Re: [RFA/testsuite] new gdb.threads testcase

[Michael Chastain <mec dot gnu at mindspring dot com>]

Hi Joel,

I'm not the maintainer of gdb.threads but I'll give you some
feedback.

-- needs a changelog entry

-- you can drop the "Please email any bugs, comments, and/or additions" line.
   we're purging those completely, not just updating the address.

-- i think it would help to add a comment with a description of the bug
   that it's testing for, just a distillation of the previous message.
   "gdb 6.1.1 on AIX 5.1 had a bug where the aix-threads code called
   getthrds() incorrectly so that gdb lost track of breakpoints" ...
   just a line or two of clues for the next guy.

-- sometimes it's okay to re-use foo.c, sometimes you need to clone it.
   the issue is that if you re-use foo.c and something in the file
   changes, then it might be hard to keep foo.c usable by foo.exp,
   bar.exp, and bletch.exp.  in this case i think it's better to clone
   it, because i'm not sure that the test would still catch the bug (on
   a broken version of gdb) if pthread_cond_wait.c got modified in the
   figure.

-- i can't think of a catchier name either.

The substance of bp-in-thread.exp looks fine to me.

Michael Snyder is the maintainer of gdb.threads/, so whatever he says
overrides anything I say.

Michael C