[comp.windows.x] Copy_planes bug in R4

devsop%fornax%netvax@CCC.NMFECC.GOV (Pothier Steve) (03/02/90)

I'm getting bitten by what I think is the copy planes bug in R4.  We
have installed a fix that reportedly fixes the problem, but we are
using R3 clients. Can anyone tell me if the copy_planes bug
is strictly a server phenomena or if it also relates to Xlib &/or
intrinsics.

The fix that I think is pertinant was given in fix-1 as:
	"CopyPlane sometimes mis-fetches source"

When I use our client with the R4 server, I get a complete server
crash (sometimes) when doing a XCopyPlane from a pixmap to pixmap.
Works ok on an R3 server. Doesn't seem to be affected by xset -bc.

I'm getting quite desperate but any work arounds I can think of will
be time consuming and our project is finishing up RSN.

Any help or suggestions will be greatly appreciated.

==============================================================================
Steve Pothier
pothiers%tuva.sainet@nmfecc.arpa
Science Applications International
Tucson, AZ
602-748-7400
==============================================================================

rws@EXPO.LCS.MIT.EDU (Bob Scheifler) (03/02/90)

    Can anyone tell me if the copy_planes bug
    is strictly a server phenomena or if it also relates to Xlib &/or
    intrinsics.

I'd say the server, but I can't really tell what your bug is.

    The fix that I think is pertinant was given in fix-1 as:
	"CopyPlane sometimes mis-fetches source"

As far as I know this didn't cause a server crash, it only caused bogus
pixels to appear on the screen.

    When I use our client with the R4 server, I get a complete server
    crash (sometimes) when doing a XCopyPlane from a pixmap to pixmap.

As usual, we can't diagnose such problems without information.  Please
supply a test program or some way of reliably reproducing the problem.
Please extract a stack trace from the server core file.  Please tell us
what machine, display, OS, and compiler you are using.  Please supply
any other information you think is relevant.

mike@tslvax.UUCP (Michael J. Tobias) (03/06/90)

in article <9003012033.AA13039@fornax.>, devsop%fornax%netvax@CCC.NMFECC.GOV (Pothier Steve) says:
> 
> I'm getting bitten by what I think is the copy planes bug in R4.  We
> have installed a fix that reportedly fixes the problem, but we are
> using R3 clients. Can anyone tell me if the copy_planes bug
> is strictly a server phenomena or if it also relates to Xlib &/or
> intrinsics.
> 
So am I, but I don't have the first fix either.  Can someone send it to
me?

Michael J. Tobias	uunet!ucf-cs!tslvax!mike
Tech-Source
(407) 830-8301