Roman Lebedev 8d487668d0 [CVP] Soften SDiv into a UDiv as long as we know domains of both of the operands. ... Yes, if operands are non-positive this comes at the extra cost of two extra negations. But a. division is already just ridiculously costly, two more subtractions can't hurt much :) and b. we have better/more analyzes/folds for an unsigned division, we could end up narrowing it's bitwidth, converting it to lshr, etc. This is essentially a take two on 0fdcca07ad, which didn't fix the potential regression i was seeing, because ValueTracking's computeKnownBits() doesn't make use of dominating conditions in it's analysis. While i could teach it that, this seems like the more general fix. This big hammer actually does catch said potential regression. Over vanilla test-suite + RawSpeed + darktable (10M IR instrs, 1M IR BB, 1M X86 ASM instrs), this fires/converts 5 more (+2%) SDiv's, the total instruction count at the end of middle-end pipeline is only +6, so out of +10 extra negations, ~half are folded away, and asm instr count is only +1, so practically speaking all extra negations are folded away and are therefore free. Sadly, all these new UDiv's remained, none folded away. But there are two less basic blocks. https://rise4fun.com/Alive/VS6 Name: v0 Pre: C0 >= 0 && C1 >= 0 %r = sdiv i8 C0, C1 => %r = udiv i8 C0, C1 Name: v1 Pre: C0 <= 0 && C1 >= 0 %r = sdiv i8 C0, C1 => %t0 = udiv i8 -C0, C1 %r = sub i8 0, %t0 Name: v2 Pre: C0 >= 0 && C1 <= 0 %r = sdiv i8 C0, C1 => %t0 = udiv i8 C0, -C1 %r = sub i8 0, %t0 Name: v3 Pre: C0 <= 0 && C1 <= 0 %r = sdiv i8 C0, C1 => %r = udiv i8 -C0, -C1		2020-07-18 17:59:56 +03:00
..
2010-09-02-Trunc.ll	…
2010-09-26-MergeConstantRange.ll	…
add.ll	…
alloca.ll	…
and.ll	…
ashr.ll	…
basic.ll	…
conflict.ll	…
crash.ll	…
deopt.ll	…
guards.ll	…
icmp.ll	[LVI][CVP] Handle (x | y) < C style conditions	2020-07-01 20:43:24 +02:00
merge-range-and-undef.ll	[ValueLattice] Distinguish between constant ranges with/without undef.	2020-03-31 12:50:20 +01:00
mul.ll	[CVP] Deduce no-wrap on `mul`	2019-10-21 08:21:44 +00:00
non-null.ll	[CVP] Add another non null test (NFC)	2020-06-20 13:05:42 +02:00
overflow_predicate.ll	…
overflows.ll	[CVP] After proving that @llvm.with.overflow()/@llvm.sat() don't overflow, also try to prove other no-wrap	2019-10-18 19:32:47 +00:00
phi-common-val.ll	Infer alignment of unmarked loads in IR/bitcode parsing.	2020-05-14 13:03:50 -07:00
pointer.ll	…
pr35807.ll	…
profmd.ll	…
range.ll	Infer alignment of unmarked loads in IR/bitcode parsing.	2020-05-14 13:03:50 -07:00
sdiv.ll	[CVP] Soften SDiv into a UDiv as long as we know domains of both of the operands.	2020-07-18 17:59:56 +03:00
select.ll	[JT][CVP] Regenerate test checks, again	2019-11-01 22:27:48 +01:00
sext.ll	…
shl.ll	[CVP] No-wrap deduction for `shl`	2019-10-21 21:31:19 +00:00
srem.ll	…
sub.ll	[ValueLattice] Merging unknown with empty CR is unknown.	2020-04-25 13:43:34 +01:00
udiv.ll	[llvm] Fix broken cases of 'CHECK[^:]*$' in tests	2020-01-28 09:52:59 -07:00
urem.ll	[llvm] Fix broken cases of 'CHECK[^:]*$' in tests	2020-01-28 09:52:59 -07:00