Skip to contentSkip to conversation

tests/perf_pmu: Avoid off-lining all CPUs

PATCH In review Tvrtko Ursulin 4 participants 8 messages Feb 19, 2018 – Feb 20, 2018

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 tests/perf_pmu: Avoid off-lining all CPUs

From: Tvrtko Ursulin <[email protected]>

Bail out from the cpu-hotplug test if we failed to bring a CPU back online.

This still leaves the machine in a quite bad state, but at least it avoids hard hanging it.

---
 tests/perf_pmu.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)
tests/perf_pmu.c +13 -2
index 7fab73e22c2d..a334d3b5770e 100644
@@ -965,6 +965,7 @@ static void cpu_hotplug(int gem_fd)
 	int link[2];
 	int fd, ret;
 	int cur = 0;
+	char buf;
 
 	igt_require(cpu0_hotplug_support());
 
@@ -1012,7 +1013,15 @@ static void cpu_hotplug(int gem_fd)
 			/* Offline followed by online a CPU. */
 			igt_assert_eq(write(cpufd, "0", 2), 2);
 			usleep(1e6);
-			igt_assert_eq(write(cpufd, "1", 2), 2);
+			ret = write(cpufd, "1", 2);
+			if (ret < 0) {
+				/*
+				 * Abort the test if we failed to bring a CPU
+				 * back online.
+				 */
+				igt_assert_eq(write(link[1], "s", 1), 1);
+				break;
+			}
 
 			close(cpufd);
 			cpu++;
@@ -1026,7 +1035,6 @@ static void cpu_hotplug(int gem_fd)
 	 * until the CPU core shuffler finishes one loop.
 	 */
 	for (;;) {
-		char buf;
 		int ret2;
 
 		usleep(500e3);
@@ -1053,6 +1061,9 @@ static void cpu_hotplug(int gem_fd)
 	close(fd);
 	close(link[0]);
 
+	/* Skip if child signals a problem with bringing a CPU back online. */
+	igt_skip_on(buf == 's');
+
 	assert_within_epsilon(val, ts[1] - ts[0], tolerance);
 }
 
Signature · 6 lines
2.14.1

_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev
Reviewed-by on v1 1/1
#1 lore raw

Quoting Tvrtko Ursulin (2018-02-19 13:40:37)

Quoted 51 lines

From: Tvrtko Ursulin <[email protected]>

Bail out from the cpu-hotplug test if we failed to bring a CPU back online.

This still leaves the machine in a quite bad state, but at least it avoids hard hanging it.

---
 tests/perf_pmu.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)
tests/perf_pmu.c +10 -2
index 7fab73e22c2d..a334d3b5770e 100644
@@ -965,6 +965,7 @@ static void cpu_hotplug(int gem_fd)
        int link[2];
        int fd, ret;
        int cur = 0;
+       char buf;
 
        igt_require(cpu0_hotplug_support());
 
@@ -1012,7 +1013,15 @@ static void cpu_hotplug(int gem_fd)
                        /* Offline followed by online a CPU. */
                        igt_assert_eq(write(cpufd, "0", 2), 2);
                        usleep(1e6);
-                       igt_assert_eq(write(cpufd, "1", 2), 2);
+                       ret = write(cpufd, "1", 2);
+                       if (ret < 0) {
+                               /*
+                                * Abort the test if we failed to bring a CPU
+                                * back online.
+                                */
+                               igt_assert_eq(write(link[1], "s", 1), 1);
+                               break;
+                       }
 
                        close(cpufd);
                        cpu++;
@@ -1026,7 +1035,6 @@ static void cpu_hotplug(int gem_fd)
         * until the CPU core shuffler finishes one loop.
         */
        for (;;) {
-               char buf;
                int ret2;
 
                usleep(500e3);

tests/perf_pmu.c static void cpu_hotplug(int gem_fd) · line 1065

Quoted 5 lines
tests/perf_pmu.c +2 -0
@@ -1053,6 +1061,9 @@ static void cpu_hotplug(int gem_fd)
        close(fd);
        close(link[0]);
 
+       /* Skip if child signals a problem with bringing a CPU back online. */
+       igt_skip_on(buf == 's');

The logic makes sense.

What happens if we try to online an already on cpu? Will that report the failure or just tell us to stop wasting its time?
-Chris
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev

#2 lore raw
Quoted 55 lines · Chris Wilson

Quoting Tvrtko Ursulin (2018-02-19 13:40:37)

Quoted 54 lines

From: Tvrtko Ursulin <[email protected]>

Bail out from the cpu-hotplug test if we failed to bring a CPU back online.

This still leaves the machine in a quite bad state, but at least it avoids hard hanging it.

---
  tests/perf_pmu.c | 15 +++++++++++++--
  1 file changed, 13 insertions(+), 2 deletions(-)
tests/perf_pmu.c +10 -2
index 7fab73e22c2d..a334d3b5770e 100644
@@ -965,6 +965,7 @@ static void cpu_hotplug(int gem_fd)
         int link[2];
         int fd, ret;
         int cur = 0;
+       char buf;
 
         igt_require(cpu0_hotplug_support());
 
@@ -1012,7 +1013,15 @@ static void cpu_hotplug(int gem_fd)
                         /* Offline followed by online a CPU. */
                         igt_assert_eq(write(cpufd, "0", 2), 2);
                         usleep(1e6);
-                       igt_assert_eq(write(cpufd, "1", 2), 2);
+                       ret = write(cpufd, "1", 2);
+                       if (ret < 0) {
+                               /*
+                                * Abort the test if we failed to bring a CPU
+                                * back online.
+                                */
+                               igt_assert_eq(write(link[1], "s", 1), 1);
+                               break;
+                       }
 
                         close(cpufd);
                         cpu++;
@@ -1026,7 +1035,6 @@ static void cpu_hotplug(int gem_fd)
          * until the CPU core shuffler finishes one loop.
          */
         for (;;) {
-               char buf;
                 int ret2;
 
                 usleep(500e3);
@@ -1053,6 +1061,9 @@ static void cpu_hotplug(int gem_fd)
         close(fd);
         close(link[0]);
 
Quoted 5 lines

tests/perf_pmu.c static void cpu_hotplug(int gem_fd) · line 1065

Quoted 2 lines
tests/perf_pmu.c +2 -0
+       /* Skip if child signals a problem with bringing a CPU back online. */
+       igt_skip_on(buf == 's');

The logic makes sense.

I wasn't so sure after I've sent it. There was an assert on write error already so that would have aborted the child. Will see what shards will say..

Quoted 2 lines

What happens if we try to online an already on cpu? Will that report the failure or just tell us to stop wasting its time?

Seems the attempt is simply ignored.

Regards,

Tvrtko
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev

#3 lore raw

[igt-dev] ✓ Fi.CI.BAT: success for tests/perf_pmu: Avoid off-lining all CPUs

In reply to Tvrtko Ursulin
== Series Details ==

== Series Details ==

Series: tests/perf_pmu: Avoid off-lining all CPUs
URL : https://patchwork.freedesktop.org/series/38513/
State : success

== Summary ==

IGT patchset tested on top of latest successful build bf777c92448d51010aba51d1f1b657b0fdc673a6 tests/gem_ctx_param: Update invalid param

with latest DRM-Tip kernel build CI_DRM_3797
282bcdb38286 drm-tip: 2018y-02m-19d-12h-01m-54s UTC integration manifest

No testlist changes.

Test gem_ctx_switch:

        Subgroup basic-default-heavy:
                pass       -> INCOMPLETE (fi-cnl-y3) fdo#105086
Test kms_chamelium:
        Subgroup dp-edid-read:
                fail       -> PASS       (fi-kbl-7500u) fdo#102505
        Subgroup hdmi-hpd-fast:
                skip       -> FAIL       (fi-kbl-7500u) fdo#102672
Test kms_pipe_crc_basic:
        Subgroup suspend-read-crc-pipe-b:
                pass       -> INCOMPLETE (fi-snb-2520m) fdo#103713

fdo#105086 https://bugs.freedesktop.org/show_bug.cgi?id=105086
fdo#102505 https://bugs.freedesktop.org/show_bug.cgi?id=102505
fdo#102672 https://bugs.freedesktop.org/show_bug.cgi?id=102672
fdo#103713 https://bugs.freedesktop.org/show_bug.cgi?id=103713

fi-bdw-5557u     total:288  pass:267  dwarn:0   dfail:0   fail:0   skip:21  time:425s
fi-bdw-gvtdvm    total:288  pass:264  dwarn:0   dfail:0   fail:0   skip:24  time:431s
fi-blb-e6850     total:288  pass:223  dwarn:1   dfail:0   fail:0   skip:64  time:375s
fi-bsw-n3050     total:288  pass:242  dwarn:0   dfail:0   fail:0   skip:46  time:491s
fi-bwr-2160      total:288  pass:183  dwarn:0   dfail:0   fail:0   skip:105 time:288s
fi-bxt-dsi       total:288  pass:258  dwarn:0   dfail:0   fail:0   skip:30  time:484s
fi-bxt-j4205     total:288  pass:259  dwarn:0   dfail:0   fail:0   skip:29  time:488s
fi-byt-j1900     total:288  pass:253  dwarn:0   dfail:0   fail:0   skip:35  time:481s
fi-byt-n2820     total:288  pass:249  dwarn:0   dfail:0   fail:0   skip:39  time:464s
fi-cfl-s2        total:288  pass:262  dwarn:0   dfail:0   fail:0   skip:26  time:566s
fi-cnl-y3        total:22   pass:21   dwarn:0   dfail:0   fail:0   skip:0  
fi-elk-e7500     total:288  pass:229  dwarn:0   dfail:0   fail:0   skip:59  time:418s
fi-gdg-551       total:288  pass:179  dwarn:0   dfail:0   fail:1   skip:108 time:286s
fi-glk-1         total:288  pass:260  dwarn:0   dfail:0   fail:0   skip:28  time:511s
fi-hsw-4770      total:288  pass:261  dwarn:0   dfail:0   fail:0   skip:27  time:392s
fi-ilk-650       total:288  pass:228  dwarn:0   dfail:0   fail:0   skip:60  time:413s
fi-ivb-3520m     total:288  pass:259  dwarn:0   dfail:0   fail:0   skip:29  time:462s
fi-ivb-3770      total:288  pass:255  dwarn:0   dfail:0   fail:0   skip:33  time:412s
fi-kbl-7500u     total:288  pass:263  dwarn:1   dfail:0   fail:1   skip:23  time:458s
fi-kbl-7560u     total:288  pass:269  dwarn:0   dfail:0   fail:0   skip:19  time:492s
fi-kbl-7567u     total:288  pass:268  dwarn:0   dfail:0   fail:0   skip:20  time:454s
fi-kbl-r         total:288  pass:261  dwarn:0   dfail:0   fail:0   skip:27  time:501s
fi-pnv-d510      total:288  pass:222  dwarn:1   dfail:0   fail:0   skip:65  time:593s
fi-skl-6260u     total:288  pass:268  dwarn:0   dfail:0   fail:0   skip:20  time:434s
fi-skl-6600u     total:288  pass:261  dwarn:0   dfail:0   fail:0   skip:27  time:512s
fi-skl-6700hq    total:288  pass:262  dwarn:0   dfail:0   fail:0   skip:26  time:525s
fi-skl-6700k2    total:288  pass:264  dwarn:0   dfail:0   fail:0   skip:24  time:496s
fi-skl-6770hq    total:288  pass:268  dwarn:0   dfail:0   fail:0   skip:20  time:489s
fi-skl-guc       total:288  pass:260  dwarn:0   dfail:0   fail:0   skip:28  time:412s
fi-skl-gvtdvm    total:288  pass:265  dwarn:0   dfail:0   fail:0   skip:23  time:430s
fi-snb-2520m     total:245  pass:211  dwarn:0   dfail:0   fail:0   skip:33 
fi-snb-2600      total:288  pass:248  dwarn:0   dfail:0   fail:0   skip:40  time:399s

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_951/issues.html
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev

#4 lore raw

[igt-dev] ✗ Fi.CI.IGT: failure for tests/perf_pmu: Avoid off-lining all CPUs

In reply to Tvrtko Ursulin
== Series Details ==

== Series Details ==

Series: tests/perf_pmu: Avoid off-lining all CPUs
URL : https://patchwork.freedesktop.org/series/38513/
State : failure

== Summary ==

Test kms_pipe_crc_basic:

        Subgroup suspend-read-crc-pipe-a:
                skip       -> PASS       (shard-snb) fdo#103375 +1
Test perf:
        Subgroup polling:
                pass       -> FAIL       (shard-hsw) fdo#102252
        Subgroup oa-exponents:
                fail       -> PASS       (shard-apl) fdo#102254
Test gem_eio:
        Subgroup in-flight-suspend:
                fail       -> PASS       (shard-hsw) fdo#104676
Test kms_cursor_crc:
        Subgroup cursor-64x64-suspend:
                incomplete -> PASS       (shard-hsw) fdo#103540
Test perf_pmu:
        Subgroup cpu-hotplug:
                incomplete -> SKIP       (shard-apl) fdo#104965
        Subgroup other-read-1:
                pass       -> FAIL       (shard-apl)
Test kms_frontbuffer_tracking:
        Subgroup fbc-1p-pri-indfb-multidraw:
                pass       -> FAIL       (shard-snb) fdo#103167
        Subgroup fbc-1p-offscren-pri-shrfb-draw-mmap-cpu:
                pass       -> FAIL       (shard-apl) fdo#101623
Test gem_exec_capture:
        Subgroup capture-vebox:
                pass       -> INCOMPLETE (shard-apl)

fdo#103375 https://bugs.freedesktop.org/show_bug.cgi?id=103375
fdo#102252 https://bugs.freedesktop.org/show_bug.cgi?id=102252
fdo#102254 https://bugs.freedesktop.org/show_bug.cgi?id=102254
fdo#104676 https://bugs.freedesktop.org/show_bug.cgi?id=104676
fdo#103540 https://bugs.freedesktop.org/show_bug.cgi?id=103540
fdo#104965 https://bugs.freedesktop.org/show_bug.cgi?id=104965
fdo#103167 https://bugs.freedesktop.org/show_bug.cgi?id=103167
fdo#101623 https://bugs.freedesktop.org/show_bug.cgi?id=101623

shard-apl        total:3429 pass:1793 dwarn:1   dfail:0   fail:14  skip:1620 time:12125s
shard-hsw        total:3434 pass:1761 dwarn:1   dfail:0   fail:3   skip:1668 time:11649s
shard-snb        total:3434 pass:1350 dwarn:1   dfail:0   fail:3   skip:2080 time:6549s

Blacklisted hosts:
shard-kbl total:3413 pass:1911 dwarn:1 dfail:0 fail:12 skip:1488 time:9581s

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_951/shards.html
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev

#5 lore raw

Re: [igt-dev] ✗ Fi.CI.IGT: failure for tests/perf_pmu: Avoid off-lining all CPUs

Quoting Patchwork (2018-02-19 15:55:19)

Quoted 5 lines

Test perf_pmu:

        Subgroup cpu-hotplug:
                incomplete -> SKIP       (shard-apl) fdo#104965
        Subgroup other-read-1:
                pass       -> FAIL       (shard-apl)

I guess that's the expected fallout from deciding to skip after trying to offline a cpu.

I also guess that makes it a no-go (without some tweaking) if we cause random fallout in subsequent tests, i.e. unpredictable flip-flops. -Chris
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev

#6 lore raw

Re: [igt-dev] ✗ Fi.CI.IGT: failure for tests/perf_pmu: Avoid off-lining all CPUs

Quoted 9 lines · Chris Wilson

Quoting Patchwork (2018-02-19 15:55:19)

Quoted 5 lines

Test perf_pmu:

         Subgroup cpu-hotplug:
                 incomplete -> SKIP       (shard-apl) fdo#104965
         Subgroup other-read-1:
                 pass       -> FAIL       (shard-apl)

I guess that's the expected fallout from deciding to skip after trying to offline a cpu.

Kind of expected because I am not quite sure why it behaves differently from the existing code which has igt_assert on the write. Maybe I am particularly blind today.

Quoted 2 lines

I also guess that makes it a no-go (without some tweaking) if we cause random fallout in subsequent tests, i.e. unpredictable flip-flops.

Yep agreed 100%.

Sounds like a case for igt_fail2 = please reboot me, or igt_wedge_machine.

A bit lighter weight, but with weaker coverage, maybe adding i915.debug_pmu_cpu=N modparam (present only when I915_DEBUG is set), and living with not testing the cpu0 offline paths?

Regards,

Tvrtko
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev

#7 lore raw

Re: [igt-dev] ✗ Fi.CI.IGT: failure for tests/perf_pmu: Avoid off-lining all CPUs

Quoting Tvrtko Ursulin (2018-02-19 17:53:06)

Quoted 16 lines · Chris Wilson
Quoted 9 lines

Quoting Patchwork (2018-02-19 15:55:19)

Quoted 5 lines

Test perf_pmu:

         Subgroup cpu-hotplug:
                 incomplete -> SKIP       (shard-apl) fdo#104965
         Subgroup other-read-1:
                 pass       -> FAIL       (shard-apl)

I guess that's the expected fallout from deciding to skip after trying to offline a cpu.

Kind of expected because I am not quite sure why it behaves differently from the existing code which has igt_assert on the write. Maybe I am particularly blind today.

Quoted 6 lines
Quoted 2 lines

I also guess that makes it a no-go (without some tweaking) if we cause random fallout in subsequent tests, i.e. unpredictable flip-flops.

Yep agreed 100%.

Sounds like a case for igt_fail2 = please reboot me, or igt_wedge_machine.

Indeed, igt_sysrq_reboot() sounds like a useful intermediate patch. -Chris
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev