public inbox for igt-dev@lists.freedesktop.org
 help / color / mirror / Atom feed
* [igt-dev] [PATCH i-g-t] igt/pm_rps: Clear previous high load on high->low transition
@ 2018-09-04 12:49 Chris Wilson
  2018-09-04 13:13 ` [igt-dev] ✗ Fi.CI.BAT: failure for " Patchwork
  2018-09-04 13:50 ` [igt-dev] [PATCH i-g-t] " Katarzyna Dec
  0 siblings, 2 replies; 3+ messages in thread
From: Chris Wilson @ 2018-09-04 12:49 UTC (permalink / raw)
  To: igt-dev; +Cc: intel-gfx

Make sure we do flush out the previous spinner and delay signaling
transition completion until we do.

References: https://bugs.freedesktop.org/show_bug.cgi?id=102250
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Katarzyna Dec <katarzyna.dec@intel.com>
---
 tests/pm_rps.c | 26 +++++++++++++++++++++-----
 1 file changed, 21 insertions(+), 5 deletions(-)

diff --git a/tests/pm_rps.c b/tests/pm_rps.c
index 84e71fa8a..23b33f41b 100644
--- a/tests/pm_rps.c
+++ b/tests/pm_rps.c
@@ -218,6 +218,7 @@ static void load_helper_set_load(enum load load)
 
 static void load_helper_run(enum load load)
 {
+	bool dummy;
 	int link[2];
 
 	/*
@@ -233,13 +234,14 @@ static void load_helper_run(enum load load)
 
 	lh.exit = false;
 	lh.load = load;
-	lh.signal = false;
+	lh.signal = true;
 
 	pipe(link);
 	lh.link = link[1];
 
 	igt_fork_helper(&lh.igt_proc) {
 		igt_spin_t *spin[2] = {};
+		bool prev_load;
 		uint32_t handle;
 
 		signal(SIGUSR1, load_helper_signal_handler);
@@ -247,10 +249,14 @@ static void load_helper_run(enum load load)
 
 		igt_debug("Applying %s load...\n", lh.load ? "high" : "low");
 
+		prev_load = lh.load == HIGH;
 		spin[0] = __igt_spin_batch_new(drm_fd);
-		if (lh.load == HIGH)
+		if (prev_load)
 			spin[1] = __igt_spin_batch_new(drm_fd);
+		prev_load = !prev_load; /* send the initial signal */
 		while (!lh.exit) {
+			bool high_load;
+
 			handle = spin[0]->handle;
 			igt_spin_batch_end(spin[0]);
 			while (gem_bo_busy(drm_fd, handle))
@@ -259,13 +265,20 @@ static void load_helper_run(enum load load)
 			igt_spin_batch_free(drm_fd, spin[0]);
 			usleep(100);
 
-			spin[0] = spin[1];
-			spin[lh.load == HIGH] = __igt_spin_batch_new(drm_fd);
+			high_load = lh.load == HIGH;
+			if (!high_load && spin[1]) {
+				igt_spin_batch_free(drm_fd, spin[1]);
+				spin[1] = NULL;
+			} else {
+				spin[0] = spin[1];
+			}
+			spin[high_load] = __igt_spin_batch_new(drm_fd);
 
-			if (lh.signal) {
+			if (lh.signal && high_load != prev_load) {
 				write(lh.link, &lh.signal, sizeof(lh.signal));
 				lh.signal = false;
 			}
+			prev_load = high_load;
 		}
 
 		handle = spin[0]->handle;
@@ -294,6 +307,9 @@ static void load_helper_run(enum load load)
 
 	close(lh.link);
 	lh.link = link[0];
+
+	/* wait for our helper to complete its first round */
+	igt_assert_eq(read(lh.link, &dummy, sizeof(dummy)), sizeof(dummy));
 }
 
 static void load_helper_stop(void)
-- 
2.19.0.rc1

_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

^ permalink raw reply related	[flat|nested] 3+ messages in thread

* [igt-dev] ✗ Fi.CI.BAT: failure for igt/pm_rps: Clear previous high load on high->low transition
  2018-09-04 12:49 [igt-dev] [PATCH i-g-t] igt/pm_rps: Clear previous high load on high->low transition Chris Wilson
@ 2018-09-04 13:13 ` Patchwork
  2018-09-04 13:50 ` [igt-dev] [PATCH i-g-t] " Katarzyna Dec
  1 sibling, 0 replies; 3+ messages in thread
From: Patchwork @ 2018-09-04 13:13 UTC (permalink / raw)
  To: Chris Wilson; +Cc: igt-dev

== Series Details ==

Series: igt/pm_rps: Clear previous high load on high->low transition
URL   : https://patchwork.freedesktop.org/series/49127/
State : failure

== Summary ==

= CI Bug Log - changes from CI_DRM_4763 -> IGTPW_1784 =

== Summary - FAILURE ==

  Serious unknown changes coming with IGTPW_1784 absolutely need to be
  verified manually.
  
  If you think the reported changes have nothing to do with the changes
  introduced in IGTPW_1784, please notify your bug team to allow them
  to document this new failure mode, which will reduce false positives in CI.

  External URL: https://patchwork.freedesktop.org/api/1.0/series/49127/revisions/1/mbox/

== Possible new issues ==

  Here are the unknown changes that may have been introduced in IGTPW_1784:

  === IGT changes ===

    ==== Possible regressions ====

    igt@gem_mmap_gtt@basic-read-no-prefault:
      fi-blb-e6850:       PASS -> INCOMPLETE

    
== Known issues ==

  Here are the changes found in IGTPW_1784 that come from known issues:

  === IGT changes ===

    ==== Issues hit ====

    igt@kms_frontbuffer_tracking@basic:
      fi-byt-clapper:     PASS -> FAIL (fdo#103167)

    igt@kms_pipe_crc_basic@hang-read-crc-pipe-a:
      fi-byt-clapper:     PASS -> FAIL (fdo#103191, fdo#107362)

    
    ==== Possible fixes ====

    igt@kms_pipe_crc_basic@read-crc-pipe-b-frame-sequence:
      fi-byt-clapper:     FAIL (fdo#103191, fdo#107362) -> PASS

    igt@pm_rpm@module-reload:
      fi-glk-dsi:         WARN (fdo#107602, fdo#107708) -> PASS
      fi-byt-clapper:     WARN (fdo#107602, fdo#107708) -> PASS

    
  fdo#103167 https://bugs.freedesktop.org/show_bug.cgi?id=103167
  fdo#103191 https://bugs.freedesktop.org/show_bug.cgi?id=103191
  fdo#107362 https://bugs.freedesktop.org/show_bug.cgi?id=107362
  fdo#107602 https://bugs.freedesktop.org/show_bug.cgi?id=107602
  fdo#107708 https://bugs.freedesktop.org/show_bug.cgi?id=107708


== Participating hosts (52 -> 25) ==

  Missing    (27): fi-kbl-soraka fi-bdw-gvtdvm fi-bxt-j4205 fi-pnv-d510 fi-skl-6700hq fi-skl-6600u fi-hsw-4770r fi-bxt-dsi fi-bsw-n3050 fi-ilk-650 fi-kbl-7500u fi-ctg-p8600 fi-ivb-3770 fi-elk-e7500 fi-ivb-3520m fi-kbl-7560u fi-kbl-7567u fi-ilk-m540 fi-skl-gvtdvm fi-skl-guc fi-cfl-s3 fi-cnl-psr fi-hsw-4200u fi-byt-squawks fi-bsw-cyan fi-kbl-guc fi-kbl-x1275 


== Build changes ==

    * IGT: IGT_4622 -> IGTPW_1784

  CI_DRM_4763: 1f8c06844acac7a349fb80471afcc09f33c6cfc0 @ git://anongit.freedesktop.org/gfx-ci/linux
  IGTPW_1784: https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_1784/
  IGT_4622: 022be555443eaa3317da6a9a451cf2c9dfcd6ab8 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_1784/issues.html
_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [igt-dev] [PATCH i-g-t] igt/pm_rps: Clear previous high load on high->low transition
  2018-09-04 12:49 [igt-dev] [PATCH i-g-t] igt/pm_rps: Clear previous high load on high->low transition Chris Wilson
  2018-09-04 13:13 ` [igt-dev] ✗ Fi.CI.BAT: failure for " Patchwork
@ 2018-09-04 13:50 ` Katarzyna Dec
  1 sibling, 0 replies; 3+ messages in thread
From: Katarzyna Dec @ 2018-09-04 13:50 UTC (permalink / raw)
  To: Chris Wilson; +Cc: igt-dev, intel-gfx

On Tue, Sep 04, 2018 at 01:49:56PM +0100, Chris Wilson wrote:
> Make sure we do flush out the previous spinner and delay signaling
> transition completion until we do.
> 
> References: https://bugs.freedesktop.org/show_bug.cgi?id=102250
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Katarzyna Dec <katarzyna.dec@intel.com>
> ---
>  tests/pm_rps.c | 26 +++++++++++++++++++++-----
>  1 file changed, 21 insertions(+), 5 deletions(-)
> 
> diff --git a/tests/pm_rps.c b/tests/pm_rps.c
> index 84e71fa8a..23b33f41b 100644
> --- a/tests/pm_rps.c
> +++ b/tests/pm_rps.c
> @@ -218,6 +218,7 @@ static void load_helper_set_load(enum load load)
>  
>  static void load_helper_run(enum load load)
>  {
> +	bool dummy;
>  	int link[2];
>  
>  	/*
> @@ -233,13 +234,14 @@ static void load_helper_run(enum load load)
>  
>  	lh.exit = false;
>  	lh.load = load;
> -	lh.signal = false;
> +	lh.signal = true;
>  
>  	pipe(link);
>  	lh.link = link[1];
>  
>  	igt_fork_helper(&lh.igt_proc) {
>  		igt_spin_t *spin[2] = {};
> +		bool prev_load;
>  		uint32_t handle;
>  
>  		signal(SIGUSR1, load_helper_signal_handler);
> @@ -247,10 +249,14 @@ static void load_helper_run(enum load load)
>  
>  		igt_debug("Applying %s load...\n", lh.load ? "high" : "low");
>  
> +		prev_load = lh.load == HIGH;
I would add parenthesis to improve clarity: 
	prev_load = (lh.load == HIGH);

Despite this:
Reviewed-by: Katarzyna Dec <katarzyna.dec@intel.com>
_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2018-09-04 13:50 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-09-04 12:49 [igt-dev] [PATCH i-g-t] igt/pm_rps: Clear previous high load on high->low transition Chris Wilson
2018-09-04 13:13 ` [igt-dev] ✗ Fi.CI.BAT: failure for " Patchwork
2018-09-04 13:50 ` [igt-dev] [PATCH i-g-t] " Katarzyna Dec

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox