Hi Matt,
On Fri, Aug 29, 2008 at 10:00 AM, Matt Helsley <[EMAIL PROTECTED]> wrote:
> Hi Rishi,
>
> When I ran the August intermediate release on a single-cpu i386 machine
> I was not able to reproduce your results. I did:
>
> $ make
> $ make install
> $ ./runltp -f fs_bind
> $ ./runltp -f fs_bind > run.out > run.err
> $ grep PASSED run.out | wc -l
> 90
> $ wc -l run.err
> 0
> $ grep -i brok run.* | wc -l
> 0
>
> Which shows all the tests passed and nothing broke. I'm trying to
> reproduce the problem on other systems but I'm having trouble getting
> other portions of LTP to compile (different bugs on different
> systems) :(.
Some build error issues were seen in some systems and astonishingly for
different test cases on different systems. They seem to the GLIBC issues. So
the same has been fixed and yet to be fixed:
1) sync_file_range() syscall tests were having this issue even after
Intermediate release. This is fixed now and available in the CVS,
2) inotify() was giving problems in IA64 running a Debian Etch. That has
been resolved and in CVS,
3) utimensat() is still giving problems on IA64 running debian Etch. There
is no fix till now. But this works very well in other archs (i386, x86_64,
IA64, PPC64).
The Problem seems with the GLIBC versions. The Debain Etch on IA64 has
GLIBC:
gcc (GCC) 4.1.2 20061115 (prerelease) (Debian 4.1.1-21)
Which is causing all the truble to build. I hope we will be fixing that
soon.
Regards--
Subrata
>
>
> So I'm going to step through the information in your log and hope this
> suggests things for you to investigate while I continue trying to
> reproduce the problem:
>
> On Tue, 2008-08-26 at 10:32 +0530, Rishikesh K Rajak wrote:
>
> <snip>
>
> > + diff -dEBb
> > -q /root/ltp-full-20080731/results/fs_bind/move/test19/mtab.before
> /root/ltp-full-20080731/results/fs_bind/move/test19/mtab.after
> > + '[' 1 -eq 0 ']'
> > + tst_resm TWARN 'move/test19: did not properly clean up its mtab
> > mounts'
> > /root/ltp-full-20080731/testscripts/test_fs_bind.sh 0 WARN :
> > move/test19: did not properly clean up its mtab mounts
>
>
> This is the first warning I see. A diff of the mtab file produced:
> +/tmp/ltp-8516/sandbox/share1 on /tmp/ltp-8516/sandbox/parent1/child1
> type none (rw,bind)
>
> So this means this mount was added by this test and not cleaned up --
> hence the test and/or others that follow may be "broken".
>
> Looking at the test19 portion of the debug trace we see what it does:
> mkdir -p "parent1"
> mount --bind "parent1" "parent1"
> mount --make-rprivate "parent1"
> mkdir -p "parent2"
> mount --bind "parent2" "parent2"
> mount --make-rprivate "parent2"
> mkdir -p "share1"
> mount --bind "share1" "share1"
> mount --make-rshared "share1"
> mount --make-rslave "parent1/child1"
>
> (not including the post-test cleanup code)
>
> So somehow in test19 /tmp/ltp-8516/sandbox/share1 was mounted
> at /tmp/ltp-8516/sandbox/parent1/child1 and the test failed to clean it
> up. The interesting thing here is I don't see where, at least according
> to this log, that mount could have occurred. It's _probably_ a move
> mount in the test19 script that's not reflected here in the log...
>
>
> Here's an example of what we should see (the test just before this one
> in the log):
>
> > + (( nsucceeded++ ))
>
> The test succeeded and now we're going to check the mounts in /proc:
>
> > + check_proc_mounts
> > move/test16 /root/ltp-full-20080731/results/fs_bind/move/test16
>
> This is just the name of the function and its arguments. It translates
> into:
>
> > + local tname=move/test16
> > + '[' '!'
> > -r /root/ltp-full-20080731/results/fs_bind/move/test16/proc_mounts.before
> ']'
> > + grep_proc_mounts
>
> So /root/ltp-full-20080731/results/fs_bind/move/test16/proc_mounts.before
> is readable. "grep" it.
>
> > + diff -dEBb
> > -q /root/ltp-full-20080731/results/fs_bind/move/test16/proc_mounts.before
> /root/ltp-full-20080731/results/fs_bind/move/test16/proc_mounts.after
> > + '[' 0 -eq 0 ']'
> > + '[' 1 -eq 1 ']'
> > + rm
> > -f /root/ltp-full-20080731/results/fs_bind/move/test16/proc_mounts.before
> /root/ltp-full-20080731/results/fs_bind/move/test16/proc_mounts.after
>
> There are no differences so we remove the .before and .after files.
>
> > + return 0
> > + check_mounts
> > move/test16 /root/ltp-full-20080731/results/fs_bind/move/test16
>
> Now we check the mtab file in a similar fashion...
>
> > + local tname=move/test16
> > + '[' '!'
> > -r /root/ltp-full-20080731/results/fs_bind/move/test16/mtab.before ']'
> > + grep_mounts
> > + diff -dEBb
> > -q /root/ltp-full-20080731/results/fs_bind/move/test16/mtab.before
> /root/ltp-full-20080731/results/fs_bind/move/test16/mtab.after
> > + '[' 0 -eq 0 ']'
> > + '[' 1 -eq 1 ']'
> > + rm
> > -f /root/ltp-full-20080731/results/fs_bind/move/test16/mtab.before
> /root/ltp-full-20080731/results/fs_bind/move/test16/mtab.after
>
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's
> challenge
> Build the coolest Linux based applications with Moblin SDK & win great
> prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Ltp-list mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/ltp-list
>
--
Regards & Thanks--
Subrata
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
