On Sat, 21 Mar 2020 16:26:42 -0700, Dixit, Ashutosh wrote:
>
> On Thu, 19 Mar 2020 15:52:01 -0700, Umesh Nerlige Ramappa wrote:
> >
> > From: Lionel Landwerlin
>
> > @@ -477,16 +468,6 @@ static bool oa_buffer_check_unlocked(struct
> > i915_perf_stream *stream)
> > */
> >
On Thu, 19 Mar 2020 15:52:01 -0700, Umesh Nerlige Ramappa wrote:
>
> From: Lionel Landwerlin
>
> We're about to introduce an options to open the perf stream, giving
> the user ability to configure how often it wants the kernel to poll
> the OA registers for available data.
>
> Right now the
== Series Details ==
Series: series starting with [1/4] drm/i915: Extend intel_wakeref to support
delayed puts (rev2)
URL : https://patchwork.freedesktop.org/series/74938/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8167_full -> Patchwork_17043_full
== Series Details ==
Series: series starting with [1/4] drm/i915: Extend intel_wakeref to support
delayed puts (rev2)
URL : https://patchwork.freedesktop.org/series/74938/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8167 -> Patchwork_17043
== Series Details ==
Series: series starting with [1/4] drm/i915: Extend intel_wakeref to support
delayed puts
URL : https://patchwork.freedesktop.org/series/74938/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_8167 -> Patchwork_17042
We dropped calling process_csb prior to handling direct submission in
order to avoid the nesting of spinlocks and lift process_csb() and the
majority of the tasklet out of irq-off. However, we do want to avoid
ksoftirqd latency in the fast path, so try and pull the interrupt-bh
local to direct
In some cases we want to hold onto the wakeref for a little after the
last user so that we can avoid having to drop and then immediately
reacquire it. Allow the last user to specify if they would like to keep
the wakeref alive for a short hysteresis.
Signed-off-by: Chris Wilson
---
Keep the engine-pm awake until the next jiffie, to avoid immediate
ping-pong under moderate load. (Forcing the idle barrier excerbates the
moderate load, dramatically increasing the driver overhead.) On the
other hand, delaying the idle-barrier slightly incurs longer rc6-off
and so more power
Drop the pretense of kicking the tasklet (used only for the defunct guc
submission backend, it should just take ownership of the submit!) and so
remove the bh-kicking from around submission.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c | 2 --
Quoting Chris Wilson (2020-03-20 17:47:45)
> We dropped calling process_csb prior to handling direct submission in
> order to avoid the nesting of spinlocks and lift process_csb() and the
> majority of the tasklet out of irq-off. However, we do want to avoid
> ksoftirqd latency in the fast path,
This is a giant CC list.
There was one version where you CC'd me on patch 6/12 but after that you
just CC'd me on the cover page. Something is messed up in your scripts
because Cc'ing me on just the cover is pointless.
regards,
dan carpenter
___
On Thu, 19 Mar 2020 15:52:03 -0700, Umesh Nerlige Ramappa wrote:
>
> From: Lionel Landwerlin
>
> This new parameter let's the application choose how often the OA
> buffer should be checked on the CPU side for data availability. Longer
> polling period tend to reduce CPU overhead if the
12 matches
Mail list logo