[Bug regex/1357] New: memory leak in re_compile_pattern()

[Bug regex/1357] New: memory leak in re_compile_pattern()

[root at liwave dot com]

Calling re_compile_pattern results in a memory leak proportional to the size of 
the Pattern String. See Example code below:
#include <regex.h>
#include <stdio.h>

int main()
{
	struct re_pattern_buffer	*patBuf;
	const char			*ErrStr;
	char				 PatStr[128];
	unsigned int			 i;
	


	patBuf = (struct re_pattern_buffer *) malloc(sizeof(struct 
re_pattern_buffer));
	if (!patBuf)
		exit(0);
	memset(patBuf, 0, sizeof(struct re_pattern_buffer));

	patBuf->translate      = 0;
	patBuf->fastmap	       = 0;
	patBuf->buffer	       = (char *) malloc(2048); //Optional
	patBuf->allocated      = 2048;
	patBuf->regs_allocated = REGS_UNALLOCATED; /Optional
	patBuf->used           = 0;

        re_set_syntax (RE_SYNTAX_POSIX_EGREP); //Optional
	for (i = 0; i < 10000; i++) {
		sprintf(PatStr, "TEST.?STRING%d", i); //This can be anything
                memset(patBuf->buffer, 0, 2048);
		ErrStr = re_compile_pattern(PatStr,
			                    strlen(PatStr),
					    patBuf);
		if (ErrStr)
		    printf("ErrStr = %s\n",ErrStr);
		} //endof for (i = 0; i < 10000; i++)
	if (patBuf->buffer)
		free(patBuf->buffer);
	if (patBuf)
		free(patBuf);
	
	exit(1);
	} //endof main()

Leak was introduced after glibc-2.2.2. This code works fine on a linux 2.4 
system running with glibc-2.2.2, leak occures in glibc 2.3.5 and 2.3.1 
glibc 2.3.5 was compiled with:
CFLAGS="-O3 -march=i686 -mcpu=i686 -funroll-loops -fomit-frame-pointer -
fexpensive-optimizations -pipe"

-- 
           Summary: memory leak in re_compile_pattern()
           Product: glibc
           Version: 2.3.5
            Status: NEW
          Severity: normal
          Priority: P2
         Component: regex
        AssignedTo: gotom at debian dot or dot jp
        ReportedBy: root at liwave dot com
                CC: glibc-bugs-regex at sources dot redhat dot com,glibc-
                    bugs at sources dot redhat dot com
  GCC host triplet: Linux gentoo 2.6.7-hardened-r18 #2 SMP Sun Jan 16
                    18:37:38 EST 2


http://sourceware.org/bugzilla/show_bug.cgi?id=1357

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug regex/1357] memory leak in re_compile_pattern()

[drepper at redhat dot com]

------- Additional Comments From drepper at redhat dot com  2005-10-14 05:49 -------
Yes, that interface doesn't work as before.  And it might remain broken.  It was
tied to the old implementation.

There might be a chance that the behavior is restored but I wouldn't hold my
breath.  Better rewrite your code to use the POSIX interfaces.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|gotom at debian dot or dot  |drepper at redhat dot com
                   |jp                          |
             Status|NEW                         |ASSIGNED


http://sourceware.org/bugzilla/show_bug.cgi?id=1357

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug regex/1357] memory leak in re_compile_pattern()

[drepper at redhat dot com]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   GCC host triplet|Linux gentoo 2.6.7-hardened-|Linux
                   |r18 #2 SMP Sun Jan 16       |
                   |18:37:38 EST 2              |


http://sourceware.org/bugzilla/show_bug.cgi?id=1357

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug regex/1357] memory leak in re_compile_pattern()

[drepper at redhat dot com]

------- Additional Comments From drepper at redhat dot com  2006-05-02 21:56 -------
I meant to suspend this bug.  Nobody I know is going to work on this.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |SUSPENDED


http://sourceware.org/bugzilla/show_bug.cgi?id=1357

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug regex/1357] memory leak in re_compile_pattern()

[jsm28 at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=1357

Joseph Myers <jsm28 at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|WONTFIX                     |---

--- Comment #4 from Joseph Myers <jsm28 at gcc dot gnu.org> ---
I'm not sure what "will does not compile" means.  If you change "/Optional" to
"//Optional" then the code given here compiles with -D_GNU_SOURCE.  Maybe
there's something wrong with the code, but you need a clearer analysis if so
before closing the bug (and WONTFIX should only apply e.g. to a bug in a port
that's been removed; INVALID is what to do if the fault is in the user program
or the requested change is otherwise undesirable, just leave the bug open if
you're not interested in fixing it).

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug regex/1357] memory leak in re_compile_pattern()

[neleai at seznam dot cz]

https://sourceware.org/bugzilla/show_bug.cgi?id=1357

Ondrej Bilka <neleai at seznam dot cz> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|SUSPENDED                   |RESOLVED
                 CC|                            |neleai at seznam dot cz
         Resolution|---                         |WONTFIX

--- Comment #3 from Ondrej Bilka <neleai at seznam dot cz> ---
Closing as mentioned interface will does not compile.

-- 
You are receiving this mail because:
You are on the CC list for the bug.