Created attachment 94271 [details] dmesg(nightly) System Environment: -------------------------- Platform: Sandybridge kernel: (drm-intel-nightly)1be8f2b4dd6d3db00af24d4891c82d2650bd282d Bug detailed description: --------------------------- New case pm_rps/min-max-config-loaded aborted on Sandybridge with -queued and -nightly kernel. It fails on -fixes kernel. output on -nightly kernel: IGT-Version: 1.5-g9597836 (i686) (Linux: 3.13.0_drm-intel-nightly_1be8f2_20140218+ i686) Test assertion failure function loaded_check, file pm_rps.c:460: Last errno: 22, Invalid argument Failed assertion: freqs[CUR] == freqs[MAX] Subtest min-max-config-loaded: FAIL Test assertion failure function igt_wait_helper, file drmtest.c:1279: Last errno: 10, No child processes Failed assertion: WIFEXITED(status) && WEXITSTATUS(status) == 0 pm_rps: drmtest.c:1113: igt_fail: Assertion `!test_with_subtests || in_fixture' failed. Aborted (core dumped) output on -fixes knernel: IGT-Version: 1.5-g9597836 (x86_64) (Linux: 3.13.0-rc8_drm-intel-fixes_f51a44_20140218+ x86_64) Test assertion failure function loaded_check, file pm_rps.c:460: Last errno: 2, No such file or directory Failed assertion: freqs[CUR] == freqs[MAX] Subtest min-max-config-loaded: FAIL Reproduce steps: ---------------------------- 1. ./pm_rps --run-subtest min-max-config-loaded
Does the below igt patch help? Please attach the new output if it still fails ... diff --git a/tests/pm_rps.c b/tests/pm_rps.c index a652cf580dc7..fd72ecad735d 100644 --- a/tests/pm_rps.c +++ b/tests/pm_rps.c @@ -439,7 +439,7 @@ static void idle_check(void) } #define LOADED_WAIT_TIMESTEP_MSEC 100 -#define LOADED_WAIT_TIMEOUT_MSEC 3000 +#define LOADED_WAIT_TIMEOUT_MSEC 10000 static void loaded_check(void) { int freqs[NUMFREQ]; @@ -457,7 +457,7 @@ static void loaded_check(void) wait += LOADED_WAIT_TIMESTEP_MSEC; } while (wait < LOADED_WAIT_TIMEOUT_MSEC); - igt_assert(freqs[CUR] == freqs[MAX]); + igt_assert_f(freqs[CUR] == freqs[MAX], "cur: %i, max: %i\n", freqs[CUR], freqs[MAX]); igt_debug("Required %d msec to reach cur=max\n", wait); }
Test this igt patch. output: IGT-Version: 1.5-g072d358 (i686) (Linux: 3.14.0-rc5_drm-intel-nightly_2bbdb4_20140304+ i686) Test assertion failure function loaded_check, file pm_rps.c:460: Last errno: 22, Invalid argument Failed assertion: freqs[CUR] == freqs[MAX] cur: 850, max: 1100 Subtest min-max-config-loaded: FAIL Test assertion failure function igt_wait_helper, file drmtest.c:1279: Last errno: 10, No child processes Failed assertion: WIFEXITED(status) && WEXITSTATUS(status) == 0 pm_rps: drmtest.c:1113: igt_fail: Assertion `!test_with_subtests || in_fixture' failed. Aborted (core dumped)
It also fails on Ivybridge and Haswell. Subcase min-max-config-idle and reset also fail.
Should be fixed in latest i-g-t: commit 5632bc81d944ae5070c403688b995ff583bb1ac7 Author: Daniel Vetter <daniel.vetter@ffwll.ch> Date: Fri Mar 14 10:26:39 2014 +0100 tests/pm_rps: load harder
It still fails BDW, file Bug 76247 to trace new platform. Verified.Fixed on SNB/IVB/HSW.
Subcase min-max-config-idle still failed on latest -nightly(19f629e96704b796efdad3cf13b1e8fb387d8ca7) [root@x-ivb9 tests]# ./pm_rps --run-subtest min-max-config-idle IGT-Version: 1.6-g4fdca96 (x86_64) (Linux: 3.14.0-rc8_drm-intel-nightly_19f629_20140331_debug+ x86_64) Test assertion failure function idle_check, file pm_rps.c:443: Last errno: 22, Invalid argument Failed assertion: freqs[CUR] == freqs[MIN] error: 650 == 350 Subtest min-max-config-idle: FAIL
Created attachment 96642 [details] dmesg about min-max-config-idle failed
(In reply to comment #6) > Subcase min-max-config-idle still failed on latest > -nightly(19f629e96704b796efdad3cf13b1e8fb387d8ca7) > > [root@x-ivb9 tests]# ./pm_rps --run-subtest min-max-config-idle > IGT-Version: 1.6-g4fdca96 (x86_64) (Linux: > 3.14.0-rc8_drm-intel-nightly_19f629_20140331_debug+ x86_64) > Test assertion failure function idle_check, file pm_rps.c:443: > Last errno: 22, Invalid argument > Failed assertion: freqs[CUR] == freqs[MIN] > error: 650 == 350 > Subtest min-max-config-idle: FAIL The failure occurred on IVB platform
Ah, that was the broken gpu turbo. Should be fix now with latest -nightly and -fixes.
Verified.Fixed.
Closing old verified.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.