projects / libfirm / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 9216d89) | patch
sparc: use 128bit long double even for softfloat
authorMatthias Braun <matze@braunis.de>
Thu, 13 Sep 2012 10:02:52 +0000 (12:02 +0200)
committerMatthias Braun <matze@braunis.de>
Thu, 13 Sep 2012 10:05:31 +0000 (12:05 +0200)
commit7af4a072cd198f7d10c3f9235d09a3922d8bb793
treeb5e9297abb6c8285680bef71e378f9da4b33018ctree | snapshot
parent9216d89f40050976f3a6999ab57198ed00304d0acommit | diff
sparc: use 128bit long double even for softfloat

We can't handle 128bit long double in softfloat yet, but most libraries
are compiled like this. It's better to have the compiler fail than to
produce code with a mismatching ABI in the default setting.
ir/be/sparc/bearch_sparc.c diff | blob | history
libfirm