[Clang] Always use --no-undefined when linking AMDGPU images

[jhuber6]

Summary

AMDGPU uses ELF shared libraries to implement their executable device
images. One downside to this method is that it disables regular warnings
on undefined symbols. This is because shared libraries expect these to
be resolves by later loads. However, the GPU images do not support
dynamic linking so any undefined symbol is going to cause a runtime
error. This patch adds --no-undefined to the ld.lld invocation to guarantee
that undefined symbols are always caught as linking errors rather than
runtime errors.

[JonChesterfield]

Adding the amdgpu reviewer group. This change might be contentious - we currently treat various broken code as calls to trap on the grounds that those functions might not be executed. By analogy, functions that call undefined functions but are not themselves called might be considered not a problem.

[jhuber6]

This can be turned off with -zundefs. So we could instruct people to use -Wl,-zundefs or -Xoffload-linker -zundefs if the old behavior is desired.

[jhuber6]

Use --no-undefined to be consistent with HIP and check for OpenCL.

[MaskRay]

While --no-undefined is an alias for -z defs, be consistent in the subject and the body.

Always use -zdefs

--no-undefined

[arsenm]

Release note change should be here

[jhuber6]

Adding release notes

clang/docs/ReleaseNotes.rst

 ^^^^^^^^^^^^^^^^^^^^^^^
 
 Miscellaneous Clang Crashes Fixed
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 Target Specific Changes
 -----------------------
 
+AMDGPU Support
+^^^^^^^^^^^^^^
+
+- Linking for AMDGPU now uses ``--no-undefined`` by default. This causes
+  undefined symbols in the created module to be a linker error. To prevent this,
+  pass ``-Wl,--undefined`` if compiling directly, or ``-Xoffload-linker
+  --undefined`` if using an offloading language.
+
 X86 Support
 ^^^^^^^^^^^
 
 Arm and AArch64 Support
 ^^^^^^^^^^^^^^^^^^^^^^^
 
 - The hard-float ABI is now available in Armv8.1-M configurations that
   have integer MVE instructions (and therefore have FP registers) but

clang/lib/Driver/ToolChains/AMDGPU.cpp

void amdgpu::Linker::ConstructJob(Compilation &C, const JobAction &JA,
 
   std::string Linker = getToolChain().GetProgramPath(getShortName());
   ArgStringList CmdArgs;
   addLinkerCompressDebugSectionsOption(getToolChain(), Args, CmdArgs);
   AddLinkerInputs(getToolChain(), Inputs, Args, CmdArgs, JA);
   if (C.getDriver().isUsingLTO())
     addLTOOptions(getToolChain(), Args, CmdArgs, Output, Inputs[0],
                   C.getDriver().getLTOMode() == LTOK_Thin);
+  CmdArgs.push_back("--no-undefined");
   CmdArgs.push_back("-shared");
   CmdArgs.push_back("-o");
   CmdArgs.push_back(Output.getFilename());
   C.addCommand(std::make_unique<Command>(
       JA, *this, ResponseFileSupport::AtFileCurCP(), Args.MakeArgString(Linker),
       CmdArgs, Inputs, Output));
 }
 

clang/test/Driver/amdgpu-toolchain-opencl.cl

 // CHECK_O4: "-cc1"{{.*}} "-O3"
 // CHECK_O5: "-cc1"{{.*}} "-O5"
 // CHECK_Og: "-cc1"{{.*}} "-Og"
 // CHECK_Ofast: "-cc1"{{.*}} "-Ofast"
 // CHECK_O_DEFAULT: "-cc1"{{.*}} "-O3"
 
 // CHK-INC: "-cc1" {{.*}}"-finclude-default-header" "-fdeclare-opencl-builtins" {{.*}}"-x" "cl"
 // CHK-INC-NOT: "-cc1" {{.*}}"-finclude-default-header" "-fdeclare-opencl-builtins" {{.*}}"-x" "cpp-output"
+
+// RUN: %clang -### --target=amdgcn-amd-amdhsa-opencl -x cl -emit-llvm -mcpu=fiji %s 2>&1 | FileCheck -check-prefix=CHK-LINK %s
+// CHK-LINK: ld.lld{{.*}} "--no-undefined" "-shared"

clang/test/Driver/amdgpu-toolchain.c

 // RUN: %clang -### --target=amdgcn--amdhsa -x assembler -mcpu=kaveri %s 2>&1 | FileCheck -check-prefix=AS_LINK %s
 // RUN: %clang -### -g --target=amdgcn--amdhsa -mcpu=kaveri %s 2>&1 | FileCheck -check-prefix=DWARF_VER %s
 // RUN: %clang -### --target=amdgcn-amd-amdpal -x assembler -mcpu=kaveri %s 2>&1 | FileCheck -check-prefix=AS_LINK %s
 // RUN: %clang -### -g --target=amdgcn-amd-amdpal -mcpu=kaveri %s 2>&1 | FileCheck -check-prefix=DWARF_VER %s
 // RUN: %clang -### --target=amdgcn-mesa-mesa3d -x assembler -mcpu=kaveri %s 2>&1 | FileCheck -check-prefix=AS_LINK %s
 // RUN: %clang -### -g --target=amdgcn-mesa-mesa3d -mcpu=kaveri %s 2>&1 | FileCheck -check-prefix=DWARF_VER %s
 
 // AS_LINK: "-cc1as"
-// AS_LINK: ld.lld{{.*}} "-shared"
+// AS_LINK: ld.lld{{.*}} "--no-undefined" "-shared"
 
 // DWARF_VER: "-dwarf-version=5"
 
 // RUN: %clang -### --target=amdgcn-amd-amdhsa -mcpu=gfx906 -nogpulib \
 // RUN:   -flto %s 2>&1 | FileCheck -check-prefix=LTO %s
 // LTO: clang{{.*}} "-flto=full"
 // LTO: ld.lld{{.*}}-plugin-opt=mcpu=gfx906