The following assertion is hit on all machines up to haswell in CI: (gem_exec_reloc:8204) CRITICAL: Test assertion failure function basic_reloc, file gem_exec_reloc.c:494: (gem_exec_reloc:8204) CRITICAL: Failed assertion: reloc.presumed_offset == offset (gem_exec_reloc:8204) CRITICAL: error: 0x7fdf7000 != 0xffffffff Full logs: https://intel-gfx-ci.01.org/CI/CI_DRM_2525/fi-byt-j1900/igt@gem_exec_reloc@basic-write-gtt-active.html
Forgot to say that is was on the following tests: - igt@gem_exec_reloc@basic-write-gtt-active - igt@gem_exec_reloc@basic-cpu-gtt - igt@gem_exec_reloc@basic-write-gtt - igt@gem_exec_reloc@basic-cpu-gtt-active And the following tests only failed on fi-ivb-3520m: - igt@gem_exec_reloc@basic-gtt-active - igt@gem_exec_reloc@basic-gtt
Caused by IGT change having patches 9372ad6 igt/gem_storedw_loop: Mark up with DOMAIN_WC ae15bfe igt/gem_mmap_wc: Mark up with DOMAIN_WC d2fe019 igt/gem_cs_tlb: Mark up with DOMAIN_WC e9cd7d5 igt/gem_concurrent_blit: Mark up with DOMAIN_WC a6138b6 igt/gem_exec_flush: Use DOMAIN_WC for access via gem_mmap__wc 4186a2d igt/gem_pwrite: Use DOMAIN_WC for access via gem_mmap__wc 3bdafec igt/gem_exec_reloc: Check interactions with WC domain
commit 60568210a5616f2961a832fcb2f15daa5be918d1 Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Thu Apr 13 16:07:33 2017 +0100 iget/gem_exec_reloc: Fix simulated relocations The contract with the kernel is that the presumed_offset matches the value written into the batch. In the case where we were creating a new object to simulate the old being relocation, we were writing some other value into the batch. It just happens that using GGTT read back on !llc was causing the original batch to migrated into the aperture, leaving a hole suitable for the new batch, and the kernel could therefore skip the relocation (causing us to complain). Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100674 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Whitelisted on CI, fixed.
Sorry for the noise.
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.