Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm Sender: cygwin-owner AT sourceware DOT cygnus DOT com Delivered-To: mailing list cygwin AT sourceware DOT cygnus DOT com Message-Id: <199902211755.LAA01815@modi.xraylith.wisc.edu> X-Authentication-Warning: modi.xraylith.wisc.edu: localhost.xraylith.wisc.edu [127.0.0.1] didn't use HELO protocol To: Christopher Faylor cc: cygwin AT sourceware DOT cygnus DOT com Subject: Re: gdb failure In-reply-to: Your message of "Sun, 21 Feb 1999 12:20:01 EST." <19990221122001 DOT A7097 AT cygnus DOT com> Date: Sun, 21 Feb 1999 11:55:05 -0600 From: Mumit Khan Christopher Faylor writes: > On Sun, Feb 21, 1999 at 12:42:45AM -0600, Mumit Khan wrote: > >Let me apologize on behalf of all who support win32 for the lack of > >support in gdb for debugging template instantiations. We need to get > >this fixed, but nobody has volunteered yet. > > Is this a win32 bug or just a generic gdb problem? This is the first > that I can recall hearing about this problem. Actually, there've been a noises about this, but not that much. It was a combination bug, in egcs debugging output (almost fixed) and something in gdb (don't have a clue, and it may very well be another egcs stabs bug). Here's a reference bug report from Xiaomao Xiao : http://www.cygnus.com/ml/gnu-win32/1998-Oct/0232.html Good news is that major problem has been fixed in egcs-1.1.2, but there's still a few nits I just can't seem to track down. With egcs-1.1.2, you can step into template instances just fine, but just don't try to step into a template instance in a library stripped of debugging info. Normally, gdb is smart enough to just step "over" these, but the cygwin version tends to say something like "Cannot access memory at address 0x2000000". I believe egcs-1.1.2 solution will be sufficient for 90% of the time. Regards, Mumit -- Want to unsubscribe from this list? Send a message to cygwin-unsubscribe AT sourceware DOT cygnus DOT com