Sparc Linux (RH 4.2) Server?

[Matt Gumbley]

Gilbert,
	Thanks for your reply... comments embedded..

"Gilbert E. Detillieux" wrote:
> I had tried building vnc under a Red Hat 6.0/sparc system a while back and
> gave up on it.  After fixing some of the compile-time problems, like the one
> in signal.h, I managed to get the code to compile, but it would fail the
> final linking, complaining about a whole bunch of unresolved references.
> Sorry, but I didn't keep a set of patches after giving up on it.

Well, at least I know that it's going to take some time to fix now!
BTW, did you mean '4.2' in the above paragraph?

> >  > I didn't really want to upgrade the machine to RedHat 6.0 yet, but it
> >  > looks like I might have to - has anyone successfully build Xvnc under RH
> >  > 6.0/Sparc?
> 
> I had more sucess there, but I did need to make some changes to get the
> thing to compile, due to a problem in one of the Makefiles generated by
> imake.  After getting things to compile, I tried it out and got some strange
> results.  The cursor didn't display properly, and moving it around left
> garbage on the screen.  The server also kept crashing on me.

I've now upgraded to 6.0 (very nice; just got to try the latest kernel
that supposedly fixes the X crashing bug) - I'll have a go at building
Xvnc too.

> Anyway, Matt, for what it's worth, here's my patch to the vnc 3.3.2r3
> source, to get it to compile under RH6.0/sparc...

Many thanks - all this imake/xmkmf stuff looks rather hairy to me!

> The problem was that imake generated Makefiles with no valid expansion for
> the LinuxMachineDefines macro, which caused a command syntax error on the
> gcc commands.  The first set of additions should have worked for the sparc,
> but didn't for some reason (SparcArchitecture wasn't being defined either).
> The second set of additions are sort of a fallback, in case the macro
> LinuxMachineDefines is still not defined at that point.

I did try botching the makefiles etc. by hand when I realised that
SparcArchitecture probably wasn't being defined, but the compile still
died with the signal.h problem.

> Please let me know if this works for you, and if so, if you get the server
> running reliably.

I will do - thanks again!

Regards,
	Matt
-- 
Matt J. Gumbley,  Software Engineer | Enigma Data Systems Ltd.
Email: mgumbley "at" enigmadata.co.uk    | Chelsea House, 8-14 The Broadway,
Tel: +44 (0)1444 476500 |  Direct:  | Haywards Heath, West Sussex.
Fax: +44 (0)1444 476501 |  476510   | RH16 3AP England.

---------------------------------------------------------------------
The VNC mailing list - see http://www.uk.research.att.com/vnc/intouch.html
---------------------------------------------------------------------