nanogui: fixed: potentional buggy of 0.88pre11:nanox


Previous by date: 4 Aug 2000 09:46:34 -0000 Re: fixed: potentional buggy of 0.88pre11:nanox - more, Greg Haerr
Next by date: 4 Aug 2000 09:46:34 -0000 Re: caret revision3 part2:edit.c, 曾昭明
Previous in thread: 4 Aug 2000 09:46:34 -0000 Re: fixed: potentional buggy of 0.88pre11:nanox, Morten Rolland
Next in thread: 4 Aug 2000 09:46:34 -0000 Re: fixed: potentional buggy of 0.88pre11:nanox, Morten Rolland

Subject: Re: fixed: potentional buggy of 0.88pre11:nanox
From: "Greg Haerr" ####@####.####
Date: 4 Aug 2000 09:46:34 -0000
Message-Id: <020701bffdd9$a9b4eb40$15320cd0@gregh>

: In the "soon to be released bug-fixes mail" by me some time back, we
: fixed this too, slightly different.  Since we know which GR_EVENT_LIST
: needs to be destroyed completely (the one associated with the client
: to be cleaned up), we did it the simple way and just free it after all
: the other code has run in GsDestroyClientResources.  I can't see how
: this might fail.

I have added this to 0.88, which I plan to release tonight.  In your patch,
I didn't however, remove the "rendundant" calls to GsWpUnmapWindow,
since the window isn't always unmapped by GsWpDestroyWindow
as the developer suggested.

I've also added Kevin's eventclient free() code, since it seems that
both patches free different things: yours frees events associated
with the client, and Kevin's frees eventclient structures associated with
the client.  I hope I got this right.  Perhaps both of you could check
the new nanox/srvnet.c.


: The "client craches => nano-X server craches IF mouse button pressed"
: problem was tracked down as the result of a more deliberate bug-hunt.

Added.


:
: I have no idea how the clipping code error was discovered -- I have not
: talked to the developer about this.  You may want to verify that it is
: a correct and relevant patch.  I'll forward questions regarding the
: clipping code patch to the developer if there are any.

I pulled the mult-rectangle regions code out from the X server
sources, and I just checked this again, and the current X server
sources don't include this bugfix.  Because the region intersect/or/xor/subtract
code is so complicated, I don't want to change this unless I can
see an example of this region bug.


: +       * nanox/srvnet.c  Remove call to GsWpUnmapWindow and
: +         GsDeliverUpdateEvent. GsWpUnmapWindow is called anyway by
: +         GsWpDestroyWindow.

Not patched.


: +       * engine/devrgn.c (REGION_SubtractO): Add a missing test for
: +         having finished.

Not patched.

Everything else was added.

Regards,

Greg



Previous by date: 4 Aug 2000 09:46:34 -0000 Re: fixed: potentional buggy of 0.88pre11:nanox - more, Greg Haerr
Next by date: 4 Aug 2000 09:46:34 -0000 Re: caret revision3 part2:edit.c, 曾昭明
Previous in thread: 4 Aug 2000 09:46:34 -0000 Re: fixed: potentional buggy of 0.88pre11:nanox, Morten Rolland
Next in thread: 4 Aug 2000 09:46:34 -0000 Re: fixed: potentional buggy of 0.88pre11:nanox, Morten Rolland


Powered by ezmlm-browse 0.20.