Skip to contentSkip to conversation

lib/Makefile

upatch Peter Samuelson 2 participants 3 messages Apr 23, 2001 – Apr 24, 2001 Library code

lore
Share & tools
lore
#0 lore raw

Introduced in 2.4.4pre4, I believe. $(export-objs) need not be conditional, and the if statement was not really correct either, although in this case it probably worked.

Peter

lib/Makefile renamed from lib/Makefile~ +3 -5
@@ -8,14 +8,12 @@
 
 L_TARGET := lib.a
 
-export-objs := cmdline.o
+export-objs := cmdline.o rwsem.o
 
 obj-y := errno.o ctype.o string.o vsprintf.o brlock.o cmdline.o
 
-ifneq ($(CONFIG_RWSEM_GENERIC_SPINLOCK)$(CONFIG_RWSEM_XCHGADD_ALGORITHM),nn)
-export-objs += rwsem.o
-obj-y += rwsem.o
-endif
+obj-$(CONFIG_RWSEM_GENERIC_SPINLOCK)	+= rwsem.o
+obj-$(CONFIG_RWSEM_XCHGADD_ALGORITHM)	+= rwsem.o
 
 ifneq ($(CONFIG_HAVE_DEC_LOCK),y) 
   obj-y += dec_and_lock.o
#1 lore raw
Quoted 3 lines · Peter Samuelson

Introduced in 2.4.4pre4, I believe. $(export-objs) need not be conditional, and the if statement was not really correct either, although in this case it probably worked.

Er, are you sure changing the test for !"nn" is correct here? I _think_ at least that is intentional and correct (since you can have one on but not the other).

Signature · 2 lines
Tom Rini (TR1265)
http://gate.crashing.org/~trini/
#2 lore raw
In reply to Peter Samuelson

[Peter Samuelson]

Quoted 3 lines
Quoted 3 lines

Introduced in 2.4.4pre4, I believe. $(export-objs) need not be conditional, and the if statement was not really correct either, although in this case it probably worked.

[Tom Rini]

Quoted 3 lines

Er, are you sure changing the test for !"nn" is correct here? I _think_ at least that is intentional and correct (since you can have one on but not the other).

I understand the intent. The point is that in our current makefiles you are not allowed to assume that a negative value is "n", because it could be "".

2.4.4pre probably works because each 'config.in' file explicitly sets these variables to "y" or "n". However, it would be perfectly legal for a config.in to unset the variable, or for that matter not even mention it. In that case the !"nn" test fails.

This is the same reason you cannot use 'ifdef CONFIG_*' in the Makefiles. Lots of people do, but each instance is a bug. They are assuming the opposite: that a variable will be "" rather than "n".

(N.B. this issue will go away with Keith's 2.5 Makefiles. And there was much rejoicing.)

Peter