On 12/06/17 16:20, [email protected] wrote:
Given what I see, you only get this when you set SAGE_TUNE_GF2X=full. Unsetting SAGE_TUNE_GF2X (which is the default) or putting it to no will work. At least it looks to me like SAGE_TUNE_GF2X is set to full from the log, let me know if this assumption is wrong.I didn't set SAGE_TUNE_GF2X, but it still attempted the tuning even with it unset. However, it built fine when I did export SAGE_TUNE_GF2X=no.
There are 3 levels of tuning for gf2x in the sage spkg: no) no tuning full) full tuning *) default if unset. I thought it was only happening on full. But obviously I am wrong. The fix must have exposed a further bug that was more prevalent than I thought. On my machine I have no problem doing a default tuning or a full tuning with both gcc and clang. And the test suite passes in full in all configurations. But given how widespread this is turning out to be, we would be better reverting https://trac.sagemath.org/ticket/18882 and wait for the new release of gf2x. The changes on git master are big enough that cherry-picking is a difficult option, and the tuning is massively faster. Francois -- You received this message because you are subscribed to the Google Groups "sage-release" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/sage-release. For more options, visit https://groups.google.com/d/optout.
