On Thu, Apr 09, 2026 at 04:26:22PM +0200, Christopher Faulet wrote:
> Le 09/04/2026 à 1:00 PM, Greg Kroah-Hartman a écrit :
> > But did you miss the check for htx_replace_blk_value() that this patch
> > had in comp_http_payload(), pasted here in a whitespace damaged way:
> >
> > @@ -330,6 +338,8 @@ comp_http_payload(struct stream *s, struct filter
> > *filter, struct http_msg *msg,
> > next = htx_remove_blk(htx, blk);
> > else {
> > blk = htx_replace_blk_value(htx,
> > blk, v, ist2(b_head(&trash), b_data(&trash)));
> > + if (!blk)
> > + goto error;
> > next = htx_get_next_blk(htx, blk);
> > }
>
> At first glance, I would say the replace cannot fail. However, it is
> probably safer to add the test. At least to prevent any bug if the API
> change. Do you want to submit a new patch ?
Sure, will do so in my next round of fixes, thanks!
greg k-h
