-
Notifications
You must be signed in to change notification settings - Fork 14.3k
[RISCV][GISel] Legalize G_CTPOP. #72005
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
The base ISA does not have an instruction for this so we need to lower. Zbb support will come in a future patch.
@llvm/pr-subscribers-llvm-globalisel Author: Craig Topper (topperc) ChangesThe base ISA does not have an instruction for this so we need to lower. Zbb support will come in a future patch. Full diff: https://github.com/llvm/llvm-project/pull/72005.diff 3 Files Affected:
diff --git a/llvm/lib/Target/RISCV/GISel/RISCVLegalizerInfo.cpp b/llvm/lib/Target/RISCV/GISel/RISCVLegalizerInfo.cpp
index 2baaadd8ae00af0..386b05e95649b5f 100644
--- a/llvm/lib/Target/RISCV/GISel/RISCVLegalizerInfo.cpp
+++ b/llvm/lib/Target/RISCV/GISel/RISCVLegalizerInfo.cpp
@@ -80,6 +80,11 @@ RISCVLegalizerInfo::RISCVLegalizerInfo(const RISCVSubtarget &ST) {
.maxScalar(0, sXLen)
.lower();
+ getActionDefinitionsBuilder(G_CTPOP)
+ .maxScalar(0, sXLen)
+ .scalarSameSizeAs(1, 0)
+ .lower();
+
getActionDefinitionsBuilder({G_CONSTANT, G_IMPLICIT_DEF})
.legalFor({s32, sXLen, p0})
.widenScalarToNextPow2(0)
diff --git a/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv32/legalize-ctpop.mir b/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv32/legalize-ctpop.mir
new file mode 100644
index 000000000000000..0ea924745d40944
--- /dev/null
+++ b/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv32/legalize-ctpop.mir
@@ -0,0 +1,198 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mattr=+m -mtriple=riscv32 -run-pass=legalizer %s -o - \
+# RUN: | FileCheck %s
+
+---
+name: ctpop_i8
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i8
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY]], [[C1]]
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[AND]], [[C]](s32)
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 85
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C2]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[COPY]], [[AND1]]
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C4]]
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[AND2]], [[C3]](s32)
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 51
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C5]]
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C5]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND3]], [[AND4]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C6]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 15
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C7]]
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND5]], [[C8]]
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C9]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[LSHR3]](s32)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s32) = COPY $x10
+ %0:_(s8) = G_TRUNC %1(s32)
+ %2:_(s8) = G_CTPOP %0(s8)
+ %3:_(s32) = G_ANYEXT %2(s8)
+ $x10 = COPY %3(s32)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i16
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i16
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY]], [[C1]]
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[AND]], [[C]](s32)
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 21845
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C2]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[COPY]], [[AND1]]
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C4]]
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[AND2]], [[C3]](s32)
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 13107
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C5]]
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C5]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND3]], [[AND4]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C6]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 3855
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C7]]
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 257
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND5]], [[C8]]
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 8
+ ; CHECK-NEXT: [[C10:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND6:%[0-9]+]]:_(s32) = G_AND [[MUL]], [[C10]]
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[AND6]], [[C9]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[LSHR3]](s32)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s32) = COPY $x10
+ %0:_(s16) = G_TRUNC %1(s32)
+ %2:_(s16) = G_CTPOP %0(s16)
+ %3:_(s32) = G_ANYEXT %2(s16)
+ $x10 = COPY %3(s32)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i32
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i32
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C]](s32)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1431655765
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C1]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[COPY]], [[AND]]
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[SUB]], [[C2]](s32)
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 858993459
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C3]]
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C3]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND1]], [[AND2]]
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C4]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 252645135
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C5]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 16843009
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND3]], [[C6]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C7]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[LSHR3]](s32)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %0:_(s32) = COPY $x10
+ %1:_(s32) = G_CTPOP %0(s32)
+ $x10 = COPY %1(s32)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i64
+body: |
+ bb.1:
+ liveins: $x10, $x11
+
+ ; CHECK-LABEL: name: ctpop_i64
+ ; CHECK: liveins: $x10, $x11
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $x10
+ ; CHECK-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $x11
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C]](s32)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1431655765
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C1]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[COPY]], [[AND]]
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[SUB]], [[C2]](s32)
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 858993459
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C3]]
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C3]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND1]], [[AND2]]
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C4]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 252645135
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C5]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 16843009
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND3]], [[C6]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C7]](s32)
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[LSHR4:%[0-9]+]]:_(s32) = G_LSHR [[COPY1]], [[C8]](s32)
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 1431655765
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[LSHR4]], [[C9]]
+ ; CHECK-NEXT: [[SUB1:%[0-9]+]]:_(s32) = G_SUB [[COPY1]], [[AND4]]
+ ; CHECK-NEXT: [[C10:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[LSHR5:%[0-9]+]]:_(s32) = G_LSHR [[SUB1]], [[C10]](s32)
+ ; CHECK-NEXT: [[C11:%[0-9]+]]:_(s32) = G_CONSTANT i32 858993459
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[LSHR5]], [[C11]]
+ ; CHECK-NEXT: [[AND6:%[0-9]+]]:_(s32) = G_AND [[SUB1]], [[C11]]
+ ; CHECK-NEXT: [[ADD2:%[0-9]+]]:_(s32) = G_ADD [[AND5]], [[AND6]]
+ ; CHECK-NEXT: [[C12:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR6:%[0-9]+]]:_(s32) = G_LSHR [[ADD2]], [[C12]](s32)
+ ; CHECK-NEXT: [[ADD3:%[0-9]+]]:_(s32) = G_ADD [[LSHR6]], [[ADD2]]
+ ; CHECK-NEXT: [[C13:%[0-9]+]]:_(s32) = G_CONSTANT i32 252645135
+ ; CHECK-NEXT: [[AND7:%[0-9]+]]:_(s32) = G_AND [[ADD3]], [[C13]]
+ ; CHECK-NEXT: [[C14:%[0-9]+]]:_(s32) = G_CONSTANT i32 16843009
+ ; CHECK-NEXT: [[MUL1:%[0-9]+]]:_(s32) = G_MUL [[AND7]], [[C14]]
+ ; CHECK-NEXT: [[C15:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
+ ; CHECK-NEXT: [[LSHR7:%[0-9]+]]:_(s32) = G_LSHR [[MUL1]], [[C15]](s32)
+ ; CHECK-NEXT: [[ADD4:%[0-9]+]]:_(s32) = G_ADD [[LSHR7]], [[LSHR3]]
+ ; CHECK-NEXT: [[C16:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
+ ; CHECK-NEXT: $x10 = COPY [[ADD4]](s32)
+ ; CHECK-NEXT: $x11 = COPY [[C16]](s32)
+ ; CHECK-NEXT: PseudoRET implicit $x10, implicit $x11
+ %1:_(s32) = COPY $x10
+ %2:_(s32) = COPY $x11
+ %0:_(s64) = G_MERGE_VALUES %1(s32), %2(s32)
+ %3:_(s64) = G_CTPOP %0(s64)
+ %4:_(s32), %5:_(s32) = G_UNMERGE_VALUES %3(s64)
+ $x10 = COPY %4(s32)
+ $x11 = COPY %5(s32)
+ PseudoRET implicit $x10, implicit $x11
+
+...
diff --git a/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv64/legalize-ctpop.mir b/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv64/legalize-ctpop.mir
new file mode 100644
index 000000000000000..16bbdb878d1e68e
--- /dev/null
+++ b/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv64/legalize-ctpop.mir
@@ -0,0 +1,180 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mattr=+m -mtriple=riscv64 -run-pass=legalizer %s -o - \
+# RUN: | FileCheck %s
+
+---
+name: ctpop_i8
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i8
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s64) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[TRUNC:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[TRUNC]], [[C1]]
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[AND]], [[C]](s32)
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 85
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C2]]
+ ; CHECK-NEXT: [[TRUNC1:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[TRUNC1]], [[AND1]]
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C4]]
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[AND2]], [[C3]](s32)
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 51
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C5]]
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C5]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND3]], [[AND4]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C6]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 15
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C7]]
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND5]], [[C8]]
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C9]](s32)
+ ; CHECK-NEXT: [[ANYEXT:%[0-9]+]]:_(s64) = G_ANYEXT [[LSHR3]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[ANYEXT]](s64)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s64) = COPY $x10
+ %0:_(s8) = G_TRUNC %1(s64)
+ %2:_(s8) = G_CTPOP %0(s8)
+ %3:_(s64) = G_ANYEXT %2(s8)
+ $x10 = COPY %3(s64)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i16
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i16
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s64) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[TRUNC:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[TRUNC]], [[C1]]
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[AND]], [[C]](s32)
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 21845
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C2]]
+ ; CHECK-NEXT: [[TRUNC1:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[TRUNC1]], [[AND1]]
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C4]]
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[AND2]], [[C3]](s32)
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 13107
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C5]]
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C5]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND3]], [[AND4]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C6]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 3855
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C7]]
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 257
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND5]], [[C8]]
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 8
+ ; CHECK-NEXT: [[C10:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND6:%[0-9]+]]:_(s32) = G_AND [[MUL]], [[C10]]
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[AND6]], [[C9]](s32)
+ ; CHECK-NEXT: [[ANYEXT:%[0-9]+]]:_(s64) = G_ANYEXT [[LSHR3]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[ANYEXT]](s64)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s64) = COPY $x10
+ %0:_(s16) = G_TRUNC %1(s64)
+ %2:_(s16) = G_CTPOP %0(s16)
+ %3:_(s64) = G_ANYEXT %2(s16)
+ $x10 = COPY %3(s64)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i32
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i32
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s64) = COPY $x10
+ ; CHECK-NEXT: [[TRUNC:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[TRUNC]], [[C]](s32)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1431655765
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C1]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[TRUNC]], [[AND]]
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[SUB]], [[C2]](s32)
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 858993459
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C3]]
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C3]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND1]], [[AND2]]
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C4]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 252645135
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C5]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 16843009
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND3]], [[C6]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C7]](s32)
+ ; CHECK-NEXT: [[ANYEXT:%[0-9]+]]:_(s64) = G_ANYEXT [[LSHR3]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[ANYEXT]](s64)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s64) = COPY $x10
+ %0:_(s32) = G_TRUNC %1(s64)
+ %2:_(s32) = G_CTPOP %0(s32)
+ %3:_(s64) = G_ANYEXT %2(s32)
+ $x10 = COPY %3(s64)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i64
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i64
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s64) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 1
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s64) = G_LSHR [[COPY]], [[C]](s64)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s64) = G_CONSTANT i64 6148914691236517205
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s64) = G_AND [[LSHR]], [[C1]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s64) = G_SUB [[COPY]], [[AND]]
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s64) = G_CONSTANT i64 2
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s64) = G_LSHR [[SUB]], [[C2]](s64)
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s64) = G_CONSTANT i64 3689348814741910323
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s64) = G_AND [[LSHR1]], [[C3]]
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s64) = G_AND [[SUB]], [[C3]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s64) = G_ADD [[AND1]], [[AND2]]
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s64) = G_CONSTANT i64 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s64) = G_LSHR [[ADD]], [[C4]](s64)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s64) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s64) = G_CONSTANT i64 1085102592571150095
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s64) = G_AND [[ADD1]], [[C5]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s64) = G_CONSTANT i64 72340172838076673
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s64) = G_MUL [[AND3]], [[C6]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s64) = G_CONSTANT i64 56
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s64) = G_LSHR [[MUL]], [[C7]](s64)
+ ; CHECK-NEXT: $x10 = COPY [[LSHR3]](s64)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %0:_(s64) = COPY $x10
+ %1:_(s64) = G_CTPOP %0(s64)
+ $x10 = COPY %1(s64)
+ PseudoRET implicit $x10
+
+...
|
@llvm/pr-subscribers-backend-risc-v Author: Craig Topper (topperc) ChangesThe base ISA does not have an instruction for this so we need to lower. Zbb support will come in a future patch. Full diff: https://github.com/llvm/llvm-project/pull/72005.diff 3 Files Affected:
diff --git a/llvm/lib/Target/RISCV/GISel/RISCVLegalizerInfo.cpp b/llvm/lib/Target/RISCV/GISel/RISCVLegalizerInfo.cpp
index 2baaadd8ae00af0..386b05e95649b5f 100644
--- a/llvm/lib/Target/RISCV/GISel/RISCVLegalizerInfo.cpp
+++ b/llvm/lib/Target/RISCV/GISel/RISCVLegalizerInfo.cpp
@@ -80,6 +80,11 @@ RISCVLegalizerInfo::RISCVLegalizerInfo(const RISCVSubtarget &ST) {
.maxScalar(0, sXLen)
.lower();
+ getActionDefinitionsBuilder(G_CTPOP)
+ .maxScalar(0, sXLen)
+ .scalarSameSizeAs(1, 0)
+ .lower();
+
getActionDefinitionsBuilder({G_CONSTANT, G_IMPLICIT_DEF})
.legalFor({s32, sXLen, p0})
.widenScalarToNextPow2(0)
diff --git a/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv32/legalize-ctpop.mir b/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv32/legalize-ctpop.mir
new file mode 100644
index 000000000000000..0ea924745d40944
--- /dev/null
+++ b/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv32/legalize-ctpop.mir
@@ -0,0 +1,198 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mattr=+m -mtriple=riscv32 -run-pass=legalizer %s -o - \
+# RUN: | FileCheck %s
+
+---
+name: ctpop_i8
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i8
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY]], [[C1]]
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[AND]], [[C]](s32)
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 85
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C2]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[COPY]], [[AND1]]
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C4]]
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[AND2]], [[C3]](s32)
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 51
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C5]]
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C5]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND3]], [[AND4]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C6]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 15
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C7]]
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND5]], [[C8]]
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C9]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[LSHR3]](s32)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s32) = COPY $x10
+ %0:_(s8) = G_TRUNC %1(s32)
+ %2:_(s8) = G_CTPOP %0(s8)
+ %3:_(s32) = G_ANYEXT %2(s8)
+ $x10 = COPY %3(s32)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i16
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i16
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY]], [[C1]]
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[AND]], [[C]](s32)
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 21845
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C2]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[COPY]], [[AND1]]
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C4]]
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[AND2]], [[C3]](s32)
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 13107
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C5]]
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C5]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND3]], [[AND4]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C6]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 3855
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C7]]
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 257
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND5]], [[C8]]
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 8
+ ; CHECK-NEXT: [[C10:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND6:%[0-9]+]]:_(s32) = G_AND [[MUL]], [[C10]]
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[AND6]], [[C9]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[LSHR3]](s32)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s32) = COPY $x10
+ %0:_(s16) = G_TRUNC %1(s32)
+ %2:_(s16) = G_CTPOP %0(s16)
+ %3:_(s32) = G_ANYEXT %2(s16)
+ $x10 = COPY %3(s32)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i32
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i32
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C]](s32)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1431655765
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C1]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[COPY]], [[AND]]
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[SUB]], [[C2]](s32)
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 858993459
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C3]]
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C3]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND1]], [[AND2]]
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C4]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 252645135
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C5]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 16843009
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND3]], [[C6]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C7]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[LSHR3]](s32)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %0:_(s32) = COPY $x10
+ %1:_(s32) = G_CTPOP %0(s32)
+ $x10 = COPY %1(s32)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i64
+body: |
+ bb.1:
+ liveins: $x10, $x11
+
+ ; CHECK-LABEL: name: ctpop_i64
+ ; CHECK: liveins: $x10, $x11
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $x10
+ ; CHECK-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $x11
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C]](s32)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1431655765
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C1]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[COPY]], [[AND]]
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[SUB]], [[C2]](s32)
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 858993459
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C3]]
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C3]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND1]], [[AND2]]
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C4]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 252645135
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C5]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 16843009
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND3]], [[C6]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C7]](s32)
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[LSHR4:%[0-9]+]]:_(s32) = G_LSHR [[COPY1]], [[C8]](s32)
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 1431655765
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[LSHR4]], [[C9]]
+ ; CHECK-NEXT: [[SUB1:%[0-9]+]]:_(s32) = G_SUB [[COPY1]], [[AND4]]
+ ; CHECK-NEXT: [[C10:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[LSHR5:%[0-9]+]]:_(s32) = G_LSHR [[SUB1]], [[C10]](s32)
+ ; CHECK-NEXT: [[C11:%[0-9]+]]:_(s32) = G_CONSTANT i32 858993459
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[LSHR5]], [[C11]]
+ ; CHECK-NEXT: [[AND6:%[0-9]+]]:_(s32) = G_AND [[SUB1]], [[C11]]
+ ; CHECK-NEXT: [[ADD2:%[0-9]+]]:_(s32) = G_ADD [[AND5]], [[AND6]]
+ ; CHECK-NEXT: [[C12:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR6:%[0-9]+]]:_(s32) = G_LSHR [[ADD2]], [[C12]](s32)
+ ; CHECK-NEXT: [[ADD3:%[0-9]+]]:_(s32) = G_ADD [[LSHR6]], [[ADD2]]
+ ; CHECK-NEXT: [[C13:%[0-9]+]]:_(s32) = G_CONSTANT i32 252645135
+ ; CHECK-NEXT: [[AND7:%[0-9]+]]:_(s32) = G_AND [[ADD3]], [[C13]]
+ ; CHECK-NEXT: [[C14:%[0-9]+]]:_(s32) = G_CONSTANT i32 16843009
+ ; CHECK-NEXT: [[MUL1:%[0-9]+]]:_(s32) = G_MUL [[AND7]], [[C14]]
+ ; CHECK-NEXT: [[C15:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
+ ; CHECK-NEXT: [[LSHR7:%[0-9]+]]:_(s32) = G_LSHR [[MUL1]], [[C15]](s32)
+ ; CHECK-NEXT: [[ADD4:%[0-9]+]]:_(s32) = G_ADD [[LSHR7]], [[LSHR3]]
+ ; CHECK-NEXT: [[C16:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
+ ; CHECK-NEXT: $x10 = COPY [[ADD4]](s32)
+ ; CHECK-NEXT: $x11 = COPY [[C16]](s32)
+ ; CHECK-NEXT: PseudoRET implicit $x10, implicit $x11
+ %1:_(s32) = COPY $x10
+ %2:_(s32) = COPY $x11
+ %0:_(s64) = G_MERGE_VALUES %1(s32), %2(s32)
+ %3:_(s64) = G_CTPOP %0(s64)
+ %4:_(s32), %5:_(s32) = G_UNMERGE_VALUES %3(s64)
+ $x10 = COPY %4(s32)
+ $x11 = COPY %5(s32)
+ PseudoRET implicit $x10, implicit $x11
+
+...
diff --git a/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv64/legalize-ctpop.mir b/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv64/legalize-ctpop.mir
new file mode 100644
index 000000000000000..16bbdb878d1e68e
--- /dev/null
+++ b/llvm/test/CodeGen/RISCV/GlobalISel/legalizer/rv64/legalize-ctpop.mir
@@ -0,0 +1,180 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mattr=+m -mtriple=riscv64 -run-pass=legalizer %s -o - \
+# RUN: | FileCheck %s
+
+---
+name: ctpop_i8
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i8
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s64) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[TRUNC:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[TRUNC]], [[C1]]
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[AND]], [[C]](s32)
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 85
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C2]]
+ ; CHECK-NEXT: [[TRUNC1:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[TRUNC1]], [[AND1]]
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C4]]
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[AND2]], [[C3]](s32)
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 51
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C5]]
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C5]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND3]], [[AND4]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C6]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 15
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C7]]
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND5]], [[C8]]
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C9]](s32)
+ ; CHECK-NEXT: [[ANYEXT:%[0-9]+]]:_(s64) = G_ANYEXT [[LSHR3]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[ANYEXT]](s64)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s64) = COPY $x10
+ %0:_(s8) = G_TRUNC %1(s64)
+ %2:_(s8) = G_CTPOP %0(s8)
+ %3:_(s64) = G_ANYEXT %2(s8)
+ $x10 = COPY %3(s64)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i16
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i16
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s64) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[TRUNC:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[TRUNC]], [[C1]]
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[AND]], [[C]](s32)
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 21845
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C2]]
+ ; CHECK-NEXT: [[TRUNC1:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[TRUNC1]], [[AND1]]
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C4]]
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[AND2]], [[C3]](s32)
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 13107
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C5]]
+ ; CHECK-NEXT: [[AND4:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C5]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND3]], [[AND4]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C6]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 3855
+ ; CHECK-NEXT: [[AND5:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C7]]
+ ; CHECK-NEXT: [[C8:%[0-9]+]]:_(s32) = G_CONSTANT i32 257
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND5]], [[C8]]
+ ; CHECK-NEXT: [[C9:%[0-9]+]]:_(s32) = G_CONSTANT i32 8
+ ; CHECK-NEXT: [[C10:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
+ ; CHECK-NEXT: [[AND6:%[0-9]+]]:_(s32) = G_AND [[MUL]], [[C10]]
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[AND6]], [[C9]](s32)
+ ; CHECK-NEXT: [[ANYEXT:%[0-9]+]]:_(s64) = G_ANYEXT [[LSHR3]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[ANYEXT]](s64)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s64) = COPY $x10
+ %0:_(s16) = G_TRUNC %1(s64)
+ %2:_(s16) = G_CTPOP %0(s16)
+ %3:_(s64) = G_ANYEXT %2(s16)
+ $x10 = COPY %3(s64)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i32
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i32
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s64) = COPY $x10
+ ; CHECK-NEXT: [[TRUNC:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[TRUNC]], [[C]](s32)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1431655765
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C1]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[TRUNC]], [[AND]]
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[SUB]], [[C2]](s32)
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 858993459
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR1]], [[C3]]
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[SUB]], [[C3]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND1]], [[AND2]]
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[ADD]], [[C4]](s32)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s32) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s32) = G_CONSTANT i32 252645135
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[ADD1]], [[C5]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s32) = G_CONSTANT i32 16843009
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[AND3]], [[C6]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[MUL]], [[C7]](s32)
+ ; CHECK-NEXT: [[ANYEXT:%[0-9]+]]:_(s64) = G_ANYEXT [[LSHR3]](s32)
+ ; CHECK-NEXT: $x10 = COPY [[ANYEXT]](s64)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %1:_(s64) = COPY $x10
+ %0:_(s32) = G_TRUNC %1(s64)
+ %2:_(s32) = G_CTPOP %0(s32)
+ %3:_(s64) = G_ANYEXT %2(s32)
+ $x10 = COPY %3(s64)
+ PseudoRET implicit $x10
+
+...
+---
+name: ctpop_i64
+body: |
+ bb.1:
+ liveins: $x10
+
+ ; CHECK-LABEL: name: ctpop_i64
+ ; CHECK: liveins: $x10
+ ; CHECK-NEXT: {{ $}}
+ ; CHECK-NEXT: [[COPY:%[0-9]+]]:_(s64) = COPY $x10
+ ; CHECK-NEXT: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 1
+ ; CHECK-NEXT: [[LSHR:%[0-9]+]]:_(s64) = G_LSHR [[COPY]], [[C]](s64)
+ ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s64) = G_CONSTANT i64 6148914691236517205
+ ; CHECK-NEXT: [[AND:%[0-9]+]]:_(s64) = G_AND [[LSHR]], [[C1]]
+ ; CHECK-NEXT: [[SUB:%[0-9]+]]:_(s64) = G_SUB [[COPY]], [[AND]]
+ ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s64) = G_CONSTANT i64 2
+ ; CHECK-NEXT: [[LSHR1:%[0-9]+]]:_(s64) = G_LSHR [[SUB]], [[C2]](s64)
+ ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s64) = G_CONSTANT i64 3689348814741910323
+ ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(s64) = G_AND [[LSHR1]], [[C3]]
+ ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(s64) = G_AND [[SUB]], [[C3]]
+ ; CHECK-NEXT: [[ADD:%[0-9]+]]:_(s64) = G_ADD [[AND1]], [[AND2]]
+ ; CHECK-NEXT: [[C4:%[0-9]+]]:_(s64) = G_CONSTANT i64 4
+ ; CHECK-NEXT: [[LSHR2:%[0-9]+]]:_(s64) = G_LSHR [[ADD]], [[C4]](s64)
+ ; CHECK-NEXT: [[ADD1:%[0-9]+]]:_(s64) = G_ADD [[LSHR2]], [[ADD]]
+ ; CHECK-NEXT: [[C5:%[0-9]+]]:_(s64) = G_CONSTANT i64 1085102592571150095
+ ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(s64) = G_AND [[ADD1]], [[C5]]
+ ; CHECK-NEXT: [[C6:%[0-9]+]]:_(s64) = G_CONSTANT i64 72340172838076673
+ ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s64) = G_MUL [[AND3]], [[C6]]
+ ; CHECK-NEXT: [[C7:%[0-9]+]]:_(s64) = G_CONSTANT i64 56
+ ; CHECK-NEXT: [[LSHR3:%[0-9]+]]:_(s64) = G_LSHR [[MUL]], [[C7]](s64)
+ ; CHECK-NEXT: $x10 = COPY [[LSHR3]](s64)
+ ; CHECK-NEXT: PseudoRET implicit $x10
+ %0:_(s64) = COPY $x10
+ %1:_(s64) = G_CTPOP %0(s64)
+ $x10 = COPY %1(s64)
+ PseudoRET implicit $x10
+
+...
|
@@ -80,6 +80,11 @@ RISCVLegalizerInfo::RISCVLegalizerInfo(const RISCVSubtarget &ST) { | |||
.maxScalar(0, sXLen) | |||
.lower(); | |||
|
|||
getActionDefinitionsBuilder(G_CTPOP) | |||
.maxScalar(0, sXLen) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There's an assumption here that the maximum bitcount for any possible value including illegal will fit in XLen bits. This should be true since LLVM's maximum integer type is i16777215.
Clamping the result type first gives better code when splitting s64 intoto two s32 ops on RV32. That will generate an s32 G_ADD to add the two ctpops together. Otherwise we would get an s64 G_ADD that also needs to split.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It might be helpful to add this information in a comment in the code?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM.
@@ -80,6 +80,11 @@ RISCVLegalizerInfo::RISCVLegalizerInfo(const RISCVSubtarget &ST) { | |||
.maxScalar(0, sXLen) | |||
.lower(); | |||
|
|||
getActionDefinitionsBuilder(G_CTPOP) | |||
.maxScalar(0, sXLen) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It might be helpful to add this information in a comment in the code?
The base ISA does not have an instruction for this so we need to lower. Zbb support will come in a future patch.
The base ISA does not have an instruction for this so we need to lower. Zbb support will come in a future patch.