[llvm] [LV] Avoid collapsing VF to zero when a loop has no memory ops or reductions (PR #216266)

[Pengcheng Wang via llvm-commits]

================
@@ -0,0 +1,53 @@
+; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py UTC_ARGS: --filter "The Smallest and Widest types" --filter "Selecting VF" --filter "Vectorization is possible but not beneficial" --version 6
+; REQUIRES: asserts
+; RUN: opt -passes=loop-vectorize -mtriple riscv64 -mattr=+v \
+; RUN:   -debug-only=loop-vectorize -disable-output -S < %s 2>&1 | \
+; RUN:   FileCheck %s --check-prefix=DEFAULT
+; RUN: opt -passes=loop-vectorize -mtriple riscv64 -mattr=+v \
+; RUN:   -vectorizer-maximize-bandwidth \
+; RUN:   -debug-only=loop-vectorize -disable-output -S < %s 2>&1 | \
+; RUN:   FileCheck %s --check-prefix=MAXBW
+
+; This is an early-exit search loop with no loads/stores and no reductions, so
+; getSmallestAndWidestTypes() cannot observe any element type. The smallest type
+; is reported as equal to the widest type instead of the -1U sentinel, so the
+; max-bandwidth VF computation does not divide the register width by the
+; sentinel and collapse the VF to zero. With and without
+; -vectorizer-maximize-bandwidth we pick the same VF.
+
+define i32 @find_first_ge(i32 %n, i32 %target) {
+; DEFAULT-LABEL: 'find_first_ge'
+; DEFAULT:  LV: The Smallest and Widest types: 8 / 8 bits.
+; DEFAULT:  LV: Selecting VF: vscale x 16.
+;
+; MAXBW-LABEL: 'find_first_ge'
+; MAXBW:  LV: The Smallest and Widest types: 8 / 8 bits.
+; MAXBW:  LV: Selecting VF: vscale x 16.
+;
----------------
wangpc-pp wrote:

In the fix commit the two RUN lines now share a single `CHECK` prefix, since after the fix both configs produce the same output. In the precommit-test commit they still differ (default vectorizes, `-vectorizer-maximize-bandwidth` bails out), so `update_analyze_test_checks.py` reports conflicting output for a shared prefix there and the test-only commit keeps `DEFAULT`/`MAXBW` to capture that difference. The final state has a single prefix.

https://github.com/llvm/llvm-project/pull/216266