Re: [gentoo-user] Gentoo system crumbling, reports of gcc death

[Kevin O'Gorman]

Thanks for all your effort, Richard.  It turns out to have been operator error
all along.  There's an old addition to my /etc/make.conf that adds the offending
LDFLAGS stuff, with a note that it was inspired by a thread in this list.  Sigh.
If I had continued reading the thread for a few more days back then, I would
have saved myself this trouble, as some wiser folks made it clear why this
was bad.  Odd that only the one package was clobbered by this.

Oh, well...

++ kevin

On 6/14/06, Richard Fish <bigfish@××××××××××.org> wrote:
> On 6/14/06, Kevin O'Gorman <kogorman@×××××.com> wrote:
> > I would guess you're out of ideas then, except in desperation, I tried
> > looking for just part
> > of that: "-W1" and came up with some stuff I hope will further inspire you:
>
> Nope, all of that is normal...
>
> > But before you go breaking a braincell on this, realize that the
> > ebuild became moot yesterday
> > when I was able to create a binary package from a backup image and
> > install it on the running
> > image.  So pursuing this may be wasted effort.  The package is very
> > old after all.
>
> Well let me take this opportunity to apologize for not giving you
> directions on how to work around this problem.  Getting around it
> probably would have been fairly simple...I think an 'LDFLAGS="" emerge
> --oneshot glib-1.2.10-r5' might have done it.
>
> I am still concerned though, since we don't know where those flags
> were coming from, they may return to bite you again.
>
> > kdevelop won't emerge because it winds up needing a *.h file that does
> > not exist, in a *.cpp file that is itself built from a *.ui file.  It
> > strikes me that this one's gonna have a huge learning curve (for me at
> > least).
>
> Sounds like a pretty common problem with MAKEOPTS=-jN (N>1).  What can
> happen with a parallel make is that if one step needs a file that is
> generated by another step, it can start executing before the file is
> complete generated by the other step.  The flow looks something like:
>
> thread-a: generating foo.h and foo.cpp from foo.ui...
> thread-b: waiting for foo.h and foo.cpp to make foo.o...
> thread-a: writing to foo.h and foo.cpp...
> thread-b: compiling foo.o from foo.cpp...
> thread-b: ERROR compiling foo.cpp!
> thread-a: finished writing to foo.h and foo.cpp...
>
> Makefiles have to be specially written to properly avoid the above scenario...
>
> The workaround is to 'MAKEOPTS=-j1 emerge --oneshot kdevelop'
>
> -Richard
> --
> gentoo-user@g.o mailing list
>
>


-- 
Kevin O'Gorman, PhD
-- 
gentoo-user@g.o mailing list