[llvm-bugs] [Bug 51518] New: 'bin/opt', needed by 'openmp/libomptarget/deviceRTLs/amdgcn/libomptarget-amdgcn-gfx701.bc', missing and no known rule to make it

via llvm-bugs llvm-bugs at lists.llvm.org
Tue Aug 17 21:22:38 PDT 2021


            Bug ID: 51518
           Summary: 'bin/opt', needed by
                    dgcn-gfx701.bc', missing and no known rule to make it
           Product: OpenMP
           Version: unspecified
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Runtime Library
          Assignee: unassignedbugs at nondot.org
          Reporter: llvm at meinersbur.de
                CC: llvm-bugs at lists.llvm.org

The OpenMP runtime build with LLVM_ENABLE_RUNTIMES occasionally with the
following message:

$ ninja runtimes
[0/3] cd
&& /usr/bin/cmake --build .
ninja: error:
needed by
'openmp/libomptarget/deviceRTLs/amdgcn/libomptarget-amdgcn-gfx701.bc', missing
and no known rule to make it
FAILED: runtimes/runtimes-stamps/runtimes-build
&& /usr/bin/cmake --build .
ninja: build stopped: subcommand failed.

Examples buildbot builds where this occurred:


This is likely a missing dependence of the 'runtimes' target on the opt
executable, i.e. the opt executable is not yet built (just the clang
executable) when the nested runtimes build is triggered.
This is supported by that fact that when manually building opt first, it does

$ ninja opt
[1/1] Linking CXX executable bin/opt
$ ninja runtimes
[0/3] cd
&& /usr/bin/cmake --build .
[3/581] Building CXX object

Note that on the CUDA buildbot AMD's ROCm SDK is not installed, yet cmake
reports it being build:

-- LIBOMPTARGET: Building AMDGPU plugin for dlopened libhsa
-- LIBOMPTARGET: Building AMDGCN device RTL. Using clang:

You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20210818/b6cc6152/attachment.html>

More information about the llvm-bugs mailing list