On Mon, Dec 04, 2023 at 03:02:45PM -0800, Richard Henderson wrote:
> On 12/4/23 12:09, Michal Suchánek wrote:
> > On Mon, Dec 04, 2023 at 02:50:17PM -0500, Stefan Hajnoczi wrote:
> > > On Mon, 4 Dec 2023 at 14:40, Philippe Mathieu-Daudé <phi...@linaro.org>
> > > wrote:
> > > > +void tcg_unregister_thread(void)
> > > > +{
> > > > + unsigned int n;
> > > > +
> > > > + n = qatomic_fetch_dec(&tcg_cur_ctxs);
> > > > + g_free(tcg_ctxs[n]);
> > > > + qatomic_set(&tcg_ctxs[n], NULL);
> > > > +}
> > >
> > > tcg_ctxs[n] may not be our context, so this looks like it could free
> > > another thread's context and lead to undefined behavior.
>
> Correct.
>
> > There is cpu->thread_id so perhaps cpu->thread_ctx could be added as
> > well. That would require a bitmap of used threads contexts rather than a
> > counter, though.
>
> Or don't free the context at all, but re-use it when incrementing and
> tcg_ctxs[n] != null (i.e. plugging in a repacement vcpu). After all, there
> can only be tcg_max_ctxs contexts.
But you would not know which contexts are in use and which aren't without
tracking the association of contexts to CPUs.
Unless there is a cpu array somewhere and you can use the same index for
both to create the association.
Thanks
Michal
