Re: 21.2.39 hangs sometimes
>>>> "Valdis" == Valdis Kletnieks
<Valdis.Kletnieks(a)vt.edu> writes:
Valdis> On Wed, 10 Jan 2001 18:54:57 GMT, Andy Piper <andy(a)xemacs.org>
said:
> We have a few of these reports. I have no idea what is causing
them. Any X
> experts have any idea?
Valdis> I have no idea offhand, but I've seen this under AIX 4.3.3 so it's
probably
Valdis> not a system or compiler issue.
> >0 0xfeb1437c in _poll () from /usr/lib/libc.so.1
> >#1 0xfeace3e4 in select () from /usr/lib/libc.so.1
> >#2 0xfeddd298 in _XWaitForReadable () from /usr/openwin/lib/libX11.so.4
> >#3 0xfeda3514 in _XRead () from /usr/openwin/lib/libX11.so.4
> >#4 0xfeda57dc in _XReply () from /usr/openwin/lib/libX11.so.4
> >#5 0xfedab74c in XAllocColor () from /usr/openwin/lib/libX11.so.4
> >#6 0x230d80 in allocate_nearest_color (display=0x68d000, colormap=52,
> >visual=0x69c700, color_def=0xffbedb68) at objects-x.c:65
> >#7 0x2378e8 in x_generate_shadow_pixels (f=0x68d000,
> >top_shadow=0xffbedc24, bottom_shadow=0xffbedc20, background=13026763,
> >core_background=13026763) at redisplay-x.c:1632
Valdis> Out of curiosity, does anybody know what visual/depth Raymond Toy was
Valdis> running at? I'm suspecting some sort of screw-up relating to a full
colormap,
Valdis> which "should only happen" on an 8-bit display, or if you've
been blessed with
Valdis> one of IBM's GXT255 cards, which does 24-bit with a subtly brokked
colormap.
I've set up my server for a default visual of 24-bit true-color, but
the Sun X server has many, many visuals available. I used to use
whatever default the server decided on but gv and ghostscript always
ended up picking the first visual which was 8-bit. Ick. I now
explicitly tell the server to use a 24-bit visual for gv and gs's
benefit.
One a good note, I haven't seen this happen since my original report,
but I've upgraded to b40 and GNOME in the meantime.
Ray