[pulseaudio-discuss] [PATCH 1/2] memblock: Detect block ID collisions in pa_memimport_get

[Peter Meerwald]

> > > Make sure we don't hand out the wrong block, in case we have the
> > > same block ID in different SHM segments.
> > 
> > this adds two checks: for block type and that the shm id matches
> > 
> > using %u in pa_log() would save the (int) casts;
> 
> Ok
> 
> > maybe both, shm_id and
> > memory.id could be logged?
> 
> This would require two logs because if b->type != PA_MEMBLOCK_IMPORTED then
> there is no memory.id.

maybe the check b->type == PA_MEMBLOCK_IMPORTED should be an assert?

anyway, I tested and the fix seems to work (no distortion anymore at the 
beginning of record stream)

thanks, p.

> > > Signed-off-by: David Henningsson <david.henningsson at canonical.com>
> > > ---
> > >   src/pulsecore/memblock.c | 5 +++++
> > >   1 file changed, 5 insertions(+)
> > > 
> > > diff --git a/src/pulsecore/memblock.c b/src/pulsecore/memblock.c
> > > index f8b008f..ca24dec 100644
> > > --- a/src/pulsecore/memblock.c
> > > +++ b/src/pulsecore/memblock.c
> > > @@ -1017,6 +1017,11 @@ pa_memblock* pa_memimport_get(pa_memimport *i,
> > > uint32_t block_id, uint32_t shm_i
> > >       pa_mutex_lock(i->mutex);
> > > 
> > >       if ((b = pa_hashmap_get(i->blocks, PA_UINT32_TO_PTR(block_id)))) {
> > > +        if (b->type != PA_MEMBLOCK_IMPORTED ||
> > > b->per_type.imported.segment->memory.id != shm_id) {
> > > +            pa_log("Cannot import memory due to to block ID collision
> > > (block %d, shm id %d)!", (int) block_id, (int) shm_id);
> > > +            b = NULL;
> > > +            goto finish;
> > > +        }
> > >           pa_memblock_ref(b);
> > >           goto finish;
> > >       }
> > > 
> > 
> 
> 

-- 

Peter Meerwald
+43-664-2444418 (mobile)