Re: [rfa/symtab] Move find_pc_section call to lookup_minimal_symbol_by_pc

[Elena Zannoni <ezannoni at redhat dot com>]

Michael Elizabeth Chastain writes:
 > The filenames are a problem because gdb1237-1280-main.c
 > and gdb1237-1280.c are not 8.3 unique.
 > 
 > You either have to make them 8.3 unique (preferred) or
 > add lines to config/djgpp/fnchange.lst.
 > 

I created a new PR to include both problems and use that new number. PR/1555.

 > For the kfail's, please use more specific patterns than just
 > ".*$gdb_prompt $".  kfail's should be for known bugs, which means
 > there should be known-bad output.  If the output doesn't match a
 > specific known-bad output, then it means that gdb has regressed
 > even more, and the KFAIL turns into a FAIL (this does happen).
 > 

I am not sure what to do here, except start adding patterns.

 > You don't need to call get_compiler_info unless you use one of
 > these functions or data symbols later: test_compiler_info,
 > gcc_compiled, hp_cc_compiler, hp_aCC_compiler, signed_keyword_not_used.
 > 

it is using gcc_compiled to figure out the flags to build the shlibs.

 > I missed the beginning of the thread, but it helps to say what
 > system(s) you tested it on (so when it breaks on a different system
 > later, it makes it easier to figure out what is wrong ... morbid
 > but true).
 > 

Adam?

I'd like to close this thread and get the test in.

elena