This is the mail archive of the gdb-testers@sourceware.cygnus.com mailing list for the GDB project. See the GDB home page for more information.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
This might be caused by the stabs added via your perl script, which might trick GDB into believing that there is a function with a prologue at _start, in which case GDB tries to set the breakpoint past the prologue. You should be able to set the desired breakpoint via `b *_start', please let me know if that doesn't work. The `breakpt lossage previously reported' applies to 4.16.85 only, 980122 doesn't have the problem. > This is observed on a gdb-980122 ( --host=i386-unknown-netbsd1.3 > --target=powerpc-foo-eabi) > > Breakpoint 1, _start () at vectors.s:34 > Current language: auto; currently asm > (gdb) i b > Num Type Disp Enb Address What > 1 breakpoint keep y 0x0004017c <_start+8 at vectors.s:34> > breakpoint already hit 1 time > 2 breakpoint keep y 0x00042940 in mainprog at cold.c:14 > (gdb) p _start > $2 = {<text variable, no debug info>} 0x40174 <_start at vectors.s:32> > (gdb) b _start > Note: breakpoint 1 also set at pc 0x4017c. > Breakpoint 3 at 0x4017c > (gdb) > > Notice start is at 0x40174, yet gdb insists on putting the breakpoint > at 0x4017c . This may well be the same breakpt lossage as previously > reported. If so, please add my vote for being able to breakpoint > arbitrary textspace addresses. > > BTW. Is there an automatic tool for adding stabs to assembly files so > one can add line numbers (like gcc -g would do for C code.) I'm using > a perl script that I hacked together to do this now, but I can't > imagine I'm the only person that likes to see their source code step > by, even if it happens to be assembly. The comments in the file are > still useful to see as one single steps. > > Here is what I'm using right now. Its not pretty and I might well be > misunderstanding the full syntax for gcc's stabs, but it seems to do > enough to let me single step. > http://www.wsrcc.com/wolfgang/ftp/asmstabs.txt > > -wolfgang > > -- Peter Schauer pes@regent.e-technik.tu-muenchen.de