[RISCV][VLOPT] Add vector narrowing integer right shift instructions to isSupportedInstr #119602
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…to isSupportedInstr
|
@llvm/pr-subscribers-backend-risc-v Author: Michael Maitland (michaelmaitland) ChangesFull diff: https://github.com/llvm/llvm-project/pull/119602.diff 2 Files Affected:
diff --git a/llvm/lib/Target/RISCV/RISCVVLOptimizer.cpp b/llvm/lib/Target/RISCV/RISCVVLOptimizer.cpp
index 52893728853e39..0df4c451894bec 100644
--- a/llvm/lib/Target/RISCV/RISCVVLOptimizer.cpp
+++ b/llvm/lib/Target/RISCV/RISCVVLOptimizer.cpp
@@ -563,8 +563,12 @@ static bool isSupportedInstr(const MachineInstr &MI) {
// Vector Integer Add-with-Carry / Subtract-with-Borrow Instructions
// FIXME: Add support
// Vector Narrowing Integer Right Shift Instructions
- // FIXME: Add support
+ case RISCV::VNSRL_WX:
case RISCV::VNSRL_WI:
+ case RISCV::VNSRL_WV:
+ case RISCV::VNSRA_WI:
+ case RISCV::VNSRA_WV:
+ case RISCV::VNSRA_WX:
// Vector Integer Compare Instructions
// FIXME: Add support
// Vector Integer Min/Max Instructions
diff --git a/llvm/test/CodeGen/RISCV/rvv/vl-opt-instrs.ll b/llvm/test/CodeGen/RISCV/rvv/vl-opt-instrs.ll
index 0215e6a80d09a4..e13482d23a26f4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vl-opt-instrs.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vl-opt-instrs.ll
@@ -982,6 +982,107 @@ define <vscale x 4 x i16> @vnsrl_wi(<vscale x 4 x i32> %a, <vscale x 4 x i16> %b
ret <vscale x 4 x i16> %2
}
+define <vscale x 4 x i16> @vnsrl_wx(<vscale x 4 x i32> %a, <vscale x 4 x i16> %b, iXLen %c, iXLen %vl) {
+; NOVLOPT-LABEL: vnsrl_wx:
+; NOVLOPT: # %bb.0:
+; NOVLOPT-NEXT: vsetvli a2, zero, e16, m1, ta, ma
+; NOVLOPT-NEXT: vnsrl.wx v11, v8, a0
+; NOVLOPT-NEXT: vsetvli zero, a1, e16, m1, ta, ma
+; NOVLOPT-NEXT: vadd.vv v8, v11, v10
+; NOVLOPT-NEXT: ret
+;
+; VLOPT-LABEL: vnsrl_wx:
+; VLOPT: # %bb.0:
+; VLOPT-NEXT: vsetvli zero, a1, e16, m1, ta, ma
+; VLOPT-NEXT: vnsrl.wx v11, v8, a0
+; VLOPT-NEXT: vadd.vv v8, v11, v10
+; VLOPT-NEXT: ret
+ %1 = call <vscale x 4 x i16> @llvm.riscv.vnsrl.nxv4i16.nxv4i32(<vscale x 4 x i16> poison, <vscale x 4 x i32> %a, iXLen %c, iXLen -1)
+ %2 = call <vscale x 4 x i16> @llvm.riscv.vadd.nxv4i16.nxv4i16(<vscale x 4 x i16> poison, <vscale x 4 x i16> %1, <vscale x 4 x i16> %b, iXLen %vl)
+ ret <vscale x 4 x i16> %2
+}
+
+define <vscale x 4 x i16> @vnsrl_wv(<vscale x 4 x i32> %a, <vscale x 4 x i16> %b, <vscale x 4 x i16> %c, iXLen %vl) {
+; NOVLOPT-LABEL: vnsrl_wv:
+; NOVLOPT: # %bb.0:
+; NOVLOPT-NEXT: vsetvli a1, zero, e16, m1, ta, ma
+; NOVLOPT-NEXT: vnsrl.wv v12, v8, v11
+; NOVLOPT-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; NOVLOPT-NEXT: vadd.vv v8, v12, v10
+; NOVLOPT-NEXT: ret
+;
+; VLOPT-LABEL: vnsrl_wv:
+; VLOPT: # %bb.0:
+; VLOPT-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; VLOPT-NEXT: vnsrl.wv v12, v8, v11
+; VLOPT-NEXT: vadd.vv v8, v12, v10
+; VLOPT-NEXT: ret
+ %1 = call <vscale x 4 x i16> @llvm.riscv.vnsrl.nxv4i16.nxv4i32.nxv4i16(<vscale x 4 x i16> poison, <vscale x 4 x i32> %a, <vscale x 4 x i16> %c, iXLen -1)
+ %2 = call <vscale x 4 x i16> @llvm.riscv.vadd.nxv4i16.nxv4i16(<vscale x 4 x i16> poison, <vscale x 4 x i16> %1, <vscale x 4 x i16> %b, iXLen %vl)
+ ret <vscale x 4 x i16> %2
+}
+
+define <vscale x 4 x i16> @vnsra_wi(<vscale x 4 x i32> %a, <vscale x 4 x i16> %b, iXLen %vl) {
+; NOVLOPT-LABEL: vnsra_wi:
+; NOVLOPT: # %bb.0:
+; NOVLOPT-NEXT: vsetvli a1, zero, e16, m1, ta, ma
+; NOVLOPT-NEXT: vnsra.wi v11, v8, 5
+; NOVLOPT-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; NOVLOPT-NEXT: vadd.vv v8, v11, v10
+; NOVLOPT-NEXT: ret
+;
+; VLOPT-LABEL: vnsra_wi:
+; VLOPT: # %bb.0:
+; VLOPT-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; VLOPT-NEXT: vnsra.wi v11, v8, 5
+; VLOPT-NEXT: vadd.vv v8, v11, v10
+; VLOPT-NEXT: ret
+ %1 = call <vscale x 4 x i16> @llvm.riscv.vnsra.nxv4i16.nxv4i32(<vscale x 4 x i16> poison, <vscale x 4 x i32> %a, iXLen 5, iXLen -1)
+ %2 = call <vscale x 4 x i16> @llvm.riscv.vadd.nxv4i16.nxv4i16(<vscale x 4 x i16> poison, <vscale x 4 x i16> %1, <vscale x 4 x i16> %b, iXLen %vl)
+ ret <vscale x 4 x i16> %2
+}
+
+define <vscale x 4 x i16> @vnsra_wx(<vscale x 4 x i32> %a, <vscale x 4 x i16> %b, iXLen %c, iXLen %vl) {
+; NOVLOPT-LABEL: vnsra_wx:
+; NOVLOPT: # %bb.0:
+; NOVLOPT-NEXT: vsetvli a2, zero, e16, m1, ta, ma
+; NOVLOPT-NEXT: vnsra.wx v11, v8, a0
+; NOVLOPT-NEXT: vsetvli zero, a1, e16, m1, ta, ma
+; NOVLOPT-NEXT: vadd.vv v8, v11, v10
+; NOVLOPT-NEXT: ret
+;
+; VLOPT-LABEL: vnsra_wx:
+; VLOPT: # %bb.0:
+; VLOPT-NEXT: vsetvli zero, a1, e16, m1, ta, ma
+; VLOPT-NEXT: vnsra.wx v11, v8, a0
+; VLOPT-NEXT: vadd.vv v8, v11, v10
+; VLOPT-NEXT: ret
+ %1 = call <vscale x 4 x i16> @llvm.riscv.vnsra.nxv4i16.nxv4i32(<vscale x 4 x i16> poison, <vscale x 4 x i32> %a, iXLen %c, iXLen -1)
+ %2 = call <vscale x 4 x i16> @llvm.riscv.vadd.nxv4i16.nxv4i16(<vscale x 4 x i16> poison, <vscale x 4 x i16> %1, <vscale x 4 x i16> %b, iXLen %vl)
+ ret <vscale x 4 x i16> %2
+}
+
+define <vscale x 4 x i16> @vnsra_wv(<vscale x 4 x i32> %a, <vscale x 4 x i16> %b, <vscale x 4 x i16> %c, iXLen %vl) {
+; NOVLOPT-LABEL: vnsra_wv:
+; NOVLOPT: # %bb.0:
+; NOVLOPT-NEXT: vsetvli a1, zero, e16, m1, ta, ma
+; NOVLOPT-NEXT: vnsra.wv v12, v8, v11
+; NOVLOPT-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; NOVLOPT-NEXT: vadd.vv v8, v12, v10
+; NOVLOPT-NEXT: ret
+;
+; VLOPT-LABEL: vnsra_wv:
+; VLOPT: # %bb.0:
+; VLOPT-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; VLOPT-NEXT: vnsra.wv v12, v8, v11
+; VLOPT-NEXT: vadd.vv v8, v12, v10
+; VLOPT-NEXT: ret
+ %1 = call <vscale x 4 x i16> @llvm.riscv.vnsra.nxv4i16.nxv4i32.nxv4i16(<vscale x 4 x i16> poison, <vscale x 4 x i32> %a, <vscale x 4 x i16> %c, iXLen -1)
+ %2 = call <vscale x 4 x i16> @llvm.riscv.vadd.nxv4i16.nxv4i16(<vscale x 4 x i16> poison, <vscale x 4 x i16> %1, <vscale x 4 x i16> %b, iXLen %vl)
+ ret <vscale x 4 x i16> %2
+}
+
+
define <vscale x 4 x i32> @vminu_vv(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b, iXLen %vl) {
; NOVLOPT-LABEL: vminu_vv:
; NOVLOPT: # %bb.0:
|
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/185/builds/10029 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/137/builds/10153 Here is the relevant piece of the build log for the reference |
michaelmaitland
added a commit
that referenced
this pull request
Dec 11, 2024
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/33/builds/8105 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/60/builds/14959 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/16/builds/10514 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/56/builds/14282 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/153/builds/17277 Here is the relevant piece of the build log for the reference |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.