Date: Sun, 07 Jan 2001 00:51:54 +0200 From: "Eli Zaretskii" Sender: halo1 AT zahav DOT net DOT il To: "Tim Van Holder" Message-Id: <9743-Sun07Jan2001005154+0200-eliz@is.elta.co.il> X-Mailer: Emacs 20.6 (via feedmail 8.3.emacs20_6 I) and Blat ver 1.8.6 CC: djgpp-workers AT delorie DOT com In-reply-to: Subject: Re: Backslashes in debug info References: Reply-To: djgpp-workers AT delorie DOT com Errors-To: nobody AT delorie DOT com X-Mailing-List: djgpp-workers AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk > From: "Tim Van Holder" > Date: Sat, 6 Jan 2001 23:08:12 +0100 > > > If you are talking about changes added a few months ago, they are > > mine, I think. > In that case: hey, you missed a spot ;-) It's possible. But I rather think that the problems, at least in part, are due to the BFD assembler, which is new to Binutils 2.10. My patches were all done a year ago, when BFD assembler was disabled; I never looked at it. (That it took so many time to get my patches into the CVS tree is another matter.) > > Where did you look? in the "gcc -S" output? That's not where I > > looked. > I checked both the assembly and the object file. Did you look in the executable with objdump? > > Breakpoint 1, main (argc=1, argv=0xfff334a0) at c:\dir1\dir2\d:5 > > 5 c:\dir1\dir2\d: No such file or directory (ENOENT) > Hmm - then the 14-char thing may be a gdb limit, unrelated to > binutils. I don't think so: try "objdump -d" on the executable, and you will see the same truncated file names. So it might be a BFD thing (since GDB uses BFD, at least to some extent). > Reading symbols from H:/Tmp/Source/web2c-7.3.3/texk/web2c/web2c/a.out...done. This is already a non-standard message. I wonder what kind of GDB is this... > [GDB] br main ??? "[GDB]" in brackets and in UPPER case? Do you have some gdb.ini file where you customize GDB's prompt? > Breakpoint 1 at 0x1d73: file h:\Tmp\Source\web2c-7.3.3\texk\web2c\web2c\foo.c, line 4. I bet you have some newer version of Binutils. If so, we cannot really compare what we see. I always try to use what the bulk of DJGPP users have, otherwise I cannot help them on c.o.m.d.