Re: Major debugging breakage

[law at redhat dot com]

  > Something has been hosed with line numbers for a long time; Geoff
  > seemed to think he knew what it was. 
Yea, it was broken before gcc-3.0, so it's been awhile.  I've cc'd Geoff
directly in the hopes that he'll chime in.


  > I do know that what CPP tells
  > the front end is definitely correct; so something, from
  > cb_file_change() in c-lex.c down to the guts of GCC, is not using
  > "lineno" correctly or getting mucked up for some reason.  Sadly, my
  > knowledge of such parts of GCC is not good enough to suggest anything
  > constructive.
Every time we parse a token we update LINENO to be equal to SRC_LINENO
(c_lex.c).   LINENO is used to initialize the line numbers in NOTEs
that are used by the debug symbol routines.  The comments in cb_line_change
seem to indicate that SRC_LINENO is meant to be used for diagnostic
line numbers.  Maybe that is the root of our problem -- mixing the two
purposes.

SRC_LINENO is updated at the start of every non-empty line via cb_line_change.


  > I can try and trace your example through GCC, but I have much less
  > time for GCC than I used to (baby...).
I understand.  I'll understand even more in a few months.


  > There are still a couple of
  > outstanding CPP things that need to be done (Cygwin mmap and the
  > driver external CPP thing that I've not got round to) which I'd rather
  > do first, seeing as how this line number thing has been around for
  > ages, and others with better understanding than me have not been able
  > to find it when they gave it a (quick?) go.
The Cygwin mmap issues are pretty serious too, so attacking that problem
makes a lot of sense.  I'd like to see the debug problems solved before
the driver external CPP thingie, but that's just my preference.

jeff