Bash-4.2 Official Patch 38
BASH PATCH REPORT
=
Bash-Release: 4.2
Patch-ID: bash42-038
Bug-Reported-by:arman...@gmail.com
Bug-Reference-ID: <20120822112810.8d14920...@windmill.latviatours.lv>
Bug-Reference-URL:
http://lists.gnu.org/archive/html/bug-bash/2012-08/msg00049.html
Bug-Description:
If a backslash-newline (which is removed) with no other input is given as
input to `read', the shell tries to dereference a null pointer and seg faults.
Patch (apply with `patch -p0'):
*** ../bash-4.2-patched/builtins/read.def 2012-03-11 17:52:44.0
-0400
--- builtins/read.def 2012-08-22 11:53:09.0 -0400
***
*** 792,796
#endif
! if (saw_escape)
{
t = dequote_string (input_string);
--- 847,851
#endif
! if (saw_escape && input_string && *input_string)
{
t = dequote_string (input_string);
*** ../bash-4.2-patched/patchlevel.hSat Jun 12 20:14:48 2010
--- patchlevel.hThu Feb 24 21:41:34 2011
***
*** 26,30
looks for to find the patch level (for the sccs version string). */
! #define PATCHLEVEL 37
#endif /* _PATCHLEVEL_H_ */
--- 26,30
looks for to find the patch level (for the sccs version string). */
! #define PATCHLEVEL 38
#endif /* _PATCHLEVEL_H_ */
--
``The lyf so short, the craft so long to lerne.'' - Chaucer
``Ars longa, vita brevis'' - Hippocrates
Chet Ramey, ITS, CWRUc...@case.eduhttp://cnswww.cns.cwru.edu/~chet/
Bash-4.2 Official Patch 39
BASH PATCH REPORT
=
Bash-Release: 4.2
Patch-ID: bash42-039
Bug-Reported-by:Dan Douglas
Bug-Reference-ID: <1498458.MpVlmOXDB7@smorgbox>
Bug-Reference-URL:
http://lists.gnu.org/archive/html/bug-bash/2012-09/msg8.html
Bug-Description:
Under certain circumstances, bash attempts to expand variables in arithmetic
expressions even when evaluation is being suppressed.
Patch (apply with `patch -p0'):
*** ../bash-4.2-patched/expr.c 2011-11-21 18:03:35.0 -0500
--- expr.c 2012-09-09 16:31:18.0 -0400
***
*** 1010,1013
--- 1073,1082
#endif
+ /*itrace("expr_streval: %s: noeval = %d", tok, noeval);*/
+ /* If we are suppressing evaluation, just short-circuit here instead of
+ going through the rest of the evaluator. */
+ if (noeval)
+ return (0);
+
/* [ */
#if defined (ARRAY_VARS)
***
*** 1183,1186
--- 1256,1263
*cp = '\0';
+ /* XXX - watch out for pointer aliasing issues here */
+ if (curlval.tokstr && curlval.tokstr == tokstr)
+ init_lvalue (&curlval);
+
FREE (tokstr);
tokstr = savestring (tp);
*** ../bash-4.2-patched/patchlevel.hSat Jun 12 20:14:48 2010
--- patchlevel.hThu Feb 24 21:41:34 2011
***
*** 26,30
looks for to find the patch level (for the sccs version string). */
! #define PATCHLEVEL 38
#endif /* _PATCHLEVEL_H_ */
--- 26,30
looks for to find the patch level (for the sccs version string). */
! #define PATCHLEVEL 39
#endif /* _PATCHLEVEL_H_ */
--
``The lyf so short, the craft so long to lerne.'' - Chaucer
``Ars longa, vita brevis'' - Hippocrates
Chet Ramey, ITS, CWRUc...@case.eduhttp://cnswww.cns.cwru.edu/~chet/
Trap variable scope
Hi everyone,
I've encountered a very strange behavior regarding variable scope and traps,
which looks very much like a bug.
This command:
echo '
set -e;
tt() { declare -r v=; };t() { tt; };
ff() { declare -r v=; false; }; f() { ff; };
trap t EXIT;
f
' | bash
produces this error message:
bash: line 3: declare: v: readonly variable
While this:
bash -c '
set -e;
tt() { declare -r v=; };t() { tt; };
ff() { declare -r v=; false; }; f() { ff; };
trap t EXIT;
f
'
doesn't. As don't these:
echo '
set -e;
tt() { declare -r v=; };t() { tt; };
ff() { declare -r v=; false; };
trap t EXIT;
ff
' | bash
echo '
set -e;
tt() { declare -r v=; };
ff() { declare -r v=; false; }; f() { ff; };
trap tt EXIT;
f
' | bash
Could this indeed be a bug? If yes, can it be fixed?
Thank you.
Sincerely,
Nick
