On 03/19/13 15:45, Laszlo Ersek wrote:
> On 03/19/13 15:04, Gerd Hoffmann wrote:
>> chardev flow control broke monitor, fix it by adding watch support.
>>
>> Signed-off-by: Gerd Hoffmann <kra...@redhat.com>
>> ---
>> v2: fix tyops
>
> Well played, Sir :)
>
>> ---
>> monitor.c | 23 +++++++++++++++++++++--
>> 1 file changed, 21 insertions(+), 2 deletions(-)
>>
>> diff --git a/monitor.c b/monitor.c
>> index 112e920..74807f9 100644
>> --- a/monitor.c
>> +++ b/monitor.c
>> @@ -261,11 +261,30 @@ int monitor_read_password(Monitor *mon, ReadLineFunc
>> *readline_func,
>> }
>> }
>>
>> +static gboolean monitor_unblocked(GIOChannel *chan, GIOCondition cond,
>> + void *opaque)
>> +{
>> + monitor_flush(opaque);
>> + return FALSE;
>> +}
>> +
>> void monitor_flush(Monitor *mon)
>> {
>> + int rc;
>> +
>> if (mon && mon->outbuf_index != 0 && !mon->mux_out) {
>> - qemu_chr_fe_write(mon->chr, mon->outbuf, mon->outbuf_index);
>> - mon->outbuf_index = 0;
>> + rc = qemu_chr_fe_write(mon->chr, mon->outbuf, mon->outbuf_index);
>> + if (rc == mon->outbuf_index) {
>> + /* all flushed */
>> + mon->outbuf_index = 0;
>> + return;
>> + }
>> + if (rc > 0) {
>> + /* partial write */
>> + memmove(mon->outbuf, mon->outbuf + rc, mon->outbuf_index - rc);
>> + mon->outbuf_index -= rc;
>> + }
>> + qemu_chr_fe_add_watch(mon->chr, G_IO_OUT, monitor_unblocked, mon);
>> }
>> }
>>
>>
>
> Looks good to me.
>
> Should we maybe forego (re)installing the (new) watch when rc<0?
> (Especially because on a persistent error the fd might immediately
> report writability.)
Yes, it is probably wise to check for errno == EAGAIN (assuming
qemu_chr_fe_write makes sure errno it set to something sensible on error).
cheers,
Gerd
