| ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py |
| ; RUN: llc < %s -mtriple=aarch64 | FileCheck %s |
| |
| ;; ============================================================================ |
| ;; Tests for canCreateUndefOrPoison = false |
| ;; These verify that freeze operations are correctly eliminated |
| ;; ============================================================================ |
| |
| ; TRUNCATE_SSAT_S: No saturation path |
| define i1 @sqxtn_no_sat_with_freeze(<4 x i32> %x) { |
| ; CHECK-LABEL: sqxtn_no_sat_with_freeze: |
| ; CHECK: // %bb.0: |
| ; CHECK-NEXT: movi v1.4s, #100 |
| ; CHECK-NEXT: and v0.16b, v0.16b, v1.16b |
| ; CHECK-NEXT: sqxtn v0.4h, v0.4s |
| ; CHECK-NEXT: umov w8, v0.h[0] |
| ; CHECK-NEXT: and w8, w8, #0xfffc |
| ; CHECK-NEXT: cmp w8, #200 |
| ; CHECK-NEXT: cset w0, gt |
| ; CHECK-NEXT: ret |
| %masked = and <4 x i32> %x, <i32 100, i32 100, i32 100, i32 100> |
| %trunc = call <4 x i16> @llvm.aarch64.neon.sqxtn.v4i16(<4 x i32> %masked) |
| %freeze = freeze <4 x i16> %trunc |
| %extract = extractelement <4 x i16> %freeze, i32 0 |
| ; Input is [0,100], so result > 200 is always false |
| %cmp = icmp sgt i16 %extract, 200 |
| ret i1 %cmp |
| } |
| |
| ; TRUNCATE_SSAT_S: Test specific known bits |
| define i16 @sqxtn_known_bits(<4 x i32> %x) { |
| ; CHECK-LABEL: sqxtn_known_bits: |
| ; CHECK: // %bb.0: |
| ; CHECK-NEXT: mov w0, wzr |
| ; CHECK-NEXT: ret |
| ; Input: [0, 32512] fits in i16 without saturation |
| %masked = and <4 x i32> %x, <i32 32512, i32 32512, i32 32512, i32 32512> |
| %trunc = call <4 x i16> @llvm.aarch64.neon.sqxtn.v4i16(<4 x i32> %masked) |
| %freeze = freeze <4 x i16> %trunc |
| %extract = extractelement <4 x i16> %freeze, i32 0 |
| ; Mask to lower 7 bits - with KnownBits, knows upper bits are already 0 |
| ; so this AND can be simplified |
| %and = and i16 %extract, 127 |
| ret i16 %and |
| } |
| |
| ;; ============================================================================ |
| ;; Tests for computeKnownBits |
| ;; These verify that known bits analysis enables optimizations |
| ;; TRUNCATE_SSAT_S: Signed to Signed Saturating Truncate (sqxtn) |
| ;; ============================================================================ |
| |
| ; Constant 32512 = 0b0111111110000000 has known zero lower bits |
| ; After truncation, ANDing with 127 (lower 7 bits) gives 0 |
| define i16 @sqxtn_known_zero_bits(<4 x i32> %x) { |
| ; CHECK-LABEL: sqxtn_known_zero_bits: |
| ; CHECK: // %bb.0: |
| ; CHECK-NEXT: mov w0, wzr |
| ; CHECK-NEXT: ret |
| %masked = and <4 x i32> %x, <i32 32512, i32 32512, i32 32512, i32 32512> |
| %trunc = call <4 x i16> @llvm.aarch64.neon.sqxtn.v4i16(<4 x i32> %masked) |
| %extract = extractelement <4 x i16> %trunc, i32 0 |
| %and = and i16 %extract, 127 |
| ret i16 %and |
| } |
| |
| ; Input range [0, 256] means result is always non-negative |
| ; Comparison with negative value should fold to false (returns 0) |
| define i1 @sqxtn_known_nonnegative(<4 x i32> %x) { |
| ; CHECK-LABEL: sqxtn_known_nonnegative: |
| ; CHECK: // %bb.0: |
| ; CHECK-NEXT: mov w0, wzr |
| ; CHECK-NEXT: ret |
| %masked = and <4 x i32> %x, <i32 256, i32 256, i32 256, i32 256> |
| %trunc = call <4 x i16> @llvm.aarch64.neon.sqxtn.v4i16(<4 x i32> %masked) |
| %extract = extractelement <4 x i16> %trunc, i32 0 |
| %cmp = icmp slt i16 %extract, 0 |
| ret i1 %cmp |
| } |
| |
| ;; ============================================================================ |
| ;; TRUNCATE_SSAT_U: Signed to Unsigned Saturating Truncate (sqxtun) |
| ;; ============================================================================ |
| |
| define i16 @sqxtun_known_zero_lower_bits(<4 x i32> %x) { |
| ; CHECK-LABEL: sqxtun_known_zero_lower_bits: |
| ; CHECK: // %bb.0: |
| ; CHECK-NEXT: mov w0, wzr |
| ; CHECK-NEXT: ret |
| %masked = and <4 x i32> %x, <i32 32512, i32 32512, i32 32512, i32 32512> |
| %trunc = call <4 x i16> @llvm.aarch64.neon.sqxtun.v4i16(<4 x i32> %masked) |
| %extract = extractelement <4 x i16> %trunc, i32 0 |
| %and = and i16 %extract, 255 |
| ret i16 %and |
| } |
| |
| define i16 @sqxtun_known_zero_upper_bits(<4 x i32> %x) { |
| ; CHECK-LABEL: sqxtun_known_zero_upper_bits: |
| ; CHECK: // %bb.0: |
| ; CHECK-NEXT: mov w0, wzr |
| ; CHECK-NEXT: ret |
| %masked = and <4 x i32> %x, <i32 127, i32 127, i32 127, i32 127> |
| %trunc = call <4 x i16> @llvm.aarch64.neon.sqxtun.v4i16(<4 x i32> %masked) |
| %extract = extractelement <4 x i16> %trunc, i32 0 |
| %shift = lshr i16 %extract, 7 |
| ret i16 %shift |
| } |
| |
| ;; ============================================================================ |
| ;; TRUNCATE_USAT_U: Unsigned to Unsigned Saturating Truncate (uqxtn) |
| ;; ============================================================================ |
| |
| define i8 @uqxtn_known_zero_upper_bits(<8 x i16> %x) { |
| ; CHECK-LABEL: uqxtn_known_zero_upper_bits: |
| ; CHECK: // %bb.0: |
| ; CHECK-NEXT: mov w0, wzr |
| ; CHECK-NEXT: ret |
| %masked = and <8 x i16> %x, <i16 63, i16 63, i16 63, i16 63, i16 63, i16 63, i16 63, i16 63> |
| %trunc = call <8 x i8> @llvm.aarch64.neon.uqxtn.v8i8(<8 x i16> %masked) |
| %extract = extractelement <8 x i8> %trunc, i32 0 |
| %shift = lshr i8 %extract, 6 |
| ret i8 %shift |
| } |
| |
| |
| declare <4 x i16> @llvm.aarch64.neon.sqxtn.v4i16(<4 x i32>) |
| declare <4 x i16> @llvm.aarch64.neon.sqxtun.v4i16(<4 x i32>) |
| declare <8 x i8> @llvm.aarch64.neon.uqxtn.v8i8(<8 x i16>) |