Hi Rusty,
> The free-outside-interrupt issue is usually dealt with by offloading to
> a wq, but your variant works (and isn't too ugly).
Ok, thanks.
>> +static void reclaim_dma_bufs(void)
>> +{
>> + unsigned long flags;
>> + struct port_buffer *buf, *tmp;
>> + LIST_HEAD(tmp_list);
>> +
>> + if (list_empty(&pending_free_dma_bufs))
>> + return;
...
> Looks like this should be an easy noop even if !is_rproc_serial.
Ok, I'll drop the superfluous check before calling reclaim_dma_bufs().
>> @@ -1415,7 +1524,16 @@ static void remove_port_data(struct port *port)
>>
>> /* Remove buffers we queued up for the Host to send us data in. */
>> while ((buf = virtqueue_detach_unused_buf(port->in_vq)))
>> - free_buf(buf);
>> + free_buf(buf, true);
>> +
>> + /*
>> + * Remove buffers from out queue for rproc-serial. We cannot afford
>> + * to leak any DMA mem, so reclaim this memory even if this might be
>> + * racy for the remote processor.
>> + */
>> + if (is_rproc_serial(port->portdev->vdev))
>> + while ((buf = virtqueue_detach_unused_buf(port->out_vq)))
>> + free_buf(buf, true);
>> }
>
> This seems wrong; either this is needed even if !is_rproc_serial(), or
> it's not necessary as the out_vq is empty.
>
> Every path I can see has the device reset (in which case the queues
> should not be active), or we got a VIRTIO_CONSOLE_PORT_REMOVE event (in
> which case, the same).
>
> I think we can have non-blocking writes which could leave buffers in
> out_vq: Amit?
Hm, the remote device could potentially freeze whenever. So I think we
should handle the situation where buffer are stuck in the out-queue for
the rproc_serial device. But I'll move this piece of code into a new
follow-up patch so we can handle this issue separately.
Thanks,
Sjur
_______________________________________________
Virtualization mailing list
Virtualization@xxxxxxxxxxxxxxxxxxxxxxxxxx
https://lists.linuxfoundation.org/mailman/listinfo/virtualization
