Hi,
On Fri, Jun 24, 2016 at 02:30:09PM -0500, Jonathon Jongsma wrote:
> On Thu, 2016-06-23 at 19:37 +0200, Victor Toso wrote:
> > file_xfer_flush_finish and file_xfer_data_flushed_cb are channel-main
> > function and should not check for SpiceFileTransferTask internal
> > errors.
>
> OK, but I'd like some additional justification of why we can skip this
> without causing problems. For example (as I understand it):
> The data will only be flushed if spice_file_transfer_task_read_async()
> was successful, and between the call to file_xfer_flush_async() and
> file_xfer_data_flushed_cb(), the task status will not change
Yes. Any pending change should affect SpiceFileTransferTask which then
will affect FileTransferOperation.
> (or is it possible that the task status can be changed by recieving
> a VD_AGENT_FILE_XFER_STATUS_ERROR message from the guest?)
In that case, we would call the task_complete() but I'm not sure if we
should cancel the flushing... Maybe it is easier to let agent ignore
the data if the cancel/error happens during the flush.
I'll try to improve the commit log!
>
> > ---
> > src/channel-main.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/src/channel-main.c b/src/channel-main.c
> > index be5a454..fef72cd 100644
> > --- a/src/channel-main.c
> > +++ b/src/channel-main.c
> > @@ -1893,7 +1893,7 @@ static void file_xfer_data_flushed_cb(GObject *source_object,
> > GError *error = NULL;
> >
> > file_xfer_flush_finish(channel, res, &error);
> > - if (error || self->error) {
> > + if (error) {
> > spice_file_transfer_task_completed(self, error);
> > return;
> > }
>
> Reviewed-by: Jonathon Jongsma <jjongsma@xxxxxxxxxx>
Thanks,
toso
>
> _______________________________________________
> Spice-devel mailing list
> Spice-devel@xxxxxxxxxxxxxxxxxxxxx
> https://lists.freedesktop.org/mailman/listinfo/spice-devel
_______________________________________________
Spice-devel mailing list
Spice-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/spice-devel