GNU bug report logs - #12322
coreutils-8.1{8,9} with /etc/mtab as real file, df/total-unprocessed fail, df/no-mtab-status is skipped

Previous Next

To reply to this bug, email your comments to 12322 AT debbugs.gnu.org.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: g.esp <at> free.fr
To: bug-coreutils <at> gnu.org
Subject: coreutils-8.1{8,9} with /etc/mtab as real file,
	df/total-unprocessed fail, df/no-mtab-status is skipped
Date: Sat, 1 Sep 2012 17:49:22 +0200 (CEST)

On the 8.0 era, I had to switch from /etc/mtab symlinked to /proc/mounts to a real /etc/mtab or one cu test fail to umount

Now depending of /etc/mtab nature, I see 2 df tests behaving differently on my linux chroot used to build

When not using /etc/mtab as symlink
df '.' does not report the intended 'no file systems processed' for total-unprocessed test and that test fail
but df with no parameter report 'no file systems processed' and an error, so df/no-mtab-status is skipped

When using /etc/mtab as symlink to /proc/mounts
df '.' report the intended 'no file systems processed' for total-unprocessed script and test pass
but df with no parameter can't read mount point existing out of chroot and df/no-mtab-status is skipped again

The log that FAIL with real /etc/mtab is as follow

FAIL: df/total-unprocessed
==========================

++ initial_cwd_=/usr/src/coreutils-8.19/tests
++ fail=0
+++ testdir_prefix_
+++ printf gt
++ pfx_=gt
+++ mktempd_ /usr/src/coreutils-8.19/tests gt-total-unprocessed.XXXX
+++ case $# in
+++ destdir_=/usr/src/coreutils-8.19/tests
+++ template_=gt-total-unprocessed.XXXX
+++ MAX_TRIES_=4
+++ case $destdir_ in
+++ case $template_ in
++++ unset TMPDIR
+++ d=/usr/src/coreutils-8.19/tests/gt-total-unprocessed.CId2
+++ case $d in
+++ test -d /usr/src/coreutils-8.19/tests/gt-total-unprocessed.CId2
++++ tr S -
++++ ls -dgo /usr/src/coreutils-8.19/tests/gt-total-unprocessed.CId2
+++ perms='drwx------ 2 4096 Aug 29 00:17 /usr/src/coreutils-8.19/tests/gt-total-unprocessed.CId2'
+++ case $perms in
+++ test 0 = 0
+++ echo /usr/src/coreutils-8.19/tests/gt-total-unprocessed.CId2
+++ return
++ test_dir_=/usr/src/coreutils-8.19/tests/gt-total-unprocessed.CId2
++ cd /usr/src/coreutils-8.19/tests/gt-total-unprocessed.CId2
++ gl_init_sh_nl_='
'
++ IFS=' 	
'
++ for sig_ in 1 2 3 13 15
+++ expr 1 + 128
++ eval 'trap '\''Exit 129'\'' 1'
+++ trap 'Exit 129' 1
++ for sig_ in 1 2 3 13 15
+++ expr 2 + 128
++ eval 'trap '\''Exit 130'\'' 2'
+++ trap 'Exit 130' 2
++ for sig_ in 1 2 3 13 15
+++ expr 3 + 128
++ eval 'trap '\''Exit 131'\'' 3'
+++ trap 'Exit 131' 3
++ for sig_ in 1 2 3 13 15
+++ expr 13 + 128
++ eval 'trap '\''Exit 141'\'' 13'
+++ trap 'Exit 141' 13
++ for sig_ in 1 2 3 13 15
+++ expr 15 + 128
++ eval 'trap '\''Exit 143'\'' 15'
+++ trap 'Exit 143' 15
++ trap remove_tmp_ 0
+ path_prepend_ ../src
+ test 1 '!=' 0
+ path_dir_=../src
+ case $path_dir_ in
+ abs_path_dir_=/usr/src/coreutils-8.19/tests/../src
+ case $abs_path_dir_ in
+ PATH=/usr/src/coreutils-8.19/tests/../src:/usr/src/coreutils-8.19/src:/bin:/usr/bin:/sbin:/usr/sbin
+ create_exe_shims_ /usr/src/coreutils-8.19/tests/../src
+ case $EXEEXT in
+ return 0
+ shift
+ test 0 '!=' 0
+ export PATH
+ print_ver_ df
+ test yes = yes
+ local i
+ for i in '$*'
+ env df --version
df (GNU coreutils) 8.19
Copyright (C) 2012 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>.
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.

Written by Torbjorn Granlund, David MacKenzie, and Paul Eggert.
+ require_mount_list_
+ local 'mount_list_fail=cannot read table of mounted file systems'
+ grep -F 'cannot read table of mounted file systems'
+ df
+ cat
+ df --local -t nfs --total .
Filesystem     1K-blocks     Used Available Use% Mounted on
-               19228308 16776508   1475048  92% /
total           19228308 16776508   1475048  92%
+ fail=1
+ compare exp out
+ compare_dev_null_ exp out
+ test 2 = 2
+ test xexp = x/dev/null
+ test xout = x/dev/null
+ return 2
+ case $? in
+ compare_ exp out
+ diff -u exp out
--- exp	2012-08-29 00:17:31.000000000 +0200
+++ out	2012-08-29 00:17:31.000000000 +0200
@@ -1 +0,0 @@
-df: no file systems processed
+ fail=1
+ cat
+ df --total _does_not_exist_
+ compare exp out
+ compare_dev_null_ exp out
+ test 2 = 2
+ test xexp = x/dev/null
+ test xout = x/dev/null
+ return 2
+ case $? in
+ compare_ exp out
+ diff -u exp out
+ Exit 1
+ set +e
+ exit 1
+ exit 1
+ remove_tmp_
+ __st=1
+ cleanup_
+ :
+ cd /usr/src/coreutils-8.19/tests
+ chmod -R u+rwx /usr/src/coreutils-8.19/tests/gt-total-unprocessed.CId2
+ rm -rf /usr/src/coreutils-8.19/tests/gt-total-unprocessed.CId2
+ exit 1

With real /etc/mtab, (on my chroot, there is many mount --bind before to enter the chroot)
[chroot-i486] root:/$ mount
/dev on /dev type none (rw)
[chroot-i486] root:/$ df
df: no file systems processed
[chroot-i486] root:/$ ln -sf /proc/mounts /etc/mtab
[chroot-i486] root:/$ mount
rootfs on / type rootfs (rw)
none on /sys type sysfs (rw,nosuid,nodev,noexec,relatime)
none on /proc type proc (rw,nosuid,nodev,noexec,relatime)
none on /dev type devtmpfs (rw,relatime,size=512760k,nr_inodes=128190,mode=755)
none on /dev/pts type devpts (rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=000)
/dev/disk/by-uuid/7a235d64-5d04-41ac-a959-70465eb74fc8 on / type ext3 (rw,relatime,errors=remount-ro,data=ordered)
tmpfs on /lib/init/rw type tmpfs (rw,nosuid,relatime,mode=755)
tmpfs on /dev/shm type tmpfs (rw,nosuid,nodev,relatime)
...
[chroot-i486] root:/$ df
df: '/lib/init/rw': No such file or directory
Filesystem     1K-blocks     Used Available Use% Mounted on
rootfs          19228308 15796388   2455168  87% /
none              512760      156    512604   1% /dev
/dev/sda3       19228308 15796388   2455168  87% /
tmpfs             517208        0    517208   0% /dev/shm
/dev/sda1       19228308 15796388   2455168  87% /boot
none              512760      156    512604   1% /dev
tmpfs             517208        0    517208   0% /dev/shm
...

So df/no-mtab-status is skipped in both case because df returned an error while reading /lib/init/rw that is out of chroot.

Maybe both tests could check if /etc/mtab is a symlink and adjust df parameter to not fail?

I made this change on the tests to see if that work
	cd $(DIR_APP) && sed -i "s/ '\.' / /" tests/df/total-unprocessed
	cd $(DIR_APP) && sed -i "s/df ||/df '.' ||/" tests/df/no-mtab-status

With /etc/mtab as a real file, both tests passed

Gilles

Message #8 received at 12322 <at> debbugs.gnu.org (full text, mbox):

From: Bernhard Voelker <mail <at> bernhard-voelker.de>
To: g.esp <at> free.fr
Cc: 12322 <at> debbugs.gnu.org
Subject: Re: bug#12322: coreutils-8.1{8, 9} with /etc/mtab as real file,
	df/total-unprocessed fail, df/no-mtab-status is skipped
Date: Sun, 02 Sep 2012 11:05:28 +0200

On 09/01/2012 05:49 PM, g.esp <at> free.fr wrote:
> Maybe both tests could check if /etc/mtab is a symlink and adjust df parameter to not fail?
> 
> I made this change on the tests to see if that work
> 	cd $(DIR_APP) && sed -i "s/ '\.' / /" tests/df/total-unprocessed
> 	cd $(DIR_APP) && sed -i "s/df ||/df '.' ||/" tests/df/no-mtab-status
> 
> With /etc/mtab as a real file, both tests passed

Thank you for the report.

To make it clear, your change suggests this:

diff --git a/tests/df/no-mtab-status.sh b/tests/df/no-mtab-status.sh
index e434f5f..6d19f8b 100755
--- a/tests/df/no-mtab-status.sh
+++ b/tests/df/no-mtab-status.sh
@@ -20,7 +20,7 @@
 . "${srcdir=.}/init.sh"; path_prepend_ ../src
 print_ver_ df

-df || skip_ "df fails"
+df '.' || skip_ "df fails"

 # Simulate "mtab" failure.
 cat > k.c <<'EOF' || framework_failure_
diff --git a/tests/df/total-unprocessed.sh b/tests/df/total-unprocessed.sh
index aa49836..dbe552e5 100755
--- a/tests/df/total-unprocessed.sh
+++ b/tests/df/total-unprocessed.sh
@@ -28,7 +28,7 @@ EOF
 # The following simply finds no match for the combination
 # of the options --local and FS-type nfs together with the
 # argument ".". It must exit non-Zero nonetheless.
-df --local -t nfs --total '.' 2>out && fail=1
+df --local -t nfs --total 2>out && fail=1
 compare exp out || fail=1

 cat <<\EOF > exp || framework_failure_


Additionally to a missing mtab, there is another reason which
can make df fail: an inaccessible mount point, e.g. when there's
a file system mounted on /root/backup which is not accessible
by a normal user.

The above tests provoke certain error conditions:
no-mtab-status tests df's behaviour when the mtab cannot be read;
total-unprocessed tests that df exits non-Zero and writes an error
message when --total is used but no file system has been processed.

While your change doesn't make the tests skip under certain
conditions, it well may trigger a different than the intended
failure. E.g. in no-mtab-status, there's a test which is supposed
to fail because we simulated the mtab failure:

  # These tests are supposed to fail:
  LD_PRELOAD=./k.so df && fail=1

That command - a simple df without arguments - could also fail due
to another reason (see above), so the failure may shadow the expected
failure from (simulated) missing mtab.

That said, I'd rather skip a test that cannot otherwise guarantee
why it fails/succeeds.

Have a nice day,
Berny

Message #11 received at 12322 <at> debbugs.gnu.org (full text, mbox):

From: g.esp <at> free.fr
To: Bernhard Voelker <mail <at> bernhard-voelker.de>
Cc: 12322 <at> debbugs.gnu.org
Subject: Re: bug#12322: coreutils-8.1{8, 9} with /etc/mtab as real file,
	df/total-unprocessed fail, df/no-mtab-status is skipped
Date: Mon, 3 Sep 2012 00:07:26 +0200 (CEST)

----- Mail original -----
> De: "Bernhard Voelker" <mail <at> bernhard-voelker.de>
> À: "g esp" <g.esp <at> free.fr>
> Cc: 12322 <at> debbugs.gnu.org
> Envoyé: Dimanche 2 Septembre 2012 11:05:28
> Objet: bug#12322: coreutils-8.1{8, 9} with /etc/mtab as real file,	df/total-unprocessed fail, df/no-mtab-status is
> skipped
> 
> On 09/01/2012 05:49 PM, g.esp <at> free.fr wrote:
> > Maybe both tests could check if /etc/mtab is a symlink and adjust
> > df parameter to not fail?
> > 
> > I made this change on the tests to see if that work
> > 	cd $(DIR_APP) && sed -i "s/ '\.' / /" tests/df/total-unprocessed
> > 	cd $(DIR_APP) && sed -i "s/df ||/df '.' ||/"
> > 	tests/df/no-mtab-status
> > 
> > With /etc/mtab as a real file, both tests passed
> 
> Thank you for the report.
> 
> To make it clear, your change suggests this:
> 
> diff --git a/tests/df/no-mtab-status.sh b/tests/df/no-mtab-status.sh
> index e434f5f..6d19f8b 100755
> --- a/tests/df/no-mtab-status.sh
> +++ b/tests/df/no-mtab-status.sh
> @@ -20,7 +20,7 @@
>  . "${srcdir=.}/init.sh"; path_prepend_ ../src
>  print_ver_ df
> 
> -df || skip_ "df fails"
> +df '.' || skip_ "df fails"
> 
>  # Simulate "mtab" failure.
>  cat > k.c <<'EOF' || framework_failure_
> diff --git a/tests/df/total-unprocessed.sh
> b/tests/df/total-unprocessed.sh
> index aa49836..dbe552e5 100755
> --- a/tests/df/total-unprocessed.sh
> +++ b/tests/df/total-unprocessed.sh
> @@ -28,7 +28,7 @@ EOF
>  # The following simply finds no match for the combination
>  # of the options --local and FS-type nfs together with the
>  # argument ".". It must exit non-Zero nonetheless.
> -df --local -t nfs --total '.' 2>out && fail=1
> +df --local -t nfs --total 2>out && fail=1
>  compare exp out || fail=1
> 
>  cat <<\EOF > exp || framework_failure_
> 
> 
Yes that should fix df/total-unprocessed.sh and df/total-unprocessed.sh in both /etc/mtab cases.

> Additionally to a missing mtab, there is another reason which
> can make df fail: an inaccessible mount point, e.g. when there's
> a file system mounted on /root/backup which is not accessible
> by a normal user.
> 
> The above tests provoke certain error conditions:
> no-mtab-status tests df's behaviour when the mtab cannot be read;
> total-unprocessed tests that df exits non-Zero and writes an error
> message when --total is used but no file system has been processed.
> 
> While your change doesn't make the tests skip under certain
> conditions, it well may trigger a different than the intended
> failure. E.g. in no-mtab-status, there's a test which is supposed
> to fail because we simulated the mtab failure:
> 
>   # These tests are supposed to fail:
>   LD_PRELOAD=./k.so df && fail=1
> 
> That command - a simple df without arguments - could also fail due
> to another reason (see above), so the failure may shadow the expected
> failure from (simulated) missing mtab.
> 
> That said, I'd rather skip a test that cannot otherwise guarantee
> why it fails/succeeds.
> 
> Have a nice day,
> Berny
> 
Yes here df command fail in real /etc/mtab case, so even if this test will not trigger for the intented reason, test will not fail.

Similary total-verify.sh may be fixed to pass in /etc/mtab real file case with this patch (hope nothing will be wrapped)
diff --git a/tests/df/total-verify.sh b/tests/df/total-verify.sh
index 948b002..317cb75 100755
--- a/tests/df/total-verify.sh
+++ b/tests/df/total-verify.sh
@@ -20,7 +20,7 @@
 print_ver_ df
 require_perl_

-df || skip_ "df fails"
+df '.' || skip_ "df fails"

 cat <<\EOF > check-df || framework_failure_
 my ($total, $used, $avail) = (0, 0, 0);
@@ -53,9 +53,9 @@ EOF

 # Use --block-size=512 to keep df from printing rounded-to-kilobyte
 # numbers which wouldn't necessarily add up to the displayed total.
-df --total -P --block-size=512 > space || framework_failure_
+df --total -P --block-size=512 '.' > space || framework_failure_
 cat space  # this helps when debugging any test failure
-df --total -i -P               > inode || framework_failure_
+df --total -i -P               '.' > inode || framework_failure_
 cat inode

 $PERL check-df space || fail=1

Gilles

Message #14 received at 12322 <at> debbugs.gnu.org (full text, mbox):

From: g.esp <at> free.fr
To: Bernhard Voelker <mail <at> bernhard-voelker.de>
Cc: 12322 <at> debbugs.gnu.org
Subject: Re: bug#12322: coreutils-8.1{8, 9} with /etc/mtab as real file,
	df/total-unprocessed fail, df/no-mtab-status is skipped
Date: Sun, 30 Sep 2012 10:29:33 +0200 (CEST)

----- Mail original -----
De: "g esp" <g.esp <at> free.fr>
À: "Bernhard Voelker" <mail <at> bernhard-voelker.de>
Cc: 12322 <at> debbugs.gnu.org
Envoyé: Lundi 3 Septembre 2012 00:07:26
Objet: bug#12322: coreutils-8.1{8, 9} with /etc/mtab as real file,	df/total-unprocessed fail, df/no-mtab-status is skipped



----- Mail original -----
> De: "Bernhard Voelker" <mail <at> bernhard-voelker.de>
> À: "g esp" <g.esp <at> free.fr>
> Cc: 12322 <at> debbugs.gnu.org
> Envoyé: Dimanche 2 Septembre 2012 11:05:28
> Objet: bug#12322: coreutils-8.1{8, 9} with /etc/mtab as real file,	df/total-unprocessed fail, df/no-mtab-status is
> skipped
> 

I fixed how /etc/mtab is filled inside my chroot using
	# Use mtab as regular file. This allow 'df' not to fail during coreutils test with mount points out
	# of the chroot that didn't exist inside (like debian /lib/init/rw)
	rm -f /etc/mtab		# In case you experiment with symlink to /proc/mounts
	# rootfs need to be on top of file or df -T -t <yourfstype> / will fail with df: no file systems processed
	grep ' / ' /proc/mounts >/etc/mtab
	# Remove the duplicate due to the mount --bind
	grep -E '/sys sysfs|/proc proc|/dev/pts|/dev/shm' /proc/mounts | sort -u >>/etc/mtab

So without changing coreutils test suite (except separate issue with test-getlogin.c isatty), tests look better

Using a correct /etc/mtab, that changed unobviously those results (as they bot use df)
-skip-seek-past-dev: skipped test: failed to determine size of -
-SKIP: dd/skip-seek-past-dev
+PASS: dd/skip-seek-past-dev

-slink: skipped test: '.' is on an XFS file system
-SKIP: du/slink
+PASS: du/slink

I was always suspicious about '.' is on an XFS file system as / is ext3 on this machine.

cp/fiemap-perf is the test that fail when on /etc/mtab, rootfs is after the /dev with / partition.
[chroot-i486] root:/$ grep ' / ' /proc/mounts | sort >/etc/mtab
[chroot-i486] root:/$ cat /etc/mtab
/dev/disk/by-uuid/7a235d64-5d04-41ac-a959-70465eb74fc8 / ext3 rw,relatime,errors=remount-ro,data=ordered 0 0
rootfs / rootfs rw 0 0
[chroot-i486] root:/$ df -T -t ext2 -t ext3 . && echo should skip with ext3
df: no file systems processed
[chroot-i486] root:/$ df -T  && echo done
Filesystem     Type   1K-blocks     Used Available Use% Mounted on
/dev/sda3      ext3    19228308 14717448   3534108  81% /
rootfs         rootfs  19228308 14717448   3534108  81% /
done

Should not some df tests be added that warn on malformed /etc/mtab?

As a notice, util-linux-2.22 libmount code may let some stray loops after coreutils tests have run with regular /etc/mtab and a running kernel before 2.6.37, git version of util-linux is fixed.

Gilles

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.