Re: [PATCH 1/3] ewah_read_mmap: bounds-check mmap reads

[Junio C Hamano <gitster@pobox.com>]

Jeff King <peff@peff.net> writes:

> diff --git a/ewah/ewah_io.c b/ewah/ewah_io.c
> index bed1994551..33c08c40f8 100644
> --- a/ewah/ewah_io.c
> +++ b/ewah/ewah_io.c
> @@ -122,16 +122,23 @@ int ewah_serialize_strbuf(struct ewah_bitmap *self, struct strbuf *sb)
>  	return ewah_serialize_to(self, write_strbuf, sb);
>  }
>  
> -int ewah_read_mmap(struct ewah_bitmap *self, const void *map, size_t len)
> +ssize_t ewah_read_mmap(struct ewah_bitmap *self, const void *map, size_t len)
>  {
>  	const uint8_t *ptr = map;
> +	size_t data_len;
>  	size_t i;
>  
> +	if (len < sizeof(uint32_t))
> +		return error("corrupt ewah bitmap: eof before bit size");
>  	self->bit_size = get_be32(ptr);
>  	ptr += sizeof(uint32_t);
> +	len -= sizeof(uint32_t);
>  
> +	if (len < sizeof(uint32_t))
> +		return error("corrupt ewah bitmap: eof before length");
>  	self->buffer_size = self->alloc_size = get_be32(ptr);
>  	ptr += sizeof(uint32_t);
> +	len -= sizeof(uint32_t);
>  
>  	REALLOC_ARRAY(self->buffer, self->alloc_size);
>  
> @@ -141,15 +148,25 @@ int ewah_read_mmap(struct ewah_bitmap *self, const void *map, size_t len)
>  	 * the endianness conversion in a separate pass to ensure
>  	 * we're loading 8-byte aligned words.
>  	 */
> -	memcpy(self->buffer, ptr, self->buffer_size * sizeof(eword_t));
> -	ptr += self->buffer_size * sizeof(eword_t);


> +	data_len = st_mult(self->buffer_size, sizeof(eword_t));

This is a faithful conversion from the original, but I somehow would
have appreciated if the latter were not sizeof(eword_t) but rather
sizeof(self->buffer_size[0]), especially as I wondered ...

> +	if (len < data_len)
> +		return error("corrupt ewah bitmap: eof in data "
> +			     "(%"PRIuMAX" bytes short)",
> +			     (uintmax_t)(data_len - len));
> +	memcpy(self->buffer, ptr, data_len);
> +	ptr += data_len;
> +	len -= data_len;
>  
>  	for (i = 0; i < self->buffer_size; ++i)
>  		self->buffer[i] = ntohll(self->buffer[i]);

... what individual datum one iteration of this loop is copying, and
then realized "buffer_size" is a misleading field name (anything that 
claims to be size and not measuring in bytes is misleading to me ;-).

> +	if (len < sizeof(uint32_t))
> +		return error("corrupt ewah bitmap: eof before rlw");
>  	self->rlw = self->buffer + get_be32(ptr);
> +	ptr += sizeof(uint32_t);
> +	len -= sizeof(uint32_t);
>  
> -	return (3 * 4) + (self->buffer_size * 8);
> +	return ptr - (const uint8_t *)map;

Much nicer; I needed to wonder what these 12 and 8 in the original are.

>  
>  int ewah_deserialize(struct ewah_bitmap *self, int fd);
> -int ewah_read_mmap(struct ewah_bitmap *self, const void *map, size_t len);
> +ssize_t ewah_read_mmap(struct ewah_bitmap *self, const void *map, size_t len);

I double checked all the callers and made sure that they are already
prepared to react sensibly to error returns, which is good.

>  
>  uint32_t ewah_checksum(struct ewah_bitmap *self);
>  
> diff --git a/t/t5310-pack-bitmaps.sh b/t/t5310-pack-bitmaps.sh
> index 423c0a475f..237ee6e5fc 100755
> --- a/t/t5310-pack-bitmaps.sh
> +++ b/t/t5310-pack-bitmaps.sh
> @@ -331,4 +331,17 @@ test_expect_success 'pack reuse respects --incremental' '
>  	git show-index <empty.idx >actual &&
>  	test_cmp expect actual
>  '
> +
> +test_expect_success 'truncated bitmap fails gracefully' '
> +	git repack -ad &&
> +	git rev-list --use-bitmap-index --count --all >expect &&
> +	bitmap=$(ls .git/objects/pack/*.bitmap) &&
> +	test_when_finished "rm -f $bitmap" &&
> +	head -c 512 <$bitmap >$bitmap.tmp &&
> +	mv $bitmap.tmp $bitmap &&
> +	git rev-list --use-bitmap-index --count --all >actual 2>stderr &&
> +	test_cmp expect actual &&
> +	test_i18ngrep corrupt stderr
> +'
> +
>  test_done