This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: When to emit C++ debug info?
Daniel Jacobowitz <drow@mvista.com> writes:
| On Mon, Jan 05, 2004 at 09:03:13PM +0100, Gabriel Dos Reis wrote:
| > Daniel Jacobowitz <drow@mvista.com> writes:
| >
| > | > We can build on the cfront virtual function table heuristic: only emit
| > | > debug information for a class in the unit that implements the virtual
| > | > function table. "But someone might give me a library with no debug
| > | > information". Fine, we can have a flag that causes full debug info
| > | > to be emitted to cover that case, but it should not be the default.
| > |
| > | I believe the summary of my arguments at this point is, the use of C++
| > | libraries without debug info is a great deal more common than GCC
| > | developers believe
| >
| > Well, I disagree with point. It has been a failing point for some
| > libraries like V3 for a long time. Other implementers (with larger
| > user base community) have managed to supply debugged libraries years
| > before us. And thanks to the recent work on ddebug mode by Benjamin,
| > Doug et al. we're hopefully going to catch up. That we've only
| > "discovered" debugged libraries recently is probably a sign of how
| > "ghettorized" we've gotten :-/
|
| Debug mode has nothing to do with this problem; we could supply a
| libstdc++ with debug mode just as well beforehand, so I'm not sure what
| you mean.
Well, I don't agree with the point that use of debug info for
libraries is rare. I mentioned the debug mode for libstdc++ just as
*an* example of how late we're in catching up. Of course, we can ship
different versions of V3, with or without debug info; but that is not
the point. The point is that we have to recognize of the importance
of the issue and try to come up with something that is not just a
special case (e.g. providing different version of V3).
-- Gaby