[lldb-dev] [Bug 45905] New: [lldb][unittest] Assertion failed : (m_replayers.find(RunID) == m_replayers.end())

via lldb-dev lldb-dev at lists.llvm.org
Wed May 13 08:44:38 PDT 2020


https://bugs.llvm.org/show_bug.cgi?id=45905

            Bug ID: 45905
           Summary: [lldb][unittest] Assertion failed :
                    (m_replayers.find(RunID) == m_replayers.end())
           Product: lldb
           Version: unspecified
          Hardware: PC
                OS: Windows NT
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: All Bugs
          Assignee: lldb-dev at lists.llvm.org
          Reporter: sylvain.audi at ubisoft.com
                CC: alexandre.ganea at ubisoft.com, jdevlieghere at apple.com,
                    llvm-bugs at lists.llvm.org

This happens in the LLDB unit tests, when building under windows.

In lldb\unittests\Utility\ReproducerInstrumentationTest.cpp:
  LLDB_REGISTER_METHOD(void, InstrumentedFoo, Validate, ());
  (...)
  LLDB_REGISTER_METHOD(void, InstrumentedBar, Validate, ());

The clashing IDs are evaluated as follows:
   &invoke<void(InstrumentedFoo::*) 
   ()>::method<(&InstrumentedFoo::Validate)>::record
  and
   &invoke<void(InstrumentedBar::*) 
   ()>::method<(&InstrumentedBar::Validate)>::record


Both "record" implementations are only calling Validate() through the vtable,
so the implementations are identical for both.

The linker does COMDAT folding (option /OPT:ICF), which merges the 2 functions,
so their address, which is used as an ID, end up being identical.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/lldb-dev/attachments/20200513/a8d13df4/attachment.html>


More information about the lldb-dev mailing list