[llvm] [LLVM][AArch64]Use load/store with consecutive registers in SME2 or S… (PR #77665)
Momchil Velikov via llvm-commits
llvm-commits at lists.llvm.org
Wed Jan 17 05:11:17 PST 2024
================
@@ -0,0 +1,295 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sme -verify-machineinstrs < %s | FileCheck %s --check-prefixes=NOPAIR
+; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sve2 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=NOPAIR
+; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sme2 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=PAIR
+; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sve2p1 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=PAIR
+
+
+declare void @my_func()
+
+define void @fbyte(<vscale x 16 x i8> %v) {
+; NOPAIR-LABEL: fbyte:
+; NOPAIR: // %bb.0:
+; NOPAIR-NEXT: stp x29, x30, [sp, #-16]! // 16-byte Folded Spill
+; NOPAIR-NEXT: addvl sp, sp, #-18
+; NOPAIR-NEXT: str p15, [sp, #4, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p14, [sp, #5, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p13, [sp, #6, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p12, [sp, #7, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p11, [sp, #8, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p10, [sp, #9, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p9, [sp, #10, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p8, [sp, #11, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p7, [sp, #12, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p6, [sp, #13, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p5, [sp, #14, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str p4, [sp, #15, mul vl] // 2-byte Folded Spill
+; NOPAIR-NEXT: str z23, [sp, #2, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z22, [sp, #3, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z21, [sp, #4, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z20, [sp, #5, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z19, [sp, #6, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z18, [sp, #7, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z17, [sp, #8, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z16, [sp, #9, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z15, [sp, #10, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z14, [sp, #11, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z13, [sp, #12, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z12, [sp, #13, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z11, [sp, #14, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z10, [sp, #15, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z9, [sp, #16, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: str z8, [sp, #17, mul vl] // 16-byte Folded Spill
+; NOPAIR-NEXT: .cfi_escape 0x0f, 0x0d, 0x8f, 0x00, 0x11, 0x10, 0x22, 0x11, 0x90, 0x01, 0x92, 0x2e, 0x00, 0x1e, 0x22 // sp + 16 + 144 * VG
+; NOPAIR-NEXT: .cfi_offset w30, -8
+; NOPAIR-NEXT: .cfi_offset w29, -16
+; NOPAIR-NEXT: .cfi_escape 0x10, 0x48, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x78, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d8 @ cfa - 16 - 8 * VG
+; NOPAIR-NEXT: .cfi_escape 0x10, 0x49, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x70, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d9 @ cfa - 16 - 16 * VG
+; NOPAIR-NEXT: .cfi_escape 0x10, 0x4a, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x68, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d10 @ cfa - 16 - 24 * VG
+; NOPAIR-NEXT: .cfi_escape 0x10, 0x4b, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x60, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d11 @ cfa - 16 - 32 * VG
+; NOPAIR-NEXT: .cfi_escape 0x10, 0x4c, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x58, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d12 @ cfa - 16 - 40 * VG
+; NOPAIR-NEXT: .cfi_escape 0x10, 0x4d, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x50, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d13 @ cfa - 16 - 48 * VG
+; NOPAIR-NEXT: .cfi_escape 0x10, 0x4e, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x48, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d14 @ cfa - 16 - 56 * VG
+; NOPAIR-NEXT: .cfi_escape 0x10, 0x4f, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x40, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d15 @ cfa - 16 - 64 * VG
+; NOPAIR-NEXT: bl my_func
+; NOPAIR-NEXT: ldr p15, [sp, #4, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr z23, [sp, #2, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr z22, [sp, #3, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr p14, [sp, #5, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr z21, [sp, #4, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr z20, [sp, #5, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr p13, [sp, #6, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr z19, [sp, #6, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr z18, [sp, #7, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr p12, [sp, #7, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr z17, [sp, #8, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr z16, [sp, #9, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr p11, [sp, #8, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr z15, [sp, #10, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr z14, [sp, #11, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr p10, [sp, #9, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr z13, [sp, #12, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr z12, [sp, #13, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr p9, [sp, #10, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr z11, [sp, #14, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr z10, [sp, #15, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr p8, [sp, #11, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr z9, [sp, #16, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr z8, [sp, #17, mul vl] // 16-byte Folded Reload
+; NOPAIR-NEXT: ldr p7, [sp, #12, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr p6, [sp, #13, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr p5, [sp, #14, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: ldr p4, [sp, #15, mul vl] // 2-byte Folded Reload
+; NOPAIR-NEXT: addvl sp, sp, #18
+; NOPAIR-NEXT: ldp x29, x30, [sp], #16 // 16-byte Folded Reload
+; NOPAIR-NEXT: ret
+;
+; PAIR-LABEL: fbyte:
+; PAIR: // %bb.0:
+; PAIR-NEXT: stp x29, x30, [sp, #-16]! // 16-byte Folded Spill
+; PAIR-NEXT: addvl sp, sp, #-18
+; PAIR-NEXT: str p8, [sp, #11, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: ptrue pn8.b
+; PAIR-NEXT: str p15, [sp, #4, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: st1b { z22.b, z23.b }, pn8, [sp, #4, mul vl] // 32-byte Folded Spill
+; PAIR-NEXT: st1b { z20.b, z21.b }, pn8, [sp, #8, mul vl] // 32-byte Folded Spill
+; PAIR-NEXT: str p14, [sp, #5, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: st1b { z18.b, z19.b }, pn8, [sp, #12, mul vl] // 32-byte Folded Spill
+; PAIR-NEXT: st1b { z16.b, z17.b }, pn8, [sp, #16, mul vl] // 32-byte Folded Spill
+; PAIR-NEXT: str p13, [sp, #6, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: st1b { z14.b, z15.b }, pn8, [sp, #20, mul vl] // 32-byte Folded Spill
+; PAIR-NEXT: st1b { z12.b, z13.b }, pn8, [sp, #24, mul vl] // 32-byte Folded Spill
+; PAIR-NEXT: str p12, [sp, #7, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: st1b { z10.b, z11.b }, pn8, [sp, #28, mul vl] // 32-byte Folded Spill
+; PAIR-NEXT: str p11, [sp, #8, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: str p10, [sp, #9, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: str p9, [sp, #10, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: str p7, [sp, #12, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: str p6, [sp, #13, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: str p5, [sp, #14, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: str p4, [sp, #15, mul vl] // 2-byte Folded Spill
+; PAIR-NEXT: st1b { z8.b, z9.b }, pn8, [sp, #32, mul vl] // 32-byte Folded Spill
+; PAIR-NEXT: .cfi_escape 0x0f, 0x0d, 0x8f, 0x00, 0x11, 0x10, 0x22, 0x11, 0x90, 0x01, 0x92, 0x2e, 0x00, 0x1e, 0x22 // sp + 16 + 144 * VG
+; PAIR-NEXT: .cfi_offset w30, -8
+; PAIR-NEXT: .cfi_offset w29, -16
+; PAIR-NEXT: .cfi_escape 0x10, 0x48, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x78, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d8 @ cfa - 16 - 8 * VG
+; PAIR-NEXT: .cfi_escape 0x10, 0x49, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x70, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d9 @ cfa - 16 - 16 * VG
+; PAIR-NEXT: .cfi_escape 0x10, 0x4a, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x68, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d10 @ cfa - 16 - 24 * VG
+; PAIR-NEXT: .cfi_escape 0x10, 0x4b, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x60, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d11 @ cfa - 16 - 32 * VG
+; PAIR-NEXT: .cfi_escape 0x10, 0x4c, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x58, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d12 @ cfa - 16 - 40 * VG
+; PAIR-NEXT: .cfi_escape 0x10, 0x4d, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x50, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d13 @ cfa - 16 - 48 * VG
+; PAIR-NEXT: .cfi_escape 0x10, 0x4e, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x48, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d14 @ cfa - 16 - 56 * VG
+; PAIR-NEXT: .cfi_escape 0x10, 0x4f, 0x0a, 0x11, 0x70, 0x22, 0x11, 0x40, 0x92, 0x2e, 0x00, 0x1e, 0x22 // $d15 @ cfa - 16 - 64 * VG
+; PAIR-NEXT: bl my_func
+; PAIR-NEXT: ldr p8, [sp, #11, mul vl] // 2-byte Folded Reload
+; PAIR-NEXT: ptrue pn8.b
----------------
momchil-velikov wrote:
This is incorrect here, the `p8` was just restored (presumably containing a value useful for the caller) and we immediately clobber it.
https://github.com/llvm/llvm-project/pull/77665
More information about the llvm-commits
mailing list