Why sim/boot in -test4?

$ cvs diff -rLINUS_260_TEST3 -rLINUS_260_TEST4 arch/ia64/Makefile 
[...]
-boot := arch/ia64/boot
+boot := arch/ia64/hp/sim/boot

Why was this change made? It's not causing me any problems, just a bit of confusion when I noticed my previous vmlinux.gz wasn't getting updated any more.

"It's not Hollywood.  War is real, war is primarily not about defeat or
victory, it is about death.  I've seen thousands and thousands of dead bodies.
Do you think I want to have an academic debate on this subject?" -- Robert Fisk

Matthew> $ cvs diff -rLINUS_260_TEST3 -rLINUS_260_TEST4 arch/ia64/Makefile Matthew> [...]
Matthew> -boot := arch/ia64/boot
Matthew> +boot := arch/ia64/hp/sim/boot

Matthew> Why was this change made? It's not causing me any Matthew> problems, just a bit of confusion when I noticed my Matthew> previous vmlinux.gz wasn't getting updated any more.

It's just a general cleanup. The bootloader ended up never being used for anything other than the simulator, so moving it to hp/sim makes sense.

Are you saying "make compressed" doesn't build vmlinux.gz anymore? It seems to work for me.

	--david

Oh, it builds it, it just builds it inside the sim directory. That's not actually a problem, it just made me think there might be a problem. Here's the messages:

LD vmlinux
OBJCOPY arch/ia64/hp/sim/boot/vmlinux.bin
GZIP arch/ia64/hp/sim/boot/vmlinux.gz
LN vmlinux.gz
Kernel: vmlinux.gz is ready

Since it used to do

OBJCOPY arch/ia64/boot/vmlinux.bin
GZIP arch/ia64/boot/vmlinux.gz
LN vmlinux.gz

I thought I'd broken something with my other makefile change.

By the way, -test3 used to take a really really long time to link each step, and -test4 doesn't. I don't know whether this was an intentional fix, but I'm certainly happy to see it.

"It's not Hollywood.  War is real, war is primarily not about defeat or
victory, it is about death.  I've seen thousands and thousands of dead bodies.
Do you think I want to have an academic debate on this subject?" -- Robert Fisk

Matthew> By the way, -test3 used to take a really really long time Matthew> to link each step, and -test4 doesn't. I don't know Matthew> whether this was an intentional fix, but I'm certainly Matthew> happy to see it.

I don't think it's a fix per se, but rather a new config option: it used to be that I turned on compilation with "-g" unconditionally (sorry, but I hate not having debug info around when you need it...). Now Sam (finally! ;-) added a CONFIG_DEBUG_KERNEL option, so if you didn't turn this on, the kernel will be compiled without debug option and the resulting object files will be a lot smaller.

BTW: someone really ought to look into why the debug info is so big and what could be done to handle it more sanely. For example, we get tons of debug-info replication due to header-files etc. It's rather embarrassing how low the state of the art is w.r.t. debug info management...

	--david