public inbox for intel-gfx@lists.freedesktop.org
 help / color / mirror / Atom feed
From: "Rogozhkin, Dmitry V" <dmitry.v.rogozhkin@intel.com>
To: "tursulin@ursulin.net" <tursulin@ursulin.net>
Cc: "peterz@infradead.org" <peterz@infradead.org>,
	"Intel-gfx@lists.freedesktop.org"
	<Intel-gfx@lists.freedesktop.org>
Subject: Re: [RFC v7 04/11] drm/i915/pmu: Expose a PMU interface for perf queries
Date: Fri, 15 Sep 2017 00:00:25 +0000	[thread overview]
Message-ID: <1505404694.18667.3.camel@intel.com> (raw)
In-Reply-To: <20170913103407.32332-1-tvrtko.ursulin@linux.intel.com>

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 != event->pmu->type)
> +               return -ENOENT;
> +
> +       /* unsupported modes and filters */
> +       if (event->attr.sample_period) /* no sampling */
> +               return -EINVAL;
> +
> +       if (has_branch_stack(event))
> +               return -EOPNOTSUPP;
> +
> +       if (event->cpu < 0)
> +               return -EINVAL;
> +
> +       cpu = cpumask_any_and(&i915_pmu_cpumask,
> +                             topology_sibling_cpumask(event->cpu));
> +       if (cpu >= nr_cpu_ids)
> +               return -ENODEV;
> +
> +       ret = 0;
> +       if (is_engine_event(event)) {
> +               ret = engine_event_init(event);
> +       } else switch (event->attr.config) {
> +       case I915_PMU_ACTUAL_FREQUENCY:
> +               if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915))
> +                       ret = -ENODEV; /* requires a mutex for
> sampling! */
> +       case I915_PMU_REQUESTED_FREQUENCY:
> +       case I915_PMU_ENERGY:
> +       case I915_PMU_RC6_RESIDENCY:
> +       case I915_PMU_RC6p_RESIDENCY:
> +       case I915_PMU_RC6pp_RESIDENCY:
> +               if (INTEL_GEN(i915) < 6)
> +                       ret = -ENODEV;
> +               break;
> +       }
> +       if (ret)
> +               return ret;

The switch for non-engine events should error out by default:

diff --git a/drivers/gpu/drm/i915/i915_pmu.c
b/drivers/gpu/drm/i915/i915_pmu.c
index d734879..3145e9a 100644
--- a/drivers/gpu/drm/i915/i915_pmu.c
+++ b/drivers/gpu/drm/i915/i915_pmu.c
@@ -329,6 +329,9 @@ static int i915_pmu_event_init(struct perf_event
*event)
                if (INTEL_GEN(i915) < 6)
                        ret = -ENODEV;
                break;
+       default:
+               ret = -ENOENT;
+               break;
        }
        if (ret)
                return ret;


Otherwise user may try to enable non-existing metric (> I915_PMU_LAST)
and eventually will be subject to kernel panic on
i915_pmu_enable/disable during refcount operations. And we need to have
an IGT test to check that.

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  reply	other threads:[~2017-09-15  0:00 UTC|newest]

Thread overview: 56+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-09-11 15:25 [RFC v3 00/11] i915 PMU and engine busy stats Tvrtko Ursulin
2017-09-11 15:25 ` [RFC 01/11] drm/i915: Convert intel_rc6_residency_us to ns Tvrtko Ursulin
2017-09-14 19:48   ` Chris Wilson
2017-09-11 15:25 ` [RFC 02/11] drm/i915: Add intel_energy_uJ Tvrtko Ursulin
2017-09-14 19:49   ` Chris Wilson
2017-09-15  9:18     ` Tvrtko Ursulin
2017-09-14 20:36   ` Ville Syrjälä
2017-09-15  6:56     ` Tvrtko Ursulin
2017-09-15  8:51       ` Chris Wilson
2017-09-15 10:07         ` Tvrtko Ursulin
2017-09-15 10:34           ` Ville Syrjälä
2017-09-15 10:38             ` Chris Wilson
2017-09-15 11:16               ` Tvrtko Ursulin
2017-09-11 15:25 ` [RFC 03/11] drm/i915: Extract intel_get_cagf Tvrtko Ursulin
2017-09-14 19:51   ` Chris Wilson
2017-09-11 15:25 ` [RFC 04/11] drm/i915/pmu: Expose a PMU interface for perf queries Tvrtko Ursulin
2017-09-12  2:06   ` Rogozhkin, Dmitry V
2017-09-12 14:59     ` Tvrtko Ursulin
2017-09-13  8:57       ` [RFC v6 " Tvrtko Ursulin
2017-09-13 10:34         ` [RFC v7 " Tvrtko Ursulin
2017-09-15  0:00           ` Rogozhkin, Dmitry V [this message]
2017-09-15  7:57             ` Tvrtko Ursulin
2017-09-14 19:46   ` [RFC " Chris Wilson
2017-09-11 15:25 ` [RFC 05/11] drm/i915/pmu: Suspend sampling when GPU is idle Tvrtko Ursulin
2017-09-13 10:34   ` [RFC v5 " Tvrtko Ursulin
2017-09-14 19:57     ` Chris Wilson
2017-09-15  9:22       ` Tvrtko Ursulin
2017-09-11 15:25 ` [RFC 06/11] drm/i915: Wrap context schedule notification Tvrtko Ursulin
2017-09-11 15:25 ` [RFC 07/11] drm/i915: Engine busy time tracking Tvrtko Ursulin
2017-09-14 20:16   ` Chris Wilson
2017-09-15  9:45     ` Tvrtko Ursulin
2017-09-11 15:25 ` [RFC 08/11] drm/i915: Export engine busy stats in debugfs Tvrtko Ursulin
2017-09-14 20:17   ` Chris Wilson
2017-09-15  9:46     ` Tvrtko Ursulin
2017-09-11 15:25 ` [RFC 09/11] drm/i915/pmu: Wire up engine busy stats to PMU Tvrtko Ursulin
2017-09-11 15:25 ` [RFC 10/11] drm/i915: Export engine stats API to other users Tvrtko Ursulin
2017-09-12 18:35   ` Ben Widawsky
2017-09-14 20:26   ` Chris Wilson
2017-09-15  9:49     ` Tvrtko Ursulin
2017-09-19 19:50       ` Ben Widawsky
2017-09-19 20:11         ` Rogozhkin, Dmitry V
2017-09-29 10:59   ` Joonas Lahtinen
2017-09-11 15:25 ` [RFC 11/11] drm/i915: Gate engine stats collection with a static key Tvrtko Ursulin
2017-09-13 12:18   ` [RFC v3 " Tvrtko Ursulin
2017-09-14 20:22     ` Chris Wilson
2017-09-15  9:51       ` Tvrtko Ursulin
2017-09-11 15:50 ` ✗ Fi.CI.BAT: warning for i915 PMU and engine busy stats (rev3) Patchwork
2017-09-12  2:03 ` [RFC v3 00/11] i915 PMU and engine busy stats Rogozhkin, Dmitry V
2017-09-12 14:54   ` Tvrtko Ursulin
2017-09-12 22:01     ` Rogozhkin, Dmitry V
2017-09-13  8:54       ` [RFC v6 04/11] drm/i915/pmu: Expose a PMU interface for perf queries Tvrtko Ursulin
2017-09-13  9:01       ` [RFC v3 00/11] i915 PMU and engine busy stats Tvrtko Ursulin
2017-09-13  9:34 ` ✗ Fi.CI.BAT: warning for i915 PMU and engine busy stats (rev4) Patchwork
2017-09-13 10:46 ` ✗ Fi.CI.BAT: failure for i915 PMU and engine busy stats (rev6) Patchwork
2017-09-13 13:27 ` ✓ Fi.CI.BAT: success for i915 PMU and engine busy stats (rev7) Patchwork
2017-09-13 21:24 ` ✓ Fi.CI.IGT: " Patchwork

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1505404694.18667.3.camel@intel.com \
    --to=dmitry.v.rogozhkin@intel.com \
    --cc=Intel-gfx@lists.freedesktop.org \
    --cc=peterz@infradead.org \
    --cc=tursulin@ursulin.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox