binfmt_flat: Fix integer overflow bug on 32 bit systems

commit 55cf2f4b945f6a6416cc2524ba740b83cc9af25a upstream.

Most of these sizes and counts are capped at 256MB so the math doesn't
result in an integer overflow.  The "relocs" count needs to be checked
as well.  Otherwise on 32bit systems the calculation of "full_data"
could be wrong.

	full_data = data_len + relocs * sizeof(unsigned long);

Fixes: c995ee28d29d ("binfmt_flat: prevent kernel dammage from corrupted executable headers")
Cc: stable@vger.kernel.org
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Acked-by: Nicolas Pitre <npitre@baylibre.com>
Link: https://lore.kernel.org/r/5be17f6c-5338-43be-91ef-650153b975cb@stanley.mountain
Signed-off-by: Kees Cook <kees@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c
index 69f4db05191a3218bebfc65574e7e6d784eb1110..c29c8e7f95d362547a6a24a742799ef260bbba5b 100644 (file)
--- a/fs/binfmt_flat.c
+++ b/fs/binfmt_flat.c
@@ -534,7 +534,7 @@ static int load_flat_file(struct linux_binprm *bprm,
         * 28 bits (256 MB) is way more than reasonable in this case.
         * If some top bits are set we have probable binary corruption.
        */
-       if ((text_len | data_len | bss_len | stack_len | full_data) >> 28) {
+       if ((text_len | data_len | bss_len | stack_len | relocs | full_data) >> 28) {
                pr_err("bad header\n");
                ret = -ENOEXEC;
                goto err;