Coreutils fails to recognize ACL libs during configure
From:
Necos Secon
Subject:
Coreutils fails to recognize ACL libs during configure
Date:
Thu, 29 Jan 2004 12:58:08 -0800
I have a system running Slackware 9.0 (upgraded to current) with the
following specs:
gcc 3.3.2
glibc 2.3.2 (with sscanf patch)
EA/ACL/NFS patch 0.85 (downloaded acl.bestbits.at) with kernel 2.4.24
acl-2.2.22 (downloaded from SGI's ftp)
attr-2.4.13 (downloaded from SGI's ftp)
As of right now, my ext3 partitions are mounted with acl enabled, and Samba
compiled just fine with acl support. However, when I tried to compile
coreutils-5.1.2 to use my acl libs and headers, I got this:
checking sys/acl.h usability... yes
checking sys/acl.h presence... yes
checking for sys/acl.h... yes
checking for acl... no
Looking into the config.log file, the failed program is:
configure:38446: checking for acl
configure:38503: gcc -o conftest -g -O2 conftest.c >&5
/tmp/cck5JOLp.o(.text+0x11): In function `main':
/tmp/coreutils-5.1.2/conftest.c:337: undefined reference to `acl'
After getting this error a couple of times, I tried applying the
coreutils-5.0 patches. There are 6 patches. I tried apply the first 3
(coreutils-5.0.dif, coreutils-acl.diff, coreutils-acl+posix.diff), but all
three failed at some point.
patching file lib/acl.c
Hunk #1 succeeded at 28 with fuzz 2 (offset 1 line).
Hunk #2 succeeded at 53 with fuzz 1 (offset -7 lines).
Hunk #3 FAILED at 87.
1 out of 3 hunks FAILED -- saving rejects to file lib/acl.c.rej
patching file m4/acl.m4
Hunk #1 FAILED at 20.
1 out of 1 hunk FAILED -- saving rejects to file m4/acl.m4.rej
patching file src/Makefile.am
Hunk #1 FAILED at 34.
1 out of 1 hunk FAILED -- saving rejects to file src/Makefile.am.rej
For the coreutils-acl.dif patch:
patching file lib/acl.c
Hunk #1 succeeded at 35 (offset 1 line).
patching file lib/acl.h
patching file src/copy.c
Hunk #1 succeeded at 41 with fuzz 2 (offset -4 lines).
Hunk #3 succeeded at 794 (offset 20 lines).
Hunk #4 succeeded at 1278 (offset 34 lines).
Hunk #5 succeeded at 1378 (offset 17 lines).
Hunk #6 succeeded at 1413 with fuzz 2 (offset 31 lines).
Hunk #7 succeeded at 1512 (offset 15 lines).
Hunk #8 succeeded at 1556 (offset 37 lines).
patching file src/copy.h
Hunk #1 succeeded at 143 (offset 4 lines).
patching file src/cp.c
Hunk #1 FAILED at 36.
Hunk #2 succeeded at 59 (offset -3 lines).
Hunk #3 succeeded at 286 with fuzz 2 (offset -3 lines).
Hunk #4 succeeded at 339 (offset -6 lines).
Hunk #5 succeeded at 367 (offset -3 lines).
Hunk #6 succeeded at 379 (offset -6 lines).
Hunk #7 succeeded at 403 (offset -4 lines).
Hunk #8 succeeded at 420 (offset -7 lines).
Hunk #9 succeeded at 453 (offset -4 lines).
Hunk #10 succeeded at 495 (offset -7 lines).
Hunk #11 succeeded at 649 with fuzz 2 (offset -4 lines).
Hunk #12 succeeded at 782 (offset -8 lines).
Hunk #13 succeeded at 1060 (offset -2 lines).
1 out of 13 hunks FAILED -- saving rejects to file src/cp.c.rej
patching file src/install.c
Hunk #1 succeeded at 162 (offset -3 lines).
patching file src/ls.c
Hunk #1 succeeded at 194 (offset -29 lines).
Hunk #2 succeeded at 2438 with fuzz 1 (offset 38 lines).
patching file src/mv.c
Hunk #1 succeeded at 132 (offset -5 lines).
After investigating the patches, I noticed that the 2 acl patches are for
coreutils 4.x. However, the patches do, indeed add acl lib checking to the
m4 files (to my knowledge, these are used with automake and m4), which
should detect the libraries and enable the acl code in coreutils. Further
investigation shows that the acl.m4 file that is included with coreutils
lacks this extra code, and as such, the patch should work. Maybe I'm missing
something with the patching. Were the patchces made with a different version
of diff (I have 2.8.1 installed)? I'm posting this email to both lists, as I
don't know where the problem really is (or if it's a combined problem).
Thank you for your time,
Theodore Charles III
Network Administrator
Los Angeles Senior High
_________________________________________________________________