gcc-4.0.3 pass 1

Ken Moffat ken at linuxfromscratch.org
Mon Jan 1 17:11:27 PST 2007


On Mon, Jan 01, 2007 at 06:56:56PM -0500, Steve Schwarz & Christina Stone wrote:
> 
> up until the first pass of gcc-4.0.3 (following the book religeously), I 

 It isn't a religion.  Learning how a linux system fits together is
far more important than that ;-)  But seriously, there is a strong
correlation between people who claim to be following the book
religiously, and those who turn out to have missed things.  You
haven't got very far yet, so there isn't a lot of scope for missing
things :-)

> have had smooth sailing. My first attempt at installing gcc failed with 
> stage 2 and stage 3 not matching.
> 
> I didn't give up, I went back into the install using:
> 
> $make clean
> 
> and then again running :
> 
> $../gcc-4.0.3/configure --prefix=/tools \
> --with-local-prefix=/tools --disable-nls --enable-shared \
> --enable-languages=c
> 
> the second time it compiled with no errors.
> 
> My question(for future reference during the build process) is why did this 
> happen? Is it a product of running from the Live CD, or just a fluke? or is 
> ther any other possiblities I should watch for?
> 
 Long ago, when most people were using pentium-II-or-older machines,
this used to happen fairly often.  ISTR that you can run 'make
bootstrap3' or even 'make bootstrap4' to add extra passes of gcc
before it compares the last two builds.

 Possible causes are flakey hardware (unreliable RAM, poor
cooling/blocked fan, bad power supply).  I was going to add "or an
old/problematic host system" but the Live CD should solve that.  The
one issue with Live CDs is that everything runs out of RAM, so if
the memory is at all problematic, you're more likely to hit
problems.  Memtest86 (or Memtest86+ for recent boards, particularly
those with x86_64 processors) are the best tools for checking your
RAM, but a meaningful test can take several hours.
 
ĸen
-- 
das eine Mal als Tragödie, das andere Mal als Farce



More information about the lfs-support mailing list