| What | Removed | Added |
|---|---|---|
| QA Contact | intel-gfx-bugs@lists.freedesktop.org | |
| Component | DRM/Intel | IGT |
| Assignee | intel-gfx-bugs@lists.freedesktop.org | dri-devel@lists.freedesktop.org |
even with 16GB of ram in APL the following tests keeps failing= : Tests cases =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D igt@gem_userptr_blits@mlocked-unsync-normal igt@gem_userptr_blits@mlocked-normal-sync output : =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D IGT-Version: 1.15-g3ce58b6 (x86_64) (Linux: 4.7.0-rc2-drm-intel-nightly-ww24-commit-55d1291+ x86_64) (gem_userptr_blits:1332) drmtest-DEBUG: Test requirement passed: fd >=3D= 0 (gem_userptr_blits:1332) ioctl-wrappers-DEBUG: Test requirement passed: !(r= et =3D=3D ENODEV && (flags & LOCAL_I915_USERPTR_UNSYNCHRONIZED) = =3D=3D 0 && !read_only) (gem_userptr_blits:1332) DEBUG: Test requirement passed: !(ret =3D=3D 0) Aperture size is 268435456 MiB Total RAM is 15,884 MiB Not enough RAM to run test, reducing buffer count. Testing unsynchronized mappings... (gem_userptr_blits:1332) igt-core-DEBUG: Starting subtest: mlocked-unsync-normal (gem_userptr_blits:1332) intel-os-DEBUG: Checking 256 surfaces of size 1,048,576 bytes (total 268,566,528) against RAM (gem_userptr_blits:1332) intel-os-DEBUG: Test requirement passed: __intel_check_memory(count, size, mode, &required, &total) (gem_userptr_blits:1332) igt-core-DEBUG: Test requirement passed: !igt_run_in_simulation() (gem_userptr_blits:1332) DEBUG: Test requirement passed: pin > sz (gem_userptr_blits:1332) DEBUG: Pinning [15,428, 15,684] MiB Killed relevant kernel messages: =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D kern :err : [Sat Jun 18 16:18:55 2016] 32 and 0 pages still available in= the bound and unbound GPU page lists. kern :err : [Sat Jun 18 16:18:55 2016] Out of memory: Kill process 1348 (gem_userptr_bli) score 1762 or sacrifice child kern :err : [Sat Jun 18 16:18:55 2016] Killed process 1348 (gem_userptr_= bli) total-vm:16169932kB, anon-rss:15988692kB, file-rss:4kB, shmem-rss:0kB looks like that is something wrong in the test itself of the test needs much more memory ?
Guys have you actually looked at the following line ? Aperture size is 268435456 MiB -- That is 256 TiB !!! That's rather impossible amount if you ask me. So there's either a bug in I= GT's gem_aperture_size() or one of the two ioctls (I915_GEM_CONTEXT_GETPARAM I915_GEM_GET_APERTURE) that it uses. With a couple of print statements you should be able to quickly track the e= xact offender. Good luck !
(In reply to Emil Velikov from comment #3) > Guys have you actually looked at the following l= ine ? > Aperture size is 268435456 MiB -- That is 256 TiB !!! >=20 > That's rather impossible amount if you ask me. So there's either a bug= in > IGT's gem_aperture_size() or one of the two ioctls > (I915_GEM_CONTEXT_GETPARAM I915_GEM_GET_APERTURE) that it uses. >=20 > With a couple of print statements you should be able to quickly track = the > exact offender. Good luck ! Yes, we saw it. The bug is reported to IGT (not to DRM/Intel). We propose t= he test should skip. I just checked on our IVB platform with 32 GB of memory. Looks like we would need now 2 TB. root@IVB102:/opt/X11R7/src/intel-gpu-tools/tests# ./gem_userptr_blits --run-subtest mlocked-normal-sync IGT-Version: 1.15-ge3abb20 (x86_64) (Linux: 4.7.0-nightly+ x86_64) Aperture size is 2048 MiB Total RAM is 31,945 MiB Testing unsynchronized mappings... Testing synchronized mappings... Killed
(In reply to Emil Velikov from comment #3) > Guys have you actually looked at the following l= ine ? > Aperture size is 268435456 MiB -- That is 256 TiB !!! >=20 > That's rather impossible amount if you ask me. So there's either a bug= in > IGT's gem_aperture_size() or one of the two ioctls It is correct. The GTT size is 1<<48 bytes.
(In reply to cprigent from comment #4) > (In reply to Emil Velikov from comment #3) > > Guys have you actually looked at the following line ? > > Aperture size is 268435456 MiB -- That is 256 TiB !!! > >=20 > > That's rather impossible amount if you ask me. So there's either = a bug in > > IGT's gem_aperture_size() or one of the two ioctls > > (I915_GEM_CONTEXT_GETPARAM I915_GEM_GET_APERTURE) that it uses. > >=20 > > With a couple of print statements you should be able to quickly t= rack the > > exact offender. Good luck ! >=20 > Yes, we saw it. The bug is reported to IGT (not to DRM/Intel). We prop= ose > the test should skip. Why? The test only allocates enough to fill RAM and then tests that the buf= fers are evicted for memory pressure. The messages are nothing to do with this t= est, just spam.
| What | Removed | Added |
|---|---|---|
| Assignee | dri-devel@lists.freedesktop.org | intel-gfx-bugs@lists.freedesktop.org |
| i915 platform | BDW | |
| QA Contact | intel-gfx-bugs@lists.freedesktop.org | |
| Component | IGT | DRM/Intel |