[Simulavr-devel] Re: Build issues with current CVS

[Joerg Wunsch]

As Joerg Wunsch wrote:

> > After configuring, when I call "gmake", I get:
> > 
> > gmake: *** No rule to make target address@hidden@', needed by 
> > `Makefile.in'.  Stop.
> 
> As I could not see why/how this might have come in, I started
> over with a fresh CVS checkout.  Now, bootstrap does no longer
> appear to even create a Makefile.in, so configure eventually
> stops with:
> 
> configure: creating ./config.status
> config.status: creating Makefile
> config.status: error: cannot find input file: Makefile.in

This is indeed caused by these messages.  Apparently, they are
*not* just warnings but fatal errors, and automake stops at
that point:

src/Makefile.am:138: patsubst %.lo,%.o,$(libavrsim_pp_la_OBJECTS: non-POSIX 
variable name
src/Makefile.am:138: (probably a GNU make extension)
src/Makefile.am:148: `:='-style assignments are not portable
src/Makefile.am:149: `:='-style assignments are not portable

If I change the := into a =, and remove the offending line 138
completely (because I cannot come up with a quick non-gmake-ish
replacement for it, I'm getting further.  Obviously, it cannot build
the Python bindings (due to the removed command), but it does build
the simulavrxx standalone binary again at least.

-- 
cheers, J"org               .-.-.   --... ...--   -.. .  DL8DTL

http://www.sax.de/~joerg/                        NIC: JW11-RIPE
Never trust an operating system you don't have sources for. ;-)