This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: *** bootstrap fails for 3.4.0 20040113 (experimental) on i586-suse-linux-gnu


On Fri, Jan 16, 2004 at 07:36:42AM -0800, Dan Kegel wrote:
> Robert Schiele wrote:
> >Well, this technical solution would not really solve the problem.  It would
> >still be possible that some vendor specific patch in my binutils, a 
> >hardware
> >bug, or cosmic radiation might cause a failure.
> 
> Then stop using a vendor-supplied binutils -- would a vanilla one really
> be hard to install?   I rather doubt hardware bugs or cosmic radiation
> will be a problem.

Well, I think you didn't get my point.  Sure I could install any binutils
package.  What I wanted to say is that it is absolutely impossible to catch
every case where some automated failure reporting mechanism misreports a
failure because you always miss one case.  And if this is annoying people I
prefer to stop automated reporting to those people and thus to the list.

In fact in the current case it was something strange which was
non-reproducable with the same sources in the same environment.  Thus it must
have been a hardware bug, a race condition or something else (which is what I
call "cosmic radiation" ;-).  So there is no indication that it was related to
a vendor specific patch and thus it most likely wouldn't have helped to
install a vanilla version and people would still have complained.

Robert

-- 
Robert Schiele			Tel.: +49-621-181-2517
Dipl.-Wirtsch.informatiker	mailto:rschiele@uni-mannheim.de

Attachment: pgp00000.pgp
Description: PGP signature


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]