strip won't strip non-executable files on XFS
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| | linux-source-2.6.15 (Ubuntu) |
Medium
|
Ben Collins | ||
Bug Description
On XFS, strip(1) refuses to strip non-executable files in some situations,
leading to major bustage in the xorg build:
dh_strip \
-Nlibdmx1-dbg \
[...]
-Nxserver-
strip: unable to copy file
'debian/
denied
strip: unable to copy file
'debian/
denied
[...]
find debian/
xargs --no-run-if-empty \
strip --strip-debug --remove-
strip: unable to copy file
'debian/
denied
[...]
This occurs on i386 and amd64, and is fixed by moving to ext3. I cannot
reproduce this outside of a build environment:
daniels@
/usr/X11R6/
daniels@
-rw-r--r-- 1 daniels daniels 281032 2005-02-23 14:56 radeon_drv.o
daniels@
daniels@
-rw-r--r-- 1 daniels daniels 179320 2005-02-23 14:56 radeon_drv.o
| Chuck Short (zulcss) wrote : | #1 |
| Matt Zimmerman (mdz) wrote : | #2 |
I think an strace of the strip command would be in order
| Daniel Stone (daniels) wrote : | #3 |
I've been talking to Eric Sandeen of SGI about this, and am running through a
xorg build now to get an strace (diverted /usr/bin/strip).
| Daniel Stone (daniels) wrote : | #4 |
I thought fakeroot may have been causing it, but no dice:
daniels@
~/canonical/
radeon_drv.o.orig
daniels@
-rw-r--r-- 1 daniels daniels 2.5M 2005-02-27 22:50 radeon_drv.o.orig
daniels@
--remove-
&& ls -lh radeon_drv.o
-rw-r--r-- 1 daniels daniels 262K 2005-02-28 00:57 radeon_drv.o
daniels@
--remove-
&& ls -lh radeon_drv.o
-rw-r--r-- 1 daniels daniels 262K 2005-02-28 00:57 radeon_drv.o
strace to come in an attachment.
| Daniel Stone (daniels) wrote : | #5 |
Created an attachment (id=1465)
strace of failing strip
| Daniel Stone (daniels) wrote : | #6 |
(From update of attachment 1465)
wrong one
| Daniel Stone (daniels) wrote : | #7 |
Created an attachment (id=1469)
the right strace this time
| Matt Zimmerman (mdz) wrote : | #8 |
It appears from the strace that the file is simply read-only:
lstat("
{st_mode=
getpid() = 19803
semop(655362, 0x131, 140737488348032) = 0
msgsnd(1179650, {0, 0x8}, 140737488348112, 0x38) = 0
msgrcv(1212419, {12314923993692
= 56
semop(655362, 0, 140737488348032) = 0
open("debian/
open("debian/
O_WRONLY|
Easily reproducible:
mizar:[/tmp] cp /bin/ls .
mizar:[/tmp] chmod 444 ls
mizar:[/tmp] strip ls
strip: unable to copy file 'ls' reason: Permission denied
| Daniel Stone (daniels) wrote : | #9 |
And yet, this only occurs on XFS for some reason ... I'm willing to bet this
isn't unrelated to Debian #296586.
dh_strip \
-Nlibdps1-dbg \
[...]
strip: unable to copy file
'debian/
strip: unable to copy file
'debian/
denied
strip: unable to copy file
'debian/
denied
[...]
| Ben Collins (ben-collins) wrote : | #10 |
Does this still affect current breezy kernels?
| Daniel Stone (daniels) wrote : | #11 |
sorry, missed the bug mail. it does appear to still be occurring, yes.
| Ben Collins (ben-collins) wrote : | #12 |
This bug has been flagged because it is old and possibly inactive. It may or may
not be fixed in the latest release (Breezy Badger 5.10). It is being marked as
"NEEDSINFO". In two weeks time, if the bug is not updated back to "NEW" and
validated against Breezy, it will be closed.
This is needed in order to help manage the current bug list for the kernel. We
would like to fix all bugs, but need users to test and help with debugging.
If this change was in error for this bug, please respond and make the
appropriate change (or email <email address hidden> if you cannot make the
change).
Thanks for your help.
| Ben Collins (ben-collins) wrote : | #13 |
Daniel, does this still affect breezy and/or dapper?
| Ben Collins (ben-collins) wrote : | #14 |
Keybuk and pitti reported that they were able to strip 644 binaries on xfs
without any problems with dapper 2.6.15 kernel.
| Daniel Stone (daniels) wrote : | #15 |
i've always been able to strip random arbitrary files normally (even copying the
same files over and running strip on them); it's only ever been the monolithic
xfree86/xorg build that's been affected, but I'm not convinced I care enough now.


I have opened up a bug upstream:
http:// bugme.osdl. org/show_ bug.cgi? id=4258
Regards
chuck