Build results: qbe.freebsd
[OK] 425a2ed09c08222e1254d93dba24c7a50e7a08b9
do not account for interferences in phi classes
Before this commit, I tried to make sure that
two interfering temporaries never ended up in
the same phi class.
This was to make sure that their register hints
were not counterproductively stepping on each
other's toes. The idea is fine, but:
* the implementation is clumsy because it
mixes the orthogonal concepts of
(i) interference and (ii) phi classes;
* the hinting process in the register
allocator is hard to understand because
the disjoint-set data structure used for
phi classes is cut in arbitrary places.
After this commit, the phi classes *really* are
phi classes represented with a proper disjoint-set
data structure.
OK: build: gmake
STDOUT
cc main.c
cc util.c
cc parse.c
cc cfg.c
cc mem.c
cc ssa.c
cc alias.c
cc load.c
cc copy.c
cc fold.c
cc live.c
cc spill.c
cc rega.c
cc gas.c
cc amd64/targ.c
cc amd64/sysv.c
cc amd64/isel.c
cc amd64/emit.c
cc arm64/targ.c
cc arm64/abi.c
cc arm64/isel.c
cc arm64/emit.c
ld obj/qbe
STDERR
OK: test: gmake check
STDOUT
tools/test.sh all
abi1.ssa... [ok]
abi2.ssa... [ok]
abi3.ssa... [ok]
abi4.ssa... [ok]
abi5.ssa... [ok]
abi6.ssa... [ok]
align.ssa... [ok]
collatz.ssa... [ok]
cprime.ssa... [ok]
cup.ssa... [ok]
dark.ssa... [ok]
double.ssa... [ok]
echo.ssa... [ok]
eucl.ssa... [ok]
euclc.ssa... [ok]
fixarg.ssa... [ok]
fpcnv.ssa... [ok]
ldbits.ssa... [ok]
ldhoist.ssa... [ok]
loop.ssa... [ok]
mandel.ssa... [ok]
max.ssa... [ok]
philv.ssa... [ok]
prime.ssa... [ok]
puts10.ssa... [ok]
queen.ssa... [ok]
strcmp.ssa... [ok]
strspn.ssa... [ok]
sum.ssa... [ok]
vararg1.ssa... [ok]
vararg2.ssa... [ok]
All is fine!
STDERR
OK: clean: gmake clean
STDOUT
rm -fr obj
STDERR