Re: [Qemu-trivial] [Qemu-devel] [PATCH] lm4549: Fix buffer overflow

[Peter Maydell]

On 1 September 2012 11:43, Stefan Weil <address@hidden> wrote:
> Report from smatch:
> lm4549.c:234 lm4549_write_samples(14) error:
>  buffer overflow 's->buffer' 1024 <= 1024
>
> There must be enough space to add two entries starting with index
> s->buffer_level, therefore the old check was wrong.
>
> Signed-off-by: Stefan Weil <address@hidden>
> ---
>  hw/lm4549.c |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/hw/lm4549.c b/hw/lm4549.c
> index 80b3ec4..e0137d5 100644
> --- a/hw/lm4549.c
> +++ b/hw/lm4549.c
> @@ -224,7 +224,7 @@ uint32_t lm4549_write_samples(lm4549_state *s, uint32_t 
> left, uint32_t right)
>         This model supports 16-bit playback.
>      */
>
> -    if (s->buffer_level >= LM4549_BUFFER_SIZE) {
> +    if (s->buffer_level > LM4549_BUFFER_SIZE - 2) {
>          DPRINTF("write_sample Buffer full\n");
>          return 0;
>      }
> --
> 1.7.10


I don't object to making the change to placate the analyser,
but I don't think this is actually a buffer overrun. We always
add and remove samples from the buffer two at a time, so it's
not possible to get here with s->buffer_level == BUFFER_SIZE-1
(which is the only case where the old and new conditions
give different answers).

-- PMM