remove .tmp_gas_check
PATCH 1/1 remove .tmp_gas_check
make ARCH=ppc tags leaves a .tmp_gas_check file behind. Just remove it after the as test.
arch/ppc/Makefile renamed from linux-2.6.11-rc3.orig/arch/ppc/Makefile +2 -2
@@ -112,10 +112,10 @@ include/asm-$(ARCH)/offsets.h: arch/$(AR TOUT := .tmp_gas_check # Ensure this is binutils 2.12.1 (or 2.12.90.0.7) or later for altivec # instructions. -AS_ALTIVEC := $(shell echo dssall | $(AS) -many -o $(TOUT) >/dev/null 2>&1 ; echo $$?) +AS_ALTIVEC := $(shell echo dssall | $(AS) -many -o $(TOUT) >/dev/null 2>&1 ; echo $$? ; rm -f $(TOUT)) # gcc-3.4 and binutils-2.14 are a fatal combination. GCC_VERSION := $(call cc-version) -BAD_GCC_AS := $(shell echo mftb 5 | $(AS) -mppc -many -o $(TOUT) >/dev/null 2>&1 && echo 0 || echo 1) +BAD_GCC_AS := $(shell echo mftb 5 | $(AS) -mppc -many -o $(TOUT) >/dev/null 2>&1 ; echo $$? ; rm -f $(TOUT)) checkbin: ifeq ($(GCC_VERSION)$(BAD_GCC_AS),03041)
Quoted 4 lines · Olaf Hering
make ARCH=ppc tags leaves a .tmp_gas_check file behind. Just remove it after the as test.
Signed-off-by Olaf Hering <[email protected]>
What is the problem exactly? .tmp_* should be removed in a normal 'clean' target ?
Signature · 2 lines
Tom Rini http://gate.crashing.org/~trini/
Quoted 9 lines · Tom Rini
Quoted 4 lines · Olaf Hering
make ARCH=ppc tags leaves a .tmp_gas_check file behind. Just remove it after the as test.
Signed-off-by Olaf Hering <[email protected]>What is the problem exactly? .tmp_* should be removed in a normal 'clean' target ?
make tags should only create the toplevel tags file, nothing else.
Quoted 7 lines · Olaf Hering
Quoted 5 lines · Tom Rini
Quoted 3 lines · Olaf Hering
make ARCH=ppc tags leaves a .tmp_gas_check file behind. Just remove it after the as test.
Quoted 6 lines
Quoted 4 lines
Quoted 1 line
Signed-off-by Olaf Hering <[email protected]>What is the problem exactly? .tmp_* should be removed in a normal 'clean' target ?
make tags should only create the toplevel tags file, nothing else.
My first guess then is there's a dependancy problem, as make tags should just invoke ctags or so, yes?
Signature · 2 lines
Tom Rini http://gate.crashing.org/~trini/
Quoted 2 lines · Tom Rini
My first guess then is there's a dependancy problem, as make tags should just invoke ctags or so, yes?
The variable is evaluated, even if it is not used. Just run 'make ARCH=ppc tags' and see yourself.
Quoted 7 lines · Olaf Hering
Quoted 2 lines · Tom Rini
My first guess then is there's a dependancy problem, as make tags should just invoke ctags or so, yes?
The variable is evaluated, even if it is not used. Just run 'make ARCH=ppc tags' and see yourself.
Right. But we only care about that if we're going to compile anything, so perhaps we need to change the test somehow to only get evaluated when needed (ala the -mcpu stuff? Or is that always run as well?).
Signature · 2 lines
Tom Rini http://gate.crashing.org/~trini/
Quoted 3 lines · Tom Rini
Right. But we only care about that if we're going to compile anything, so perhaps we need to change the test somehow to only get evaluated when needed (ala the -mcpu stuff? Or is that always run as well?).
I dont want to see that file because its just a tmpfile. mkdir $foo
make ARCH=ppc tags &
make ARCH=ppc O=$foo -j menuconfig
make ARCH=ppc O=$foo -j all
Its just in the way when doing diffs. and make clean may remove the tags file. But its just silly to expect 'make tags; make clean' to get rid of it.
Quoted 6 lines · Olaf Hering
Quoted 3 lines · Tom Rini
Right. But we only care about that if we're going to compile anything, so perhaps we need to change the test somehow to only get evaluated when needed (ala the -mcpu stuff? Or is that always run as well?).
Quoted 8 lines
I dont want to see that file because its just a tmpfile. mkdir $foo
make ARCH=ppc tags &
make ARCH=ppc O=$foo -j menuconfig
make ARCH=ppc O=$foo -j allIts just in the way when doing diffs. and make clean may remove the tags file. But its just silly to expect 'make tags; make clean' to get rid of it.
I agree you shouldn't have to see it. I'm saying the problem is the variable shouldn't be evaluated.
Signature · 2 lines
Tom Rini http://gate.crashing.org/~trini/
Quoted 2 lines · Tom Rini
I agree you shouldn't have to see it. I'm saying the problem is the variable shouldn't be evaluated.
This patch seems to work for me.
arch/ppc/Makefile renamed from linux-2.6.11-rc3.orig/arch/ppc/Makefile +17 -16
@@ -112,26 +112,27 @@ include/asm-$(ARCH)/offsets.h: arch/$(AR TOUT := .tmp_gas_check # Ensure this is binutils 2.12.1 (or 2.12.90.0.7) or later for altivec # instructions. -AS_ALTIVEC := $(shell echo dssall | $(AS) -many -o $(TOUT) >/dev/null 2>&1 ; echo $$?) # gcc-3.4 and binutils-2.14 are a fatal combination. GCC_VERSION := $(call cc-version) -BAD_GCC_AS := $(shell echo mftb 5 | $(AS) -mppc -many -o $(TOUT) >/dev/null 2>&1 && echo 0 || echo 1) checkbin: -ifeq ($(GCC_VERSION)$(BAD_GCC_AS),03041) - @echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build ' - @echo 'correctly with gcc-3.4 and your version of binutils.' - @echo '*** Please upgrade your binutils or downgrade your gcc' - @false -endif -ifneq ($(AS_ALTIVEC),0) - echo $(AS_ALTIVEC) - @echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build ' - @echo 'correctly with old versions of binutils.' - @echo '*** Please upgrade your binutils to 2.12.1 or newer' - @false -endif - @true + @if test "$(GCC_VERSION)" = "0304" ; then \ + if ! /bin/echo mftb 5 | $(AS) -v -mppc -many -o $(TOUT) >/dev/null 2>&1 ; then \ + echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build '; \ + echo 'correctly with gcc-3.4 and your version of binutils.'; \ + echo '*** Please upgrade your binutils or downgrade your gcc'; \ + rm -f $(TOUT) ;\ + false; \ + fi ; \ + fi + @if ! /bin/echo dssall | $(AS) -many -o $(TOUT) >/dev/null 2>&1 ; then \ + echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build ' ; \ + echo 'correctly with old versions of binutils.' ; \ + echo '*** Please upgrade your binutils to 2.12.1 or newer' ; \ + rm -f $(TOUT) ; \ + false ; \ + fi + @rm -f $(TOUT) CLEAN_FILES += include/asm-$(ARCH)/offsets.h \ arch/$(ARCH)/kernel/asm-offsets.s \
Quoted 6 lines · Olaf Hering
Quoted 2 lines · Tom Rini
I agree you shouldn't have to see it. I'm saying the problem is the variable shouldn't be evaluated.
This patch seems to work for me.
What I don't like is that checkbin shouldn't be evaluated for 'tags' or other non-compile targets, so you don't need those explicit rm's. Also, you forgot the signed-off-by line :)
Signature · 2 lines
Tom Rini http://gate.crashing.org/~trini/
Quoted 4 lines · Tom Rini
Quoted 2 lines · Olaf Hering
On Sun, Feb 06, Tom Rini wrote:
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
I agree you shouldn't have to see it. I'm saying the problem is the variable shouldn't be evaluated.
This patch seems to work for me.
What I don't like is that checkbin shouldn't be evaluated for 'tags' or other non-compile targets, so you don't need those explicit rm's. Also, you forgot the signed-off-by line :)
Why should make tags or oldconfig ever care about the installed toolchain? How should the tmpfile disappear if one calls make checkbin with success? Maybe I misunderstood you.
Quoted 9 lines · Olaf Hering
Quoted 7 lines · Tom Rini
Quoted 5 lines · Olaf Hering
Quoted 2 lines · Tom Rini
I agree you shouldn't have to see it. I'm saying the problem is the variable shouldn't be evaluated.
Quoted 7 lines
Quoted 5 lines
Quoted 1 line
This patch seems to work for me.
What I don't like is that checkbin shouldn't be evaluated for 'tags' or other non-compile targets, so you don't need those explicit rm's. Also, you forgot the signed-off-by line :)
Why should make tags or oldconfig ever care about the installed toolchain?
It shouldn't, and 'tags' at least shouldn't depend on 'checkbin' (oldconfig I _believe_ actually invokes HOSTCC and creates other .tmp files, but I don't have a tree handy).
Quoted 1 line
How should the tmpfile disappear if one calls make checkbin with success?
On make clean, like the other .tmp files that are around at the end of a kernel compile (kallsyms/modules/maybe something else).
Quoted 1 line
Maybe I misunderstood you.
I think so. We both agree there should be no ppc-specific .tmp files around after certain make targets. I'm saying that there's always some .tmp files generated on every arch for certain targets, and removed with a clean. PPC should only create .tmp files (And leave them alone) in the cases where every arch generates some .tmp files.
Signature · 2 lines
Tom Rini http://gate.crashing.org/~trini/
Quoted 12 lines · Tom Rini
Quoted 11 lines · Olaf Hering
Quoted 9 lines · Tom Rini
Quoted 6 lines · Olaf Hering
Quoted 2 lines · Tom Rini
I agree you shouldn't have to see it. I'm saying the problem is the variable shouldn't be evaluated.
This patch seems to work for me.
Quoted 7 lines
Quoted 5 lines
Quoted 3 lines
What I don't like is that checkbin shouldn't be evaluated for 'tags' or other non-compile targets, so you don't need those explicit rm's. Also, you forgot the signed-off-by line :)
Why should make tags or oldconfig ever care about the installed toolchain?
It shouldn't, and 'tags' at least shouldn't depend on 'checkbin'
tags does not call checkbin, so this part is ok.
Quoted 4 lines
Quoted 1 line
How should the tmpfile disappear if one calls make checkbin with success?
On make clean, like the other .tmp files that are around at the end of a kernel compile (kallsyms/modules/maybe something else).
Thats ok now.
Quoted 7 lines
Quoted 1 line
Maybe I misunderstood you.
I think so. We both agree there should be no ppc-specific .tmp files around after certain make targets. I'm saying that there's always some .tmp files generated on every arch for certain targets, and removed with a clean. PPC should only create .tmp files (And leave them alone) in the cases where every arch generates some .tmp files.
arch/ppc/Makefile renamed from linux-2.6.11-rc3.orig/arch/ppc/Makefile +14 -16
@@ -112,26 +112,24 @@ include/asm-$(ARCH)/offsets.h: arch/$(AR TOUT := .tmp_gas_check # Ensure this is binutils 2.12.1 (or 2.12.90.0.7) or later for altivec # instructions. -AS_ALTIVEC := $(shell echo dssall | $(AS) -many -o $(TOUT) >/dev/null 2>&1 ; echo $$?) # gcc-3.4 and binutils-2.14 are a fatal combination. GCC_VERSION := $(call cc-version) -BAD_GCC_AS := $(shell echo mftb 5 | $(AS) -mppc -many -o $(TOUT) >/dev/null 2>&1 && echo 0 || echo 1) checkbin: -ifeq ($(GCC_VERSION)$(BAD_GCC_AS),03041) - @echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build ' - @echo 'correctly with gcc-3.4 and your version of binutils.' - @echo '*** Please upgrade your binutils or downgrade your gcc' - @false -endif -ifneq ($(AS_ALTIVEC),0) - echo $(AS_ALTIVEC) - @echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build ' - @echo 'correctly with old versions of binutils.' - @echo '*** Please upgrade your binutils to 2.12.1 or newer' - @false -endif - @true + @if test "$(GCC_VERSION)" = "0304" ; then \ + if ! /bin/echo mftb 5 | $(AS) -v -mppc -many -o $(TOUT) >/dev/null 2>&1 ; then \ + echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build '; \ + echo 'correctly with gcc-3.4 and your version of binutils.'; \ + echo '*** Please upgrade your binutils or downgrade your gcc'; \ + false; \ + fi ; \ + fi + @if ! /bin/echo dssall | $(AS) -many -o $(TOUT) >/dev/null 2>&1 ; then \ + echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build ' ; \ + echo 'correctly with old versions of binutils.' ; \ + echo '*** Please upgrade your binutils to 2.12.1 or newer' ; \ + false ; \ + fi CLEAN_FILES += include/asm-$(ARCH)/offsets.h \ arch/$(ARCH)/kernel/asm-offsets.s \