Intel-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Robert Bragg <robert@sixbynine.org>
To: intel-gfx@lists.freedesktop.org
Cc: David Airlie <airlied@linux.ie>,
	dri-devel@lists.freedesktop.org,
	Sourab Gupta <sourab.gupta@intel.com>,
	Deepak S <deepak.s@intel.com>,
	Daniel Vetter <daniel.vetter@intel.com>
Subject: [PATCH 8/9] drm/i915: add oa_event_min_timer_exponent sysctl
Date: Wed, 20 Apr 2016 15:23:13 +0100	[thread overview]
Message-ID: <1461162194-1424-9-git-send-email-robert@sixbynine.org> (raw)
In-Reply-To: <1461162194-1424-1-git-send-email-robert@sixbynine.org>

The minimal sampling period is now configurable via a
dev.i915.oa_min_timer_exponent sysctl parameter.

Following the precedent set by perf, the default is the minimum that
won't (on its own) exceed the default kernel.perf_event_max_sample_rate
default of 100000 samples/s.

Signed-off-by: Robert Bragg <robert@sixbynine.org>
---
 drivers/gpu/drm/i915/i915_perf.c | 42 ++++++++++++++++++++++++++++------------
 1 file changed, 30 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index c2ba16a..cc1cd52 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -71,6 +71,23 @@ static u32 i915_perf_stream_paranoid = true;
  */
 #define OA_EXPONENT_MAX 31
 
+/* for sysctl proc_dointvec_minmax of i915_oa_min_timer_exponent */
+static int zero;
+static int oa_exponent_max = OA_EXPONENT_MAX;
+
+/* Theoretically we can program the OA unit to sample every 160ns but don't
+ * allow that by default unless root...
+ *
+ * The period is derived from the exponent as:
+ *
+ *   period = 80ns * 2^(exponent + 1)
+ *
+ * Referring to perf's kernel.perf_event_max_sample_rate for a precedent
+ * (100000 by default); with an OA exponent of 6 we get a period of 10.240
+ * microseconds - just under 100000Hz
+ */
+static u32 i915_oa_min_timer_exponent = 6;
+
 /* XXX: beware if future OA HW adds new report formats that the current
  * code assumes all reports have a power-of-two size and ~(size - 1) can
  * be used as a mask to align the OA tail pointer.
@@ -1250,21 +1267,13 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv,
 				return -EINVAL;
 			}
 
-			/* NB: The exponent represents a period as follows:
-			 *
-			 *   80ns * 2^(period_exponent + 1)
-			 *
-			 * Theoretically we can program the OA unit to sample
+			/* Theoretically we can program the OA unit to sample
 			 * every 160ns but don't allow that by default unless
 			 * root.
-			 *
-			 * Referring to perf's
-			 * kernel.perf_event_max_sample_rate for a precedent
-			 * (100000 by default); with an OA exponent of 6 we get
-			 * a period of 10.240 microseconds -just under 100000Hz
 			 */
-			if (value < 6 && !capable(CAP_SYS_ADMIN)) {
-				DRM_ERROR("Sampling period too high without root privileges\n");
+			if (value < i915_oa_min_timer_exponent &&
+			    !capable(CAP_SYS_ADMIN)) {
+				DRM_ERROR("OA timer exponent too low without root privileges\n");
 				return -EACCES;
 			}
 
@@ -1326,6 +1335,15 @@ static struct ctl_table oa_table[] = {
 	 .mode = 0644,
 	 .proc_handler = proc_dointvec,
 	 },
+	{
+	 .procname = "oa_min_timer_exponent",
+	 .data = &i915_oa_min_timer_exponent,
+	 .maxlen = sizeof(i915_oa_min_timer_exponent),
+	 .mode = 0644,
+	 .proc_handler = proc_dointvec_minmax,
+	 .extra1 = &zero,
+	 .extra2 = &oa_exponent_max,
+	 },
 	{}
 };
 
-- 
2.7.1

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

  parent reply	other threads:[~2016-04-20 14:23 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-04-20 14:23 [PATCH 0/9] Enable Gen 7 Observation Architecture Robert Bragg
2016-04-20 14:23 ` [PATCH 1/9] drm/i915: Add i915 perf infrastructure Robert Bragg
2016-04-20 22:41   ` Chris Wilson
2016-04-20 14:23 ` [PATCH 2/9] drm/i915: rename OACONTROL GEN7_OACONTROL Robert Bragg
2016-04-20 14:23 ` [PATCH 3/9] drm/i915: don't whitelist oacontrol in cmd parser Robert Bragg
2016-04-20 14:23 ` [PATCH 4/9] drm/i915: Add 'render basic' Haswell OA unit config Robert Bragg
2016-04-20 14:23 ` [PATCH 5/9] drm/i915: Enable i915 perf stream for Haswell OA unit Robert Bragg
2016-04-20 16:16   ` kbuild test robot
2016-04-20 20:30   ` [Intel-gfx] " kbuild test robot
2016-04-20 21:11   ` Chris Wilson
2016-04-21 16:15     ` Robert Bragg
2016-04-23  8:48       ` Chris Wilson
2016-04-20 22:15   ` Chris Wilson
2016-04-20 22:46   ` Chris Wilson
2016-04-22 11:04     ` Robert Bragg
2016-04-22 11:18       ` Chris Wilson
2016-04-20 22:52   ` Chris Wilson
2016-04-21 15:43     ` Robert Bragg
2016-04-21 16:21       ` Chris Wilson
2016-04-20 23:09   ` Chris Wilson
2016-04-21 15:18     ` Robert Bragg
2016-04-22  1:10       ` Robert Bragg
2016-04-20 23:16   ` Chris Wilson
2016-04-21 15:01     ` Robert Bragg
2016-04-23 10:34   ` Martin Peres
2016-05-03 19:34     ` Robert Bragg
2016-05-03 20:03       ` Robert Bragg
2016-05-04  9:09         ` Martin Peres
2016-05-04  9:49           ` Robert Bragg
2016-05-04 12:24             ` Daniel Vetter
2016-05-04 13:24               ` Robert Bragg
2016-05-04 13:33                 ` Robert Bragg
2016-05-04 13:51                 ` Daniel Vetter
2016-05-04  9:04       ` [Intel-gfx] " Martin Peres
2016-05-04 11:15         ` Robert Bragg
2016-04-20 14:23 ` [PATCH 6/9] drm/i915: advertise available metrics via sysfs Robert Bragg
2016-04-20 14:23 ` [PATCH 7/9] drm/i915: Add dev.i915.perf_event_paranoid sysctl option Robert Bragg
2016-04-20 14:23 ` Robert Bragg [this message]
2016-04-20 14:23 ` [PATCH 9/9] drm/i915: Add more Haswell OA metric sets Robert Bragg
2016-04-20 14:56 ` [PATCH 0/9] Enable Gen 7 Observation Architecture Robert Bragg
2016-04-21  7:46 ` ✓ Fi.CI.BAT: success for Enable Gen 7 Observation Architecture (rev3) Patchwork
2016-04-21 12:41 ` ✗ Fi.CI.BAT: failure " Patchwork
2016-04-23  8:31 ` ✗ Fi.CI.BAT: warning " Patchwork
2016-04-24 17:23 ` ✓ Fi.CI.BAT: success " 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=1461162194-1424-9-git-send-email-robert@sixbynine.org \
    --to=robert@sixbynine.org \
    --cc=airlied@linux.ie \
    --cc=daniel.vetter@intel.com \
    --cc=deepak.s@intel.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=sourab.gupta@intel.com \
    /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