[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [RFC] avoid gc_in_progress in redisplay

From: Eli Zaretskii
Subject: Re: [RFC] avoid gc_in_progress in redisplay
Date: Fri, 19 Oct 2012 12:19:04 +0200

> Date: Fri, 19 Oct 2012 09:27:29 +0400
> From: Dmitry Antipov <address@hidden>
> IIUC if we call to unblock_input _after_ clearing gc_in_progress in 
> Fgarbage_collect,
> we can avoid gc_in_progress == true cases in redisplay routines; thus, some 
> things
> may be simplified. Comments?

I think this is a red herring: mouse event cannot possibly come in
while GC is in progress, since we have switched to SYNC_INPUT.  Any
mouse event that is detected during GC will just set a flag, which
will be handled either as part of QUIT or when Emacs becomes idle.
And I don't believe GC can QUIT, can it?

Like I wrote when we switched all platforms to SYNC_INPUT: there's lot
of code all over the place that still tries to cater to asynchronous
input events (mouse highlight being one of them) that can no longer
happen.  We need to remove that code and delete the corresponding data
structure members which support that code.

But I wouldn't recommend doing that during the feature freeze.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]