[PATCH] D81861: [HIP] Do not use llvm-link/opt/llc for -fgpu-rdc

[Yaxun Liu via Phabricator via cfe-commits]

yaxunl marked 5 inline comments as done.
yaxunl added inline comments.


================
Comment at: clang/test/Driver/hip-phases.hip:34
+// RDC-DAG: [[P6:[0-9]+]]: backend, {[[P5]]}, ir, (device-[[T]], [[ARCH]])
+// RDC-DAG: [[P10:[0-9]+]]: linker, {[[P6]]}, image, (device-hip, [[ARCH]])
 
----------------
ashi1 wrote:
> should also use device-[[T]]? I think there are several locations here using device-hip instead of device[[T]].
will fix


================
Comment at: clang/test/Driver/hip-phases.hip:253
+// L2-DAG: [[P0:[0-9]+]]: input, "{{.*}}obj1.o", object, (host-[[T:hip]])
+// RL2-DAG: [[P1:[0-9]+]]: clang-offload-unbundler, {[[P0]]}, object, (host-[[T:hip]])
+// L2-DAG: [[P2:[0-9]+]]: input, "{{.*}}obj2.o", object, (host-[[T:hip]])
----------------
ashi1 wrote:
> [[T]] is defined everytime here, should probably use [[T]] in subsequent checks?
But those [[T]] are defined through check prefixes not used by this run line, so we have to define it here.


================
Comment at: clang/test/Driver/hip-save-temps.hip:58
+
+// -fgpu-rdc link
+// RUN: touch %T/obj1.o
----------------
ashi1 wrote:
> Should this link test be moved into a separate file? Also to cover without -o? I may also be adding more tests to this case for static lib generation too.
will do


================
Comment at: clang/test/Driver/hip-toolchain-rdc.hip:10
 // RUN:   --hip-device-lib-path=%S/Inputs/hip_multiple_inputs/lib2 \
-// RUN:   -fuse-ld=lld -fgpu-rdc \
+// RUN:   -fuse-ld=lld -fgpu-rdc -nogpuinc \
 // RUN:   %S/Inputs/hip_multiple_inputs/a.cu \
----------------
ashi1 wrote:
> What is the difference between this file and hip-toolchain-rdc-separate.hip?
this one do compile and link together. that one do compile and link with separate commands.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D81861/new/

https://reviews.llvm.org/D81861