problem cross-compiling 9.1

[Warner Losh]

On Oct 3, 2012, at 3:41 AM, Daniel Braniss wrote:

> reposting to hackers, maybe better luck here?
> 
> When using an amd64 host to 'make TARGET_ARCH=i386 buildworld' it seems that
> it's using the wrong cpp, at least when building ioctl.c via mkioctl in
> usr.bin/ktrace and having set WITHOUT_CPP(*). This used to work with
> previous releases.
> 
> ...
> ===> usr.bin/kdump (depend)
> env CPP="cpp"  sh /r+d/stable/9/usr.bin/kdump/mkioctls 
> /home/obj/rnd/alix/i386.i386/r+d/stable/9/tmp/usr/include > ioctl.c
> ...
> cc -O2 -pipe  -I/r+d/stable/9/usr.bin/kdump/../ktrace 
> -I/r+d/stable/9/usr.bin/kdump -I/r+d/stable/9/usr.bin/kdump/../.. -DNDEBUG 
> -std=gnu99 -fstack-protector -Wno-pointer-sign -c ioctl.c
> ioctl.c: In function 'ioctlname':
> ioctl.c:1216: error: 'MPTIO_RAID_ACTION32' undeclared (first use in this 
> function)
> ioctl.c:1216: error: (Each undeclared identifier is reported only once
> ioctl.c:1216: error: for each function it appears in.)
> ioctl.c:1292: error: 'MPTIO_READ_EXT_CFG_HEADER32' undeclared (first use in 
> this function)
> ioctl.c:1632: error: 'MPTIO_READ_EXT_CFG_PAGE32' undeclared (first use in this 
> function)
> ioctl.c:1772: error: 'CCISS_PASSTHRU32' undeclared (first use in this function)
> ioctl.c:2010: error: 'IPMICTL_RECEIVE_MSG_TRUNC_32' undeclared (first use in 
> this function)
> ioctl.c:2082: error: 'IPMICTL_RECEIVE_MSG_32' undeclared (first use in this 
> function)
> ioctl.c:2300: error: 'MPTIO_READ_CFG_PAGE32' undeclared (first use in this 
> function)
> ioctl.c:2870: error: 'MPTIO_READ_CFG_HEADER32' undeclared (first use in this 
> function)
> ioctl.c:2878: error: 'IPMICTL_SEND_COMMAND_32' undeclared (first use in this 
> function)
> ioctl.c:2938: error: 'MPTIO_WRITE_CFG_PAGE32' undeclared (first use in this 
> function)
> *** [ioctl.o] Error code 1
> 
> *: Im compiling for an embedded system, and hence I want the minimum stuff.
> 
> at the moment the work around is to remove the WITHOUT_CPP, but it got me
> worried.
> 
> any fix?

You have found the fix.  Remove the WITHOUT_XXXX options from the build that keep it from completing.  You'll be able to add them at installworld time w/o a hassle.  nanobsd uses this to keep things down, while still being able to build the system.

Warner