https://ci.freebsd.org/job/FreeBSD-head-i386-test/10638/testReport/junit/sys.fs.tmpfs/times_test/empty/ https://ci.freebsd.org/job/FreeBSD-head-i386-test/10638/testReport/junit/sys.fs.tmpfs/times_test/non_empty/ Changes: https://ci.freebsd.org/job/FreeBSD-head-i386-test/10638/
A commit references this bug: Author: lwhsu Date: Wed Sep 16 09:58:16 UTC 2020 New revision: 365793 URL: https://svnweb.freebsd.org/changeset/base/365793 Log: Temporarily skip sys.fs.tmpfs.times_test.{empty,non_empty} in CI PR: 249362 Sponsored by: The FreeBSD Foundation Changes: head/contrib/netbsd-tests/fs/tmpfs/t_times.sh
kib, can you help to check if this is related to r365783 to r365787? Should the codes or the tests be fixed? Thanks!
A commit references this bug: Author: kib Date: Wed Sep 16 21:28:19 UTC 2020 New revision: 365810 URL: https://svnweb.freebsd.org/changeset/base/365810 Log: tmpfs: restore atime updates for reads from page cache. Split TMPFS_NODE_ACCCESSED bit into dedicated byte that can be updated atomically without locks or (locked) atomics. tn_update_getattr() change also contains unrelated bug fix. Reported by: lwhsu PR: 249362 Reviewed by: markj (previous version) Discussed with: mjg Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26451 Changes: head/sys/fs/tmpfs/tmpfs.h head/sys/fs/tmpfs/tmpfs_fifoops.c head/sys/fs/tmpfs/tmpfs_subr.c head/sys/fs/tmpfs/tmpfs_vnops.c
A commit references this bug: Author: lwhsu Date: Thu Sep 17 07:13:42 UTC 2020 New revision: 365831 URL: https://svnweb.freebsd.org/changeset/base/365831 Log: Revert r249362, atime update in tmpfs is fixed in r365810 PR: 249362 Sponsored by: The FreeBSD Foundation Changes: head/contrib/netbsd-tests/fs/tmpfs/t_times.sh
The test sys/fs/tmpfs/times_test:empty has started filing since: https://ci.freebsd.org/view/Test/job/FreeBSD-main-amd64-test/23814/testReport/junit/sys.fs.tmpfs/times_test/empty/ I have narrowed it down to: https://github.com/jlduran/freebsd-src/pull/36/commits/263c95e6744614000126e5fae450495a3fcbaf0e
NOTE: That not the fix, it is just to illustrate that both values are always equal, and therefore failing the test.