number of regs (e.g. most riscs), many functions won't need to use callee
clobbered registers. Do a speculative check to see if we can get a free
register without processing the fixed list (which has all of these). This
saves a lot of time on machines with lots of callee clobbered regs (e.g.
ppc and itanium, also x86).
This reduces ppc llc compile time from 184s -> 172s on kc++. This is probably
worth FAR FAR more on itanium though.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22972 91177308-0d34-0410-b5e6-96231b3b80d8
we spill out of the fast path. The scan of active_ and the calls to
updateSpillWeights don't need to happen unless a spill occurs. This reduces
debug llc time of kc++ with ppc from 187.3s to 183.2s.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22971 91177308-0d34-0410-b5e6-96231b3b80d8
simplifies BRTWOWAY into BR if one of the results is a fall-through.
Unless I'm missing something, there is no reason to duplicate this
in the target-specific code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22952 91177308-0d34-0410-b5e6-96231b3b80d8
the old condition to a one bit value. The incoming value must have been
promoted, and the top bits are undefined. This causes us to generate:
_test:
rlwinm r2, r3, 0, 31, 31
li r3, 17
cmpwi cr0, r2, 0
bne .LBB_test_2 ;
.LBB_test_1: ;
li r3, 1
.LBB_test_2: ;
blr
instead of:
_test:
rlwinm r2, r3, 0, 31, 31
li r2, 17
cmpwi cr0, r3, 0
bne .LBB_test_2 ;
.LBB_test_1: ;
li r2, 1
.LBB_test_2: ;
or r3, r2, r2
blr
for:
int %test(bool %c) {
%retval = select bool %c, int 17, int 1
ret int %retval
}
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22947 91177308-0d34-0410-b5e6-96231b3b80d8
it in the block. This codegens:
int %test(bool %c) {
%retval = select bool %c, int 17, int 1
ret int %retval
}
as:
_test:
rlwinm r2, r3, 0, 31, 31
li r2, 17
cmpwi cr0, r3, 0
bne .LBB_test_2 ;
.LBB_test_1: ;
li r2, 1
.LBB_test_2: ;
or r3, r2, r2
blr
instead of:
_test:
rlwinm r2, r3, 0, 31, 31
li r2, 17
li r4, 1
cmpwi cr0, r3, 0
bne .LBB_test_2 ;
.LBB_test_1: ;
or r2, r4, r4
.LBB_test_2: ;
or r3, r2, r2
blr
... which is one fewer instruction. The savings are more significant for
global address and constantfp nodes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22946 91177308-0d34-0410-b5e6-96231b3b80d8
This gets us this for the previous testcase:
_test:
lis r2, 0
ori r3, r2, 65535
blr
Note that we actually write to r3 (the return reg) correctly now :)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22933 91177308-0d34-0410-b5e6-96231b3b80d8
temporary registers for things that define a register. This allows dag->dag
isel to compile this:
int %test() { ret int 65535 }
into:
_test:
lis r2, 0
ori r2, r2, 65535
blr
Next up, getting CopyFromReg to work, allowing arguments and cross-bb values.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22932 91177308-0d34-0410-b5e6-96231b3b80d8
namespace instead of the reg class namespace. Update getRegClassForType()
to use modified names due to tblgen change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22923 91177308-0d34-0410-b5e6-96231b3b80d8
(which died about a week ago) so we're back to load-(2^n-1)-then-AND
sequences. slow, but things should now be Almost Completely Working,
modulo those pesky alignment/ABI issues.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22904 91177308-0d34-0410-b5e6-96231b3b80d8