Bash-4.2 Official Patch 43
BASH PATCH REPORT
=
Bash-Release: 4.2
Patch-ID: bash42-043
Bug-Reported-by:konsolebox
Bug-Reference-ID:
Bug-Reference-URL:
http://lists.gnu.org/archive/html/bug-bash/2013-01/msg00138.html
Bug-Description:
When SIGCHLD is trapped, and a SIGCHLD trap handler runs when a pending
`read -t' invocation times out and generates SIGALRM, bash can crash with
a segmentation fault.
Patch (apply with `patch -p0'):
*** ../bash-4.2-patched/builtins/read.def 2012-10-31 21:22:51.000517000
-0400
--- builtins/read.def 2013-01-25 10:28:16.38000 -0500
***
*** 386,393
/* Tricky. The top of the unwind-protect stack is the free of
input_string. We want to run all the rest and use input_string,
!so we have to remove it from the stack. */
! remove_unwind_protect ();
! run_unwind_frame ("read_builtin");
input_string[i] = '\0'; /* make sure it's terminated */
retval = 128+SIGALRM;
goto assign_vars;
--- 386,403
/* Tricky. The top of the unwind-protect stack is the free of
input_string. We want to run all the rest and use input_string,
!so we have to save input_string temporarily, run the unwind-
!protects, then restore input_string so we can use it later. */
!
input_string[i] = '\0'; /* make sure it's terminated */
+ if (i == 0)
+ {
+ t = (char *)xmalloc (1);
+ t[0] = 0;
+ }
+ else
+ t = savestring (input_string);
+
+ run_unwind_frame ("read_builtin");
+ input_string = t;
retval = 128+SIGALRM;
goto assign_vars;
*** ../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 42
#endif /* _PATCHLEVEL_H_ */
--- 26,30
looks for to find the patch level (for the sccs version string). */
! #define PATCHLEVEL 43
#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 44
BASH PATCH REPORT
=
Bash-Release: 4.2
Patch-ID: bash42-044
Bug-Reported-by:"Dashing"
Bug-Reference-ID: <20130211175049.d90786f...@smtp.hushmail.com>
Bug-Reference-URL:
http://lists.gnu.org/archive/html/bug-bash/2013-02/msg00030.html
Bug-Description:
When converting a multibyte string to a wide character string as part of
pattern matching, bash does not handle the end of the string correctly,
causing the search for the NUL to go beyond the end of the string and
reference random memory. Depending on the contents of that memory, bash
can produce errors or crash.
Patch (apply with `patch -p0'):
*** ../bash-4.2-patched/lib/glob/xmbsrtowcs.c 2012-07-08 21:53:19.0
-0400
--- lib/glob/xmbsrtowcs.c 2013-02-12 12:00:39.0 -0500
***
*** 217,220
--- 217,226
n = mbsnrtowcs(wsbuf+wcnum, &p, nms, wsbuf_size-wcnum, &state);
+ if (n == 0 && p == 0)
+ {
+ wsbuf[wcnum] = L'\0';
+ break;
+ }
+
/* Compensate for taking single byte on wcs conversion failure above. */
if (wcslength == 1 && (n == 0 || n == (size_t)-1))
***
*** 222,226
state = tmp_state;
p = tmp_p;
! wsbuf[wcnum++] = *p++;
}
else
--- 228,238
state = tmp_state;
p = tmp_p;
! wsbuf[wcnum] = *p;
! if (*p == 0)
! break;
! else
! {
! wcnum++; p++;
! }
}
else
*** ../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 43
#endif /* _PATCHLEVEL_H_ */
--- 26,30
looks for to find the patch level (for the sccs version string). */
! #define PATCHLEVEL 44
#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 45
BASH PATCH REPORT
=
Bash-Release: 4.2
Patch-ID: bash42-045
Bug-Reported-by:Stephane Chazelas
Bug-Reference-ID: <20130218195539.ga9...@chaz.gmail.com>
Bug-Reference-URL:
http://lists.gnu.org/archive/html/bug-bash/2013-02/msg00080.html
Bug-Description:
The <&n- and >&n- redirections, which move one file descriptor to another,
leave the file descriptor closed when applied to builtins or compound
commands.
Patch (apply with `patch -p0'):
*** ../bash-4.2-patched/redir.c 2013-01-30 11:56:09.0 -0500
--- redir.c 2013-02-19 09:38:36.0 -0500
***
*** 1008,1011
--- 1008,1021
REDIRECTION_ERROR (r, errno, -1);
}
+ if ((flags & RX_UNDOABLE) && (ri == r_move_input || ri ==
r_move_output))
+ {
+ /* r_move_input and r_move_output add an additional close()
+that needs to be undone */
+ if (fcntl (redirector, F_GETFD, 0) != -1)
+ {
+ r = add_undo_redirect (redir_fd, r_close_this, -1);
+ REDIRECTION_ERROR (r, errno, -1);
+ }
+ }
#if defined (BUFFERED_INPUT)
check_bash_input (redirector);
*** ../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 44
#endif /* _PATCHLEVEL_H_ */
--- 26,30
looks for to find the patch level (for the sccs version string). */
! #define PATCHLEVEL 45
#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/
show-all-if-ambiguous inserts spurious backslashes
If you have "set show-all-if-ambiguous on" in your .inputrc, typing this sequence in bash (in a directory containing some non-hidden files): echo * Results in the list of possible matches being displayed (as expected), but the command line is then edited to: echo \* Which is probably not what the user wanted. ~Felix.
Re: show-all-if-ambiguous inserts spurious backslashes
On 3/10/13 5:26 PM, Marcel (Felix) Giannelia wrote: > If you have "set show-all-if-ambiguous on" in your .inputrc, typing > this sequence in bash (in a directory containing some non-hidden files): > > echo * > > Results in the list of possible matches being displayed (as expected), > but the command line is then edited to: > > echo \* > > Which is probably not what the user wanted. Thanks for the report. The bash default completion code needs to treat TAB with show-all-if-ambigous on the same as straight TAB. This will be fixed in the next release. Chet -- ``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/
