Skip to contentSkip to conversation

offsets.h generation

Linux-ia64 Applied Peter Chubb 3 participants 12 messages May 19, 2003 – May 31, 2003

lore
Share & tools
lore
#0 lore raw

At present, if you change something that means offsets.h should be regenerated, the automatic dependencies mechanism doesn't do it.

Herewith a patch that adds proper dependency generation for
printoffsets.s (and hence offsets.h)

# This is a BitKeeper generated patch for the following project:
# Project Name: Linux kernel tree
# This patch format is intended for GNU patch command version 2.5 or higher. # This patch includes the following deltas:
# ChangeSet 1.1091 -> 1.1092
# arch/ia64/Makefile 1.48 -> 1.49
# arch/ia64/tools/Makefile 1.12 -> 1.13
#
# The following is the BitKeeper ChangeSet Log
# --------------------------------------------
# 03/05/19 [email protected] 1.1092
# [IA64] Add dependencies so that offsets.h gets regenerated if it has to be. # --------------------------------------------
#

arch/ia64/Makefile +3 -1
diff -Nru a/arch/ia64/Makefile b/arch/ia64/Makefile
@@ -86,10 +86,12 @@
 
 prepare: include/asm-ia64/offsets.h
 
+include/asm-ia64/offsets.h : arch/ia64/tools/offsets.h
+
 boot:	lib/lib.a vmlinux
 	$(Q)$(MAKE) $(build)=$(boot) $@
 
-include/asm-ia64/offsets.h:
+arch/ia64/tools/offsets.h include/asm-ia64/offsets.h:
 	$(Q)$(MAKE) $(build)=$(tools) $@
 
 define archhelp
arch/ia64/tools/Makefile +4 -3
diff -Nru a/arch/ia64/tools/Makefile b/arch/ia64/tools/Makefile
@@ -32,8 +32,7 @@
 
 $(obj)/print_offsets: $(src)/print_offsets.c FORCE
 	[ -r $(TARGET) ] || echo "#define IA64_TASK_SIZE 0" > $(TARGET)
-	$(CC) $(CFLAGS) -DKBUILD_BASENAME=$(subst $(comma),_,$(subst -,_,$(*F))) \
-		$(src)/print_offsets.c -o $@
+	$(CC) $(c_flags) $(src)/print_offsets.c -o $@
 
 FORCE:
 
@@ -44,7 +43,9 @@
 
 $(obj)/print_offsets.s: $(src)/print_offsets.c
 	[ -r $(TARGET) ] || echo "#define IA64_TASK_SIZE 0" > $(TARGET)
-	$(CC) $(CFLAGS) -DKBUILD_BASENAME=$(subst $(comma),_,$(subst -,_,$(*F))) -S $^ -o $@
+	$(CC) $(c_flags) -MT $@ -S $^ -o $@
+	scripts/fixdep $(depfile) $@ "$(CC) $(c_flags) -MT $@ -S $^ -o $@" > $(obj)/.print_offsets.s.cmd
+	rm $(depfile)
 
 endif
 
#1 lore raw

On Mon, May 19, 2003 at 02:30:57PM +1000, Peter Chubb wrote:
Content-Description: message body text

Quoted 6 lines

At present, if you change something that means offsets.h should be regenerated, the automatic dependencies mechanism doesn't do it.

Herewith a patch that adds proper dependency generation for
printoffsets.s (and hence offsets.h)

I agree with the intention, but not the solution.
Other architectures generate a .s file from a .c file, and filechk_gen-asm-offsets in the top-level Makefile is used to retreive the asm offset parameters.

In this way you use the kbuild mecanishm to find dependencies, and ia64 shift to do it the same way as other architectures. See i386, sparc and other for a template how to do it. Note that most of the architectures list too much dependencies in the arch specific makefile. I will clean that up sometime.

Would you mind trying to look into this solution?

Sam

#2 lore raw
In reply to Peter Chubb
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Sam" == Sam Ravnborg <[email protected]> writes:

Sam> On Mon, May 19, 2003 at 02:30:57PM +1000, Peter Chubb wrote:
Sam> Content-Description: message body text

Quoted 5 lines
Quoted 5 lines

At present, if you change something that means offsets.h should be regenerated, the automatic dependencies mechanism doesn't do it.

Herewith a patch that adds proper dependency generation for printoffsets.s (and hence offsets.h)

Sam> I agree with the intention, but not the solution. Other Sam> architectures generate a .s file from a .c file, and Sam> filechk_gen-asm-offsets in the top-level Makefile is used to Sam> retreive the asm offset parameters.

OK, here's a patch that does it that way. Because some of the needed header files in turn depend on a symbol from offsets.h, I had to create a dummy one `by hand' if it didn't previously exist. A timestamp file, include/asm-ia64/.offsets.h.stamp, tracks when this was done so that 'make' after 'make clean' works; and 'make' after 'make' works.

# This is a BitKeeper generated patch for the following project:
# Project Name: Linux kernel tree
# This patch format is intended for GNU patch command version 2.5 or higher. # This patch includes the following deltas:

#	           ChangeSet	1.1091  -> 1.1092 
#	  arch/ia64/Makefile	1.48    -> 1.49   
#	               (new)	        -> 1.1     arch/ia64/kernel/asm-offsets.c

#
# The following is the BitKeeper ChangeSet Log
# --------------------------------------------
# 03/05/20 [email protected] 1.1092
# Generate include/asm/offsets.h in a way compatible with the other architectures, so that
# dependencies are maintained correctly.
# --------------------------------------------
#

arch/ia64/Makefile +12 -6
diff -Nru a/arch/ia64/Makefile b/arch/ia64/Makefile
@@ -66,7 +66,6 @@
 drivers-$(CONFIG_IA64_GENERIC)	+= arch/ia64/hp/common/ arch/ia64/hp/zx1/ arch/ia64/hp/sim/
 
 boot := arch/ia64/boot
-tools := arch/ia64/tools
 
 .PHONY: boot compressed check
 
@@ -80,17 +79,24 @@
 
 archclean:
 	$(Q)$(MAKE) $(clean)=$(boot)
-	$(Q)$(MAKE) $(clean)=$(tools)
 
-CLEAN_FILES += include/asm-ia64/offsets.h vmlinux.gz bootloader
+CLEAN_FILES += include/asm-ia64/offsets.h include/asm-ia64/.offsets.h.stamp vmlinux.gz bootloader
 
-prepare: include/asm-ia64/offsets.h
+prepare: include/asm-ia64/offsets.h 
+arch/$(ARCH)/kernel/asm-offsets.s: include/asm include/linux/version.h \
+				   include/config/MARKER include/asm-$(ARCH)/.offsets.h.stamp
+
+include/asm-$(ARCH)/.offsets.h.stamp: 
+	[ -f include/asm-$(ARCH)/offsets.h ] || echo "#define IA64_TASK_SIZE 0" > include/asm-$(ARCH)/offsets.h
+	touch $@
+
+
+include/asm-$(ARCH)/offsets.h: arch/$(ARCH)/kernel/asm-offsets.s
+	$(call filechk,gen-asm-offsets)
 
 boot:	lib/lib.a vmlinux
 	$(Q)$(MAKE) $(build)=$(boot) $@
 
-include/asm-ia64/offsets.h:
-	$(Q)$(MAKE) $(build)=$(tools) $@
 
 define archhelp
   echo '  compressed	- Build compressed kernel image'
arch/ia64/kernel/asm-offsets.c new file +182 -0

182 lines not on this page · raw message

#3 lore raw
In reply to Peter Chubb
Quoted 2 lines · Peter Chubb

OK, here's a patch that does it that way.

Much better, thanks. A few more comments.
I assume that deletion of files in arch/ia64/tools will follow in seperate cset.

Quoted 6 lines

Because some of the needed
header files in turn depend on a symbol from offsets.h, I had to create a dummy one `by hand' if it didn't previously exist. A timestamp file, include/asm-ia64/.offsets.h.stamp, tracks when this was done so that 'make' after 'make clean' works; and 'make' after 'make' works.

David: Can we remove this circular dependency to get rid of this ugly workaround?

Review

Quoted 2 lines
(quoted diff) +1 -1
-CLEAN_FILES += include/asm-ia64/offsets.h vmlinux.gz bootloader
+CLEAN_FILES += include/asm-ia64/offsets.h include/asm-ia64/.offsets.h.stamp vmlinux.gz bootloader

Break in two lines.

Review

Quoted 5 lines
(quoted diff) +3 -1
-prepare: include/asm-ia64/offsets.h
+prepare: include/asm-ia64/offsets.h 
 
+arch/$(ARCH)/kernel/asm-offsets.s: include/asm include/linux/version.h \
+				   include/config/MARKER include/asm-$(ARCH)/.offsets.h.stamp

Dependency on include/asm include/linux/version.h and include/config/MARKER is not needed.
I would prefer to have dependencies for arch/$(ARCH)/kernel/asm-offsets.s located in kernel/Makefile from where the .s fil is actually build. Note that it is the general %.s : %.c rule from the top-level Makefile that kicks in and make sure the .s file get build when needed.

Review

Quoted 4 lines
(quoted diff) +4 -0
+
+include/asm-$(ARCH)/.offsets.h.stamp: 
+	[ -f include/asm-$(ARCH)/offsets.h ] || echo "#define IA64_TASK_SIZE 0" > include/asm-$(ARCH)/offsets.h
+	touch $@

So this part should go in kernel/MAkefile as well.

Quoted 1 line

+#include "../kernel/sigframe.h"

This include should be without "../kernel".

Thanks for cleaning this up.

Sam

#4 lore raw
In reply to Peter Chubb
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line · Peter Chubb

Peter> OK, here's a patch that does it that way. Because some of Peter> the needed header files in turn depend on a symbol from Peter> offsets.h, I had to create a dummy one `by hand' if it didn't Peter> previously exist. A timestamp file,
Peter> include/asm-ia64/.offsets.h.stamp, tracks when this was done Peter> so that 'make' after 'make clean' works; and 'make' after Peter> 'make' works.

Is processor.h the only cause of this recursive dependency? As far as I remember, we don't even need to include offsets.h in processor.h anymore. It used to be necessary for get_wchan(), but that's now an out-of-line function.

	--david
#5 lore raw
In reply to Peter Chubb
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

"David" = David Mosberger <[email protected]> writes:

On Tue, 20 May 2003 11:26:31 +1000, Peter Chubb <[email protected]> said:

Peter> OK, here's a patch that does it that way. Because some of the Peter> needed header files in turn depend on a symbol from offsets.h, Peter> I had to create a dummy one `by hand' if it didn't previously Peter> exist. A timestamp file, include/asm-ia64/.offsets.h.stamp, Peter> tracks when this was done so that 'make' after 'make clean' Peter> works; and 'make' after 'make' works.

David> Is processor.h the only cause of this recursive dependency? As David> far as I remember, we don't even need to include offsets.h in David> processor.h anymore. It used to be necessary for get_wchan(), David> but that's now an out-of-line function.

I'll try deleting that one tomorrow when I'm in at work... if we can get rid of it, everything becomes *much* cleaner.

Peter C

#6 lore raw
In reply to Peter Chubb

David> Is processor.h the only cause of this recursive dependency? As David> far as I remember, we don't even need to include offsets.h in David> processor.h anymore. It used to be necessary for get_wchan(), David> but that's now an out-of-line function.

Now there's also a dependency in thread_info.h that unfortunately can't be worked around, and in ptrace.h

Anyway, here's hopefully the final patch, against the IA64-2.5 BK tree as of this morning.

arch/ia64/Makefile +4 -5
diff -Nur --exclude=RCS --exclude=CVS --exclude=SCCS --exclude=BitKeeper --exclude=ChangeSet linux-ia64-2.5/arch/ia64/Makefile linux-ia64-2.5-wip/arch/ia64/Makefile
@@ -66,7 +66,6 @@
 drivers-$(CONFIG_IA64_GENERIC)	+= arch/ia64/hp/common/ arch/ia64/hp/zx1/ arch/ia64/hp/sim/
 
 boot := arch/ia64/boot
-tools := arch/ia64/tools
 
 .PHONY: boot compressed check
 
@@ -80,17 +79,17 @@
 
 archclean:
 	$(Q)$(MAKE) $(clean)=$(boot)
-	$(Q)$(MAKE) $(clean)=$(tools)
 
 CLEAN_FILES += include/asm-ia64/offsets.h vmlinux.gz bootloader
 
-prepare: include/asm-ia64/offsets.h
+prepare: include/asm-ia64/offsets.h
+
+include/asm-$(ARCH)/offsets.h: arch/$(ARCH)/kernel/asm-offsets.s
+	$(call filechk,gen-asm-offsets)
 
 boot:	lib/lib.a vmlinux
 	$(Q)$(MAKE) $(build)=$(boot) $@
 
-include/asm-ia64/offsets.h:
-	$(Q)$(MAKE) $(build)=$(tools) $@
 
 define archhelp
   echo '  compressed	- Build compressed kernel image'
arch/ia64/kernel/Makefile +8 -0
diff -Nur --exclude=RCS --exclude=CVS --exclude=SCCS --exclude=BitKeeper --exclude=ChangeSet linux-ia64-2.5/arch/ia64/kernel/Makefile linux-ia64-2.5-wip/arch/ia64/kernel/Makefile
@@ -18,3 +18,11 @@
 obj-$(CONFIG_IOSAPIC)		+= iosapic.o
 obj-$(CONFIG_MODULES)		+= module.o
 obj-$(CONFIG_SMP)		+= smp.o smpboot.o
+
+arch/$(ARCH)/kernel/asm-offsets.s: include/asm-$(ARCH)/.offsets.h.stamp
+
+include/asm-$(ARCH)/.offsets.h.stamp: 
+	[ -s include/asm-$(ARCH)/offsets.h ] || echo "#define IA64_TASK_SIZE 0" > include/asm-$(ARCH)/offsets.h
+	touch $@
+
+CLEAN_FILES += include/asm-ia64/.offsets.h.stamp
arch/ia64/kernel/asm-offsets.c +182 -0

182 lines not on this page · raw message

arch/ia64/tools/Makefile +0 -51

51 lines not on this page · raw message

arch/ia64/tools/print_offsets.awk +0 -72

72 lines not on this page · raw message

arch/ia64/tools/print_offsets.c +0 -220

220 lines not on this page · raw message

#7 lore raw
In reply to Peter Chubb
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line · Peter Chubb

Peter> Anyway, here's hopefully the final patch, against the Peter> IA64-2.5 BK tree as of this morning.

Argh, bad timing: I just commited Rohit's patch for the streamlined syscall path, which of course makes a number of changes to print_offsets.c ;-(

I'd push it right now, except that I need to twists bk to let me do it ("bk undo" just never works quite as well as I'd like...).

	--david
#8 lore raw
In reply to Peter Chubb

OK, here's a patch against this morning's BK tree. (or would you prefer I stuck a BK tree in a publicly accessible place, that you could pull from?)

# This is a BitKeeper generated patch for the following project:
# Project Name: Linux kernel tree
# This patch format is intended for GNU patch command version 2.5 or higher. # This patch includes the following deltas:
# ChangeSet 1.1099 -> 1.1101
# arch/ia64/Makefile 1.49 -> 1.50
# arch/ia64/tools/print_offsets.c 1.18 -> (deleted)
# arch/ia64/kernel/Makefile 1.17 -> 1.18
# (new) -> 1.1 arch/ia64/kernel/asm-offsets.c
#
# The following is the BitKeeper ChangeSet Log
# --------------------------------------------
# 03/05/26 [email protected] 1.1100
# IA64: generate include/asm/offsets.h in a way similar to all other architectures. #
# Delete references to arch/ia64/tools; use standard sed script to generate offsets.h. #
# This gets the dependencies right for offsets.h
#
# (Next change set removes arch/ia64/tools/ directory)
# --------------------------------------------
# 03/05/26 [email protected] 1.1101
# IA64: Remove now-unneeded arch/ia64/tools subdirectory
# --------------------------------------------
#

arch/ia64/Makefile +4 -5
diff -Nru a/arch/ia64/Makefile b/arch/ia64/Makefile
@@ -66,7 +66,6 @@
 drivers-$(CONFIG_IA64_GENERIC)	+= arch/ia64/hp/common/ arch/ia64/hp/zx1/ arch/ia64/hp/sim/
 
 boot := arch/ia64/boot
-tools := arch/ia64/tools
 
 .PHONY: boot compressed check
 
@@ -80,17 +79,17 @@
 
 archclean:
 	$(Q)$(MAKE) $(clean)=$(boot)
-	$(Q)$(MAKE) $(clean)=$(tools)
 
 CLEAN_FILES += include/asm-ia64/offsets.h vmlinux.gz bootloader
 
-prepare: include/asm-ia64/offsets.h
+prepare: include/asm-ia64/offsets.h 
+
+include/asm-$(ARCH)/offsets.h: arch/$(ARCH)/kernel/asm-offsets.s
+	$(call filechk,gen-asm-offsets)
 
 boot:	lib/lib.a vmlinux
 	$(Q)$(MAKE) $(build)=$(boot) $@
 
-include/asm-ia64/offsets.h:
-	$(Q)$(MAKE) $(build)=$(tools) $@
 
 define archhelp
   echo '  compressed	- Build compressed kernel image'
arch/ia64/kernel/Makefile +8 -0
diff -Nru a/arch/ia64/kernel/Makefile b/arch/ia64/kernel/Makefile
@@ -18,3 +18,11 @@
 obj-$(CONFIG_IOSAPIC)		+= iosapic.o
 obj-$(CONFIG_MODULES)		+= module.o
 obj-$(CONFIG_SMP)		+= smp.o smpboot.o
+
+arch/$(ARCH)/kernel/asm-offsets.s: include/asm-$(ARCH)/.offsets.h.stamp
+
+include/asm-$(ARCH)/.offsets.h.stamp: 
+	[ -s include/asm-$(ARCH)/offsets.h ] || echo "#define IA64_TASK_SIZE 0" > include/asm-$(ARCH)/offsets.h
+	touch $@
+
+CLEAN_FILES += include/asm-ia64/.offsets.h.stamp
arch/ia64/kernel/asm-offsets.c new file +189 -0

189 lines not on this page · raw message

arch/ia64/tools/print_offsets.c deleted +0 -224

224 lines not on this page · raw message

#9 lore raw
In reply to Peter Chubb

And here's the next changeset to get rid of the rest of arch/ia64/tools/*:

# This is a BitKeeper generated patch for the following project:
# Project Name: Linux kernel tree
# This patch format is intended for GNU patch command version 2.5 or higher. # This patch includes the following deltas:
# ChangeSet 1.1101 -> 1.1102
# arch/ia64/tools/Makefile 1.12 -> (deleted)
# arch/ia64/tools/print_offsets.awk 1.6 -> (deleted)
#
# The following is the BitKeeper ChangeSet Log
# --------------------------------------------
# 03/05/27 [email protected] 1.1102
# Delete the rest of arch/ia64/tools.
#
# .del-print_offsets.awk~ce325580e04f9929:
# Delete: arch/ia64/tools/print_offsets.awk
# .del-Makefile~90bde6a95198c56c:
# Delete: arch/ia64/tools/Makefile
# --------------------------------------------
#

arch/ia64/tools/Makefile deleted +0 -51
diff -Nru a/arch/ia64/tools/Makefile b/arch/ia64/tools/Makefile
@@ -1,51 +0,0 @@
-CFLAGS	= -g -O2 -Wall $(CPPFLAGS)
-
-TARGET	= include/asm-ia64/offsets.h
-
-src = $(obj)
-
-clean-files := print_offsets.s print_offsets offsets.h
-
-$(TARGET): $(obj)/offsets.h
-	@if ! cmp -s $(obj)/offsets.h ${TARGET}; then	\
-		echo -e "*** Updating ${TARGET}...";	\
-		cp $(obj)/offsets.h ${TARGET};		\
-	else						\
-		echo "*** ${TARGET} is up to date";	\
-	fi
-
-#
-# If we're cross-compiling, we use the cross-compiler to translate
-# print_offsets.c into an assembly file and then awk to translate this
-# file into offsets.h.  This avoids having to use a simulator to
-# generate this file.  This is based on an idea suggested by Asit
-# Mallick.  If we're running natively, we can of course just build
-# print_offsets and run it. --davidm
-#
-
-ifeq ($(CROSS_COMPILE),)
-
-$(obj)/offsets.h: $(obj)/print_offsets
-	$(obj)/print_offsets > $(obj)/offsets.h
-
-comma	:= ,
-
-$(obj)/print_offsets: $(src)/print_offsets.c FORCE
-	[ -r $(TARGET) ] || echo "#define IA64_TASK_SIZE 0" > $(TARGET)
-	$(CC) $(CFLAGS) -DKBUILD_BASENAME=$(subst $(comma),_,$(subst -,_,$(*F))) \
-		$(src)/print_offsets.c -o $@
-
-FORCE:
-
-else
-
-$(obj)/offsets.h: $(obj)/print_offsets.s
-	$(AWK) -f $(src)/print_offsets.awk $^ > $@
-
-$(obj)/print_offsets.s: $(src)/print_offsets.c
-	[ -r $(TARGET) ] || echo "#define IA64_TASK_SIZE 0" > $(TARGET)
-	$(CC) $(CFLAGS) -DKBUILD_BASENAME=$(subst $(comma),_,$(subst -,_,$(*F))) -S $^ -o $@
-
-endif
-
-.PHONY: all modules modules_install
arch/ia64/tools/print_offsets.awk deleted +0 -72

72 lines not on this page · raw message

#10 lore raw
In reply to Peter Chubb

And here's the next changeset to get rid of the rest of arch/ia64/tools/*:

# This is a BitKeeper generated patch for the following project:
# Project Name: Linux kernel tree
# This patch format is intended for GNU patch command version 2.5 or higher. # This patch includes the following deltas:
# ChangeSet 1.1101 -> 1.1102
# arch/ia64/tools/Makefile 1.12 -> (deleted)
# arch/ia64/tools/print_offsets.awk 1.6 -> (deleted)
#
# The following is the BitKeeper ChangeSet Log
# --------------------------------------------
# 03/05/27 [email protected] 1.1102
# Delete the rest of arch/ia64/tools.
#
# .del-print_offsets.awk~ce325580e04f9929:
# Delete: arch/ia64/tools/print_offsets.awk
# .del-Makefile~90bde6a95198c56c:
# Delete: arch/ia64/tools/Makefile
# --------------------------------------------
#

arch/ia64/tools/Makefile deleted +0 -51
diff -Nru a/arch/ia64/tools/Makefile b/arch/ia64/tools/Makefile
@@ -1,51 +0,0 @@
-CFLAGS	= -g -O2 -Wall $(CPPFLAGS)
-
-TARGET	= include/asm-ia64/offsets.h
-
-src = $(obj)
-
-clean-files := print_offsets.s print_offsets offsets.h
-
-$(TARGET): $(obj)/offsets.h
-	@if ! cmp -s $(obj)/offsets.h ${TARGET}; then	\
-		echo -e "*** Updating ${TARGET}...";	\
-		cp $(obj)/offsets.h ${TARGET};		\
-	else						\
-		echo "*** ${TARGET} is up to date";	\
-	fi
-
-#
-# If we're cross-compiling, we use the cross-compiler to translate
-# print_offsets.c into an assembly file and then awk to translate this
-# file into offsets.h.  This avoids having to use a simulator to
-# generate this file.  This is based on an idea suggested by Asit
-# Mallick.  If we're running natively, we can of course just build
-# print_offsets and run it. --davidm
-#
-
-ifeq ($(CROSS_COMPILE),)
-
-$(obj)/offsets.h: $(obj)/print_offsets
-	$(obj)/print_offsets > $(obj)/offsets.h
-
-comma	:= ,
-
-$(obj)/print_offsets: $(src)/print_offsets.c FORCE
-	[ -r $(TARGET) ] || echo "#define IA64_TASK_SIZE 0" > $(TARGET)
-	$(CC) $(CFLAGS) -DKBUILD_BASENAME=$(subst $(comma),_,$(subst -,_,$(*F))) \
-		$(src)/print_offsets.c -o $@
-
-FORCE:
-
-else
-
-$(obj)/offsets.h: $(obj)/print_offsets.s
-	$(AWK) -f $(src)/print_offsets.awk $^ > $@
-
-$(obj)/print_offsets.s: $(src)/print_offsets.c
-	[ -r $(TARGET) ] || echo "#define IA64_TASK_SIZE 0" > $(TARGET)
-	$(CC) $(CFLAGS) -DKBUILD_BASENAME=$(subst $(comma),_,$(subst -,_,$(*F))) -S $^ -o $@
-
-endif
-
-.PHONY: all modules modules_install
arch/ia64/tools/print_offsets.awk deleted +0 -72

72 lines not on this page · raw message

#11 lore raw
In reply to Peter Chubb

Peter> OK, here's a patch against this morning's BK tree.

Applied, thanks.

Peter> or would you prefer I stuck a BK tree in a publicly Peter> accessible place, that you could pull from?)

Nah, that's OK. With patches it's much easier to make trivial changes, such as deleting the trailing whitespace that everybody seems to love so much... ;-)

	--david