Subject: Anybodt get gre working with 1.5 or 1.5.1 kernel source?
To: None <port-amiga@netbsd.org>
From: Rob Healey <rhealey@norstar.com>
List: port-amiga
Date: 07/10/2001 15:01:54
	I have been having a problem trying to get netinet/ip_gre.c to
	compile for a while, specifically:

cc  -Werror -Wall -Wstrict-prototypes -Wmissing-prototypes  -Wpointer-arith -Wno-uninitialized -Wno-main -O2 -m68020 -msoft-float -I. -I../../../../arch -I../../../.. -nostdinc -DDRACO -DM68030 -DM68020 -DFPU_EMULATE -DLKM -DDIAGNOSTIC -DSERIBUF_SIZE="0x1000" -DSEROBUF_SIZE="0x20" -DMAXUSERS=16 -D_KERNEL -Damiga -DFPCOPROC  -c ../../../../netinet/ip_gre.c
cc1: warnings being treated as errors
../../../../netinet/ip_gre.c: In function `gre_input2':
../../../../netinet/ip_gre.c:194: warning: implicit declaration of function `setsoftnet'

	This error is from the schednetisr() macro which appears to define
	a schednetisr() macro which is not being defined.

	It looks like maybe the machine/cpu.h file should be included to
	haul in the proper macro definitions but why does the Amiga port need
	this when the other arch's seem to compile the file just fine
	without it?

	I happen to use AppleTalk so just undefining that to get rid of the
	schednetisr() invokation isn't going to help...

	Thoughts?

	This is on a fresh 1.5.1 install on a newfs'd partition so no left
	over droppings from old NetBSD's exist.

	I noticed GENERIC has gre commented out, is it because it no workee?

	-Rob