underlay" from Oct 19, 2017, leads to the following
static checker warning:
drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:1838
fill_plane_attributes_from_fb()
warn: cas
Am 06.11.2017 um 10:33 schrieb Evan Quan:
Change-Id: I437e3e08cd48943de277c5d3eefdbaf21fd6a489
Signed-off-by: Evan Quan
Tested-and-Acked-by: Christian König
---
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 4
1 file changed, 4 insertions(+)
diff --git a/drivers/gpu/drm/amd/
What benefits will the larger VRAM bar bring in, or are there new features
relying on larger VRAM bar under development? SRIOV wants to leverage this sort
of optimization too.
I think it’s better to try something which can keep this change for SRIOV and
don’t introduce too much latency.
—
Sin
Change-Id: I437e3e08cd48943de277c5d3eefdbaf21fd6a489
Signed-off-by: Evan Quan
---
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 4
1 file changed, 4 insertions(+)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
ind
Am 06.11.2017 um 10:20 schrieb Gary Sun:
remove debugfs file in amdgpu_device_finish
NAK, the debugfs files are removed automatically by drm_debugfs_cleanup().
So that patch is unnecessary.
Regards,
Christian.
Signed-off-by: Gary Sun
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h|1
Am 06.11.2017 um 08:21 schrieb Ding, Pixel:
Hi Christian,
The latest driver fails to load on SRIOV VF with xen hypervisor driver.
+int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev)
+{
+ u64 space_needed = roundup_pow_of_two(adev->mc.real_vram_size);
+ u32 rbar_size = order_bas
Am 06.11.2017 um 08:07 schrieb Dan Carpenter:
Static analysis tools get annoyed that we don't indent this if
statement. Actually, the if statement isn't required because kfree()
can handle NULL pointers just fine. The DCE110STRENC_FROM_STRENC()
macro is a wrapper around container_of() but it's
Hello Harry Wentland,
The patch 4562236b3bc0: "drm/amd/dc: Add dc display driver (v2)" from
Sep 12, 2017, leads to the following static checker warning:
drivers/gpu/drm/amd/amdgpu/../display/dc/dm_services.h:117
get_reg_field_value_ex()
warn: mask and shift to zero
drivers/gpu/d
Hi Felix,
Please review.
[PATCH 1/2] drm/amdkfd: initialise kfd inside amdgpu_device_init
As you suggested, move kfd init/fini inside amdgpu_device_init.
Other changes for KFD interfaces are dropped.
[PATCH 2/2] drm/amdgpu: release exclusive mode after hw_init
__
From: pding
Signed-off-by: pding
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4
drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c| 3 ---
2 files changed, 4 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
i
From: pding
Also finalize kfd inside amdgpu_device_fini. kfd device_init needs
SRIOV exclusive accessing. Try to gather exclusive accessing to
reduce time consuming.
Signed-off-by: pding
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4
drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c| 5 ---
63 matches
Mail list logo
