On 08/30/2011 07:02 AM, Lluís wrote:
Stefan Hajnoczi writes:
On Mon, Aug 29, 2011 at 8:27 PM, Michael Roth<mdr...@linux.vnet.ibm.com> wrote:
@@ -380,7 +381,6 @@ else
trace-obj-y = trace.o
ifeq ($(TRACE_BACKEND),simple)
trace-obj-y += simpletrace.o
-user-obj-y += qemu-timer-common.o
endif
endif
Now that we have a concrete patch to look at I think this approach is
problematic. There are several subsystems in QEMU which might be
built outside the main qemu binary for qemu-io, qemu-img, qemu-ga,
etc.
[...]
If QEMU is split up into libraries then having an explicit list of
dependencies for each subsystem will be very useful, whereas the
CONFIG_* approach doesn't collect that information in one place.
So I think explicit subsys-obj-y += qemu-timer-common.o together with
$(sort) during the link stage actually allows for a cleaner build
system. I prefer that approach.
I couldn't agree more. The only problem I see with '$(sort)' is that it
will invariably change the order of object files, which can influence
code placement.
Whether or not the spatial locality among compilation units is
important, I don't know. Although I believe it won't have much of a
performance penalty.
In any case, I tried to find a straightforward way of filtering-out
repeated words in a list with make, but couldn't find any solution other
than '$(sort)' or calling an external command with '$(shell)'.
Hmm, looking again I'm confused why we need to do this in the first
place...the rule is:
LINK = $(call quiet-command,$(CC) $(QEMU_CFLAGS) $(CFLAGS) $(LDFLAGS) -o
$@ $(1) $(LIBS)," LINK $(TARGET_DIR)$@")
%$(EXESUF): %.o
$(call LINK,$^)
According to the documentation $^ should remove duplicate dependencies,
so I'm not getting why we need to de-dupe them once they get passed to LINK:
http://www.gnu.org/software/make/manual/make.html#index-g_t_0024_005e-948
Is there some distinction between duplicate dependencies and duplicate
words in the dependency list?
Lluis
