Reviewed-by: Evan Quan
-Original Message-
From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Rex
Zhu
Sent: Friday, March 16, 2018 12:49 PM
To: amd-gfx@lists.freedesktop.org
Cc: Zhu, Rex
Subject: [PATCH] drm/amdgpu: Delete dead code when early init
Change-Id: I464
issue:
Game F1 performance drops 13% when per vm bo is enabled.
root cause:
if some BOs are fallback to allowed domain, they will never be validated if no
eviction happens,
that means they always exist in allowed domain.
Fix:
maintain a per vm allowed domain BOs list, then try to validated them
Change-Id: I2cf672ad36b8b4cc1a6b2e704f786bf6a155d9ce
Signed-off-by: Chunming Zhou
---
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c| 5 -
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 12 ++--
2 files changed, 10 insertions(+), 7 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amd
Rex Zhu (6):
drm/amd/pp: Remove dead functions in vega10_smumgr.c
drm/amd/pp: Mark bunches of functins in vega10_smumgr.c static
drm/amd/pp: Move functions to smu backend table for vega10
drm/amd/pp: Clean up header file for Vega10
drm/amd/pp: Delete get_xclk function in powerplay
drm/
use smc_table_manager function to copy/save tables to/from smu.
Change-Id: I94a43436dec9a71411fe5e8c2b888b306c9bbacd
Signed-off-by: Rex Zhu
---
.../gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c | 38 --
.../gpu/drm/amd/powerplay/smumgr/vega10_smumgr.h | 2 --
2 files chan
Change-Id: I28b53e6312eaf995eb5091cf66d659a3022b3e84
---
drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c | 10 +-
drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.h | 3 ---
2 files changed, 5 insertions(+), 8 deletions(-)
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_s
Signed-off-by: Rex Zhu
Change-Id: I12f61309036edbeb8d0f4c9a557778c7a227a58b
---
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 36 -
.../gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c | 7 ++--
drivers/gpu/drm/amd/powerplay/inc/smumgr.h | 8
drivers/gpu/drm/a
Change-Id: If0c33c6277308d73d4ffa5134b67b1e22a90fe53
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 11 ++-
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h | 2 ++
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c | 1 -
drivers/gpu/drm/a
use asic's callback function get_xclk
Change-Id: Id4ed442e9832ab9d680dc709106a5b7cb67e94c4
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c| 5 +---
drivers/gpu/drm/amd/include/cgs_common.h | 1 -
drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 27 +
Removes unneeded void * casts for the following pointers:
hwmgr->backend
hwmgr->smu_backend
Change-Id: I9887d39c27dadf953ab32e7ecd7e5b90c73f9a5f
Signed-off-by: Rex Zhu
---
drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 184 -
.../gpu/drm/amd/powerplay/hwmgr/vega10_power
Am 16.03.2018 um 10:56 schrieb Chunming Zhou:
Change-Id: I2cf672ad36b8b4cc1a6b2e704f786bf6a155d9ce
Signed-off-by: Chunming Zhou
---
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c| 5 -
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 12 ++--
2 files changed, 10 insertions(+), 7 deleti
There seems one unnecessary change in vega10_fan_ctrl_set_fan_speed_rpm().
Other than that, the series is
Reviewed-by: Evan Quan
-Original Message-
From: amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.org] On Behalf Of Rex
Zhu
Sent: Friday, March 16, 2018 5:40 PM
To: amd-gfx@lists.fre
Yes， Thanks for pointing out this error. that is used for testing the ref clk
value on vega10.
Best Regards
Rex
From: Quan, Evan
Sent: Friday, March 16, 2018 7:24:47 PM
To: Zhu, Rex; amd-gfx@lists.freedesktop.org
Cc: Zhu, Rex
Subject: RE: [PATCH 5/6] drm
Hi everybody,
since I've got positive feedback from Daniel I continued working on this
approach.
A few issues are still open:
1. Daniel suggested that I make the invalidate_mappings callback a parameter of
dma_buf_attach().
This approach unfortunately won't work because when the attachment is
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
Signed-off-by: Christian König
---
Make sure the transfered BO is never destroy before the transfer BO.
Signed-off-by: Christian König
---
drivers/gpu/drm/ttm/ttm_bo_util.c | 50 +++
1 file changed, 30 insertions(+), 20 deletions(-)
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c
b/drivers/gpu
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| 1 -
drivers/gpu/drm
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 b/drivers/gpu/drm/ttm/ttm_bo.c
index 98e06f8bf23
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 se
Quoting Christian König (2018-03-16 13:20:45)
> @@ -326,6 +338,29 @@ struct dma_buf_attachment {
> struct device *dev;
> struct list_head node;
> void *priv;
> +
> + /**
> +* @invalidate_mappings:
> +*
> +* Optional callback provided by the impo
NP, makes sense what you did.
Andrey
On 03/16/2018 01:44 AM, Zhu, Rex wrote:
Hi Andrey,
Sorry for the confusion.
I just need to push a bunch of patches that conflict with your patch.
So apply your patch first.
Best Regards
Rex
*From:*amd-gfx [mailto:amd-gfx-boun...@lists.freedesktop.o
Am 16.03.2018 um 14:51 schrieb Chris Wilson:
Quoting Christian König (2018-03-16 13:20:45)
@@ -326,6 +338,29 @@ struct dma_buf_attachment {
struct device *dev;
struct list_head node;
void *priv;
+
+ /**
+* @invalidate_mappings:
+*
+* Optio
Need to be able to query the VCN firmware version from
userspace to determine supported features, etc.
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 12
include/uapi/drm/amdgpu_drm.h | 2 ++
2 files changed, 14 insertions(+)
diff --git a/drive
Since GPU reset is not enabled yet anyway, a timeout will just print a message,
can we just change amdgpu_lockup_timeout to MAX_SCHEDULE_TIMEOUT until we
enable GPU reset?
Alex
From: Evan Quan
Sent: Friday, March 16, 2018 12:52:32 AM
To: amd-gfx@lists.freedesk
On 2018-03-16 05:14 PM, Deucher, Alexander wrote:
> Since GPU reset is not enabled yet anyway, a timeout will just print a
> message, can we just change amdgpu_lockup_timeout to
> MAX_SCHEDULE_TIMEOUT until we enable GPU reset?
Wouldn't that be rather surprising for somebody who sets
amdgpu.gpu_re
How about something like this:
/* Disable timeout when GPU reset is disabled to avoid confusing timeout
messages in the kernel log */
If (amdgpu_gpu_recovery == 0 || amdgpu_gpu_recovery == -1)
amdgpu_lockup_timeout = MAX_SCHEDULE_TIMEOUT;
Alex
From: Miche
On Fri, Mar 16, 2018 at 5:39 AM, Rex Zhu wrote:
> Removes unneeded void * casts for the following pointers:
> hwmgr->backend
> hwmgr->smu_backend
>
> Change-Id: I9887d39c27dadf953ab32e7ecd7e5b90c73f9a5f
> Signed-off-by: Rex Zhu
With Evan's comment on patch 5 addressed, the series is:
Reviewed-by
Add kernel doc for the functions in amdgpu_device.c
Signed-off-by: Alex Deucher
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 439 -
1 file changed, 427 insertions(+), 12 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
b/drivers/gpu/drm/amd/a
Hi Dave,
Updates for 4.17. Highlights:
- Continued cleanup and restructuring of powerplay
- Fetch VRAM type from vbios rather than hardcoding for SOC15 asics
- Allow ttm to drop its backing store when drivers don't need it
- DC bandwidth calc updates
- Enable DC backlight control pre-DCE11 asics
29 matches
Mail list logo
