This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
Re: patch to add QNX NTO i386 support
> From: "Kris Warkentin" <kewarken@qnx.com>
> Date: Wed, 12 Feb 2003 17:18:35 -0500
>
> Okay. I've got all our config files down to what's below. Not
> much left to them. Andrew had wanted me to submit separate native
> and remote patches but I can't figure out how to do that. We would
> never build a native only gdb - we always support the remote
> protocol as well - so I'm at a loss for how to get it to NOT
> compile the remote-nto.c file.
>
> I think we'd want the remote stuff to be optional. One way to make it
> such is having a configure option, say --enable-remote-nto, to enable
> it (which would be the default on QNX NTO systems, where you'd be able
> to use --disable-remote-nto to disable it). GDB should compile either
> way (but the remote-nto stuff probably wouldn't be of any use except
> for connecting to QNX NTO systems).
How would this work? If the option were supplied would I append
remote-qnx.o to TDEPFILES?
> As you can see, I've got the common stuff separated into nto-tdep.c and
> i386-nto-tdep.c which will be used by both remote-nto.c and
nto-procfs.c.
> The problem seems to be that when I build a native debugger, it
includes the
> stuff from i386nto.mt as well. I'm thinking that I want NATDEPFILES
and
> TDEPFILES to both have the *nto-tdep in them but then where does the
> remote-nto.c go? Any help?
>
> Sorry I'm not following you here. There's no point in listing files
> both in NATDEPFILES and TDEPFILES, since a native GDB by definition
> also includes the target stuff for the host its being built for.
Yes. Daniel J. explained this to me.
> I'm missing config/tm-qnxnto.h from the files you listed. And what
> purpose fulfills the tm-nto.h file you listed? Anyway, show me the
> complete target-stuff and let's get that integrated first!
The missing tm-qnxnto.h was an oversight - not necessary for the question I
was asking. The tm-nto.h file is necessary because configure will not build
a native debugger without it, even if it's empty. I'm trying to solve a few
annoying bugs (SOLIB_BKPT_NAME and SVR$_EXEC_EMU) first before I submit.
cheers,
Kris