> It's hard to get a testcase without > http://gcc.gnu.org/viewcvs/gcc?view=revision&revision=201856 because > none of these *INTERNAL* symbols will be emitted in debug info.
The original code was in there as a form of assembly-time assertion that the symbol would never get output. Now that you want to emit it in the debug info, the original intent is still valid, but without that extra kruft, the "assertion" is now gone. I think Jason is suggesting that you replace that with a test case, which would explicitly materialize one or more of the inlined functions that would get these mangled names, and then check the assembler output to verify that the (now valid) mangled name doesn't appear anywhere as an assembler label. -cary
