RFC: example fix to pr 2360

Daniel Jacobowitz drow@false.org
Thu Dec 20 20:23:00 GMT 2007


On Thu, Dec 20, 2007 at 11:37:13AM -0800, Doug Evans wrote:
> On Dec 16, 2007 4:01 PM, Daniel Jacobowitz <drow@false.org> wrote:
> > Are there any examples of this without #line?
> 
> Thanks for the review.  I'm not aware of any examples without #line.
> 
> While dwarf doesn't understand inodes the paths do have some canonical
> behaviour.  e.g. As I understand it, posix says //tmp/foo == /tmp/foo.
>  I recognize the paths are from the compilation host not the debug
> host, we could write a path canonicalizer.

Actually, I believe this is one of the few places where / and // are
separate, and POSIX allows //tmp to be different from /tmp.  /tmp//foo
is another story.

> But no matter, I can certainly pursue a different fix for the problem
> at hand, perhaps by adding an option to gcc to set DW_AT_comp_dir.

Does -fworking-directory help?  That's the default if you preprocess
with -g, if I understand correctly.

-- 
Daniel Jacobowitz
CodeSourcery



More information about the Gdb-patches mailing list