On Tue, 24 Oct 2017, Tvrtko Ursulin wrote:
> On 24/10/17 18:48, Jani Nikula wrote:
>> On Tue, 24 Oct 2017, Chris Wilson wrote:
>>> Quoting Sagar Arun Kamble (2017-10-24 11:41:13)
diff --git a/drivers/gpu/drm/i915/intel_device_info.c
b/drivers/gpu/drm/i915/intel_device_info.c
index
Op 25-10-17 om 03:36 schreef Pandiyan, Dhinakaran:
> On Wed, 2017-10-11 at 12:28 +0200, Maarten Lankhorst wrote:
>> Op 10-10-17 om 22:12 schreef Chris Wilson:
>>> Quoting Maarten Lankhorst (2017-10-10 17:04:27)
Make sure read_all_entries has all outputs possible enabled, but also
add a te
Do not close fd 0 by accident.
Signed-off-by: Maarten Lankhorst
---
lib/igt_debugfs.c | 40 +++-
1 file changed, 11 insertions(+), 29 deletions(-)
diff --git a/lib/igt_debugfs.c b/lib/igt_debugfs.c
index 8b33b478a9a9..c2a3a1091644 100644
--- a/lib/igt_debugfs
On 10/24/17 9:01 AM, SF Markus Elfring wrote:
>>> Do you prefer to delegate the proposed software refactoring
>>> only to a corresponding optimiser?
>>
>> yes.
>
> Will any applications around the semantic patch language
> (Coccinelle software) fit also in the preferred tool category?
What do you
On Fri, Oct 20, 2017 at 03:20:22PM +0100, Chris Wilson wrote:
> We only suppress the error state from the context, iff we triggering the
> GPU reset by hangcheck (igt_hang_ring). If we trigger an immediate reset
> via igt_force_gpu_reset(), we will populate the error-state but not free
> it because
== Series Details ==
Series: HAX: Find CRC failures more reliably (rev2)
URL : https://patchwork.freedesktop.org/series/32555/
State : failure
== Summary ==
IGT patchset tested on top of latest successful build
45a253e34b063426d18cdf02319eb6f6bc238e6b igt/prime_mmap_coherency: Remove
manual g
From: Tvrtko Ursulin
Code to be shared between debugfs and the PMU implementation.
v2: Checkpatch cleanup.
v3: Also consolidate i915_sysfs.c/gt_act_freq_mhz_show.
v4: Rebase.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson (v2)
---
drivers/gpu/drm/i915/i915_debugfs.c | 9 ++---
From: Tvrtko Ursulin
Track total time requests have been executing on the hardware.
We add new kernel API to allow software tracking of time GPU
engines are spending executing requests.
Both per-engine and global API is added with the latter also
being exported for use by external users.
v2:
From: Tvrtko Ursulin
For clients like intel-gpu-overlay it is easier to read the
count via the perf API than having to parse /proc.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_pmu.c | 23 +++
include/uapi/drm/i915_drm.h | 4 ++
From: Tvrtko Ursulin
CPU hotplug core bug has been fixed (1f7c70d6b2bc ("cpu/hotplug: Reset node
state after operation")) so sending out a re-spin with the relevant workaround
removed. Plus a couple of rebases to catch up with latest drm-tip.
Tvrtko Ursulin (9):
drm/i915: Extract intel_get_cag
From: Tvrtko Ursulin
Will be used for exposing the PMU counters.
v2:
* Move intel_runtime_pm_get/put to the callers. (Chris Wilson)
* Restore full unit conversion precision.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_drv.h | 8 +++-
drivers
From: Tvrtko Ursulin
We can use engine busy stats instead of the sampling timer for
better accuracy.
By doing this we replace the stohastic sampling with busyness
metric derived directly from engine activity. This is context
switch interrupt driven, so as accurate as we can get from
software tra
From: Tvrtko Ursulin
From: Chris Wilson
From: Tvrtko Ursulin
From: Dmitry Rogozhkin
The first goal is to be able to measure GPU (and invidual ring) busyness
without having to poll registers from userspace. (Which not only incurs
holding the forcewake lock indefinitely, perturbing the system,
From: Tvrtko Ursulin
For clients like intel-gpu-overlay it is easier to read the
counters via the perf API than having to parse sysfs.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_pmu.c | 31 +++
include/uapi/drm/i915_drm.h
From: Tvrtko Ursulin
If only a subset of events is enabled we can afford to suspend
the sampling timer when the GPU is idle and so save some cycles
and power.
v2: Rebase and limit timer even more.
v3: Rebase.
v4: Rebase.
v5: Skip action if perf PMU failed to register.
v6: Checkpatch cleanup.
v7:
From: Tvrtko Ursulin
No functional change just something which will be handy in the
following patch.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/intel_lrc.c | 17 ++---
1 file changed, 14 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/
Quoting Tvrtko Ursulin (2017-10-25 10:06:01)
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index bafe1cdd57d8..08bbceaeb9b8 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -3349,6 +3349,7 @@ i915_gem_idle_work_handler(
== Series Details ==
Series: HAX: Find CRC failures more reliably (rev2)
URL : https://patchwork.freedesktop.org/series/32555/
State : warning
== Summary ==
IGT patchset tested on top of latest successful build
45a253e34b063426d18cdf02319eb6f6bc238e6b igt/prime_mmap_coherency: Remove
manual g
== Series Details ==
Series: HAX: Find CRC failures more reliably
URL : https://patchwork.freedesktop.org/series/32555/
State : failure
== Summary ==
Test kms_busy:
Subgroup extended-modeset-hang-oldfb-with-reset-render-A:
pass -> DMESG-WARN (shard-hsw) fdo#102249
On Tue, 2017-10-24 at 21:50 +0100, Chris Wilson wrote:
> Insert a breakpoint, a chance to escape back to the scheduler and run
> something else for a bit, if we find that the GGTT is full and needs to
> be idled in order to make some room. In practice, this should only be an
> issue in stress tests
== Series Details ==
Series: i915 PMU and engine busy stats (rev20)
URL : https://patchwork.freedesktop.org/series/27488/
State : success
== Summary ==
Series 27488v20 i915 PMU and engine busy stats
https://patchwork.freedesktop.org/api/1.0/series/27488/revisions/20/mbox/
Test kms_pipe_crc_ba
== Series Details ==
Series: lib/gt: Always eat the unwanted error state
URL : https://patchwork.freedesktop.org/series/32364/
State : failure
== Summary ==
Series 32364 revision 1 was fully merged or fully failed: no git log
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/
Instead of trying to create a timer with zero delay (i.e. with expires
set to the current jiffies and not the future, an already expired
timer), execute that request immediately.
v2: Refactor list_del_init+signal into its own little function.
Signed-off-by: Chris Wilson
Link:
https://patchwork.
Quoting Chris Wilson (2017-10-25 11:06:01)
> Instead of trying to create a timer with zero delay (i.e. with expires
> set to the current jiffies and not the future, an already expired
> timer), execute that request immediately.
>
> v2: Refactor list_del_init+signal into its own little function.
>
== Series Details ==
Series: HAX: Find CRC failures more reliably (rev2)
URL : https://patchwork.freedesktop.org/series/32555/
State : failure
== Summary ==
IGT patchset tested on top of latest successful build
e2598d4ec8358388d2fc2d65171665d0eb5329ec lib/gt: Always eat the unwanted error
sta
Instead of trying to create a timer with zero delay (i.e. with expires
set to the current jiffies and not the future, an already expired
timer), execute that request immediately.
v2: Refactor list_del_init+signal into its own little function.
v3: Reorder testing so as not to immediately signal a d
On Wed, 2017-10-25 at 11:17 +0100, Chris Wilson wrote:
> Instead of trying to create a timer with zero delay (i.e. with expires
> set to the current jiffies and not the future, an already expired
> timer), execute that request immediately.
>
> v2: Refactor list_del_init+signal into its own little
On Thu, Oct 19, 2017 at 05:13:40PM +0200, Maarten Lankhorst wrote:
> The original intent was to preserve watermarks as much as possible
> in intel_pipe_wm.raw_wm, and put the validated ones in intel_pipe_wm.wm.
>
> It seems this approach is insufficient and we don't always preserve
> the raw water
Quoting Chris Wilson (2017-10-24 17:17:09)
> Quoting Kees Cook (2017-10-24 16:13:44)
> > In preparation for unconditionally passing the struct timer_list pointer to
> > all timer callbacks, switch to using the new timer_setup() and from_timer()
> > to pass the timer pointer explicitly.
> >
> > Cc:
== Series Details ==
Series: series starting with [v2] drm/i915/selftests: Don't try to queue a
request with zero delay (rev2)
URL : https://patchwork.freedesktop.org/series/32582/
State : warning
== Summary ==
Series 32582v2 series starting with [v2] drm/i915/selftests: Don't try to queue
a
On Fri, Oct 20, 2017 at 11:24:19AM +0100, Chris Wilson wrote:
> __gem_create() doesn't touch the outparam *handle on failure, so we can
> no longer assert that it zero. This is reasonable to remove as it is just
> testing the library itself and not the kernel, so no loss in coverage.
> We already h
Include GuC and HuC firmware details in captured error state
to provide additional debug information. To reuse existing
uc firmware pretty printer, introduce new drm-printer variant
that works with our i915_error_state_buf output. Also update
uc firmware pretty printer to accept const input.
v2: d
We keep details of GuC and HuC in separate error state struct.
Make GuC log part of it to group all related data together.
Signed-off-by: Michal Wajdeczko
Cc: Chris Wilson
Cc: Joonas Lahtinen
---
drivers/gpu/drm/i915/i915_drv.h | 2 +-
drivers/gpu/drm/i915/i915_gpu_error.c | 18 +++-
Quoting Michał Winiarski (2017-10-25 11:34:27)
> On Fri, Oct 20, 2017 at 11:24:19AM +0100, Chris Wilson wrote:
> > __gem_create() doesn't touch the outparam *handle on failure, so we can
> > no longer assert that it zero. This is reasonable to remove as it is just
> > testing the library itself and
== Series Details ==
Series: series starting with [v3] drm/i915/selftests: Don't try to queue a
request with zero delay (rev3)
URL : https://patchwork.freedesktop.org/series/32582/
State : warning
== Summary ==
Series 32582v3 series starting with [v3] drm/i915/selftests: Don't try to queue
a
== Series Details ==
Series: i915 PMU and engine busy stats (rev20)
URL : https://patchwork.freedesktop.org/series/27488/
State : success
== Summary ==
Test kms_flip:
Subgroup plain-flip-fb-recreate-interruptible:
fail -> PASS (shard-hsw) fdo#100368
Test drv
On Tue, 2017-10-24 at 23:08 +0100, Chris Wilson wrote:
> During evict, we wish to idle the GPU if we see that the GGTT is full.
> However, our test for idle in i915_gem_evict_something() and in
> i915_gem_switch_to_kernel_context() do not match leading to
> disappointment - we never believe that we
== Series Details ==
Series: HAX: Find CRC failures more reliably (rev2)
URL : https://patchwork.freedesktop.org/series/32555/
State : success
== Summary ==
Test kms_flip:
Subgroup plain-flip-fb-recreate-interruptible:
fail -> PASS (shard-hsw) fdo#100368
Tes
Quoting Michal Wajdeczko (2017-10-25 11:41:12)
> We keep details of GuC and HuC in separate error state struct.
> Make GuC log part of it to group all related data together.
The only significant change would appear to placement in error-state; it
now comes at the end. Maybe that's ok; what do we f
On Tue, Oct 24, 2017 at 07:24:58PM +0100, Chris Wilson wrote:
> Since the headless subtest wants to compare execution latency of a
> headless mode vs a single head, it needs to be able to do a modeset and
> ergo it requires DRM_MASTER.
>
> Signed-off-by: Chris Wilson
Reviewed-by: Michał Winiarsk
== Series Details ==
Series: series starting with [v5,1/2] drm/i915: Add Guc/HuC firmware details to
error state
URL : https://patchwork.freedesktop.org/series/32615/
State : failure
== Summary ==
Series 32615v1 series starting with [v5,1/2] drm/i915: Add Guc/HuC firmware
details to error st
On 25/10/2017 10:13, Chris Wilson wrote:
Quoting Tvrtko Ursulin (2017-10-25 10:06:01)
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index bafe1cdd57d8..08bbceaeb9b8 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3349,6 +33
Quoting Tvrtko Ursulin (2017-10-25 13:44:24)
>
> On 25/10/2017 10:13, Chris Wilson wrote:
> > Quoting Tvrtko Ursulin (2017-10-25 10:06:01)
> >> diff --git a/drivers/gpu/drm/i915/i915_gem.c
> >> b/drivers/gpu/drm/i915/i915_gem.c
> >> index bafe1cdd57d8..08bbceaeb9b8 100644
> >> --- a/drivers/gpu/d
On Mon, 2017-10-23 at 16:19 -0700, Kenneth Graunke wrote:
> On Monday, October 23, 2017 3:53:15 PM PDT Rodrigo Vivi wrote:
> > On Mon, Oct 23, 2017 at 10:32:43PM +, Jordan Justen wrote:
> > > On 2017-10-19 16:30:44, Kristian Høgsberg wrote:
> > > > On Thu, Oct 19, 2017 at 4:18 PM, Kenneth Graun
Quoting Chris Wilson (2017-10-25 11:24:19)
> Quoting Chris Wilson (2017-10-24 17:17:09)
> > Quoting Kees Cook (2017-10-24 16:13:44)
> > > In preparation for unconditionally passing the struct timer_list pointer
> > > to
> > > all timer callbacks, switch to using the new timer_setup() and
> > > fr
This reverts commit 6d0dbd309687113009a276886628c7c74c848d3c.
timer_setup_on_stack() does not yet exist:
In file included from drivers/gpu/drm/i915/i915_sw_fence.c:517:0:
drivers/gpu/drm/i915/selftests/lib_sw_fence.c: In function ‘timed_fence_init’:
drivers/gpu/drm/i915/selftests/lib_sw_fence.c:6
On Wed, 2017-10-25 at 14:13 +0100, Chris Wilson wrote:
> This reverts commit 6d0dbd309687113009a276886628c7c74c848d3c.
>
> timer_setup_on_stack() does not yet exist:
>
> In file included from drivers/gpu/drm/i915/i915_sw_fence.c:517:0:
> drivers/gpu/drm/i915/selftests/lib_sw_fence.c: In function
Quoting Chris Wilson (2017-10-25 14:13:36)
> This reverts commit 6d0dbd309687113009a276886628c7c74c848d3c.
>
> timer_setup_on_stack() does not yet exist:
>
> In file included from drivers/gpu/drm/i915/i915_sw_fence.c:517:0:
> drivers/gpu/drm/i915/selftests/lib_sw_fence.c: In function ‘timed_fence
== Series Details ==
Series: drm/i915: Call cond_resched() before repeating
i915_gem_evict_something()
URL : https://patchwork.freedesktop.org/series/32576/
State : failure
== Summary ==
Series 32576 revision 1 was fully merged or fully failed: no git log
== Logs ==
For more details see: ht
On Tue, Oct 24, 2017 at 12:52:06PM +0300, Ville Syrjala wrote:
> From: Ville Syrjälä
>
> New version of the CNL DVFS series. Everything from the previous series
> (first 8 patches) are already reviewed. I slapped on two additional patches
> that fell out from the review of the first series.
>
>
On Tue, 24 Oct 2017 19:21:20 +0200, Sujaritha Sundaresan
wrote:
Unifying the various seq_puts messages in debugfs to the simplest one for
feature support.
v2: Clarifying the commit message (Anusha)
v3: Re-factoring code as per review (Michal)
v4: Rebase
v5: Split from following patch
v6:
== Series Details ==
Series: Revert "drm/i915/selftests: Convert timers to use timer_setup()"
URL : https://patchwork.freedesktop.org/series/32637/
State : warning
== Summary ==
Series 32637v1 Revert "drm/i915/selftests: Convert timers to use timer_setup()"
https://patchwork.freedesktop.org/ap
On Wed, Oct 25, 2017 at 3:11 PM, Chris Wilson wrote:
> Quoting Chris Wilson (2017-10-25 11:24:19)
>> Quoting Chris Wilson (2017-10-24 17:17:09)
>> > Quoting Kees Cook (2017-10-24 16:13:44)
>> > > In preparation for unconditionally passing the struct timer_list pointer
>> > > to
>> > > all timer c
Quoting Kees Cook (2017-10-25 15:05:13)
> On Wed, Oct 25, 2017 at 3:11 PM, Chris Wilson
> wrote:
> > Quoting Chris Wilson (2017-10-25 11:24:19)
> >> Quoting Chris Wilson (2017-10-24 17:17:09)
> >> > Quoting Kees Cook (2017-10-24 16:13:44)
> >> > > In preparation for unconditionally passing the st
On Wed, Oct 25, 2017 at 4:16 PM, Chris Wilson wrote:
> Quoting Kees Cook (2017-10-25 15:05:13)
>> On Wed, Oct 25, 2017 at 3:11 PM, Chris Wilson
>> wrote:
>> > Quoting Chris Wilson (2017-10-25 11:24:19)
>> >> Quoting Chris Wilson (2017-10-24 17:17:09)
>> >> > Quoting Kees Cook (2017-10-24 16:13:4
From: Tvrtko Ursulin
If only a subset of events is enabled we can afford to suspend
the sampling timer when the GPU is idle and so save some cycles
and power.
v2: Rebase and limit timer even more.
v3: Rebase.
v4: Rebase.
v5: Skip action if perf PMU failed to register.
v6: Checkpatch cleanup.
v7:
On Wed, 25 Oct 2017, Kees Cook wrote:
> On Wed, Oct 25, 2017 at 4:16 PM, Chris Wilson
> wrote:
>> Quoting Kees Cook (2017-10-25 15:05:13)
>>> On Wed, Oct 25, 2017 at 3:11 PM, Chris Wilson
>>> wrote:
>>> > Quoting Chris Wilson (2017-10-25 11:24:19)
>>> >> Quoting Chris Wilson (2017-10-24 17:17:
On October 25, 2017 06:05:16 Joonas Lahtinen
wrote:
On Mon, 2017-10-23 at 16:19 -0700, Kenneth Graunke wrote:
On Monday, October 23, 2017 3:53:15 PM PDT Rodrigo Vivi wrote:
> On Mon, Oct 23, 2017 at 10:32:43PM +, Jordan Justen wrote:
> > On 2017-10-19 16:30:44, Kristian Høgsberg wrote:
>
== Series Details ==
Series: Revert "drm/i915/selftests: Convert timers to use timer_setup()"
URL : https://patchwork.freedesktop.org/series/32637/
State : failure
== Summary ==
Series 32637 revision 1 was fully merged or fully failed: no git log
== Logs ==
For more details see: https://inte
From: Michał Winiarski
Also:
Revert "drm/i915/guc: Assert that we switch between known ggtt->invalidate
functions"
This reverts commit 04f7b24eccdfae680a36e9825fe0d61dcd5ed528.
---
drivers/gpu/drm/i915/i915_gem_gtt.c | 8 ++--
drivers/gpu/drm/i915/i915_params.h | 4 ++--
2 files changed,
In the next patch, we will want to install a callback when the engines
(GT as a whole) become idle and similarly when they first become busy.
To enable that callback, first rename intel_engines_mark_idle() to
intel_engines_park() and provide the companion intel_engines_unpark().
Signed-off-by: Chr
The execlists emulation on top of the GuC (used for scheduling and
preemption) depends on the MI_USER_INTERRUPT for its notifications and
tasklet action. As we always employ the irq, there is no advantage in
ever disabling it while we are using the GuC, so allow us to arm the
breadcrumb irq when en
Recently W=1 on gcc-7.2 (-Wunused-const-variable) caught a regression
that had been lurking for 6 months, so lets try enabling the full set of
warnings for CI builds. This means more patches will be rejected early
that contain trivial and sometimes not so trivial bugs. However, our
code does not ye
== Series Details ==
Series: drm/i915: Enhanced for initialize partially filled pagetables
URL : https://patchwork.freedesktop.org/series/31029/
State : failure
== Summary ==
Series 31029v1 drm/i915: Enhanced for initialize partially filled pagetables
https://patchwork.freedesktop.org/api/1.0/
--
CONFIDENTIALITY NOTICE:
The information in this communication, including any attachments, is
intended only for the person or entity to which it is addressed and may
contain confidential, proprietary, and/or privileged material. If you are
not the addressee or if you received this communi
== Series Details ==
Series: i915 PMU and engine busy stats (rev21)
URL : https://patchwork.freedesktop.org/series/27488/
State : success
== Summary ==
Series 27488v21 i915 PMU and engine busy stats
https://patchwork.freedesktop.org/api/1.0/series/27488/revisions/21/mbox/
Test kms_busy:
If the system has bsd2, we do not know which ring the kernel will alias
I915_EXEC_BSD onto and so we do not what the matching string should be.
Skip the unknown.
v2: Deny the aliased I915_EXEC_BSD exists at all; be specific!
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103324
Signed-off
Hi All,
On 24 October 2017 at 22:18, PrasannaKumar Muralidharan
wrote:
> In i915 driver unload drm_vblank_get is added to test whether
> drm_vblank_cleanup refcount validation patch is working.
>
> Signed-off-by: PrasannaKumar Muralidharan
> ---
> Changes in v2:
> Use drm_crtc_vblank_get instead
On Wed, Oct 25, 2017 at 04:13:58PM +0100, Chris Wilson wrote:
> If the system has bsd2, we do not know which ring the kernel will alias
> I915_EXEC_BSD onto and so we do not what the matching string should be.
> Skip the unknown.
>
> v2: Deny the aliased I915_EXEC_BSD exists at all; be specific!
>
On Tue, 24 Oct 2017 19:21:21 +0200, Sujaritha Sundaresan
wrote:
We currently have two module parameters that control GuC:
"enable_guc_loading" and "enable_guc_submission". Whenever
we need submission=1, we also need loading=1.We also need
loading=1 when we want to want to verify the HuC, whic
== Series Details ==
Series: series starting with [CI,1/4] drm/i915: Add -Wall -Wextra to our build,
set warnings to full
URL : https://patchwork.freedesktop.org/series/32642/
State : failure
== Summary ==
Series 32642v1 series starting with [CI,1/4] drm/i915: Add -Wall -Wextra to our
build,
Quoting Joonas Lahtinen (2017-10-25 12:07:59)
> On Tue, 2017-10-24 at 23:08 +0100, Chris Wilson wrote:
> > During evict, we wish to idle the GPU if we see that the GGTT is full.
> > However, our test for idle in i915_gem_evict_something() and in
> > i915_gem_switch_to_kernel_context() do not match
From: Tvrtko Ursulin
If only a subset of events is enabled we can afford to suspend
the sampling timer when the GPU is idle and so save some cycles
and power.
v2: Rebase and limit timer even more.
v3: Rebase.
v4: Rebase.
v5: Skip action if perf PMU failed to register.
v6: Checkpatch cleanup.
v7:
From: Tvrtko Ursulin
Lightly tested (apart from patch 13) series to start the discussion early.
Please skip patches 1-9, and probably 10-11 as well, those ones are the current
PMU effort which is yet unmerged but needed as a basis for patches 12-17.
Customer ask is to allow clients to query how
From: Tvrtko Ursulin
For clients like intel-gpu-overlay it is easier to read the
count via the perf API than having to parse /proc.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_pmu.c | 23 +++
include/uapi/drm/i915_drm.h | 4 ++
From: Tvrtko Ursulin
Keep a per-engine number of runnable (waiting for GPU time) requests.
Signed-off-by: Tvrtko Ursulin
---
drivers/gpu/drm/i915/i915_gem_request.c | 5 +
drivers/gpu/drm/i915/intel_engine_cs.c | 6 --
drivers/gpu/drm/i915/intel_lrc.c| 1 +
drivers/gpu/drm/i91
From: Tvrtko Ursulin
From: Chris Wilson
From: Tvrtko Ursulin
From: Dmitry Rogozhkin
The first goal is to be able to measure GPU (and invidual ring) busyness
without having to poll registers from userspace. (Which not only incurs
holding the forcewake lock indefinitely, perturbing the system,
From: Tvrtko Ursulin
Some clients have the DRM fd passed to them over a socket by the X server.
Grab the real client and pid when they create their first context and
update the exposed data for more useful enumeration.
Signed-off-by: Tvrtko Ursulin
---
drivers/gpu/drm/i915/i915_drv.h
From: Tvrtko Ursulin
Code to be shared between debugfs and the PMU implementation.
v2: Checkpatch cleanup.
v3: Also consolidate i915_sysfs.c/gt_act_freq_mhz_show.
v4: Rebase.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson (v2)
---
drivers/gpu/drm/i915/i915_debugfs.c | 9 ++---
From: Tvrtko Ursulin
No functional change just something which will be handy in the
following patch.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/intel_lrc.c | 17 ++---
1 file changed, 14 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/
From: Tvrtko Ursulin
Track total time requests have been executing on the hardware.
We add new kernel API to allow software tracking of time GPU
engines are spending executing requests.
Both per-engine and global API is added with the latter also
being exported for use by external users.
v2:
From: Tvrtko Ursulin
We can use engine busy stats instead of the sampling timer for
better accuracy.
By doing this we replace the stohastic sampling with busyness
metric derived directly from engine activity. This is context
switch interrupt driven, so as accurate as we can get from
software tra
From: Tvrtko Ursulin
For clients like intel-gpu-overlay it is easier to read the
counters via the perf API than having to parse sysfs.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_pmu.c | 31 +++
include/uapi/drm/i915_drm.h
From: Tvrtko Ursulin
Expose per-client and per-engine busyness under the previously added sysfs
client root.
The new file is named 'busy' and contains a list of, one line for each
engine, monotonically increasing nano-second resolution times each
client's jobs were executing on the GPU.
$ cat /
From: Tvrtko Ursulin
Some customers want to know how much of the GPU time are their clients
using in order to make dynamic load balancing decisions.
With the hooks already in place which track the overall engine busyness,
we can extend that slightly to split that time between contexts.
Signed-o
From: Tvrtko Ursulin
Expose a list of clients with open file handles in sysfs.
This will be a basis for a top-like utility showing per-client and per-
engine GPU load.
Currently we only expose each client's pid and name under opaque numbered
directories in /sys/class/drm/card0/clients/.
For in
From: Tvrtko Ursulin
Some customers want to know how much of the GPU time are their clients
using in order to make dynamic load balancing decisions.
With the accounting infrastructure in place in the previous patch, we add
a new context param (I915_CONTEXT_GET_ENGINE_BUSY) which takes a class an
From: Tvrtko Ursulin
Will be used for exposing the PMU counters.
v2:
* Move intel_runtime_pm_get/put to the callers. (Chris Wilson)
* Restore full unit conversion precision.
Signed-off-by: Tvrtko Ursulin
Reviewed-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_drv.h | 8 +++-
drivers
Some tests are designed to exercise the limits of the HW and may trigger
unintended side-effects making the machine unusable. This should not be
executed by default, but are still useful for early platform validation.
References: https://bugs.freedesktop.org/show_bug.cgi?id=103453
Signed-off-by: C
Hi,
> -Original Message-
> From: Intel-gfx [mailto:intel-gfx-boun...@lists.freedesktop.org] On Behalf
> Of Patchwork
> Sent: keskiviikko 25. lokakuuta 2017 18.29
> To: Chris Wilson
> Cc: intel-gfx@lists.freedesktop.org
> Subject: [Intel-gfx] ✗ Fi.CI.BAT: failure for series starting with [
From: Tvrtko Ursulin
We add a PMU counter to expose the number of requests currently submitted
to the GPU, plus the number of runnable requests waiting on GPU time.
This is useful to analyze the overall load of the system.
Signed-off-by: Tvrtko Ursulin
---
drivers/gpu/drm/i915/i915_pmu.c | 30
From: Tvrtko Ursulin
By default we are not collecting any per-engine and per-context
statistcs.
Add a new sysfs toggle to enable this facility:
$ echo 1 >/sys/class/drm/card0/clients/enable_stats
Signed-off-by: Tvrtko Ursulin
---
drivers/gpu/drm/i915/i915_drv.h | 2 ++
drivers/gpu/drm/i91
Quoting Saarinen, Jani (2017-10-25 16:34:31)
> Hi,
> > -Original Message-
> > From: Intel-gfx [mailto:intel-gfx-boun...@lists.freedesktop.org] On Behalf
> > Of Patchwork
> > Sent: keskiviikko 25. lokakuuta 2017 18.29
> > To: Chris Wilson
> > Cc: intel-gfx@lists.freedesktop.org
> > Subject
Quoting Tvrtko Ursulin (2017-10-25 16:36:15)
> From: Tvrtko Ursulin
>
> Lightly tested (apart from patch 13) series to start the discussion early.
>
> Please skip patches 1-9, and probably 10-11 as well, those ones are the
> current
> PMU effort which is yet unmerged but needed as a basis for p
On Tue, 24 Oct 2017 19:21:23 +0200, Sujaritha Sundaresan
wrote:
Updating GuC and HuC firmware select function to support removing
i915_modparams.enable_guc_loading module parameter.
Hmm, is it correct order of patches ? this modparam was already removed in
2/6
v2: Clarifying the commi
== Series Details ==
Series: drm/i915/selftests: Hide dangerous tests
URL : https://patchwork.freedesktop.org/series/32644/
State : success
== Summary ==
Series 32644v1 drm/i915/selftests: Hide dangerous tests
https://patchwork.freedesktop.org/api/1.0/series/32644/revisions/1/mbox/
Test kms_b
== Series Details ==
Series: i915 PMU and engine busy stats (rev21)
URL : https://patchwork.freedesktop.org/series/27488/
State : warning
== Summary ==
Test kms_cursor_crc:
Subgroup cursor-256x85-offscreen:
pass -> SKIP (shard-hsw)
Test kms_plane_lowres:
On Tue, 24 Oct 2017 19:21:25 +0200, Sujaritha Sundaresan
wrote:
The Additional Data Struct (ADS) contains objects that are required by
guc post FW load and are not necessarily submission-only (although that's
our current only use-case). If in the future we load GuC with submission
disabled to
Warn when refcount > 0 in drm_vblank_cleanup.
Signed-off-by: PrasannaKumar Muralidharan
---
Changes in v3:
* Dropped i915 patch that is used for testing this.
No changes in v2.
drivers/gpu/drm/drm_vblank.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/drm_vblank.c b/dri
Quoting Maarten Lankhorst (2017-10-09 11:33:21)
> Op 09-10-17 om 11:59 schreef Chris Wilson:
> > Quoting Maarten Lankhorst (2017-10-09 10:41:29)
> >> Op 06-10-17 om 19:18 schreef Chris Wilson:
> >>> When we unregister the connector, we may have a pending hotplug work.
> >>> This needs to be cancel
1 - 100 of 182 matches
Mail list logo
