[llvm] [KnownBits] Remove `hasConflict()` assertions (PR #94568)
via llvm-commits
llvm-commits at lists.llvm.org
Fri Jun 7 04:57:17 PDT 2024
================
@@ -403,8 +391,7 @@ KnownBits KnownBits::lshr(const KnownBits &LHS, const KnownBits &RHS,
if (Exact) {
unsigned FirstOne = LHS.countMaxTrailingZeros();
if (FirstOne < MinShiftAmount) {
- // Always poison. Return zero because we don't like returning conflict.
- Known.setAllZero();
+ // Always poison.
----------------
c8ef wrote:
It appears that in the current version(
cfe2878), the testcase `llvm/test/Analysis/ScalarEvolution/pr76234.ll` is still experiencing a regression from `[0, 1)` to `full-set`. Is this the intended behavior?
```
_bk;t=1717760175285# | <stdin>:4:2: note: possible intended match here
_bk;t=1717760175285# | --> %B9 U: full-set S: full-set
_bk;t=1717760175285# | ^
_bk;t=1717760175285# |
_bk;t=1717760175285# | Input file: <stdin>
_bk;t=1717760175285# | Check file: C:\ws\src\llvm\test\Analysis\ScalarEvolution\pr76234.ll
_bk;t=1717760175285# |
_bk;t=1717760175285# | -dump-input=help explains the following input dump.
_bk;t=1717760175285# |
_bk;t=1717760175285# | Input was:
_bk;t=1717760175285# | <<<<<<
_bk;t=1717760175285# | 1: Printing analysis 'Scalar Evolution Analysis' for function 'PR76234':
_bk;t=1717760175285# | 2: Classifying expressions for: @PR76234
_bk;t=1717760175285# | 3: %B9 = shl i896 0, -264147265567832623176169892458258303259423663018060761063980354513336951278362429737208627943828593947337197496628564339441173779751342768625269489231469788454193341999502542084365758838213220526512116454105594202074014146375780869419198449383518238244769290448868999168
_bk;t=1717760175285# | next:9'0 X error: no match found
_bk;t=1717760175285# | 4: --> %B9 U: full-set S: full-set
_bk;t=1717760175285# | next:9'0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
_bk;t=1717760175285# | next:9'1 ? possible intended match
_bk;t=1717760175285# | 5: %B39 = ashr i896 %B9, 1
_bk;t=1717760175285# | next:9'0 ~~~~~~~~~~~~~~~~~~~~~~~~~
```
https://github.com/llvm/llvm-project/pull/94568
More information about the llvm-commits
mailing list