Re: PR/49033 CVS commit: src/tests/fs/vfs

[Andreas Gustafsson <gson%gson.org@localhost>]

The following reply was made to PR kern/49033; it has been noted by GNATS.

From: Andreas Gustafsson <gson%gson.org@localhost>
To: <martin%netbsd.org@localhost>
Cc: gnats-bugs%NetBSD.org@localhost
Subject: Re: PR/49033 CVS commit: src/tests/fs/vfs
Date: Fri, 25 Jul 2014 21:37:31 +0300

 >  PR kern/49033: try to provide better diagnostics if the udf_lookup_complex
 >  test randomly fails.
 
 Thanks :)
 
 Output from 
http://www.gson.org/netbsd/bugs/build/i386/2014/2014.07.25.12.18.13/test.html#fs_vfs_t_vnops_udf_lookup_complex
 :
 
 what   sb1             sb2
 st_dev 50432           50432
 st_mode                        16877   16877
 st_ino                 6       6
 st_nlink               2       2
 st_uid                 0       0
 st_gid                 0       0
 st_rdev                        -1      -1
 st_atim                        1406308232.27096000     1406308232.47096000
 st_mtim                        1406308232.27096000     1406308232.27096000
 st_ctim                        1406308232.27096000     1406308232.27096000
 st_birthtim            1406308232.27096000     1406308232.27096000
 st_size                        40                      40
 st_blocks              0                       0
 st_flags               0                       0
 st_gen                 0                       0
 
 Looks like the mismatch is in st_atim.  Does this mean UDF updates the
 st_atime on inode accesses and not just on data accesses, and if so,
 should it be considered a bug or is it a deliberate departure from
 standard POSIX semantics due to some filesystem limitation?
 -- 
 Andreas Gustafsson, gson%gson.org@localhost