[M3commit] CVS Update: cm3

[Jay Krell]

CVSROOT:	/usr/cvs
Changes by:	jkrell at birch.	12/06/15 11:33:33

Modified files:
	cm3/m3-sys/m3cc/gcc/gcc/m3cg/: parse.c 
	./: parse.c 
	cm3/m3-sys/m3cc/gcc-4.7/gcc/: calls.c fold-const.c gimple.c 
	                              gimplify.c targhooks.c tree-cfg.c 
	                              tree-dump.c tree-nested.c 
	                              tree-pretty-print.c tree.def 
	./: calls.c fold-const.c gimple.c gimplify.c targhooks.c 
	    tree-cfg.c tree-dump.c tree-nested.c tree-pretty-print.c 
	    tree.def 
	cm3/m3-sys/m3cc/gcc-4.7/gcc/config/alpha/: alpha.c 
	./: alpha.c 
	cm3/m3-sys/m3cc/gcc-4.7/gcc/config/i386/: i386.c 
	./: i386.c 
	cm3/m3-sys/m3cc/src/: clean_marker.txt 

Log message:
	move I386_DARWIN and AMD64_DARWIN to gcc 4.7 backend (4.7.0)
	I disable the sra optimization. It doesn't play nicely with our bitfield abuse.
	I allow flag_tree_forwprop that I disabled with 4.5.
	
	The following platforms successfully cross build a bootstrap:
	
	ALPHA{32,64}_VMS
	ALPHA_OSF
	AMD64_{DARWIN,FREEBSD,LINUX,NETBSD,SOLARIS}
	I386_{CYGWIN,DARWIN,FREEBSD,INTERIX,LINUX,NETBSD,OPENBSD,SOLARIS}
	PPC64_DARWIN
	PPC_{DARWIN,LINUX}
	SPARC32_{LINUX,SOLARIS}
	SPARC64_{LINUX,OPENBSD,SOLARIS}
	
	I will move some more to 4.7 upon testing, others without testing.
	
	Optimizer is looking good.
	We disable scalar replacement of aggregate (sra).
	It doesn't like bit_field_refs I think.
	
	ARM_DARWIN is deliberately held back at 4.2 -- where Apple forked
	(but I'll see about merging their changes)
	
	OpenBSD patches are yet to be applied, which will fix AMD64_OPENBSD
	and maybe others.
	
	MIPS* are yet broken. I'll get to that.
	Nobody uses them.
	MIPS64EL_{OPENBSD,LINUX} do run on recent modern hardware though.