Re: [Intel-gfx] [PATCH i-g-t 6/6] intel_gpu_top: Add a sanity check discovered busy metric is per engine

On Fri, 2021-11-19 at 12:59 +, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Adding a cross-check with ABI config name space and not just relying > on > sysfs names. > > Signed-off-by: Tvrtko Ursulin > Cc: Dmitry Rogozhkin > --- > tools/intel_gpu_top.c | 6 ++ > 1 file changed, 6

Re: [Intel-gfx] [PATCH 4/5] drm/i915: Increase initial busyspin limit

On Thu, 2018-08-02 at 15:54 +0100, Chris Wilson wrote: > Quoting Tvrtko Ursulin (2018-08-02 15:47:37) > > > > On 28/07/2018 17:46, Chris Wilson wrote: > > > Here we bump the busyspin for the current request to avoid > > > sleeping and > > > the cost of both idling and downclocking the CPU. > > >

Re: [Intel-gfx] [PATCH v9 7/7] drm/i915: add a sysfs entry to let users set sseu configs

On Thu, 2018-08-02 at 11:00 +0100, Tvrtko Ursulin wrote: > [Picking this point in the thread to reply on some points mentioned > by  > folks in the whole thread.] > > I don't remember if any patches from Lionel's series actually had r- > b's,  > but a few people including myself have certainly

Re: [Intel-gfx] [PATCH v9 7/7] drm/i915: add a sysfs entry to let users set sseu configs

On Tue, 2018-07-24 at 21:50 +, Bloomfield, Jon wrote: > Gratuitous top posting to re-kick the thread. > > For Gen11 we can't have an on/off switch anyway (media simply won't > run > with an oncompatible slice config), so let's agree on an api to allow > userland > to select the slice

Re: [Intel-gfx] [PATCH 8/8] drm/i915: Expose RPCS (SSEU) configuration to userspace

:tvrtko.ursu...@linux.intel.com] Sent: Tuesday, May 8, 2018 1:25 AM To: Rogozhkin, Dmitry V <dmitry.v.rogozh...@intel.com>; Landwerlin, Lionel G <lionel.g.landwer...@intel.com>; intel-gfx@lists.freedesktop.org Subject: Re: [Intel-gfx] [PATCH 8/8] drm/i915: Expose RPCS (SSEU) configurati

Re: [Intel-gfx] [PATCH 8/8] drm/i915: Expose RPCS (SSEU) configuration to userspace

>> I'm pretty sure Dmitry wants dynamic configurations. Yes, I afraid we really need dynamic slice configurations for media. From: Landwerlin, Lionel G Sent: Friday, May 4, 2018 9:25 AM To: Tvrtko Ursulin <tvrtko.ursu...@linux.intel.com>; intel-gfx@lists.freedesktop.org; Rogozhk

Re: [Intel-gfx] vlv punit and sideband tidy

On Wed, 2018-03-07 at 19:41 +, Chris Wilson wrote: > Mika > believes that if we keep the cpu in C0 whilst the gpu is busy, then > it > behaves much better -- but that is a very tough sell Chris, Mika, I wonder does i915 driver tries to keep CPU in C0 at the moment already or you just consider

Re: [Intel-gfx] [RFC] drm/i915: Add a new modparam for customized ring multiplier

at's the talk about Gen9 only. Dmitry. -Original Message- From: Intel-gfx [mailto:intel-gfx-boun...@lists.freedesktop.org] On Behalf Of Rogozhkin, Dmitry V Sent: Tuesday, December 26, 2017 9:39 AM To: Chris Wilson <ch...@chris-wilson.co.uk>; Li, Yaodong <yaodong...@intel.com>; inte

Re: [Intel-gfx] [RFC] drm/i915: Add a new modparam for customized ring multiplier

ris-wilson.co.uk] Sent: Tuesday, December 26, 2017 8:59 AM To: Rogozhkin, Dmitry V <dmitry.v.rogozh...@intel.com>; Li, Yaodong <yaodong...@intel.com>; intel-gfx@lists.freedesktop.org Cc: Gong, Zhipeng <zhipeng.g...@intel.com>; Widawsky, Benjamin <benjamin.widaw...@intel.co

Re: [Intel-gfx] [RFC] drm/i915: Add a new modparam for customized ring multiplier

;sagar.a.kam...@intel.com>; Rogozhkin, Dmitry V <dmitry.v.rogozh...@intel.com>; Li, Yaodong <yaodong...@intel.com> Subject: Re: [RFC] drm/i915: Add a new modparam for customized ring multiplier Quoting Chris Wilson (2017-12-18 21:47:25) > Quoting Jackie Li (2017-12-18 21:22:08) >

Re: [Intel-gfx] [PATCH v5] drm/i915: Restore GT performance in headless mode with DMC loaded

On Tue, 2017-12-05 at 15:09 +0200, Imre Deak wrote: > On Sat, Dec 02, 2017 at 02:05:42AM +0200, Rogozhkin, Dmitry V wrote: > > On Thu, 2017-11-30 at 13:19 +0200, Imre Deak wrote: > > > > > +#define NEEDS_CSR_GT_PERF_WA(dev_priv) \ > > > > > +

Re: [Intel-gfx] [PATCH v5] drm/i915: Restore GT performance in headless mode with DMC loaded

On Thu, 2017-11-30 at 13:19 +0200, Imre Deak wrote: > > > +#define NEEDS_CSR_GT_PERF_WA(dev_priv) \ > > > + (HAS_CSR(dev_priv) && IS_GEN9(dev_priv) && ! > IS_SKYLAKE(dev_priv)) > > Nitpick: could be just !IS_SKYLAKE(), but works in the above way too. > For all other platforms the GT_IRQ

Re: [Intel-gfx] [PATCH v5] drm/i915: Restore GT performance in headless mode with DMC loaded

On Thu, 2017-11-30 at 13:19 +0200, Imre Deak wrote: > On Thu, Nov 30, 2017 at 09:45:25AM +, Chris Wilson wrote: > > Quoting Tvrtko Ursulin (2017-11-30 09:18:20) > > > From: Tvrtko Ursulin > > > > > > It seems that the DMC likes to transition between the DC states a

Re: [Intel-gfx] [PATCH 2/2] drm/i915/pmu: Add queued counter

On Wed, 2017-11-22 at 12:46 +, Tvrtko Ursulin wrote: > 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

Re: [Intel-gfx] [PATCH 06/10] drm/i915/pmu: Wire up engine busy stats to PMU

Nice data, thank you! -Original Message- From: Tvrtko Ursulin [mailto:tvrtko.ursu...@linux.intel.com] Sent: Wednesday, October 4, 2017 10:35 AM To: Tvrtko Ursulin <tursu...@ursulin.net>; Intel-gfx@lists.freedesktop.org Cc: Chris Wilson <ch...@chris-wilson.co.uk>; Rogozhk

Re: [Intel-gfx] [RFC 04/10] drm/i915: Expose a PMU interface for perf queries

Hi Peter, friendly reminder. Could you, please, respond? -Original Message- From: Rogozhkin, Dmitry V Sent: Wednesday, September 20, 2017 1:15 PM To: Rogozhkin, Dmitry V <dmitry.v.rogozh...@intel.com>; pet...@infradead.org Cc: Intel-gfx@lists.freedesktop.org Subject: RE: [Intel-gfx

Re: [Intel-gfx] [PATCH 7/8] drm/i915/pmu: Wire up engine busy stats to PMU

On Tue, 2017-09-26 at 13:32 +0100, Tvrtko Ursulin wrote: > On 25/09/2017 18:48, Chris Wilson wrote: > > Quoting Tvrtko Ursulin (2017-09-25 16:15:42) > >> From: Tvrtko Ursulin > >> > >> We can use engine busy stats instead of the MMIO sampling timer > >> for better

Re: [Intel-gfx] [RFC 04/10] drm/i915: Expose a PMU interface for perf queries

Hi Peter, could you, please, comment on below? -Original Message- From: Intel-gfx [mailto:intel-gfx-boun...@lists.freedesktop.org] On Behalf Of Rogozhkin, Dmitry V Sent: Wednesday, September 13, 2017 4:06 PM To: pet...@infradead.org Cc: Intel-gfx@lists.freedesktop.org Subject: Re: [Intel

Re: [Intel-gfx] [RFC 10/11] drm/i915: Export engine stats API to other users

On Tue, 2017-09-19 at 12:50 -0700, Ben Widawsky wrote: > On 17-09-15 10:49:56, Tvrtko Ursulin wrote: > > > >On 14/09/2017 21:26, Chris Wilson wrote: > >>Quoting Tvrtko Ursulin (2017-09-11 16:25:58) > >>>From: Tvrtko Ursulin > >>> > >>>Other kernel users might want to

Re: [Intel-gfx] [PATCH 8/8] drm/i915: Gate engine stats collection with a static key

On Mon, 2017-09-18 at 12:38 +0100, Tvrtko Ursulin wrote: > static inline void intel_engine_context_out(struct intel_engine_cs > *engine) > { > unsigned long flags; > > - if (READ_ONCE(engine->stats.enabled) == 0) > - return; > - > -

Re: [Intel-gfx] [PATCH i-g-t 5/5] tests/perf_pmu: Tests for i915 PMU API

Did you try tests on the system with 2 VDBOX engines? On my side 2 tests are failing on SKL GT4e NUC: (perf_pmu:5414) CRITICAL: Test assertion failure function busy_check_all, file perf_pmu.c:164: (perf_pmu:5414) CRITICAL: Failed assertion: (double)(val[i]) <= (1.0 + tolerance) * (double)0.0f &&

Re: [Intel-gfx] [RFC v7 04/11] drm/i915/pmu: Expose a PMU interface for perf queries

On Wed, 2017-09-13 at 11:34 +0100, Tvrtko Ursulin wrote: > +static int i915_pmu_event_init(struct perf_event *event) > +{ > + struct drm_i915_private *i915 = > + container_of(event->pmu, typeof(*i915), pmu.base); > + int cpu, ret; > + > + if (event->attr.type !=

Re: [Intel-gfx] [RFC 04/10] drm/i915: Expose a PMU interface for perf queries

On Tue, 2017-08-29 at 21:21 +0200, Peter Zijlstra wrote: > On Tue, Aug 29, 2017 at 07:16:31PM +0000, Rogozhkin, Dmitry V wrote: > > > Pretty strict, people tend to get fairly upset every time we leak stuff. > > > In fact Debian and Android carry a perf_event_paranoid pat

Re: [Intel-gfx] [RFC v3 00/11] i915 PMU and engine busy stats

On Tue, 2017-09-12 at 15:54 +0100, Tvrtko Ursulin wrote: > On 12/09/2017 03:03, Rogozhkin, Dmitry V wrote: > > Hi, > > > > Just tried v3 series. perf-stat works fine. From the IGT tests which I > > wrote for i915 PMU > > (https://patchwork.freedesktop.org/

Re: [Intel-gfx] [RFC 04/11] drm/i915/pmu: Expose a PMU interface for perf queries

On Mon, 2017-09-11 at 16:25 +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > From: Chris Wilson > From: Tvrtko Ursulin > From: Dmitry Rogozhkin > > The first goal is to be able

Re: [Intel-gfx] [RFC v3 00/11] i915 PMU and engine busy stats

Hi, Just tried v3 series. perf-stat works fine. From the IGT tests which I wrote for i915 PMU (https://patchwork.freedesktop.org/series/29313/) all pass (assuming pmu.enabled will be exposed in debugfs) except cpu_online subtest. And this is pretty interesting - see details below. Ok, be

Re: [Intel-gfx] [PATCH 1/2] drm/i915/tracepoints: Don't compile-out low-level tracepoints

On Mon, 2017-09-11 at 20:52 +0100, Chris Wilson wrote: > Quoting Tvrtko Ursulin (2017-09-11 16:34:08) > > > > On 11/09/2017 14:09, Michał Winiarski wrote: > > > There's no reason to hide those tracepoints. > > > Let's also remove the DRM_I915_LOW_LEVEL_TRACEPOINTS Kconfig option. > > > > No

Re: [Intel-gfx] [RFC v2 2/3] drm/i915/pmu: serve global events and support perf stat

On Tue, 2017-08-29 at 11:28 +0200, Peter Zijlstra wrote: > On Wed, Aug 23, 2017 at 11:38:43PM +0000, Rogozhkin, Dmitry V wrote: > > On Wed, 2017-08-23 at 08:26 -0700, Dmitry Rogozhkin wrote: > > > +static cpumask_t i915_pmu_cpumask = CPU_MASK_CPU0; > > > > Peter, thi

Re: [Intel-gfx] [RFC 04/10] drm/i915: Expose a PMU interface for perf queries

On Tue, 2017-08-29 at 11:30 +0200, Peter Zijlstra wrote: > On Mon, Aug 28, 2017 at 10:43:17PM +0000, Rogozhkin, Dmitry V wrote: > > > Hi Peter, > > > > I have updated my fixes to Tvrtko's PMU, they are here: > > https://patchwork.freedesktop.org/series/28842/, and

Re: [Intel-gfx] [RFC 04/10] drm/i915: Expose a PMU interface for perf queries

On Wed, 2017-08-23 at 20:22 +0200, Peter Zijlstra wrote: > On Wed, Aug 23, 2017 at 05:51:38PM +0000, Rogozhkin, Dmitry V wrote: > > > Anyhow, returning to the metrics i915 exposes. Some metrics are just > > exposure of some counters supported already inside i915 PMU which do n

Re: [Intel-gfx] [RFC v2 2/3] drm/i915/pmu: serve global events and support perf stat

Peter, any comments? On Wed, 2017-08-23 at 23:38 +, Rogozhkin, Dmitry V wrote: > On Wed, 2017-08-23 at 08:26 -0700, Dmitry Rogozhkin wrote: > > +static cpumask_t i915_pmu_cpumask = CPU_MASK_CPU0; > > > Peter, this hardcoding of cpumask to use CPU0 works, but should I >

Re: [Intel-gfx] [RFC v2 3/3] drm/i915/pmu: deny perf driver level sampling of i915 PMU

y, August 25, 2017 11:19 AM To: Rogozhkin, Dmitry V <dmitry.v.rogozh...@intel.com>; Wilson, Chris <chris.wil...@intel.com> Subject: RE: [RFC v2 3/3] drm/i915/pmu: deny perf driver level sampling of i915 PMU Quoting Rogozhkin, Dmitry V (2017-08-25 19:06:13) > Hi Chris, not sure yo

Re: [Intel-gfx] [RFC v2 3/3] drm/i915/pmu: deny perf driver level sampling of i915 PMU

Hi Chris, Why we had event->hw->hrtimer in i915 PMU? Was there any particular reason? You had some use case which did not work? According to Peter we should not expose the timer out of our pmu, and I do not see the reason why we need it at the first place. So, I went forward and wiped it out and

Re: [Intel-gfx] [RFC v2 2/3] drm/i915/pmu: serve global events and support perf stat

On Wed, 2017-08-23 at 08:26 -0700, Dmitry Rogozhkin wrote: > +static cpumask_t i915_pmu_cpumask = CPU_MASK_CPU0; Peter, this hardcoding of cpumask to use CPU0 works, but should I implement something smarter or this will be sufficient? I see that cstate.c you have pointed me to tries to track CPUs

Re: [Intel-gfx] [RFC 04/10] drm/i915: Expose a PMU interface for perf queries

On Wed, 2017-08-23 at 20:22 +0200, Peter Zijlstra wrote: > On Wed, Aug 23, 2017 at 05:51:38PM +0000, Rogozhkin, Dmitry V wrote: > > > Anyhow, returning to the metrics i915 exposes. Some metrics are just > > exposure of some counters supported already inside i915 PMU which do n

Re: [Intel-gfx] [RFC 04/10] drm/i915: Expose a PMU interface for perf queries

On Wed, 2017-08-23 at 20:01 +0200, Peter Zijlstra wrote: > On Wed, Aug 23, 2017 at 05:51:38PM +0000, Rogozhkin, Dmitry V wrote: > > > https://patchwork.freedesktop.org/patch/171953/. This patch makes 'perf > > stat -e i915/rcs0-busy/' to error out and supports 'perf stat -e

Re: [Intel-gfx] [RFC 04/10] drm/i915: Expose a PMU interface for perf queries

On Wed, 2017-08-23 at 20:04 +0200, Peter Zijlstra wrote: > On Wed, Aug 23, 2017 at 05:51:38PM +0000, Rogozhkin, Dmitry V wrote: > > > > The above command tries to add an event 'i915/rcs0-busy/' to a task. How > > > are i915 resource associated to any one particular

Re: [Intel-gfx] [RFC 04/10] drm/i915: Expose a PMU interface for perf queries

On Tue, 2017-08-22 at 20:17 +0200, Peter Zijlstra wrote: > On Sat, Aug 12, 2017 at 02:15:13AM +0000, Rogozhkin, Dmitry V wrote: > > $ perf stat -e instructions,i915/rcs0-busy/ workload.sh > > <... wrokload.sh output...> > > > > Performance counter stats for 'wor

Re: [Intel-gfx] [RFC 2/2] drm/i915/pmu: serve global events and support perf stat

On Tue, 2017-08-15 at 10:56 -0700, Dmitry Rogozhkin wrote: > This patch should probably be squashed with Tvrtko's PMU enabling > patch... > > i915 events don't have a CPU context to work with, so i915 > PMU should work in global mode, i.e. expose perf_invalid_context. > This will make the

Re: [Intel-gfx] [PATCH] RFC drm/i915: Slaughter the thundering i915_wait_request herd

want us to experiment with? Dmitry. -Original Message- From: Chris Wilson [mailto:ch...@chris-wilson.co.uk] Sent: Wednesday, November 4, 2015 5:48 PM To: Gong, Zhipeng Cc: intel-gfx@lists.freedesktop.org; Rogozhkin, Dmitry V Subject: Re: [Intel-gfx] [PATCH] RFC drm/i915: Slaughter

Re: [Intel-gfx] [PATCH] RFC drm/i915: Slaughter the thundering i915_wait_request herd

: intel-gfx@lists.freedesktop.org; Rogozhkin, Dmitry V Subject: Re: [PATCH] RFC drm/i915: Slaughter the thundering i915_wait_request herd On Mon, Nov 02, 2015 at 05:39:54AM +, Gong, Zhipeng wrote: > Chris- > > The patch cannot be applied on the latest drm-intel-nightly directly. >