[Bug target/79487] Invalid _Decimal32 comparison on s390x

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79487

--- Comment #18 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Created attachment 40741
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=40741&action=edit
gcc7-pr79487.patch

Untested fix.  Seems at least from eyeballing the #c0 testcase (which is not
executable) this does the right thing.  Haven't tried it on
float-cast-overflow-10.c with the -8.c volatile hack removed though.