restorecond going crazy on 2.6.31.4 - inotify regression?

restorecond going crazy on 2.6.31.4 - inotify regression?

[Robert Hancock]

I just upgraded to 2.6.31.4 on a CentOS 5.3 x86_64 machine, and noticed 
the SELinux restorecond daemon using 100% CPU. Previously it was running 
2.6.30.7 which didn't have this problem. It looks like it's triggered by 
either opening or closing a gnome-terminal. strace on restorecond shows 
the following repeating continuously

inotify_add_watch(3, "/var/run/utmp", IN_MODIFY|IN_MOVED_FROM) = 9
read(3, "\t\0\0\0\0\200\0\0\0\0\0\0\0\0\0\0", 32768) = 16
write(1, "wd=9 mask=32768 cookie=0 len=0\n", 31wd=9 mask=32768 cookie=0 
len=0
) = 31
open("/var/run/utmp", O_RDONLY)         = 4
fstat(4, {st_mode=S_IFREG|0664, st_size=5760, ...}) = 0
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) 
= 0x7fe0d938c000
read(4, 
"\10\0\0\0\177\2\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4096) 
= 4096
read(4, 
"\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 
4096) = 1664
read(4, "", 4096)                       = 0
close(4)                                = 0
munmap(0x7fe0d938c000, 4096)            = 0
inotify_rm_watch(3, 9)                  = 0
inotify_add_watch(3, "/var/run/utmp", IN_MODIFY|IN_MOVED_FROM) = 9

It looks like it's getting continuous notifications that /var/run/utmp 
has changed, which seems a bit odd. I know there were some inotify 
issues in the 2.6.31 series, is there something still amiss?