summaryrefslogtreecommitdiff
path: root/llvm/test/CodeGen/AMDGPU/bitcast_vector_bigint.ll
diff options
context:
space:
mode:
Diffstat (limited to 'llvm/test/CodeGen/AMDGPU/bitcast_vector_bigint.ll')
-rw-r--r--llvm/test/CodeGen/AMDGPU/bitcast_vector_bigint.ll351
1 files changed, 351 insertions, 0 deletions
diff --git a/llvm/test/CodeGen/AMDGPU/bitcast_vector_bigint.ll b/llvm/test/CodeGen/AMDGPU/bitcast_vector_bigint.ll
new file mode 100644
index 000000000000..2c6aabec7633
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/bitcast_vector_bigint.ll
@@ -0,0 +1,351 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 < %s | FileCheck -check-prefix=GFX9 %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1200 < %s | FileCheck -check-prefix=GFX12 %s
+
+; Make sure stack use isn't introduced for these bitcasts.
+
+define i160 @bitcast_v5i32_to_i160(<5 x i32> %vec) {
+; GFX9-LABEL: bitcast_v5i32_to_i160:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_v5i32_to_i160:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast <5 x i32> %vec to i160
+ ret i160 %bitcast
+}
+
+define i192 @bitcast_v6i32_to_i192(<6 x i32> %vec) {
+; GFX9-LABEL: bitcast_v6i32_to_i192:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_v6i32_to_i192:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast <6 x i32> %vec to i192
+ ret i192 %bitcast
+}
+
+define i224 @bitcast_v7i32_to_i224(<7 x i32> %vec) {
+; GFX9-LABEL: bitcast_v7i32_to_i224:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_v7i32_to_i224:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast <7 x i32> %vec to i224
+ ret i224 %bitcast
+}
+
+define i256 @bitcast_v8i32_to_i256(<8 x i32> %vec) {
+; GFX9-LABEL: bitcast_v8i32_to_i256:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_v8i32_to_i256:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast <8 x i32> %vec to i256
+ ret i256 %bitcast
+}
+
+define <5 x i32> @bitcast_i160_to_v5i32(i160 %int) {
+; GFX9-LABEL: bitcast_i160_to_v5i32:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_i160_to_v5i32:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast i160 %int to <5 x i32>
+ ret <5 x i32> %bitcast
+}
+
+define <6 x i32> @bitcast_i192_to_v6i32(i192 %int) {
+; GFX9-LABEL: bitcast_i192_to_v6i32:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_i192_to_v6i32:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast i192 %int to <6 x i32>
+ ret <6 x i32> %bitcast
+}
+
+define <7 x i32> @bitcast_i224_to_v7i32(i224 %int) {
+; GFX9-LABEL: bitcast_i224_to_v7i32:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_i224_to_v7i32:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast i224 %int to <7 x i32>
+ ret <7 x i32> %bitcast
+}
+
+define <8 x i32> @bitcast_i256_to_v8i32(i256 %int) {
+; GFX9-LABEL: bitcast_i256_to_v8i32:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_i256_to_v8i32:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast i256 %int to <8 x i32>
+ ret <8 x i32> %bitcast
+}
+
+define i192 @bitcast_v3i64_to_i192(<3 x i64> %vec) {
+; GFX9-LABEL: bitcast_v3i64_to_i192:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_v3i64_to_i192:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast <3 x i64> %vec to i192
+ ret i192 %bitcast
+}
+
+define <3 x i64> @bitcast_i192_to_v3i64(i192 %int) {
+; GFX9-LABEL: bitcast_i192_to_v3i64:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_i192_to_v3i64:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast i192 %int to <3 x i64>
+ ret <3 x i64> %bitcast
+}
+
+define <10 x i16> @bitcast_i160_to_v10i16(i160 %int) {
+; GFX9-LABEL: bitcast_i160_to_v10i16:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_mov_b32 s4, 0xffff
+; GFX9-NEXT: v_bfi_b32 v0, s4, v0, v0
+; GFX9-NEXT: v_bfi_b32 v2, s4, v2, v2
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_i160_to_v10i16:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: v_bfi_b32 v0, 0xffff, v0, v0
+; GFX12-NEXT: v_bfi_b32 v2, 0xffff, v2, v2
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast i160 %int to <10 x i16>
+ ret <10 x i16> %bitcast
+}
+
+define i160 @bitcast_v10i16_to_i160(<10 x i16> %vec) {
+; GFX9-LABEL: bitcast_v10i16_to_i160:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_v10i16_to_i160:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast <10 x i16> %vec to i160
+ ret i160 %bitcast
+}
+
+define i12 @bitcast_v2i6_to_i12(<2 x i6> %vec) {
+; GFX9-LABEL: bitcast_v2i6_to_i12:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: v_lshlrev_b16_e32 v1, 6, v1
+; GFX9-NEXT: v_and_b32_e32 v0, 63, v0
+; GFX9-NEXT: v_or_b32_e32 v0, v0, v1
+; GFX9-NEXT: v_and_b32_e32 v0, 0xfff, v0
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_v2i6_to_i12:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: v_lshlrev_b16 v1, 6, v1
+; GFX12-NEXT: v_and_b32_e32 v0, 63, v0
+; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1)
+; GFX12-NEXT: v_or_b32_e32 v0, v0, v1
+; GFX12-NEXT: v_and_b32_e32 v0, 0xfff, v0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast <2 x i6> %vec to i12
+ ret i12 %bitcast
+}
+
+define <2 x i6> @bitcast_i12_to_v2i6(i12 %int) {
+; GFX9-LABEL: bitcast_i12_to_v2i6:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: v_and_b32_e32 v2, 63, v0
+; GFX9-NEXT: v_lshrrev_b16_e32 v0, 6, v0
+; GFX9-NEXT: v_and_b32_e32 v1, 63, v0
+; GFX9-NEXT: v_mov_b32_e32 v0, v2
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_i12_to_v2i6:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: v_lshrrev_b16 v1, 6, v0
+; GFX12-NEXT: v_and_b32_e32 v0, 63, v0
+; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_2)
+; GFX12-NEXT: v_and_b32_e32 v1, 63, v1
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast i12 %int to <2 x i6>
+ ret <2 x i6> %bitcast
+}
+
+define i160 @bitcast_v5f32_to_i160(<5 x float> %vec) {
+; GFX9-LABEL: bitcast_v5f32_to_i160:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_v5f32_to_i160:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast <5 x float> %vec to i160
+ ret i160 %bitcast
+}
+
+define <5 x float> @bitcast_i160_to_v5f32(i160 %int) {
+; GFX9-LABEL: bitcast_i160_to_v5f32:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_i160_to_v5f32:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast i160 %int to <5 x float>
+ ret <5 x float> %bitcast
+}
+
+define <6 x float> @bitcast_i192_to_v6f32(i192 %int) {
+; GFX9-LABEL: bitcast_i192_to_v6f32:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_i192_to_v6f32:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast i192 %int to <6 x float>
+ ret <6 x float> %bitcast
+}
+
+define i192 @bitcast_v6f32_to_i192(<6 x float> %vec) {
+; GFX9-LABEL: bitcast_v6f32_to_i192:
+; GFX9: ; %bb.0:
+; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX9-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX12-LABEL: bitcast_v6f32_to_i192:
+; GFX12: ; %bb.0:
+; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0
+; GFX12-NEXT: s_wait_expcnt 0x0
+; GFX12-NEXT: s_wait_samplecnt 0x0
+; GFX12-NEXT: s_wait_bvhcnt 0x0
+; GFX12-NEXT: s_wait_kmcnt 0x0
+; GFX12-NEXT: s_setpc_b64 s[30:31]
+ %bitcast = bitcast <6 x float> %vec to i192
+ ret i192 %bitcast
+}
generated by cgit v1.2.3 (git 2.25.1) at 2026-05-31 15:44:44 +0000