unsubscribe

[Mory Kapustianyk]

----- Original Message ----- 
From: "Steve Strobel" <steve at link-comm.com>
To: "RTEMS Users" <rtems-users at rtems.com>
Cc: "Gene Smith" <gds at chartertn.net>; "Ralf Corsepius"
<ralf_corsepius at rtems.org>
Sent: Tuesday, August 24, 2004 10:27 AM
Subject: Re: cygwin configure problem


> At 06:37 AM 8/19/2004 +0200, Ralf Corsepius wrote:
> >[snip]
> >configure: error: invalid variable name: build_alias
> >[snip]
> >Makes me wonder if we might be hitting some "max. processes", "max.
> >files/buffers", "max. stack size" or similar issues on Cygwin/Win.
> >[I am thinking along the lines of processes temporarily not properly
> >deallocating resources - something like "transient zombies"]
>
> That is my suspicion too, though I haven't been able to identify which
limit I am exceeding.  When I was able to build about half the time (getting
"build_alias" the other half), I could reboot, open bash in Cygwin, and
start the configure/make, then not touch the computer until it finished or
failed and get different results each time.
>
> Adding more memory didn't seem to help.  The number of handles used isn't
especially high when it fails.  Sometimes it would succeed with a bunch of
other programs running and heavy processor use.  It reminds me of the stray
pointer bugs I have fought on embedded projects, where adding a character to
a printf can cause a problem to go away for a while, until the stray pointer
hammers something different in a later version.
>
> > Interestingly, I built a test m68k executable of "main() {}" and
> >> tried to run it with NAV enabled and it was allowed with no
> >> NAV messages.
> >built? Do you mean you invoking the compiler worked?
>
> Though it seemed that disabling Symantec anti-virus helped at times, I
never ran enough builds to get good statistics (once I got one successful
build, I always made changes, so it was hard to compare).  Succeeding one
time out of two or one time out of four isn't very conclusive.
>
> >Well, this spawns far less processes and involves far less tools than
> >configuring the source tree :-)
>
> I have never seen the build_alias problem except when building RTEMS
itself.  Our application uses the same tools and builds 100% of the time.
It is also fairly large, but not nearly as deeply nested (in the recursive
build sense).
>
> >Ralf.
>
> ---
> Steve Strobel
> Link Communications, Inc.
> 1035 Cerise Rd
> Billings, MT 59101-7378
> (406) 245-5002 ext 102
> (406) 245-4889 (fax)
> WWW: http://www.link-comm.com
> MailTo:steve at link-comm.com
>
>