Re: Assert failure on 'list_member_ptr(rel->joininfo, restrictinfo)'

On Wed, Nov 15, 2023 at 8:04 AM Andres Freund <andres@anarazel.de> wrote:
>
> On 2023-11-14 14:42:13 +0200, Alexander Korotkov wrote:
> > It's possibly dumb option, but what about just removing the assert?
>
> That's not at all an option - the in-place bms_* functions can free their
> input. So a dangling pointer to the "old" version is a use-after-free waiting
> to happen - you just need a query that actually gets to bitmapsets that are a
> bit larger.

Yeah, now I got it, thank you.  I was under the wrong impression that
bitmapset has the level of indirection, so the pointer remains valid.
Now, I see that bitmapset manipulation functions can do free/repalloc
making the previous bitmapset pointer invalid.

------
Regards,
Alexander Korotkov