diff options
| author |   Siddhesh Poyarekar <siddhesh@redhat.com> | 2013-09-23 11:24:30 +0530 |
|---|---|---|
| committer | Siddhesh Poyarekar <siddhesh@redhat.com> | 2013-09-23 11:29:53 +0530 |
| commit | 303e567a8062200dc06acde7c76fc34679f08d8f (patch) | |
| tree | 8e0c198956de9addb51216c5cfccd47d7c4be69b /string/strcoll_l.c | |
| parent | Fall back to non-cached sequence traversal and comparison on malloc fail (diff) | |
| download | glibc-303e567a8062200dc06acde7c76fc34679f08d8f.tar.gz glibc-303e567a8062200dc06acde7c76fc34679f08d8f.tar.bz2 glibc-303e567a8062200dc06acde7c76fc34679f08d8f.zip | |
Check for integer overflow in cache size computation in strcoll
strcoll is implemented using a cache for indices and weights of
collation sequences in the strings so that subsequent passes do not
have to search through collation data again. For very large string
inputs, the cache size computation could overflow. In such a case,
use the fallback function that does not cache indices and weights of
collation sequences.
Fixes CVE-2012-4412.
Diffstat (limited to 'string/strcoll_l.c')
| -rw-r--r-- | string/strcoll_l.c | 10 |
1 files changed, 9 insertions, 1 deletions
diff --git a/string/strcoll_l.c b/string/strcoll_l.c index eb042ff2ec..4ee101a118 100644 --- a/string/strcoll_l.c +++ b/string/strcoll_l.c @@ -524,7 +524,15 @@ STRCOLL (const STRING_TYPE *s1, const STRING_TYPE *s2, __locale_t l) memset (&seq1, 0, sizeof (seq1)); seq2 = seq1; - if (! __libc_use_alloca ((s1len + s2len) * (sizeof (int32_t) + 1))) + size_t size_max = SIZE_MAX / (sizeof (int32_t) + 1); + + if (MIN (s1len, s2len) > size_max + || MAX (s1len, s2len) > size_max - MIN (s1len, s2len)) + { + /* If the strings are long enough to cause overflow in the size request, + then skip the allocation and proceed with the non-cached routines. */ + } + else if (! __libc_use_alloca ((s1len + s2len) * (sizeof (int32_t) + 1))) { seq1.idxarr = (int32_t *) malloc ((s1len + s2len) * (sizeof (int32_t) + 1)); |