On Tue, 14 May 2019, Nathan Sidwell wrote: > This patch nadgers the driver's subprocess names to include the driver name. > It results in more informative error messages. For instance, rather than: > > >./xg++ -B./ frob.cc -c -fdump-tree-nope > cc1plus: error: unrecognized command line option '-fdump-tree-nope' > > we get: > > >./xg++ -B./ frob.cc -c -fdump-tree-nope > xg++(cc1plus): error: unrecognized command line option '-fdump-tree-nope' > > Thereby cluing the user into this being a compiler error. (When this error is > buried inside a build log, the poor user can be more confused as to what this > cc1plus thing might be).
My inclination is that, whatever you do with argv[0], it's a bug for the message to mention cc1plus at all. The fact that there is an executable called cc1plus is an implementation detail, which shouldn't be visible in diagnostics; the diagnostics should just mention the program the user called (e.g. g++) and not cc1plus. (So make toplev.c use the value of COLLECT_GCC, if set, instead of argv[0], to determine the name for diagnostic purposes, I suppose. I believe COLLECT_GCC should always be set except when cc1plus was called manually, and if it was called manually then it's reasonable to mention it in diagnostics and the person calling it is probably a GCC developer anyway.) -- Joseph S. Myers jos...@codesourcery.com
