================
@@ -0,0 +1,258 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
UTC_ARGS: --version 5
+; RUN: llc < %s -o - -mcpu=sm_90 -march=nvptx64 -mattr=+ptx80 | FileCheck %s
+; RUN: %if ptxas-12.0 %{ llc < %s -mtriple=nvptx64 -mcpu=sm_90 -mattr=+ptx80|
%ptxas-verify -arch=sm_90 %}
+
+target triple = "nvptx64-nvidia-cuda"
+
+@llvm.used = appending global [1 x ptr] [ptr
@test_distributed_shared_cluster], section "llvm.metadata"
+
+declare ptr addrspace(7) @llvm.nvvm.mapa.shared.cluster(ptr addrspace(3), i32)
+declare i1 @llvm.nvvm.isspacep.shared.cluster(ptr)
+declare i32 @llvm.nvvm.read.ptx.sreg.ctaid.x()
+declare ptr @llvm.nvvm.mapa(ptr, i32)
+
+define i32 @test_distributed_shared_cluster(ptr %ptr, ptr addrspace(3)
%smem_ptr) local_unnamed_addr {
----------------
AlexMaclean wrote:
Please break this up into more than one big function. Perhaps group some of the
atomic cases together but otherwise each instruction/intrinsic should have it's
own function. This will make the test easier to debug if something starts
failing at some point.
https://github.com/llvm/llvm-project/pull/135444
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
