Fix conversion from float/double to unsigned int
commitd6c9669c3c83ce9f570a9c3528110b666aba88f2
authorMichael Forney <mforney@mforney.org>
Fri, 25 Aug 2023 22:04:47 +0000 (25 15:04 -0700)
committerQuentin Carbonneaux <quentin@c9x.me>
Sat, 26 Aug 2023 13:16:45 +0000 (26 15:16 +0200)
tree53ff27d39d2929d1d52b307697f1d89bbabf61f4
parentd41d91ddd851798349f37d9bb75ed2718bbe28f5
Fix conversion from float/double to unsigned int

signed int can't represent all the values of unsigned int, so we
need to do the conversion to signed long, and use the lower 32 bits
as the result.
amd64/isel.c
test/fpcnv.ssa