[comp.protocols.tcp-ip.ibmpc] Compiling newer versions of KA9Q with huge memory model

ebarria@magnus.acs.ohio-state.edu (Enrique A Barria) (05/02/91)

This message is empty.

ebarria@magnus.acs.ohio-state.edu (Enrique A Barria) (05/02/91)

I've been using portions of KA9Q recompiled with huge memory model
(v890816 NOS) without problems. I'm trying to upgrade KA9Q, but when
using KA9Q NOS version 910423 I get 'Stack violation' error messages.
Does anyone have any idea on how to solve this problem? Any help will
be greatly appreciated.

The error messages I get are like this:
	Stack violation, process display
	SP = 4ece0450, legal stack range [63f60008, 63f601fc)
Errors are produced either by the display or keyboard processes.


	Enrique Barria

-----------------------------------------------------------------------------
Academic Computing Services,  The Ohio State University, Columbus
Phone: (614) 292-4843 Ext. 4-0353          email: Barria@osu.edu
Fax:   (614) 292-7081                             Barria@ohstmail  (bitnet)

rich@skybridge.SCL.CWRU.Edu (Rich Warwick) (05/03/91)

In article <1991May2.152354.25971@magnus.acs.ohio-state.edu> ebarria@magnus.acs.ohio-state.edu (Enrique A Barria) writes:
>I've been using portions of KA9Q recompiled with huge memory model
>(v890816 NOS) without problems. I'm trying to upgrade KA9Q, but when
>using KA9Q NOS version 910423 I get 'Stack violation' error messages.
>Does anyone have any idea on how to solve this problem? Any help will
>be greatly appreciated.
>
>The error messages I get are like this:
>	Stack violation, process display
>	SP = 4ece0450, legal stack range [63f60008, 63f601fc)
>Errors are produced either by the display or keyboard processes.
>
Well, I've had similar problems with G1EMM v.113016. Enlarging the 
stack space for the processes seems to help a little, but the same problem
keeps recurring. I'd be interested to hear what changes you made to the
code. I #undef-ed all the other transport code except the the ehternet/
packet driver stuff, which caused no problems at all. I added a few things
to the display driver, and KA-BOOM! Stack violation city!! I tried increasing
the stack space for a bunch of the processes (keyboard, display, most of the 
daemons & clients) and that kept it from choking on the display or keyboard
drivers, and the clients seem to work just fine, but start up a server and
BLAMMO! Stack violations...very strange...Anybody out there (Phil Karn?
Kelvin Hill?) have any idea why this happens?
		
						Rich




--
/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
Richard Warwick              Internet:  rich@b62707.student.CWRU.Edu
Case Western Reserve U.	      o o       rich@SCL.CWRU.Edu
Cleveland (YUCK!!!!), OH       -        rmw3@po.CWRU.Edu

ebarria@magnus.acs.ohio-state.edu (Enrique A Barria) (05/06/91)

Thanks to everyone who has e-mailed me regarding this problem. In my
original posting I failed to mention clearly that I experience stack
violation problems without adding any of my own code. I just recompiled
the original distribution from thumper with huge memory model. Since then,
at least one other person has told me about experiencing similar problems 
even working with large memory model.
	All this seems to me like an obscure bug deep inside the KA9Q kernel.
Any comments? (from the authors?).

	Enrique Barria

PS. Please respond through e-mail

-----------------------------------------------------------------------------
Academic Computing Services,  The Ohio State University, Columbus
Phone: (614) 292-4843 Ext. 4-0353          email: Barria@osu.edu
Fax:   (614) 292-7081                             Barria@ohstmail  (bitnet)