Eric Blake <ebb9 <at> byu.net> writes:
> It looks like this commit is the culprit:
> http://git.sv.gnu.org/gitweb/?p=coreutils.git;a=commitdiff;h=0a74437
>
> It looks like a true regression. The change was introduced because of this
> thread:
> http://lists.gnu.org/archive/html/bug-coreutils/2007-06/msg00004.html
> http://lists.gnu.org/archive/html/bug-coreutils/2007-07/msg00079.html
>
> It looks like when -l is NOT in effect, you should not stat the result of a
> symlink. But when -l IS in effect, and C_EXEC is enabled, then the result of
> stat'ing the symling IS used, in order to color the symlink target
differently
> than the color of the symlink.
>
And I believe this patch fixes things. I don't know if Corinna wants to be
mentioned in THANKS (she's been generally reluctant to expose her email address
on the cygwin lists); hence the obfuscated address in the log message for now.
>From ff17266b281a93adafd9ac2484d9836e8d1c97a1 Mon Sep 17 00:00:00 2001
From: Eric Blake <[EMAIL PROTECTED]>
Date: Mon, 14 Apr 2008 11:14:00 -0600
Subject: [PATCH] Fix ls -l coloring regression from 2007-07-15.
* src/ls.c (main): Stat symlinks for long format if the target
might need to be colored as an executable.
* tests/ls/stat-free-symlinks (LS_COLORS): Enhance test.
* NEWS: Mention the fix.
Reported by Corinna Vinschen <corinna-cygwin AT cygwin DOT com>.
Signed-off-by: Eric Blake <[EMAIL PROTECTED]>
---
NEWS | 3 +++
src/ls.c | 3 ++-
tests/ls/stat-free-symlinks | 21 +++++++++++++++++++--
3 files changed, 24 insertions(+), 3 deletions(-)
diff --git a/NEWS b/NEWS
index e208b30..e86503e 100644
--- a/NEWS
+++ b/NEWS
@@ -20,6 +20,9 @@ GNU coreutils NEWS -*-
outline -*-
ls no longer segfaults on files in /proc when linked with an older version
of libselinux. E.g., ls -l /proc/sys would dereference a NULL pointer.
+ ls -l --color would mistakenly fail to color the target of a symlink.
+ [introduced in coreutils-6.9]
+
"mkdir -Z x dir" no longer segfaults when diagnosing invalid context "x"
mkfifo and mknod would fail similarly. Now they're fixed.
diff --git a/src/ls.c b/src/ls.c
index e029fe0..d474ecc 100644
--- a/src/ls.c
+++ b/src/ls.c
@@ -1177,7 +1177,8 @@ main (int argc, char **argv)
{
/* Avoid following symbolic links when possible. */
if (is_colored (C_ORPHAN)
- || (is_colored (C_EXEC) && color_symlink_as_referent)
+ || (is_colored (C_EXEC)
+ && (color_symlink_as_referent || format == long_format))
|| (is_colored (C_MISSING) && format == long_format))
check_symlink_color = true;
diff --git a/tests/ls/stat-free-symlinks b/tests/ls/stat-free-symlinks
index 07fb38e..f419410 100755
--- a/tests/ls/stat-free-symlinks
+++ b/tests/ls/stat-free-symlinks
@@ -32,8 +32,8 @@ fail=0
# ls from coreutils 6.9 would unnecessarily stat a symlink in an unusual case:
# When not coloring orphan and missing entries, and without ln=target,
-# ensure that ls -F (or -d, or -l: i.e., when not dereferencing)
-# does not stat a symlink to directory, and does still color that
+# ensure that ls -F (or -d: i.e., when not dereferencing)
+# does not stat a symlink, and does still color that
# symlink and an executable file properly.
LS_COLORS='or=0:mi=0:ex=01;32:ln=01;35' \
@@ -52,4 +52,21 @@ grep '^stat("x"' err && fail=1
compare out exp || fail=1
+# ls from coreutils 6.10 would mistakenly not stat a symlink in a counter
+# case: When coloring executable files and without ln=target, ensure that
+# ls -l correctly colors both halves of the link listing.
+
+LS_COLORS='or=0:mi=0:ex=01;32:ln=01;35' \
+ ls -l --color=always link-to-x > out 2> err || fail=1
+
+# Check that output is colorized as requested.
+{
+ ls -l link-to-x | sed 's/link-to-x -> x//' | tr -d '\n'
+ printf '\033[0m\033[01;35mlink-to-x\033[0m -> \033[01;32mx\033[0m\n'
+ printf '\033[m'
+} > exp 2> err || fail=1
+
+test -s err || fail=1
+compare out exp || fail=1
+
(exit $fail); exit $fail
--
1.5.5
_______________________________________________
Bug-coreutils mailing list
Bug-coreutils@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-coreutils
