Debian-sparc folks, I am looking for help to solve the FTBFS problem that I discussed below with Dirk Eddelbuettel. Thanks in advance,
Rafael ----- Forwarded message from Dirk Eddelbuettel <[EMAIL PROTECTED]> ----- From: Dirk Eddelbuettel <[EMAIL PROTECTED]> Subject: Re: PLplot FTBFS on sparc due to Octave problem Date: Thu, 15 Jan 2004 16:47:55 -0600 To: Rafael Laboissiere <[EMAIL PROTECTED]> Cc: [EMAIL PROTECTED] On Thu, Jan 15, 2004 at 11:37:09PM +0100, Rafael Laboissiere wrote: > Hi Dirk, > > Here we go again. My plplot packages FTBFS on sparc because Octave is > segfaulting. Here are the relevant parts of the buildd log: > (http://buildd.debian.org/fetch.php?&pkg=plplot&ver=5.2.1.cvs.20040115-3&arch=sparc&stamp=1074204496&file=log&as=raw) > > [...] > Setting up octave2.1 (2.1.52-1) ... > [...] > ./configure [...] > [...] > checking for octave... yes > checking Octave version... panic: Bus error -- stopping myself... > attempting to save variables to `octave-core'... > save to `octave-core' complete > > In other architectures, the last line read: > > checking Octave version... 2.1.52 > > The configure script runs something very simple, like this: > > octave -q -f 2>&1 <<EOF > printf(octave_config_info("version")); > EOF > > The last successful buildd of plplot on sparc was done with > octave2.1_2.1.50-3. Are you aware of any problems regarding Octave 2.1.52 > on sparc? No, not really. I think it took it a while to get built there, but then it looks as if the entire buildd daemon system was lagging after the security incident. As for Octave, I don't think it has any new code in which Sparc is treated differently as a platform. Could it be the gcc toolchain that is acting up? Dirk -- Mr. O'Neill comes off as obnoxious and arrogant, but smart. It seems the president never got beyond the first two characteristics. -- David Wessel, Wall Street Journal, 15 Jan 2004 ----- End forwarded message ----- -- Rafael