Fix memory leak in __printf_fp_l (bug 26215)

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Fix memory leak in __printf_fp_l (bug 26215)

Joseph Myers
__printf_fp_l has a memory leak in the case of some I/O errors, where
both buffer and wbuffer have been malloced but the handling of I/O
errors only frees wbuffer.  This patch fixes this by moving the
declaration of buffer to an outer scope and ensuring that it is freed
when wbuffer is freed.

Tested for x86_64 and x86.

---

This patch is relative to a tree with my previous fix for bug 26214
<https://sourceware.org/pipermail/libc-alpha/2020-July/116043.html>
(pending review) applied.

diff --git a/stdio-common/Makefile b/stdio-common/Makefile
index dc3fd38a19..8475fd1f09 100644
--- a/stdio-common/Makefile
+++ b/stdio-common/Makefile
@@ -68,7 +68,8 @@ tests := tstscanf test_rdwr test-popen tstgetln test-fseek \
  scanf14a scanf16a \
  tst-printf-bz25691 \
  tst-vfprintf-width-prec-alloc \
- tst-printf-fp-free
+ tst-printf-fp-free \
+ tst-printf-fp-leak
 
 
 test-srcs = tst-unbputc tst-printf tst-printfsz-islongdouble
@@ -80,12 +81,14 @@ tests-special += $(objpfx)tst-unbputc.out $(objpfx)tst-printf.out \
  $(objpfx)tst-vfprintf-width-prec-mem.out \
  $(objpfx)tst-printfsz-islongdouble.out \
  $(objpfx)tst-printf-bz25691-mem.out \
- $(objpfx)tst-printf-fp-free-mem.out
+ $(objpfx)tst-printf-fp-free-mem.out \
+ $(objpfx)tst-printf-fp-leak-mem.out
 generated += tst-printf-bz18872.c tst-printf-bz18872.mtrace \
      tst-printf-bz18872-mem.out \
      tst-vfprintf-width-prec.mtrace tst-vfprintf-width-prec-mem.out \
      tst-printf-bz25691.mtrace tst-printf-bz25691-mem.out \
-     tst-printf-fp-free.mtrace tst-printf-fp-free-mem.out
+     tst-printf-fp-free.mtrace tst-printf-fp-free-mem.out \
+     tst-printf-fp-leak.mtrace tst-printf-fp-leak-mem.out
 endif
 
 tests-special += $(objpfx)tst-errno-manual.out
@@ -113,6 +116,8 @@ tst-printf-bz25691-ENV = \
   MALLOC_TRACE=$(objpfx)tst-printf-bz25691.mtrace
 tst-printf-fp-free-ENV = \
   MALLOC_TRACE=$(objpfx)tst-printf-fp-free.mtrace
+tst-printf-fp-leak-ENV = \
+  MALLOC_TRACE=$(objpfx)tst-printf-fp-leak.mtrace
 
 $(objpfx)tst-unbputc.out: tst-unbputc.sh $(objpfx)tst-unbputc
  $(SHELL) $< $(common-objpfx) '$(test-program-prefix)' > $@; \
diff --git a/stdio-common/printf_fp.c b/stdio-common/printf_fp.c
index 49c693575e..c0b79f2184 100644
--- a/stdio-common/printf_fp.c
+++ b/stdio-common/printf_fp.c
@@ -72,7 +72,10 @@
       if (putc (outc, fp) == EOF)      \
  {      \
   if (buffer_malloced)      \
-    free (wbuffer);      \
+    {      \
+      free (buffer);      \
+      free (wbuffer);      \
+    }      \
   return -1;      \
  }      \
       ++done;      \
@@ -87,7 +90,10 @@
   if (PUT (fp, wide ? (const char *) wptr : ptr, outlen) != outlen)   \
     {      \
       if (buffer_malloced)      \
- free (wbuffer);      \
+ {      \
+  free (buffer);      \
+  free (wbuffer);      \
+ }      \
       return -1;      \
     }      \
   ptr += outlen;      \
@@ -110,7 +116,10 @@
       if (PAD (fp, ch, len) != len)      \
  {      \
   if (buffer_malloced)      \
-    free (wbuffer);      \
+    {      \
+      free (buffer);      \
+      free (wbuffer);      \
+    }      \
   return -1;      \
  }      \
       done += len;      \
@@ -259,7 +268,8 @@ __printf_fp_l (FILE *fp, locale_t loc,
 
   /* Buffer in which we produce the output.  */
   wchar_t *wbuffer = NULL;
-  /* Flag whether wbuffer is malloc'ed or not.  */
+  char *buffer = NULL;
+  /* Flag whether wbuffer and buffer are malloc'ed or not.  */
   int buffer_malloced = 0;
 
   p.expsign = 0;
@@ -1172,7 +1182,6 @@ __printf_fp_l (FILE *fp, locale_t loc,
       PADN ('0', width);
 
     {
-      char *buffer = NULL;
       char *buffer_end = NULL;
       char *cp = NULL;
       char *tmpptr;
@@ -1252,6 +1261,7 @@ __printf_fp_l (FILE *fp, locale_t loc,
   free (wbuffer);
   /* Avoid a double free if the subsequent PADN encounters an
      I/O error.  */
+  buffer = NULL;
   wbuffer = NULL;
  }
     }
diff --git a/stdio-common/tst-printf-fp-leak.c b/stdio-common/tst-printf-fp-leak.c
new file mode 100644
index 0000000000..da11a486dd
--- /dev/null
+++ b/stdio-common/tst-printf-fp-leak.c
@@ -0,0 +1,34 @@
+/* Test memory leak in __printf_fp_l (bug 26215).
+   Copyright (C) 2020 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 2.1 of the License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with the GNU C Library; if not, see
+   <https://www.gnu.org/licenses/>.  */
+
+#include <mcheck.h>
+#include <stdio.h>
+#include <support/check.h>
+
+static int
+do_test (void)
+{
+  mtrace ();
+  FILE *fp = fopen ("/dev/full", "w");
+  TEST_VERIFY_EXIT (fp != NULL);
+  TEST_COMPARE (fprintf (fp, "%.65536f", 1.0), -1);
+  fclose (fp);
+  return 0;
+}
+
+#include <support/test-driver.c>

--
Joseph S. Myers
[hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: Fix memory leak in __printf_fp_l (bug 26215)

Sourceware - libc-alpha mailing list


On 09/07/2020 15:07, Joseph Myers wrote:
> __printf_fp_l has a memory leak in the case of some I/O errors, where
> both buffer and wbuffer have been malloced but the handling of I/O
> errors only frees wbuffer.  This patch fixes this by moving the
> declaration of buffer to an outer scope and ensuring that it is freed
> when wbuffer is freed.
>
> Tested for x86_64 and x86.

LGTM, thanks.

Reviewed-by: Adhemerval Zanella <[hidden email]>

>
> ---
>
> This patch is relative to a tree with my previous fix for bug 26214
> <https://sourceware.org/pipermail/libc-alpha/2020-July/116043.html>
> (pending review) applied.
>
> diff --git a/stdio-common/Makefile b/stdio-common/Makefile
> index dc3fd38a19..8475fd1f09 100644
> --- a/stdio-common/Makefile
> +++ b/stdio-common/Makefile
> @@ -68,7 +68,8 @@ tests := tstscanf test_rdwr test-popen tstgetln test-fseek \
>   scanf14a scanf16a \
>   tst-printf-bz25691 \
>   tst-vfprintf-width-prec-alloc \
> - tst-printf-fp-free
> + tst-printf-fp-free \
> + tst-printf-fp-leak
>  
>  
>  test-srcs = tst-unbputc tst-printf tst-printfsz-islongdouble

Ok.

> @@ -80,12 +81,14 @@ tests-special += $(objpfx)tst-unbputc.out $(objpfx)tst-printf.out \
>   $(objpfx)tst-vfprintf-width-prec-mem.out \
>   $(objpfx)tst-printfsz-islongdouble.out \
>   $(objpfx)tst-printf-bz25691-mem.out \
> - $(objpfx)tst-printf-fp-free-mem.out
> + $(objpfx)tst-printf-fp-free-mem.out \
> + $(objpfx)tst-printf-fp-leak-mem.out
>  generated += tst-printf-bz18872.c tst-printf-bz18872.mtrace \
>       tst-printf-bz18872-mem.out \
>       tst-vfprintf-width-prec.mtrace tst-vfprintf-width-prec-mem.out \
>       tst-printf-bz25691.mtrace tst-printf-bz25691-mem.out \
> -     tst-printf-fp-free.mtrace tst-printf-fp-free-mem.out
> +     tst-printf-fp-free.mtrace tst-printf-fp-free-mem.out \
> +     tst-printf-fp-leak.mtrace tst-printf-fp-leak-mem.out
>  endif
>  
>  tests-special += $(objpfx)tst-errno-manual.out
> @@ -113,6 +116,8 @@ tst-printf-bz25691-ENV = \
>    MALLOC_TRACE=$(objpfx)tst-printf-bz25691.mtrace
>  tst-printf-fp-free-ENV = \
>    MALLOC_TRACE=$(objpfx)tst-printf-fp-free.mtrace
> +tst-printf-fp-leak-ENV = \
> +  MALLOC_TRACE=$(objpfx)tst-printf-fp-leak.mtrace
>  
>  $(objpfx)tst-unbputc.out: tst-unbputc.sh $(objpfx)tst-unbputc
>   $(SHELL) $< $(common-objpfx) '$(test-program-prefix)' > $@; \

Ok.

> diff --git a/stdio-common/printf_fp.c b/stdio-common/printf_fp.c
> index 49c693575e..c0b79f2184 100644
> --- a/stdio-common/printf_fp.c
> +++ b/stdio-common/printf_fp.c
> @@ -72,7 +72,10 @@
>        if (putc (outc, fp) == EOF)      \
>   {      \
>    if (buffer_malloced)      \
> -    free (wbuffer);      \
> +    {      \
> +      free (buffer);      \
> +      free (wbuffer);      \
> +    }      \
>    return -1;      \
>   }      \
>        ++done;      \
> @@ -87,7 +90,10 @@
>    if (PUT (fp, wide ? (const char *) wptr : ptr, outlen) != outlen)   \
>      {      \
>        if (buffer_malloced)      \
> - free (wbuffer);      \
> + {      \
> +  free (buffer);      \
> +  free (wbuffer);      \
> + }      \
>        return -1;      \
>      }      \
>    ptr += outlen;      \
> @@ -110,7 +116,10 @@
>        if (PAD (fp, ch, len) != len)      \
>   {      \
>    if (buffer_malloced)      \
> -    free (wbuffer);      \
> +    {      \
> +      free (buffer);      \
> +      free (wbuffer);      \
> +    }      \
>    return -1;      \
>   }      \
>        done += len;      \

Ok.

> @@ -259,7 +268,8 @@ __printf_fp_l (FILE *fp, locale_t loc,
>  
>    /* Buffer in which we produce the output.  */
>    wchar_t *wbuffer = NULL;
> -  /* Flag whether wbuffer is malloc'ed or not.  */
> +  char *buffer = NULL;
> +  /* Flag whether wbuffer and buffer are malloc'ed or not.  */
>    int buffer_malloced = 0;
>  
>    p.expsign = 0;
> @@ -1172,7 +1182,6 @@ __printf_fp_l (FILE *fp, locale_t loc,
>        PADN ('0', width);
>  
>      {
> -      char *buffer = NULL;
>        char *buffer_end = NULL;
>        char *cp = NULL;
>        char *tmpptr;
> @@ -1252,6 +1261,7 @@ __printf_fp_l (FILE *fp, locale_t loc,
>    free (wbuffer);
>    /* Avoid a double free if the subsequent PADN encounters an
>       I/O error.  */
> +  buffer = NULL;
>    wbuffer = NULL;
>   }
>      }

Ok.

> diff --git a/stdio-common/tst-printf-fp-leak.c b/stdio-common/tst-printf-fp-leak.c
> new file mode 100644
> index 0000000000..da11a486dd
> --- /dev/null
> +++ b/stdio-common/tst-printf-fp-leak.c
> @@ -0,0 +1,34 @@
> +/* Test memory leak in __printf_fp_l (bug 26215).
> +   Copyright (C) 2020 Free Software Foundation, Inc.
> +   This file is part of the GNU C Library.
> +
> +   The GNU C Library is free software; you can redistribute it and/or
> +   modify it under the terms of the GNU Lesser General Public
> +   License as published by the Free Software Foundation; either
> +   version 2.1 of the License, or (at your option) any later version.
> +
> +   The GNU C Library is distributed in the hope that it will be useful,
> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> +   Lesser General Public License for more details.
> +
> +   You should have received a copy of the GNU Lesser General Public
> +   License along with the GNU C Library; if not, see
> +   <https://www.gnu.org/licenses/>.  */
> +
> +#include <mcheck.h>
> +#include <stdio.h>
> +#include <support/check.h>
> +
> +static int
> +do_test (void)
> +{
> +  mtrace ();
> +  FILE *fp = fopen ("/dev/full", "w");
> +  TEST_VERIFY_EXIT (fp != NULL);
> +  TEST_COMPARE (fprintf (fp, "%.65536f", 1.0), -1);
> +  fclose (fp);
> +  return 0;
> +}
> +
> +#include <support/test-driver.c>
>

Ok.
Reply | Threaded
Open this post in threaded view
|

Re: Fix memory leak in __printf_fp_l (bug 26215)

Sourceware - libc-alpha mailing list
In reply to this post by Joseph Myers
* Joseph Myers:

> __printf_fp_l has a memory leak in the case of some I/O errors, where
> both buffer and wbuffer have been malloced but the handling of I/O
> errors only frees wbuffer.  This patch fixes this by moving the
> declaration of buffer to an outer scope and ensuring that it is freed
> when wbuffer is freed.

Do we need to treat this as a securityh vulnerability?  I don't think
so, because there are no impacted applications as far as I can tell.

Thanks,
Florian

Reply | Threaded
Open this post in threaded view
|

Re: Fix memory leak in __printf_fp_l (bug 26215)

Joseph Myers
On Fri, 10 Jul 2020, Florian Weimer via Libc-alpha wrote:

> * Joseph Myers:
>
> > __printf_fp_l has a memory leak in the case of some I/O errors, where
> > both buffer and wbuffer have been malloced but the handling of I/O
> > errors only frees wbuffer.  This patch fixes this by moving the
> > declaration of buffer to an outer scope and ensuring that it is freed
> > when wbuffer is freed.
>
> Do we need to treat this as a securityh vulnerability?  I don't think
> so, because there are no impacted applications as far as I can tell.

I don't know whether any of bugs 26201, 26211, 26214, 26215 (all memory
issues in printf with valid but unusual format strings) should be
considered as security issues.

--
Joseph S. Myers
[hidden email]