Alex Ghitza wrote:
On Sun, 31 Jan 2010 09:47:00 +0000, "Dr. David Kirkby"
<david.kir...@onetel.net> wrote:
The issue reported on sage-support makes me think we should insist that updates
are performed with the same version of gcc as that which was used to build gcc
in the first placed. Basically, two people have found
1) Install Sage with older gcc.
2) Install the latest gcc 4.4.3
3) Try to update Sage, and it fails
4) Build Sage from the start, using gcc 4.4.3 and it works.
Again, that's not what happened (but I guess I didn't explain it
properly). Here's the actual run of events:
Sorry. I should have read the thread more carefully - it is not your fault.
1) Build sage-4.3.1 with gcc-4.4.2 on Arch Linux; Sage runs fine, passes
doctests, etc. Someone else gets the binary and can use it to run Sage
2) Upgrade gcc to gcc-4.4.3
3) Try to run sage, get errors that the original poster listed
So you are saying that simply updating gcc, causes Sage to no longer work? That
seems very naughty of the shipped gcc/libc if that is happening. I could
understand if one removed a new version of gcc, and replaced it with an old
version this might happen.
But I think when shipping binaries, we should include ALL the libraries used by
gcc with Sage. That means:
* libgcc_s
* libgfortran
* libstdc++.
I've said this before, but nobody has really agreed with me. I think there could
be issues if Sage is built with a new gcc, and someone has only an old one.
I'm just in the process of building a SPARC binary package. I decided to use
LZMA compression, which is going to use 2 hours of CPU time on my oldest SPARC
machine. But I have included those libraries.
On Solaris at least, only a very old gcc is shipped, so I suspect the libraries
will be too out of date. If I build Sage with gcc 4.4.1, and someone has only
gcc 3.4.3 (shipped with Solaris), then I think their library is likely to be too
old. So it's better if we ship one with the same version of gcc as used to build
Sage.
4) Rebuild sage-4.3.1 (so the same version) from scratch with gcc-4.4.3,
and everything is fine again (for me, but not for the other person who
couldn't use my binary because of a cpu flag issue)
Fair enough. I mis understood this. But I still think we should not let someone
upgrade Sage with a version different to what they used to build it.
Dave
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org
