On Thu, 2007-05-17 at 02:10 +0800, David Woodhouse wrote:
> The reason I say 'maybe a little less' is because I concentrated on
> modules at the time because the Kbuild stuff was relatively easy for
> those -- and I have a vague recollection that the built-in stuff didn't
> show as big a difference. Searching for 'built-in.o' in the above-linked
> CSV file seems to confirm that recollection.
Actually I'm full of crap. The reason we don't see much improvement in
built-in.o in that list because I didn't _do_ it for built-in.o yet --
it's still only happening for the few multi-obj .o files which get
linked into built-on.o (e.g. mounts-y in init/, etc.).
I lack the wit to make that one work right now. It might look something
like this...
Actually, built-in.o probably ought to be just another case of
$(multi-obj-y). And I need to filter out the non-C sources of those too,
to fix md and crypto stuff (which uses .S files) so I don't need to add
'CONFIG_COMBINED_COMPILE=' to their makefiles.
ifdef CONFIG_COMBINED_COMPILEx
builtin-cfiles-y := $(wildcard $(obj-y:.o=.c))
multi-objs-y += $(obj)/builtin-cfiles.o
multi-used-y += $(obj)/builtin-cfiles.o
$(builtin-target): $(filter-out $(builtin-cfiles-y:.c=.o),$(obj-y))
$(obj)/builtin-cfiles.o FORCE
$(call if_changed,link_o_target)
else
$(builtin-target): $(obj-y) FORCE
$(call if_changed,link_o_target)
endif # CONFIG_COMBINED_COMPILE
--
dwmw2
-
To unsubscribe from this list: send the line "unsubscribe linux-arch" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
