-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_drv.h | 6 --
drivers/gpu/drm/i915/i915_perf.c | 16 +---
2 files changed, 1 insertion(+), 21 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index db7480831e52..75607450ba00 100644
to i915_perf_stream (Lionel)
v6: Add missing struct declarations in i915_perf.h
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson (v4)
---
drivers/gpu/drm/i915/gt/intel_gpu_commands.h | 24 ++
drivers/gpu/drm/i915/gt/intel_gt_types.h | 5 +
drivers/gpu/drm/i915/i915_debugfs.c
An upcoming change needs not to be interrupted.
Signed-off-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_active.c | 4 +++-
drivers/gpu/drm/i915/i915_active.h | 5 ++---
2 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_active.c
b/drivers/gpu/drm
,
Lionel Landwerlin (12):
drm/i915: introduce a mechanism to extend execbuf2
drm/i915: add syncobj timeline support
drm/i915/perf: drop list of streams
drm/i915/perf: store the associated engine of a stream
drm/i915/perf: introduce a versioning of the i915-perf uapi
drm/i915/perf: allow
(Lionel)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson (v4)
---
drivers/gpu/drm/i915/gt/intel_gpu_commands.h | 1 +
drivers/gpu/drm/i915/i915_drv.h | 19 +-
drivers/gpu/drm/i915/i915_perf.c | 270 ---
drivers/gpu/drm/i915/i915_perf.h
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
v3: Rename I915_EXEC_EXT -> I915_EXEC_USE_EXTENSIONS (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
Reporting this version will help application figure out what level of
the support the running kernel provides.
v2: Add i915_perf_ioctl_version() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_getparam.c | 4
drivers/gpu/drm/i915
-----
From: Lionel Landwerlin
Sent: Wednesday, August 28, 2019 10:33 PM
To: intel-gfx@lists.freedesktop.org
Cc: Lionel Landwerlin ; Zhou, David(ChunMing)
; Koenig, Christian ; Jason Ekstrand
Subject: [PATCH v11 01/10] drm/syncobj: add sideband payload
The Vulkan timeline semaphores allow
to userspace (Chris)
v6: Drop lock when copying config list to userspace (Chris)
Fix deadlock when calling i915_perf_get_oa_config() under
perf.metrics_lock (Lionel)
Add i915_oa_config_get() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915
to i915_perf_stream (Lionel)
v6: Add missing struct declarations in i915_perf.h
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson (v4)
---
drivers/gpu/drm/i915/gt/intel_gpu_commands.h | 24 ++
drivers/gpu/drm/i915/gt/intel_gt_types.h | 5 +
drivers/gpu/drm/i915/i915_debugfs.c
We'll use this information later to verify that a client trying to
reconfigure the stream does so on the right engine.
Signed-off-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_drv.h | 5 +
drivers/gpu/drm/i915/i915_perf.c | 7 +++
2 files changed, 12 insertions(+)
diff --git
Reporting this version will help application figure out what level of
the support the running kernel provides.
v2: Add i915_perf_ioctl_version() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_getparam.c | 4
drivers/gpu/drm/i915
-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_drv.h | 6 --
drivers/gpu/drm/i915/i915_perf.c | 16 +---
2 files changed, 1 insertion(+), 21 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index db7480831e52..75607450ba00 100644
)
v8: Remove active request from engine (Lionel)
v9: Move fetching OA configuration pass engine pinning (Lionel)
Lock VMA before moving to active (Chris)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 146 +-
drivers/gpu/drm/i915
(Lionel)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson (v4)
---
drivers/gpu/drm/i915/gt/intel_gpu_commands.h | 1 +
drivers/gpu/drm/i915/i915_drv.h | 19 +-
drivers/gpu/drm/i915/i915_perf.c | 270 ---
drivers/gpu/drm/i915/i915_perf.h
v4: Create i915 request in emit_oa_config() to avoid deadlocks (Lionel)
v5: Move locking to the stream (Lionel)
v6: Move active reconfiguration request into i915_perf_stream (Lionel)
v7: Pin VMA outside request creation (Chris)
Lock VMA before move to active (Chris)
Signed-off-by: Lionel
priorities to avoid preemption rather than the HW mechanism
v4: Just modify the port priority reporting function
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 8 +
drivers/gpu/drm/i915/i915_drv.h | 8 +
drivers/gpu
://lists.freedesktop.org/archives/dri-devel/2019-August/229287.html
drop the ordering check from v8 (Lionel)
v10: Set first extension enum item to 1 (Jason)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 307 ++
drivers/gpu/drm/i915/i915_drv.c
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
v3: Rename I915_EXEC_EXT -> I915_EXEC_USE_EXTENSIONS (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
Hi all,
Some missing locks of the VMA that Chris spotted in the review. This
also simplifies a bit the execbuf code, again as recommended by Chris.
Cheers,
Lionel Landwerlin (11):
drm/i915: introduce a mechanism to extend execbuf2
drm/i915: add syncobj timeline support
drm/i915/perf: drop
On 30/08/2019 18:48, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-30 15:47:23)
err_unpin:
- __i915_vma_unpin(vma);
+ mutex_lock(>drm.struct_mutex);
+ i915_vma_unpin_and_release(, 0);
+ mutex_unlock(>drm.struct_mutex);
Strangely unpin_and_release() d
On 30/08/2019 18:48, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-30 15:47:23)
err_unpin:
- __i915_vma_unpin(vma);
+ mutex_lock(>drm.struct_mutex);
+ i915_vma_unpin_and_release(, 0);
+ mutex_unlock(>drm.struct_mutex);
Strangely unpin_and_release() d
On 30/08/2019 18:32, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-30 15:47:24)
+static int
+get_execbuf_oa_config(struct i915_execbuffer *eb)
+{
+ int err = 0;
+
+ eb->perf_file = NULL;
+ eb->oa_config = NULL;
+ eb->oa_vma = NULL;
+ eb->
)
v8: Remove active request from engine (Lionel)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 151 +-
drivers/gpu/drm/i915/i915_getparam.c | 4 +
include/uapi/drm/i915_drm.h | 39 +
3 files changed, 193
://lists.freedesktop.org/archives/dri-devel/2019-August/229287.html
drop the ordering check from v8 (Lionel)
v10: Set first extension enum item to 1 (Jason)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 307 ++
drivers/gpu/drm/i915/i915_drv.c
(Lionel)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson (v4)
---
drivers/gpu/drm/i915/gt/intel_gpu_commands.h | 1 +
drivers/gpu/drm/i915/i915_drv.h | 19 +-
drivers/gpu/drm/i915/i915_perf.c | 270 ---
drivers/gpu/drm/i915/i915_perf.h
We'll use this information later to verify that a client trying to
reconfigure the stream does so on the right engine.
Signed-off-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_drv.h | 5 +
drivers/gpu/drm/i915/i915_perf.c | 7 +++
2 files changed, 12 insertions(+)
diff --git
to userspace (Chris)
v6: Drop lock when copying config list to userspace (Chris)
Fix deadlock when calling i915_perf_get_oa_config() under
perf.metrics_lock (Lionel)
Add i915_oa_config_get() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915
-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_drv.h | 6 --
drivers/gpu/drm/i915/i915_perf.c | 16 +---
2 files changed, 1 insertion(+), 21 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index db7480831e52..75607450ba00 100644
priorities to avoid preemption rather than the HW mechanism
v4: Just modify the port priority reporting function
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 8 +
drivers/gpu/drm/i915/i915_drv.h | 8 +
drivers/gpu
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
v3: Rename I915_EXEC_EXT -> I915_EXEC_USE_EXTENSIONS (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
Hi all,
This moves locking a bit, hopefully in the right direction so that it
helps dropping the use of struct_mutex.
Cheers,
Lionel Landwerlin (11):
drm/i915: introduce a mechanism to extend execbuf2
drm/i915: add syncobj timeline support
drm/i915/perf: drop list of streams
drm/i915
Reporting this version will help application figure out what level of
the support the running kernel provides.
v2: Add i915_perf_ioctl_version() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_getparam.c | 4
drivers/gpu/drm/i915
to i915_perf_stream (Lionel)
v6: Add missing struct declarations in i915_perf.h
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson (v4)
---
drivers/gpu/drm/i915/gt/intel_gpu_commands.h | 24 ++
drivers/gpu/drm/i915/gt/intel_gt_types.h | 5 +
drivers/gpu/drm/i915/i915_debugfs.c
v4: Create i915 request in emit_oa_config() to avoid deadlocks (Lionel)
v5: Move locking to the stream (Lionel)
v6: Move active reconfiguration request into i915_perf_stream (Lionel)
Signed-off-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_drv.h | 13 ++-
drivers/gpu/drm/i915
ialise updates to the OA config
itself.
Signed-off-by: Chris Wilson
Cc: Lionel Landwerlin
Thanks, that's helpful to remember why it works :)
Reviewed-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_perf.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/gpu/drm/i915/i915_pe
FYI, apart from a couple of pre-existing issues, the Mesa series that
makes use of this i915 change is reviewed :
https://gitlab.freedesktop.org/mesa/mesa/merge_requests/1525
Cheers,
-Lionel
On 22/08/2019 15:12, Lionel Landwerlin wrote:
Hi,
Just fiddling with enums to avoid default 0
On 29/08/2019 09:45, Lionel Landwerlin wrote:
On 28/08/2019 22:41, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-28 15:33:26)
+ rq = i915_request_create(i915->engine[RCS0]->kernel_context);
+ if (IS_ERR(rq))
+ return PTR_ERR(rq);
+
+
On 28/08/2019 22:41, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-28 15:33:26)
+ rq = i915_request_create(i915->engine[RCS0]->kernel_context);
+ if (IS_ERR(rq))
+ return PTR_ERR(rq);
+
+ err = i915_active_request_set(>engine[RCS0]->la
tree, please drop us a note to help
improve the system]
url:
https://github.com/0day-ci/linux/commits/Lionel-Landwerlin/drm-i915-Vulkan-performance-query-support/20190829-020841
base: git://anongit.freedesktop.org/drm-intel for-linux-next
config: i386-defconfig (attached as .config)
compiler
I included this patch into it.
-Lionel
-David
-Original Message-----
From: Lionel Landwerlin
Sent: Wednesday, August 28, 2019 10:33 PM
To: intel-gfx@lists.freedesktop.org
Cc: Lionel Landwerlin ; Zhou, David(ChunMing)
; Koenig, Christian ; Jason Ekstrand
Subject: [PATCH v11 01/10] drm/sy
On 29/08/2019 00:02, Rodrigo Vivi wrote:
From: Lionel Landwerlin
commit 6a67a20366f894c172734f28c5646bdbe48a46e3 upstream.
subslice_mask is an array indexed by slice, not subslice.
Signed-off-by: Lionel Landwerlin
Fixes: 8cc7669355136f ("drm/i915: store all subslice masks")
Bugzi
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
v3: Rename I915_EXEC_EXT -> I915_EXEC_USE_EXTENSIONS (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
v4: Create i915 request in emit_oa_config() to avoid deadlocks (Lionel)
Signed-off-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_drv.h | 12 +++
drivers/gpu/drm/i915/i915_perf.c | 146 +++
2 files changed, 104 insertions(+), 54 deletions(-)
diff --git
to userspace (Chris)
v6: Drop lock when copying config list to userspace (Chris)
Fix deadlock when calling i915_perf_get_oa_config() under
perf.metrics_lock (Lionel)
Add i915_oa_config_get() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915
)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 133 +-
drivers/gpu/drm/i915/gt/intel_engine_cs.c | 4 +
drivers/gpu/drm/i915/gt/intel_engine_types.h | 9 ++
drivers/gpu/drm/i915/gt/intel_lrc.c | 4 +-
drivers/gpu/drm/i915
://lists.freedesktop.org/archives/dri-devel/2019-August/229287.html
drop the ordering check from v8 (Lionel)
v10: Set first extension enum item to 1 (Jason)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 307 ++
drivers/gpu/drm/i915/i915_drv.c
mapping of OA config (Chris)
v3: Properly deal with the perf_metric lock (Chris/Lionel)
v4: Fix oa config unref/put when not found (Lionel)
v5: Allocate BOs for configurations on the stream instead of globally
(Lionel)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson (v4)
---
drivers
to i915_perf_stream (Lionel)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson (v4)
---
drivers/gpu/drm/i915/gt/intel_gpu_commands.h | 24 ++
drivers/gpu/drm/i915/gt/intel_gt_types.h | 5 +
drivers/gpu/drm/i915/i915_debugfs.c | 31 +++
drivers/gpu/drm/i915/i915_drv.h
priorities to avoid preemption rather than the HW mechanism
v4: Just modify the port priority reporting function
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 8 +
drivers/gpu/drm/i915/i915_drv.h | 8 +
drivers/gpu
Reporting this version will help application figure out what level of
the support the running kernel provides.
v2: Add i915_perf_ioctl_version() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_getparam.c | 4
drivers/gpu/drm/i915
, this also includes 3
patches (the first 3) from the timeline semaphore work. You can safely
ignore them, they're getting reviewed on a different series. It's just
unfortunate they interact because of uAPI changes.
Cheers,
Lionel Landwerlin (10):
drm/syncobj: add sideband payload
drm/i915: introduce
of blattant mistakes
Store payload atomically (Chris)
v6: Only touch atomic value once (Jason)
Signed-off-by: Lionel Landwerlin
Reviewed-by: David Zhou (v5)
Cc: Christian Koenig
Cc: Jason Ekstrand
Cc: David(ChunMing) Zhou
---
drivers/gpu/drm/drm_internal.h | 2 ++
drivers/gpu/drm
On 23/08/2019 10:20, Lucas De Marchi wrote:
From: Michel Thierry
Compared to Icelake, Tigerlake's MAX_CONTEXT_HW_ID is smaller by one, but
since we just use the upper 32 bits of the lrc_desc, it's guaranteed OA
will use the correct one.
Cc: Lionel Landwerlin
Signed-off-by: Michel Thierry
On 22/08/2019 12:43, Joonas Lahtinen wrote:
Quoting Lucas De Marchi (2019-08-16 11:04:57)
From: Lionel Landwerlin
The way our hardware is designed doesn't seem to let us use the
MI_RECORD_PERF_COUNT command without setting up a circular buffer.
In the case where the user didn't request OA
://lists.freedesktop.org/archives/dri-devel/2019-August/229287.html
drop the ordering check from v8 (Lionel)
v10: Set first extension enum item to 1 (Jason)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 307 ++
drivers/gpu/drm/i915/i915_drv.c
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
v3: Rename I915_EXEC_EXT -> I915_EXEC_USE_EXTENSIONS (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
Hi,
Just fiddling with enums to avoid default 0 values as recommended by
Jason.
Cheers,
Lionel Landwerlin (2):
drm/i915: introduce a mechanism to extend execbuf2
drm/i915: add syncobj timeline support
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 346 +++---
drivers/gpu/drm
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
v3: Rename I915_EXEC_EXT -> I915_EXEC_USE_EXTENSIONS (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
Hi,
Just fiddling with enums to avoid default 0 values as recommended by
Jason.
Cheers,
Lionel Landwerlin (2):
drm/i915: introduce a mechanism to extend execbuf2
drm/i915: add syncobj timeline support
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 346 +++---
drivers/gpu/drm
://lists.freedesktop.org/archives/dri-devel/2019-August/229287.html
drop the ordering check from v8 (Lionel)
v10: Set first extension enum item to 1 (Jason)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 307 ++
drivers/gpu/drm/i915/i915_drv.c
://lists.freedesktop.org/archives/dri-devel/2019-August/229287.html
drop the ordering check from v8 (Lionel)
v10: Set first extension enum item to 1 (Jason)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 307 ++
drivers/gpu/drm/i915/i915_drv.c
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
v3: Rename I915_EXEC_EXT -> I915_EXEC_USE_EXTENSIONS (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
Hi,
Just fiddling with enums to avoid default 0 values as recommended by
Jason.
Cheers,
Lionel Landwerlin (2):
drm/i915: introduce a mechanism to extend execbuf2
drm/i915: add syncobj timeline support
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 346 +++---
drivers/gpu/drm
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
v3: Rename I915_EXEC_EXT -> I915_EXEC_USE_EXTENSIONS (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
://lists.freedesktop.org/archives/dri-devel/2019-August/229287.html
drop the ordering check from v8 (Lionel)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 307 ++
drivers/gpu/drm/i915/i915_drv.c | 3 +-
drivers/gpu/drm/i915/i915_getparam.c
Hi all,
Just a rebase and a change added then reversed that puts us right back
to v3.
Cheers,
Lionel Landwerlin (2):
drm/i915: introduce a mechanism to extend execbuf2
drm/i915: add syncobj timeline support
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 346 +++---
drivers
On 17/08/2019 11:38, Lucas De Marchi wrote:
From: Michel Thierry
Compared to Icelake, Tigerlake's MAX_CONTEXT_HW_ID is smaller by one, but
since we just use the upper 32 bits of the lrc_desc, it's guaranteed OA
will use the correct one.
Cc: Lionel Landwerlin
Signed-off-by: Michel Thierry
, Lucas De Marchi wrote:
From: Lionel Landwerlin
The way our hardware is designed doesn't seem to let us use the
MI_RECORD_PERF_COUNT command without setting up a circular buffer.
In the case where the user didn't request OA reports to be available
through the i915 perf stream, we can set the OA
On 01/08/2019 17:16, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-01 15:29:34)
On 31/07/2019 23:03, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-07-31 15:07:33)
-static struct drm_syncobj **
-get_fence_array(struct drm_i915_gem_execbuffer2 *args,
- struct
On 01/08/2019 18:16, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-01 15:29:34)
On 31/07/2019 23:03, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-07-31 15:07:33)
...
I think I have convinced myself that with the split between wait before,
signal after combined with the rule
On 31/07/2019 23:03, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-07-31 15:07:33)
-static struct drm_syncobj **
-get_fence_array(struct drm_i915_gem_execbuffer2 *args,
- struct drm_file *file)
+static struct i915_eb_fences *
+get_timeline_fence_array(struct i915_execbuffer
On 01/08/2019 12:22, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-01 10:01:44)
On 01/08/2019 11:08, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-01 08:43:24)
On 31/07/2019 23:03, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-07-31 15:07:33)
I think I have convinced
On 01/08/2019 11:08, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-08-01 08:43:24)
On 31/07/2019 23:03, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-07-31 15:07:33)
I think I have convinced myself that with the split between wait before,
signal after combined with the rule
On 31/07/2019 23:03, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-07-31 15:07:33)
-static struct drm_syncobj **
-get_fence_array(struct drm_i915_gem_execbuffer2 *args,
- struct drm_file *file)
+static struct i915_eb_fences *
+get_timeline_fence_array(struct i915_execbuffer
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
v2: Check for invalid flags in execbuffer2 (Lionel)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson (v1)
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 38
BIT_ULL (Chris)
v6: Fix issue with already signaled timeline points,
dma_fence_chain_find_seqno() setting fence to NULL (Chris)
v7: Report ENOENT with invalid syncobj handle (Lionel)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 297
://gitlab.freedesktop.org/mesa/mesa/merge_requests/1525
IGT tests are available in the vulkan perf query series :
https://patchwork.freedesktop.org/series/64220/
Cheers,
Lionel Landwerlin (2):
drm/i915: introduce a mechanism to extend execbuf2
drm/i915: add syncobj timeline support
.../gpu/drm/i915/gem
On 17/07/2019 21:09, Tvrtko Ursulin wrote:
On 17/07/2019 15:06, Chris Wilson wrote:
Quoting Tvrtko Ursulin (2019-07-17 14:46:15)
On 17/07/2019 14:35, Chris Wilson wrote:
Quoting Tvrtko Ursulin (2019-07-17 14:23:55)
On 17/07/2019 14:17, Chris Wilson wrote:
Quoting Tvrtko Ursulin
igure contexts on the fly")
Signed-off-by: Chris Wilson
Cc: Lionel Landwerlin
Cc: Tvrtko Ursulin
Reviewed-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_perf.c | 61 ++--
1 file changed, 35 insertions(+), 26 deletions(-)
diff --git a/drivers/gpu/drm/i915/i91
gem_engine(ce,
i915_gem_context_lock_engines(ctx),
it) {
Could almost reduce the scope of variable. Regardless :
Reviewed-by: Lionel Landwerlin
Thanks,
-Lionel
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedeskto
priorities to avoid preemption rather than the HW mechanism
v4: Just modify the port priority reporting function
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 8 +
drivers/gpu/drm/i915/i915_drv.h | 8 +
drivers/gpu
to userspace (Chris)
v6: Drop lock when copying config list to userspace (Chris)
Fix deadlock when calling i915_perf_get_oa_config() under
perf.metrics_lock (Lionel)
Add i915_oa_config_get() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915
v4: Create i915 request in emit_oa_config() to avoid deadlocks (Lionel)
Signed-off-by: Lionel Landwerlin
---
drivers/gpu/drm/i915/i915_drv.h | 7 ++
drivers/gpu/drm/i915/i915_perf.c | 204 +++
2 files changed, 134 insertions(+), 77 deletions(-)
diff --git
)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 133 +-
drivers/gpu/drm/i915/gt/intel_engine_cs.c | 3 +
drivers/gpu/drm/i915/gt/intel_engine_types.h | 9 ++
drivers/gpu/drm/i915/gt/intel_lrc.c | 1 +
drivers/gpu/drm/i915/gt
We're planning to use this for a couple of new feature where we need
to provide additional parameters to execbuf.
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 32 ++-
include/uapi/drm/i915_drm.h
Reporting this version will help application figure out what level of
the support the running kernel provides.
v2: Add i915_perf_ioctl_version() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_drv.c | 3 +++
drivers/gpu/drm/i915/i915_drv.h
BIT_ULL (Chris)
v6: Fix issue with already signaled timeline points,
dma_fence_chain_find_seqno() setting fence to NULL (Chris)
v7: Report ENOENT with invalid syncobj handle (Lionel)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 297
Hi all,
Just posted some tests : https://patchwork.freedesktop.org/series/64220/
And shockingly it found a few bugs.
This series is also rebased on top of Chris' on the fly OA
reconfiguration of contexts.
Cheers,
Lionel Landwerlin (9):
drm/i915/perf: introduce a versioning of the i915-perf
be over the required amount of
time (according to HW architects).
v2: Don't forget to save/restore registers used for the wait (Chris)
v3: Name used CS_GPR registers (Chris)
Fix compile issue due to rebase (Lionel)
v4: Fix save/restore helpers (Umesh)
Signed-off-by: Lionel Landwerlin
mapping of OA config (Chris)
v3: Properly deal with the perf_metric lock (Chris/Lionel)
v4: Fix oa config unref/put when not found (Lionel)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/gt/intel_gpu_commands.h | 1 +
drivers/gpu/drm/i915/i915_drv.h
or take we may issue
multiple idle barriers across multiple engines).
Reported-by: Lionel Landwerlin
Fixes: ce476c80b8bf ("drm/i915: Keep contexts pinned until after the next kernel
context switch")
Fixes: a9877da2d629 ("drm/i915/oa: Reconfigure contexts on the fly")
Signed-off-
On 11/07/2019 12:46, Lionel Landwerlin wrote:
+
+ syncobj = drm_syncobj_find(eb->file, user_fence.handle);
+ if (!syncobj) {
+ DRM_DEBUG("Invalid syncobj handle provided\n");
+ err = -EINVAL;
+
On 18/07/2019 17:54, Chris Wilson wrote:
Since userspace has the ability to bypass the CPU cache from within its
unprivileged command stream, we have to flush the CPU cache to memory
in order to overwrite the previous contents on creation.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc:
On 15/07/2019 14:30, Koenig, Christian wrote:
Hi Lionel,
sorry for the delayed response, I'm just back from vacation.
Am 03.07.19 um 11:17 schrieb Lionel Landwerlin:
On 03/07/2019 11:56, Chris Wilson wrote:
Quoting Lionel Landwerlin (2019-07-01 12:34:33)
+ syncobj
On 12/07/2019 19:23, John Harrison wrote:
On 7/12/2019 01:53, Tvrtko Ursulin wrote:
On 12/07/2019 08:07, john.c.harri...@intel.com wrote:
From: John Harrison
Newer hardware supports extra feature in the whitelist registers. This
patch updates the selftest to test that entries marked as read
to userspace (Chris)
v6: Drop lock when copying config list to userspace (Chris)
Fix deadlock when calling i915_perf_get_oa_config() under
perf.metrics_lock (Lionel)
Add i915_oa_config_get() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915
Reporting this version will help application figure out what level of
the support the running kernel provides.
v2: Add i915_perf_ioctl_version() (Chris)
Signed-off-by: Lionel Landwerlin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_drv.c | 3 +++
drivers/gpu/drm/i915/i915_drv.h
)
Signed-off-by: Lionel Landwerlin
---
.../gpu/drm/i915/gem/i915_gem_execbuffer.c| 126 +-
drivers/gpu/drm/i915/gt/intel_engine_cs.c | 2 +
drivers/gpu/drm/i915/gt/intel_engine_types.h | 9 ++
drivers/gpu/drm/i915/gt/intel_lrc.c | 1 +
drivers/gpu/drm/i915/gt
Hi,
A few changes :
- Don't call copy_to_user under lock in i915_query
- Fix save/restore helper for noa wait
Many thanks to the reviewers!
Cheers,
Lionel Landwerlin (9):
drm/i915/perf: introduce a versioning of the i915-perf uapi
drm/i915/perf: allow for CS OA configs
501 - 600 of 2230 matches
Mail list logo