On Sun, 11 Mar 2012 21:26:42 +0200
Alon Levy <al...@redhat.com> wrote:

> With this change ppm_save is called after rendering, and not before.
> There are two lose ends:
>  hmp: monitor will be active before ppm_save is complete.

The plan is to lock hmp's shell until rendering completes and the file is
saved. Looks ok to me.

>  qmp: return will be emitted before ppm_save is complete.

Let see if I got this right, please correct me with I'm wrong:

 o Before this commit: when screendump returns, there's an out of date
   screendump file available in the FS

 o After this commit: when screendump returns, there's no screendump
   file available yet

There's a behavior change, which is better done via a new command, which
would need to be async and we don't support that today. Also, having an
out of date screendump is not exactly useful anyway.

Honestly, I don't know what's the best thing to do in this case, but I'll be
fine with it if Gerd acks this patch.

Reply via email to