[llvm] Allow Specifying SGMasks for Inline Asm (PR #155491)

[Jan Patrick Lehr via llvm-commits]

================
@@ -0,0 +1,71 @@
+; RUN: llc -O3 -mcpu=gfx942 < %s | FileCheck %s
+; CHECK: v_add_f32_e32
+; CHECK-NEXT: ;;#ASMSTART
+; CHECK-NEXT: v_add_f32_e32
+; CHECK-NEXT: ;;#ASMEND
+; CHECK: v_add_f32_e32
+; ModuleID = '<stdin>'
+source_filename = "llvm-link"
+target datalayout = "e-p:64:64-p1:64:64-p2:32:32-p3:32:32-p4:64:64-p5:32:32-p6:32:32-p7:160:256:256:32-p8:128:128:128:48-p9:192:256:256:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64-S32-A5-G1-ni:7:8:9"
+target triple = "amdgcn-amd-amdhsa"
+
+ at llvm.compiler.used = appending addrspace(1) global [1 x ptr] [ptr addrspacecast (ptr addrspace(1) @__hip_cuid_bffb86447932ec40 to ptr)], section "llvm.metadata"
+ at __hip_cuid_bffb86447932ec40 = addrspace(1) global i8 0
----------------
jplehr wrote:

Is this needed for the test to work?

https://github.com/llvm/llvm-project/pull/155491