Eric Blake wrote:
On 06/01/2010 01:10 PM, Cole Robinson wrote:
> If VM startup fails early enough (can't find a referenced USB device),
> libvirtd will crash trying to clear the VNC port bit, since port = 0,
> which overflows us out of the bitmap bounds.
>
> Fix this by being more defensive in the bitmap operations, and only
> clearing a previously set VNC port.
>
> +++ b/src/util/bitmap.c
> @@ -118,7 +118,7 @@ int virBitmapSetBit(virBitmapPtr bitmap, size_t b)
> */
> int virBitmapClearBit(virBitmapPtr bitmap, size_t b)
> {
> - if (b > bitmap->size - 1)
> + if (bitmap->size != 0 && b > bitmap->size - 1)
>
I think this could use a v2: virBitmapSetBit and virBitmapGetBit should
get the same treatment for bounds checking.
Meanwhile, we already reject attempts to create a bitmap with SIZE_MAX
bits. Therefore, since b is unsigned, we can safely avoid the && and
instead do the computation via a single comparison:
if (bitmap->size <= b)
return -1;
For that matter, should virBitmapAlloc(0) return NULL, instead of it's
current behavior of allocating an (empty) bitmap?
Yes, you are right - especially since there is no grow operation :-). I
should have returned NULL for size 0 request in the original version.
Regards,
Jim