On Thu, Apr 22, 2010 at 04:38:12PM -0700, Henry Yei wrote:
> Bryan, I've confirmed that the linkat01 test passes with this fix on
> x86 with 2.6.21 kernel and arm on 2.6.33 kernel. Thanks!
> 
> Henry Yei <[email protected]>

Thanks Bryan & Henry.

Merged into tree.
23e7806fec4c6be42fdf577559df5498cce45118

-Rishi
> 
> On Thu, Apr 22, 2010 at 2:40 PM, Bryan Freed <[email protected]> wrote:
> > I have a fix for a failing LTP testcase.  Section 6 of the online LTP doc
> > http://ltp.sourceforge.net/documentation/how-to/ltp.php#_7 says I can
> > request a test be added by sending to this list, and it looks from the
> > archive like the subject line for patches should contain "[LTP][PATCH]", so
> > here goes.  There appears to be no standard diff format, so I will use what
> > is convenient...
> > Running the ltp suite on ChromeOS (x86 with a 2.6.32 kernel) fails linkat01
> > testcase 21:
> >
> > linkat01   21  TFAIL  :  linkat() failed: TEST_ERRNO=EXDEV(18): Invalid
> > cross-device link
> > This is probably because my /tmp is mounted from someplace other than the
> > filesystem that contains newdirfd.
> > The goal of testcase 21 is to demonstrate that linking a directory will fail
> > EPERM.  The fix is to use olddir/. instead of /tmp.
> > diff -urN
> > ltp-full-20091231_vanilla/testcases/kernel/syscalls/linkat/linkat01.c
> > ltp-full-20091231/testcases/kernel/syscalls/linkat/linkat01.c
> > --- ltp-full-20091231_vanilla/testcases/kernel/syscalls/linkat/linkat01.c
> >     2009-11-26 23:43:12.000000000 -0800
> > +++ ltp-full-20091231/testcases/kernel/syscalls/linkat/linkat01.c
> > 2010-04-12 16:28:59.000000000 -0700
> > @@ -158,7 +158,7 @@
> >         { &cwd_fd, "/proc/cpuinfo", &newdirfd, TEST_FILE1, 0,
> >           0, 0, EXDEV },
> >         /* 21. directory link */
> > -       { &cwd_fd, "/tmp", &newdirfd, TEST_FILE1, 0,
> > +       { &olddirfd, ".", &newdirfd, TEST_FILE1, 0,
> >           0, 0, EPERM },
> >         /* 22. invalid flag */
> >         { &olddirfd, TEST_FILE1, &newdirfd, TEST_FILE1, 1,
> >
> > Let me know if this is the incorrect process for upstreaming LTP patches.
> > bryan.
> > ------------------------------------------------------------------------------
> >
> > _______________________________________________
> > Ltp-list mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/ltp-list
> >
> >
> 
> ------------------------------------------------------------------------------
> _______________________________________________
> Ltp-list mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/ltp-list

-- 
Thanks & Regards
Rishi
LTP Maintainer
IBM, LTC, Bangalore
Please join IRC #ltp @ irc.freenode.net

------------------------------------------------------------------------------
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

Reply via email to