* [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 check
@ 2004-07-28 15:46 Tom Rini
2004-07-28 22:00 ` Andrew Morton
0 siblings, 1 reply; 27+ messages in thread
From: Tom Rini @ 2004-07-28 15:46 UTC (permalink / raw)
To: Andrew Morton; +Cc: Olaf Hering, Kernel Mailing List, linuxppc-dev
The following patch does three things. First, it removes all instances
of:
ifeq ($(CONFIG_FOO),y)
AFLAGS += -Wa,-mfoo
endif
and makes us set them once in arch/ppc/Makefile, via
aflags-$(CONFIG_FOO), just like we do for CFLAGS. Next it adds a test
for gcc-3.4 and binutils-2.14. The problem with this combination is
that the -many flag is broken in binutils-2.14 and gcc-3.4 will pass it
down, causing other flags to be overridden and the compile to fail.
Changing gcc or binutils versions fixes this. Finally, it changes
places in the Makefiles where we did:
ifeq ($(CONFIG_FOO),y)
obj-$(CONFIG_BAR) += foo_bar.o
endif
into
obj-$(CONFIG_FOO) += $(bar-y)
bar-$(CONFIG_BAR) += foo_bar.o
Signed-off-by: Tom Rini <trini@kernel.crashing.org>
arch/ppc/Makefile | 23 ++++++++++++++++-----
arch/ppc/kernel/Makefile | 16 +-------------
arch/ppc/mm/Makefile | 4 ---
arch/ppc/platforms/Makefile | 33 ++++++++++--------------------
arch/ppc/syslib/Makefile | 47 ++++++++++++++------------------------------
5 files changed, 46 insertions(+), 77 deletions(-)
--- 1.55/arch/ppc/Makefile 2004-07-05 03:27:10 -07:00
+++ edited/arch/ppc/Makefile 2004-07-15 07:20:40 -07:00
@@ -22,7 +22,7 @@
LDFLAGS_vmlinux := -Ttext $(KERNELLOAD) -Bstatic
CPPFLAGS += -Iarch/$(ARCH)
-AFLAGS += -Iarch/$(ARCH)
+aflags-y += -Iarch/$(ARCH)
cflags-y += -Iarch/$(ARCH) -msoft-float -pipe \
-ffixed-r2 -Wno-uninitialized -mmultiple
CPP = $(CC) -E $(CFLAGS)
@@ -31,10 +31,16 @@
cflags-y += -mstring
endif
+aflags-$(CONFIG_4xx) += -m405
cflags-$(CONFIG_4xx) += -Wa,-m405
+aflags-$(CONFIG_6xx) += -maltivec
+cflags-$(CONFIG_6xx) += -Wa,-maltivec
+aflags-$(CONFIG_E500) += -me500
cflags-$(CONFIG_E500) += -Wa,-me500
+aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
cflags-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
+AFLAGS += $(aflags-y)
CFLAGS += $(cflags-y)
head-y := arch/ppc/kernel/head.o
@@ -106,17 +112,24 @@
else
NEW_AS := 0
endif
+# gcc-3.4 and binutils-2.14 are a fatal combination.
+GCC_VERSION := $(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-version.sh $(CC))
+BAD_GCC_AS := $(shell echo mftb 5 | $(AS) -mppc -many -o /dev/null >/dev/null 2>&1 && echo 0 || echo 1)
-ifneq ($(NEW_AS),0)
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 ($(NEW_AS),0)
@echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build '
@echo 'correctly with old versions of binutils.'
@echo '*** Please upgrade your binutils to ${GOODVER} or newer'
@false
-else
-checkbin:
- @true
endif
+ @true
CLEAN_FILES += include/asm-$(ARCH)/offsets.h \
arch/$(ARCH)/kernel/asm-offsets.s
--- 1.47/arch/ppc/kernel/Makefile 2004-06-17 23:41:08 -07:00
+++ edited/arch/ppc/kernel/Makefile 2004-07-12 08:22:12 -07:00
@@ -2,16 +2,6 @@
# Makefile for the linux kernel.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-ifdef CONFIG_4xx
-EXTRA_AFLAGS := -Wa,-m405
-endif
-ifdef CONFIG_E500
-EXTRA_AFLAGS := -Wa,-me500
-endif
-
extra-$(CONFIG_PPC_STD_MMU) := head.o
extra-$(CONFIG_40x) := head_4xx.o
extra-$(CONFIG_44x) := head_44x.o
@@ -35,7 +25,5 @@
obj-$(CONFIG_TAU) += temp.o
obj-$(CONFIG_ALTIVEC) += vecemu.o vector.o
-ifdef CONFIG_MATH_EMULATION
-obj-$(CONFIG_8xx) += softemu8xx.o
-endif
-
+sw-math-emu-$(CONFIG_8xx) += softemu8xx.o
+obj-$(CONFIG_MATH_EMULATION) += $(sw-math-emu-y)
--- 1.17/arch/ppc/mm/Makefile 2004-06-17 23:41:08 -07:00
+++ edited/arch/ppc/mm/Makefile 2004-07-12 08:22:21 -07:00
@@ -2,10 +2,6 @@
# Makefile for the linux ppc-specific parts of the memory manager.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-
obj-y := fault.o init.o mem_pieces.o \
mmu_context.o pgtable.o
--- 1.31/arch/ppc/platforms/Makefile 2004-06-16 11:22:35 -07:00
+++ edited/arch/ppc/platforms/Makefile 2004-07-12 08:22:27 -07:00
@@ -2,29 +2,23 @@
# Makefile for the linux kernel.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-ifdef CONFIG_40x
-EXTRA_AFLAGS := -Wa,-m405
-endif
-
# Extra CFLAGS so we don't have to do relative includes
CFLAGS_pmac_setup.o += -Iarch/$(ARCH)/mm
-obj-$(CONFIG_APUS) += apus_setup.o
-ifeq ($(CONFIG_APUS),y)
-obj-$(CONFIG_PCI) += apus_pci.o
-endif
+# Multiple dependancies
+apus-$(CONFIG_PCI) += apus_pci.o
+chrp-$(CONFIG_SMP) += chrp_smp.o
+pmac-$(CONFIG_NVRAM) += pmac_nvram.o
+pmac-$(CONFIG_CPU_FREQ_PMAC) += pmac_cpufreq.o
+pmac-$(CONFIG_SMP) += pmac_smp.o
+
+obj-$(CONFIG_APUS) += apus_setup.o $(apus-y)
obj-$(CONFIG_PPC_PMAC) += pmac_pic.o pmac_setup.o pmac_time.o \
pmac_feature.o pmac_pci.o pmac_sleep.o \
- pmac_low_i2c.o
-obj-$(CONFIG_PPC_CHRP) += chrp_setup.o chrp_time.o chrp_pci.o
+ pmac_low_i2c.o $(pmac-y)
+obj-$(CONFIG_PPC_CHRP) += chrp_setup.o chrp_time.o chrp_pci.o \
+ $(chrp-y)
obj-$(CONFIG_PPC_PREP) += prep_pci.o prep_setup.o
-ifeq ($(CONFIG_PPC_PMAC),y)
-obj-$(CONFIG_NVRAM) += pmac_nvram.o
-obj-$(CONFIG_CPU_FREQ_PMAC) += pmac_cpufreq.o
-endif
obj-$(CONFIG_PMAC_BACKLIGHT) += pmac_backlight.o
obj-$(CONFIG_PPC_RTAS) += error_log.o proc_rtas.o
obj-$(CONFIG_PREP_RESIDUAL) += residual.o
@@ -48,8 +42,3 @@
obj-$(CONFIG_SANDPOINT) += sandpoint.o
obj-$(CONFIG_SBC82xx) += sbc82xx.o
obj-$(CONFIG_SPRUCE) += spruce.o
-
-ifeq ($(CONFIG_SMP),y)
-obj-$(CONFIG_PPC_PMAC) += pmac_smp.o
-obj-$(CONFIG_PPC_CHRP) += chrp_smp.o
-endif
--- 1.29/arch/ppc/syslib/Makefile 2004-07-01 22:23:46 -07:00
+++ edited/arch/ppc/syslib/Makefile 2004-07-12 08:22:30 -07:00
@@ -2,39 +2,27 @@
# Makefile for the linux kernel.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-ifdef CONFIG_4xx
-EXTRA_AFLAGS := -Wa,-m405
-endif
-ifdef CONFIG_E500
-EXTRA_AFLAGS := -Wa,-me500
-endif
-
CFLAGS_prom_init.o += -fPIC
CFLAGS_btext.o += -fPIC
+# Multiple dependancies
+ibm-4xx-$(CONFIG_GEN_RTC) += todc_time.o
+ibm-4xx-$(CONFIG_KGDB) += ppc4xx_kgdb.o
+ibm-40x-$(CONFIG_PCI) += indirect_pci.o pci_auto.o ppc405_pci.o
+mpc-8xx-$(CONFIG_PCI) += qspan_pci.o i8259.o
+gen550-$(CONFIG_KGDB) += gen550_kgdb.o gen550_dbg.o
+gen550-$(CONFIG_SERIAL_TEXT_DEBUG) += gen550_dbg.o
+mpc-85xx-$(CONFIG_PCI) += indirect_pci.o pci_auto.o
+
obj-$(CONFIG_PPCBUG_NVRAM) += prep_nvram.o
obj-$(CONFIG_PPC_OCP) += ocp.o
obj-$(CONFIG_IBM_OCP) += ibm_ocp.o
obj-$(CONFIG_44x) += ibm44x_common.o
obj-$(CONFIG_440GP) += ibm440gp_common.o
obj-$(CONFIG_440GX) += ibm440gx_common.o
-ifeq ($(CONFIG_4xx),y)
-obj-$(CONFIG_4xx) += ppc4xx_pic.o
-obj-$(CONFIG_40x) += ppc4xx_setup.o
-obj-$(CONFIG_GEN_RTC) += todc_time.o
-obj-$(CONFIG_KGDB) += ppc4xx_kgdb.o
-ifeq ($(CONFIG_40x),y)
-obj-$(CONFIG_KGDB) += ppc4xx_kgdb.o
-obj-$(CONFIG_PCI) += indirect_pci.o pci_auto.o ppc405_pci.o
-endif
-endif
-obj-$(CONFIG_8xx) += m8xx_setup.o ppc8xx_pic.o
-ifeq ($(CONFIG_8xx),y)
-obj-$(CONFIG_PCI) += qspan_pci.o i8259.o
-endif
+obj-$(CONFIG_4xx) += ppc4xx_pic.o $(ibm-4xx-y)
+obj-$(CONFIG_40x) += ppc4xx_setup.o $(ibm-40x-y)
+obj-$(CONFIG_8xx) += m8xx_setup.o ppc8xx_pic.o $(mpc-8xx-y)
obj-$(CONFIG_PPC_OF) += prom_init.o prom.o of_device.o
obj-$(CONFIG_PPC_PMAC) += open_pic.o indirect_pci.o
obj-$(CONFIG_POWER4) += open_pic2.o
@@ -73,16 +61,11 @@
obj-$(CONFIG_PCI_8260) += m8260_pci.o indirect_pci.o
obj-$(CONFIG_8260_PCI9) += m8260_pci_erratum9.o
obj-$(CONFIG_CPM2) += cpm2_common.o
-ifeq ($(CONFIG_PPC_GEN550),y)
-obj-$(CONFIG_KGDB) += gen550_kgdb.o gen550_dbg.o
-obj-$(CONFIG_SERIAL_TEXT_DEBUG) += gen550_dbg.o
-endif
+obj-$(CONFIG_PPC_GEN550) += $(gen550-y)
obj-$(CONFIG_BOOTX_TEXT) += btext.o
obj-$(CONFIG_MPC10X_BRIDGE) += mpc10x_common.o indirect_pci.o
obj-$(CONFIG_MPC10X_OPENPIC) += open_pic.o
obj-$(CONFIG_40x) += dcr.o
obj-$(CONFIG_BOOKE) += dcr.o
-obj-$(CONFIG_85xx) += open_pic.o ppc85xx_common.o ppc85xx_setup.o
-ifeq ($(CONFIG_85xx),y)
-obj-$(CONFIG_PCI) += indirect_pci.o pci_auto.o
-endif
+obj-$(CONFIG_85xx) += open_pic.o ppc85xx_common.o ppc85xx_setup.o \
+ $(mpc-85xx-y)
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 check
2004-07-28 15:46 [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 check Tom Rini
@ 2004-07-28 22:00 ` Andrew Morton
2004-07-28 22:07 ` Tom Rini
0 siblings, 1 reply; 27+ messages in thread
From: Andrew Morton @ 2004-07-28 22:00 UTC (permalink / raw)
To: Tom Rini; +Cc: olh, linux-kernel, linuxppc-dev, Sylvain Munaut
Tom Rini <trini@kernel.crashing.org> wrote:
>
> The following patch does three things. First, it removes all instances
> of:
> ifeq ($(CONFIG_FOO),y)
> AFLAGS += -Wa,-mfoo
> endif
>
> and makes us set them once in arch/ppc/Makefile, via
> aflags-$(CONFIG_FOO), just like we do for CFLAGS. Next it adds a test
> for gcc-3.4 and binutils-2.14. The problem with this combination is
> that the -many flag is broken in binutils-2.14 and gcc-3.4 will pass it
> down, causing other flags to be overridden and the compile to fail.
> Changing gcc or binutils versions fixes this. Finally, it changes
> places in the Makefiles where we did:
> ifeq ($(CONFIG_FOO),y)
> obj-$(CONFIG_BAR) += foo_bar.o
> endif
> into
> obj-$(CONFIG_FOO) += $(bar-y)
> bar-$(CONFIG_BAR) += foo_bar.o
Unfortunately this has significant clashes with the mpc52xx bk tree
which I'm carrying.
That patch has been sitting around for several weeks now - can we get it
merged up first?
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 check
2004-07-28 22:00 ` Andrew Morton
@ 2004-07-28 22:07 ` Tom Rini
2004-07-28 22:51 ` Andrew Morton
2004-07-29 8:05 ` [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c Giuliano Pochini
0 siblings, 2 replies; 27+ messages in thread
From: Tom Rini @ 2004-07-28 22:07 UTC (permalink / raw)
To: Andrew Morton; +Cc: olh, linux-kernel, linuxppc-dev, Sylvain Munaut, Kumar Gala
On Wed, Jul 28, 2004 at 03:00:52PM -0700, Andrew Morton wrote:
> Tom Rini <trini@kernel.crashing.org> wrote:
> >
> > The following patch does three things. First, it removes all instances
> > of:
> > ifeq ($(CONFIG_FOO),y)
> > AFLAGS += -Wa,-mfoo
> > endif
> >
> > and makes us set them once in arch/ppc/Makefile, via
> > aflags-$(CONFIG_FOO), just like we do for CFLAGS. Next it adds a test
> > for gcc-3.4 and binutils-2.14. The problem with this combination is
> > that the -many flag is broken in binutils-2.14 and gcc-3.4 will pass it
> > down, causing other flags to be overridden and the compile to fail.
> > Changing gcc or binutils versions fixes this. Finally, it changes
> > places in the Makefiles where we did:
> > ifeq ($(CONFIG_FOO),y)
> > obj-$(CONFIG_BAR) += foo_bar.o
> > endif
> > into
> > obj-$(CONFIG_FOO) += $(bar-y)
> > bar-$(CONFIG_BAR) += foo_bar.o
>
> Unfortunately this has significant clashes with the mpc52xx bk tree
> which I'm carrying.
>
> That patch has been sitting around for several weeks now - can we get it
> merged up first?
I talked with Kumar Gala at OLS, and I believe everyone (ppc) is happy
with mpc52xx stuff. So yes, that can go out.
I've taken the binutils-2.14+gcc-3.4 bit out (and none of the other
cleanups) as it seems like we get 1-2 reports a week from this bad tools
combination:
Signed-off-by: Tom Rini <trini@kernel.crashing.org>
===== arch/ppc/Makefile 1.55 vs edited =====
--- 1.55/arch/ppc/Makefile 2004-07-05 03:27:10 -07:00
+++ edited/arch/ppc/Makefile 2004-07-15 07:20:40 -07:00
@@ -106,17 +106,24 @@
else
NEW_AS := 0
endif
+# gcc-3.4 and binutils-2.14 are a fatal combination.
+GCC_VERSION := $(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-version.sh $(CC))
+BAD_GCC_AS := $(shell echo mftb 5 | $(AS) -mppc -many -o /dev/null >/dev/null 2>&1 && echo 0 || echo 1)
-ifneq ($(NEW_AS),0)
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 ($(NEW_AS),0)
@echo -n '*** ${VERSION}.${PATCHLEVEL} kernels no longer build '
@echo 'correctly with old versions of binutils.'
@echo '*** Please upgrade your binutils to ${GOODVER} or newer'
@false
-else
-checkbin:
- @true
endif
+ @true
CLEAN_FILES += include/asm-$(ARCH)/offsets.h \
arch/$(ARCH)/kernel/asm-offsets.s
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 check
2004-07-28 22:07 ` Tom Rini
@ 2004-07-28 22:51 ` Andrew Morton
2004-07-29 15:04 ` Sylvain Munaut
2004-07-29 8:05 ` [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c Giuliano Pochini
1 sibling, 1 reply; 27+ messages in thread
From: Andrew Morton @ 2004-07-28 22:51 UTC (permalink / raw)
To: Tom Rini; +Cc: olh, linux-kernel, linuxppc-dev, tnt, kumar.gala
Tom Rini <trini@kernel.crashing.org> wrote:
>
> I talked with Kumar Gala at OLS, and I believe everyone (ppc) is happy
> with mpc52xx stuff. So yes, that can go out.
OK, well could someone pease send Linus a pull request on that tree?
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 check
2004-07-28 22:51 ` Andrew Morton
@ 2004-07-29 15:04 ` Sylvain Munaut
0 siblings, 0 replies; 27+ messages in thread
From: Sylvain Munaut @ 2004-07-29 15:04 UTC (permalink / raw)
To: Andrew Morton; +Cc: Tom Rini, linux-kernel, linuxppc-dev
Hi
Andrew Morton wrote:
>Tom Rini <trini@kernel.crashing.org> wrote:
>
>
>>I talked with Kumar Gala at OLS, and I believe everyone (ppc) is happy
>>with mpc52xx stuff. So yes, that can go out.
>>
>>
>
>OK, well could someone pease send Linus a pull request on that tree?
>
>
Tom is pushing it into for-linus-ppc and will try to get Linus to pull
it from there.
Originally when I saw -rc1 was out I just thought I would wait unti
2.6.8 to post the update and get it merged.
Sylvain Munaut
PS: Sorry for the delay, bad mail filters made me miss the messages ;(
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-28 22:07 ` Tom Rini
2004-07-28 22:51 ` Andrew Morton
@ 2004-07-29 8:05 ` Giuliano Pochini
2004-07-29 14:43 ` Tom Rini
1 sibling, 1 reply; 27+ messages in thread
From: Giuliano Pochini @ 2004-07-29 8:05 UTC (permalink / raw)
To: Tom Rini
Cc: Kumar Gala, Sylvain Munaut, linuxppc-dev, linux-kernel, olh,
Andrew Morton
On 28-Jul-2004 Tom Rini wrote:
> I've taken the binutils-2.14+gcc-3.4 bit out (and none of the other
> cleanups) as it seems like we get 1-2 reports a week from this bad tools
> combination:
I had no time to do a lot of testing, but it seems that binutils 2.15 +
gcc 3.3.3 is a bad one too. I didn't try to compile the kernel (which
may also break), but at least I couldn't compile gcc 3.4.1 with the
above combination. It seems that as doesn't get the -mxxx parameter
required to compile altivec stuff. Hacking the Makefile to make it
pass -Wa,-m7455 helped a little, but it eventually failed in another
weird way. I hadn't time to investigate further, sorry.
--
Giuliano.
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-29 8:05 ` [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c Giuliano Pochini
@ 2004-07-29 14:43 ` Tom Rini
2004-07-30 18:59 ` Giuliano Pochini
0 siblings, 1 reply; 27+ messages in thread
From: Tom Rini @ 2004-07-29 14:43 UTC (permalink / raw)
To: Giuliano Pochini
Cc: Kumar Gala, Sylvain Munaut, linuxppc-dev, linux-kernel, olh,
Andrew Morton
On Thu, Jul 29, 2004 at 10:05:49AM +0200, Giuliano Pochini wrote:
> On 28-Jul-2004 Tom Rini wrote:
>
> > I've taken the binutils-2.14+gcc-3.4 bit out (and none of the other
> > cleanups) as it seems like we get 1-2 reports a week from this bad tools
> > combination:
>
> I had no time to do a lot of testing, but it seems that binutils 2.15 +
> gcc 3.3.3 is a bad one too. I didn't try to compile the kernel (which
> may also break), but at least I couldn't compile gcc 3.4.1 with the
> above combination. It seems that as doesn't get the -mxxx parameter
> required to compile altivec stuff. Hacking the Makefile to make it
> pass -Wa,-m7455 helped a little, but it eventually failed in another
> weird way. I hadn't time to investigate further, sorry.
Stock gcc-3.3.3 or from the hammer branch? There is, I think, a second
problem that was left out. The problem with gcc-3.4 + binutils-2.14 is
that -many gets passed, which zeros out previous flags. -many is fine
in binutils-2.15 (and 2.13 and 2.12 and 2.12.1 it seems), but 2.15 does
require -maltivec to be passed in order to handle altivec instructions.
Getting this right was part of the cleanup that conflicted with the
mpc52xx changes (Andrew: trying to take care of getting this into Linus'
tree now).
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-29 14:43 ` Tom Rini
@ 2004-07-30 18:59 ` Giuliano Pochini
2004-07-30 19:07 ` Tom Rini
0 siblings, 1 reply; 27+ messages in thread
From: Giuliano Pochini @ 2004-07-30 18:59 UTC (permalink / raw)
To: Tom Rini; +Cc: kumar.gala, tnt, linuxppc-dev, linux-kernel, olh, akpm
On Thu, 29 Jul 2004 07:43:47 -0700
Tom Rini <trini@kernel.crashing.org> wrote:
> > I had no time to do a lot of testing, but it seems that binutils 2.15 +
> > gcc 3.3.3 is a bad one too. I didn't try to compile the kernel (which
> > may also break), but at least I couldn't compile gcc 3.4.1 with the
> > above combination. It seems that as doesn't get the -mxxx parameter
> > required to compile altivec stuff. Hacking the Makefile to make it
> > pass -Wa,-m7455 helped a little, but it eventually failed in another
> > weird way. I hadn't time to investigate further, sorry.
>
> Stock gcc-3.3.3 or from the hammer branch ?
Stock.
> There is, I think, a second
> problem that was left out. The problem with gcc-3.4 + binutils-2.14 is
> that -many gets passed, which zeros out previous flags. -many is fine
> in binutils-2.15 (and 2.13 and 2.12 and 2.12.1 it seems), but 2.15 does
> require -maltivec to be passed in order to handle altivec instructions.
> Getting this right was part of the cleanup that conflicted with the
> mpc52xx changes (Andrew: trying to take care of getting this into Linus'
> tree now).
Ah, ok, this is useful to know. Btw, I compiled linux 2.6.7 with gcc 3.4.1
and binutils 2.15 without problems.
--
Giuliano.
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-30 18:59 ` Giuliano Pochini
@ 2004-07-30 19:07 ` Tom Rini
2004-07-30 20:48 ` Giuliano Pochini
2004-08-02 7:45 ` [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c Olaf Hering
0 siblings, 2 replies; 27+ messages in thread
From: Tom Rini @ 2004-07-30 19:07 UTC (permalink / raw)
To: Giuliano Pochini; +Cc: kumar.gala, tnt, linuxppc-dev, linux-kernel, olh, akpm
On Fri, Jul 30, 2004 at 08:59:01PM +0200, Giuliano Pochini wrote:
> On Thu, 29 Jul 2004 07:43:47 -0700
> Tom Rini <trini@kernel.crashing.org> wrote:
> > > I had no time to do a lot of testing, but it seems that binutils 2.15 +
> > > gcc 3.3.3 is a bad one too. I didn't try to compile the kernel (which
> > > may also break), but at least I couldn't compile gcc 3.4.1 with the
> > > above combination. It seems that as doesn't get the -mxxx parameter
> > > required to compile altivec stuff. Hacking the Makefile to make it
> > > pass -Wa,-m7455 helped a little, but it eventually failed in another
> > > weird way. I hadn't time to investigate further, sorry.
> >
> > Stock gcc-3.3.3 or from the hammer branch ?
>
> Stock.
That is interesting. Olaf, is gcc-3.3.x + binutils-2.15 one of the
combinations you've got in your matrix of toolchains?
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-30 19:07 ` Tom Rini
@ 2004-07-30 20:48 ` Giuliano Pochini
2004-07-30 21:03 ` Tom Rini
2004-08-02 7:45 ` [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c Olaf Hering
1 sibling, 1 reply; 27+ messages in thread
From: Giuliano Pochini @ 2004-07-30 20:48 UTC (permalink / raw)
To: Tom Rini; +Cc: kumar.gala, tnt, linuxppc-dev, linux-kernel, olh, akpm
On Fri, 30 Jul 2004 12:07:31 -0700
Tom Rini <trini@kernel.crashing.org> wrote:
>
> On Fri, Jul 30, 2004 at 08:59:01PM +0200, Giuliano Pochini wrote:
>
> > On Thu, 29 Jul 2004 07:43:47 -0700
> > Tom Rini <trini@kernel.crashing.org> wrote:
> > > > I had no time to do a lot of testing, but it seems that binutils 2.15 +
> > > > gcc 3.3.3 is a bad one too. I didn't try to compile the kernel (which
> > > > may also break), but at least I couldn't compile gcc 3.4.1 with the
> > > > above combination. It seems that as doesn't get the -mxxx parameter
> > > > required to compile altivec stuff. Hacking the Makefile to make it
> > > > pass -Wa,-m7455 helped a little, but it eventually failed in another
> > > > weird way. I hadn't time to investigate further, sorry.
> > >
> > > Stock gcc-3.3.3 or from the hammer branch ?
> >
> > Stock.
>
> That is interesting.
Actually, I don't know what the "hammer" thing is. I downloaded gcc from one
of the gcc.gnu.org mirrors: IIRC ftp://ftp.irisa.fr/pub/mirrors/gcc.gnu.org/gcc/releases/gcc-3.4.1
> Olaf, is gcc-3.3.x + binutils-2.15 one of the
> combinations you've got in your matrix of toolchains?
gcc 3.3.3 + binutils 2.15 fails quite soon here:
AS arch/ppc/kernel/l2cr.o
arch/ppc/kernel/l2cr.S: Assembler messages:
arch/ppc/kernel/l2cr.S:110: Error: Unrecognized opcode: `dssall'
arch/ppc/kernel/l2cr.S:278: Error: Unrecognized opcode: `dssall'
arch/ppc/kernel/l2cr.S:387: Error: Unrecognized opcode: `dssall'
make[1]: *** [arch/ppc/kernel/l2cr.o] Error 1
--
Giuliano.
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-30 20:48 ` Giuliano Pochini
@ 2004-07-30 21:03 ` Tom Rini
2004-08-04 10:22 ` Wolfram Quester
` (2 more replies)
0 siblings, 3 replies; 27+ messages in thread
From: Tom Rini @ 2004-07-30 21:03 UTC (permalink / raw)
To: Giuliano Pochini; +Cc: kumar.gala, tnt, linuxppc-dev, linux-kernel, olh, akpm
On Fri, Jul 30, 2004 at 10:48:28PM +0200, Giuliano Pochini wrote:
[snip]
> gcc 3.3.3 + binutils 2.15 fails quite soon here:
>
> AS arch/ppc/kernel/l2cr.o
> arch/ppc/kernel/l2cr.S: Assembler messages:
> arch/ppc/kernel/l2cr.S:110: Error: Unrecognized opcode: `dssall'
> arch/ppc/kernel/l2cr.S:278: Error: Unrecognized opcode: `dssall'
> arch/ppc/kernel/l2cr.S:387: Error: Unrecognized opcode: `dssall'
> make[1]: *** [arch/ppc/kernel/l2cr.o] Error 1
Can you try with the following?
Signed-off-by: Tom Rini <trini@kernel.crashing.org>
===== arch/ppc/Makefile 1.57 vs edited =====
--- 1.57/arch/ppc/Makefile 2004-07-28 21:58:36 -07:00
+++ edited/arch/ppc/Makefile 2004-07-29 12:21:33 -07:00
@@ -22,7 +22,7 @@
LDFLAGS_vmlinux := -Ttext $(KERNELLOAD) -Bstatic
CPPFLAGS += -Iarch/$(ARCH)
-AFLAGS += -Iarch/$(ARCH)
+aflags-y += -Iarch/$(ARCH)
cflags-y += -Iarch/$(ARCH) -msoft-float -pipe \
-ffixed-r2 -Wno-uninitialized -mmultiple
CPP = $(CC) -E $(CFLAGS)
@@ -33,10 +33,16 @@
cflags-y += -mstring
endif
+aflags-$(CONFIG_4xx) += -m405
cflags-$(CONFIG_4xx) += -Wa,-m405
+aflags-$(CONFIG_6xx) += -maltivec
+cflags-$(CONFIG_6xx) += -Wa,-maltivec
+aflags-$(CONFIG_E500) += -me500
cflags-$(CONFIG_E500) += -Wa,-me500
+aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
cflags-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
+AFLAGS += $(aflags-y)
CFLAGS += $(cflags-y)
head-y := arch/ppc/kernel/head.o
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-30 21:03 ` Tom Rini
@ 2004-08-04 10:22 ` Wolfram Quester
2004-08-04 10:41 ` Guido Guenther
2004-08-05 14:12 ` Olaf Hering
2 siblings, 0 replies; 27+ messages in thread
From: Wolfram Quester @ 2004-08-04 10:22 UTC (permalink / raw)
To: Tom Rini
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
olh, akpm
[-- Attachment #1: Type: text/plain, Size: 1956 bytes --]
Hi alltogether!
On Fri, Jul 30, 2004 at 02:03:18PM -0700, Tom Rini wrote:
>
> On Fri, Jul 30, 2004 at 10:48:28PM +0200, Giuliano Pochini wrote:
>
> [snip]
> > gcc 3.3.3 + binutils 2.15 fails quite soon here:
> >
> > AS arch/ppc/kernel/l2cr.o
> > arch/ppc/kernel/l2cr.S: Assembler messages:
> > arch/ppc/kernel/l2cr.S:110: Error: Unrecognized opcode: `dssall'
> > arch/ppc/kernel/l2cr.S:278: Error: Unrecognized opcode: `dssall'
> > arch/ppc/kernel/l2cr.S:387: Error: Unrecognized opcode: `dssall'
> > make[1]: *** [arch/ppc/kernel/l2cr.o] Error 1
>
> Can you try with the following?
I had to apply this patch to get 2.6.8-rc2 compiled on debian unstable
(3.3.4-2 and binutils 2.15-1). I think ths patch should make it into
2.6.8.
With best regards,
Wolfi
>
> Signed-off-by: Tom Rini <trini@kernel.crashing.org>
>
> ===== arch/ppc/Makefile 1.57 vs edited =====
> --- 1.57/arch/ppc/Makefile 2004-07-28 21:58:36 -07:00
> +++ edited/arch/ppc/Makefile 2004-07-29 12:21:33 -07:00
> @@ -22,7 +22,7 @@
>
> LDFLAGS_vmlinux := -Ttext $(KERNELLOAD) -Bstatic
> CPPFLAGS += -Iarch/$(ARCH)
> -AFLAGS += -Iarch/$(ARCH)
> +aflags-y += -Iarch/$(ARCH)
> cflags-y += -Iarch/$(ARCH) -msoft-float -pipe \
> -ffixed-r2 -Wno-uninitialized -mmultiple
> CPP = $(CC) -E $(CFLAGS)
> @@ -33,10 +33,16 @@
> cflags-y += -mstring
> endif
>
> +aflags-$(CONFIG_4xx) += -m405
> cflags-$(CONFIG_4xx) += -Wa,-m405
> +aflags-$(CONFIG_6xx) += -maltivec
> +cflags-$(CONFIG_6xx) += -Wa,-maltivec
> +aflags-$(CONFIG_E500) += -me500
> cflags-$(CONFIG_E500) += -Wa,-me500
> +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
> cflags-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
>
> +AFLAGS += $(aflags-y)
> CFLAGS += $(cflags-y)
>
> head-y := arch/ppc/kernel/head.o
>
> --
> Tom Rini
> http://gate.crashing.org/~trini/
>
> ** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-30 21:03 ` Tom Rini
2004-08-04 10:22 ` Wolfram Quester
@ 2004-08-04 10:41 ` Guido Guenther
2004-08-04 12:37 ` Guido Guenther
2004-08-05 14:12 ` Olaf Hering
2 siblings, 1 reply; 27+ messages in thread
From: Guido Guenther @ 2004-08-04 10:41 UTC (permalink / raw)
To: Tom Rini
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
olh, akpm
[-- Attachment #1.1: Type: text/plain, Size: 702 bytes --]
On Fri, Jul 30, 2004 at 02:03:18PM -0700, Tom Rini wrote:
>
> On Fri, Jul 30, 2004 at 10:48:28PM +0200, Giuliano Pochini wrote:
>
> [snip]
> > gcc 3.3.3 + binutils 2.15 fails quite soon here:
> >
> > AS arch/ppc/kernel/l2cr.o
> > arch/ppc/kernel/l2cr.S: Assembler messages:
> > arch/ppc/kernel/l2cr.S:110: Error: Unrecognized opcode: `dssall'
> > arch/ppc/kernel/l2cr.S:278: Error: Unrecognized opcode: `dssall'
> > arch/ppc/kernel/l2cr.S:387: Error: Unrecognized opcode: `dssall'
> > make[1]: *** [arch/ppc/kernel/l2cr.o] Error 1
>
> Can you try with the following?
Doesn't apply against 2.6.8-rc3. For what it's worth, I've attached a version that does.
Cheers,
-- Guido
[-- Attachment #1.2: introduce-aflags.diff --]
[-- Type: text/plain, Size: 778 bytes --]
--- ../linux-2.6.8-rc3.orig/arch/ppc/Makefile 2004-06-16 07:18:57.000000000 +0200
+++ arch/ppc/Makefile 2004-08-04 11:39:26.518867808 +0200
@@ -15,14 +15,20 @@
LDFLAGS_vmlinux := -Ttext $(KERNELLOAD) -Bstatic
CPPFLAGS += -Iarch/$(ARCH)
-AFLAGS += -Iarch/$(ARCH)
+aflags-y += -Iarch/$(ARCH)
cflags-y += -Iarch/$(ARCH) -msoft-float -pipe \
-ffixed-r2 -Wno-uninitialized -mmultiple -mstring
CPP = $(CC) -E $(CFLAGS)
+aflags-$(CONFIG_4xx) += -m405
cflags-$(CONFIG_4xx) += -Wa,-m405
+aflags-$(CONFIG_6xx) += -maltivec
+cflags-$(CONFIG_6xx) += -Wa,-maltivec
+aflags-$(CONFIG_E500) += -me500
+aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
cflags-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
+AFLAGS += $(aflags-y)
CFLAGS += $(cflags-y)
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-04 10:41 ` Guido Guenther
@ 2004-08-04 12:37 ` Guido Guenther
0 siblings, 0 replies; 27+ messages in thread
From: Guido Guenther @ 2004-08-04 12:37 UTC (permalink / raw)
To: Tom Rini
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
olh, akpm
[-- Attachment #1: Type: text/plain, Size: 442 bytes --]
On Wed, Aug 04, 2004 at 12:41:57PM +0200, Guido Guenther wrote:
> On Fri, Jul 30, 2004 at 02:03:18PM -0700, Tom Rini wrote:
> > On Fri, Jul 30, 2004 at 10:48:28PM +0200, Giuliano Pochini wrote:
> > > gcc 3.3.3 + binutils 2.15 fails quite soon here:
[..snip..]
> > Can you try with the following?
> Doesn't apply against 2.6.8-rc3. For what it's worth, I've attached a version that does.
Forget about that one, wrong tree.
-- Guido
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-30 21:03 ` Tom Rini
2004-08-04 10:22 ` Wolfram Quester
2004-08-04 10:41 ` Guido Guenther
@ 2004-08-05 14:12 ` Olaf Hering
2004-08-05 16:54 ` Tom Rini
2 siblings, 1 reply; 27+ messages in thread
From: Olaf Hering @ 2004-08-05 14:12 UTC (permalink / raw)
To: Tom Rini
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Fri, Jul 30, Tom Rini wrote:
>
> +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
this should be -Wa,-mppc64bridge for some reasons.
--
USB is for mice, FireWire is for men!
sUse lINUX ag, nÜRNBERG
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 14:12 ` Olaf Hering
@ 2004-08-05 16:54 ` Tom Rini
2004-08-05 17:00 ` Olaf Hering
2004-08-05 18:00 ` Olaf Hering
0 siblings, 2 replies; 27+ messages in thread
From: Tom Rini @ 2004-08-05 16:54 UTC (permalink / raw)
To: Olaf Hering
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Thu, Aug 05, 2004 at 04:12:57PM +0200, Olaf Hering wrote:
> On Fri, Jul 30, Tom Rini wrote:
>
> >
> > +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
>
> this should be -Wa,-mppc64bridge for some reasons.
That, er, doesn't make sense. The assembler needs -Wa,?
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 16:54 ` Tom Rini
@ 2004-08-05 17:00 ` Olaf Hering
2004-08-05 17:20 ` Tom Rini
2004-08-05 18:00 ` Olaf Hering
1 sibling, 1 reply; 27+ messages in thread
From: Olaf Hering @ 2004-08-05 17:00 UTC (permalink / raw)
To: Tom Rini
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Thu, Aug 05, Tom Rini wrote:
> On Thu, Aug 05, 2004 at 04:12:57PM +0200, Olaf Hering wrote:
> > On Fri, Jul 30, Tom Rini wrote:
> >
> > >
> > > +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
> >
> > this should be -Wa,-mppc64bridge for some reasons.
>
> That, er, doesn't make sense. The assembler needs -Wa,?
The cmdline was 'gcc .. -mppc64bridge ..'
But there is more breakage with g5 32bit, I'm looking at it right now.
--
USB is for mice, FireWire is for men!
sUse lINUX ag, nÜRNBERG
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 17:00 ` Olaf Hering
@ 2004-08-05 17:20 ` Tom Rini
2004-08-05 17:39 ` Olaf Hering
0 siblings, 1 reply; 27+ messages in thread
From: Tom Rini @ 2004-08-05 17:20 UTC (permalink / raw)
To: Olaf Hering
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Thu, Aug 05, 2004 at 07:00:44PM +0200, Olaf Hering wrote:
> On Thu, Aug 05, Tom Rini wrote:
>
> > On Thu, Aug 05, 2004 at 04:12:57PM +0200, Olaf Hering wrote:
> > > On Fri, Jul 30, Tom Rini wrote:
> > >
> > > >
> > > > +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
> > >
> > > this should be -Wa,-mppc64bridge for some reasons.
> >
> > That, er, doesn't make sense. The assembler needs -Wa,?
>
> The cmdline was 'gcc .. -mppc64bridge ..'
> But there is more breakage with g5 32bit, I'm looking at it right now.
Hmm. Was cflags-... done correctly?
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 17:20 ` Tom Rini
@ 2004-08-05 17:39 ` Olaf Hering
2004-08-05 18:03 ` Tom Rini
0 siblings, 1 reply; 27+ messages in thread
From: Olaf Hering @ 2004-08-05 17:39 UTC (permalink / raw)
To: Tom Rini
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Thu, Aug 05, Tom Rini wrote:
> On Thu, Aug 05, 2004 at 07:00:44PM +0200, Olaf Hering wrote:
> > On Thu, Aug 05, Tom Rini wrote:
> >
> > > On Thu, Aug 05, 2004 at 04:12:57PM +0200, Olaf Hering wrote:
> > > > On Fri, Jul 30, Tom Rini wrote:
> > > >
> > > > >
> > > > > +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
> > > >
> > > > this should be -Wa,-mppc64bridge for some reasons.
> > >
> > > That, er, doesn't make sense. The assembler needs -Wa,?
> >
> > The cmdline was 'gcc .. -mppc64bridge ..'
> > But there is more breakage with g5 32bit, I'm looking at it right now.
>
> Hmm. Was cflags-... done correctly?
I have currently no idea whats going on with the cflags. power3 and g5
fails for me with current binutils and hammer branch.
g5 needs the altivec option, because -maltivec -mppc64bridge will disable
altivec again in as. And arch/ppc/kernel/Makefile adds -mppc64bridge
after the cflags in arch/ppc/Makefile. Maybe the EXTRA_CFLAGS can be
removed from arch/ppc/kernel/Makefile?
plain rc3 with g5 32bit (and power3) .config gives this:
gcc -m32 -Wp,-MD,usr/.initramfs_data.o.d -nostdinc -iwithprefix include -D__KERNEL__ -Iinclude -Iarch/ppc -D__ASSEMBLY__ -Iarch/ppc -mppc64bridge -c -o usr/initramfs_data.o usr/initramfs_data.S
cc1: error: invalid option `ppc64bridge'
this comes from aflags-$(CONFIG_PPC64BRIDGE) in arch/ppc/Makefile
--
USB is for mice, FireWire is for men!
sUse lINUX ag, nÜRNBERG
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 17:39 ` Olaf Hering
@ 2004-08-05 18:03 ` Tom Rini
0 siblings, 0 replies; 27+ messages in thread
From: Tom Rini @ 2004-08-05 18:03 UTC (permalink / raw)
To: Olaf Hering
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Thu, Aug 05, 2004 at 07:39:26PM +0200, Olaf Hering wrote:
> On Thu, Aug 05, Tom Rini wrote:
>
> > On Thu, Aug 05, 2004 at 07:00:44PM +0200, Olaf Hering wrote:
> > > On Thu, Aug 05, Tom Rini wrote:
> > >
> > > > On Thu, Aug 05, 2004 at 04:12:57PM +0200, Olaf Hering wrote:
> > > > > On Fri, Jul 30, Tom Rini wrote:
> > > > >
> > > > > >
> > > > > > +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
> > > > >
> > > > > this should be -Wa,-mppc64bridge for some reasons.
> > > >
> > > > That, er, doesn't make sense. The assembler needs -Wa,?
> > >
> > > The cmdline was 'gcc .. -mppc64bridge ..'
> > > But there is more breakage with g5 32bit, I'm looking at it right now.
> >
> > Hmm. Was cflags-... done correctly?
>
> I have currently no idea whats going on with the cflags. power3 and g5
> fails for me with current binutils and hammer branch.
>
> g5 needs the altivec option, because -maltivec -mppc64bridge will disable
> altivec again in as. And arch/ppc/kernel/Makefile adds -mppc64bridge
> after the cflags in arch/ppc/Makefile. Maybe the EXTRA_CFLAGS can be
> removed from arch/ppc/kernel/Makefile?
EXTRA_CFLAGS can go, yes (I was assuming -rc2 then -final, or I would
have done this already_. The problem looks like AFLAGS being used when
CFLAGS are wanted to be used.
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 16:54 ` Tom Rini
2004-08-05 17:00 ` Olaf Hering
@ 2004-08-05 18:00 ` Olaf Hering
2004-08-05 18:14 ` Tom Rini
1 sibling, 1 reply; 27+ messages in thread
From: Olaf Hering @ 2004-08-05 18:00 UTC (permalink / raw)
To: Tom Rini
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Thu, Aug 05, Tom Rini wrote:
> On Thu, Aug 05, 2004 at 04:12:57PM +0200, Olaf Hering wrote:
> > On Fri, Jul 30, Tom Rini wrote:
> >
> > >
> > > +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
> >
> > this should be -Wa,-mppc64bridge for some reasons.
>
> That, er, doesn't make sense. The assembler needs -Wa,?
This makes g5 32bit happy. aflags- is used with 'gcc $options', not for as
I'm not sure if the other aflags- should stay.
g5 has altivec, some include files use dsall in asm().
head.S will not compile without -maltivevec
diff -purNX /tmp/kernel_exclude.txt linux-2.6.8-rc3.orig/arch/ppc/Makefile linux-2.6.8-rc3/arch/ppc/Makefile
--- linux-2.6.8-rc3.orig/arch/ppc/Makefile 2004-08-05 17:35:10.636631000 +0000
+++ linux-2.6.8-rc3/arch/ppc/Makefile 2004-08-05 17:51:09.847355812 +0000
@@ -39,8 +39,8 @@ aflags-$(CONFIG_6xx) += -maltivec
cflags-$(CONFIG_6xx) += -Wa,-maltivec
aflags-$(CONFIG_E500) += -me500
cflags-$(CONFIG_E500) += -Wa,-me500
-aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
cflags-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
+cflags-$(CONFIG_POWER4) += -Wa,-maltivec
AFLAGS += $(aflags-y)
CFLAGS += $(cflags-y)
diff -purNX /tmp/kernel_exclude.txt linux-2.6.8-rc3.orig/arch/ppc/kernel/Makefile linux-2.6.8-rc3/arch/ppc/kernel/Makefile
--- linux-2.6.8-rc3.orig/arch/ppc/kernel/Makefile 2004-08-05 17:35:10.665626000 +0000
+++ linux-2.6.8-rc3/arch/ppc/kernel/Makefile 2004-08-05 17:50:19.532770245 +0000
@@ -5,6 +5,9 @@
ifdef CONFIG_PPC64BRIDGE
EXTRA_AFLAGS := -Wa,-mppc64bridge
endif
+ifdef CONFIG_POWER4
+EXTRA_AFLAGS := -Wa,-maltivec
+endif
ifdef CONFIG_4xx
EXTRA_AFLAGS := -Wa,-m405
endif
--
USB is for mice, FireWire is for men!
sUse lINUX ag, nÜRNBERG
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 18:00 ` Olaf Hering
@ 2004-08-05 18:14 ` Tom Rini
2004-08-05 19:04 ` Sam Ravnborg
` (3 more replies)
0 siblings, 4 replies; 27+ messages in thread
From: Tom Rini @ 2004-08-05 18:14 UTC (permalink / raw)
To: Olaf Hering
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Thu, Aug 05, 2004 at 08:00:25PM +0200, Olaf Hering wrote:
> On Thu, Aug 05, Tom Rini wrote:
>
> > On Thu, Aug 05, 2004 at 04:12:57PM +0200, Olaf Hering wrote:
> > > On Fri, Jul 30, Tom Rini wrote:
> > >
> > > >
> > > > +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
> > >
> > > this should be -Wa,-mppc64bridge for some reasons.
> >
> > That, er, doesn't make sense. The assembler needs -Wa,?
>
> This makes g5 32bit happy. aflags- is used with 'gcc $options', not for as
> I'm not sure if the other aflags- should stay.
I mistook AFLAGS for being always invoked with gas, which is not the
case. Lets do the following:
Signed-off-by: Tom Rini <trini@kernel.crashing.org>
--- 1.58/arch/ppc/Makefile 2004-07-31 12:16:29 -07:00
+++ edited/arch/ppc/Makefile 2004-08-05 11:13:05 -07:00
@@ -22,28 +22,25 @@
LDFLAGS_vmlinux := -Ttext $(KERNELLOAD) -Bstatic
CPPFLAGS += -Iarch/$(ARCH)
-aflags-y += -Iarch/$(ARCH)
-cflags-y += -Iarch/$(ARCH) -msoft-float -pipe \
+AFLAGS += -Iarch/$(ARCH)
+CFLAGS += -Iarch/$(ARCH) -msoft-float -pipe \
-ffixed-r2 -Wno-uninitialized -mmultiple
CPP = $(CC) -E $(CFLAGS)
CHECK := $(CHECK) -D__powerpc__=1
ifndef CONFIG_E500
-cflags-y += -mstring
+CFLAGS += -mstring
endif
-aflags-$(CONFIG_4xx) += -m405
-cflags-$(CONFIG_4xx) += -Wa,-m405
-aflags-$(CONFIG_6xx) += -maltivec
-cflags-$(CONFIG_6xx) += -Wa,-maltivec
-aflags-$(CONFIG_E500) += -me500
-cflags-$(CONFIG_E500) += -Wa,-me500
-aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
-cflags-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
+cpu-as-$(CONFIG_4xx) += -Wa,-m405
+cpu-as-$(CONFIG_6xx) += -Wa,-maltivec
+cpu-as-$(CONFIG_POWER4) += -Wa,-maltivec
+cpu-as-$(CONFIG_E500) += -Wa,-me500
+cpu-as-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
-AFLAGS += $(aflags-y)
-CFLAGS += $(cflags-y)
+AFLAGS += $(cpu-as-y)
+CFLAGS += $(cpu-as-y)
head-y := arch/ppc/kernel/head.o
head-$(CONFIG_8xx) := arch/ppc/kernel/head_8xx.o
--- 1.47/arch/ppc/kernel/Makefile 2004-06-17 23:41:08 -07:00
+++ edited/arch/ppc/kernel/Makefile 2004-08-05 11:10:48 -07:00
@@ -2,16 +2,6 @@
# Makefile for the linux kernel.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-ifdef CONFIG_4xx
-EXTRA_AFLAGS := -Wa,-m405
-endif
-ifdef CONFIG_E500
-EXTRA_AFLAGS := -Wa,-me500
-endif
-
extra-$(CONFIG_PPC_STD_MMU) := head.o
extra-$(CONFIG_40x) := head_4xx.o
extra-$(CONFIG_44x) := head_44x.o
--- 1.17/arch/ppc/mm/Makefile 2004-06-17 23:41:08 -07:00
+++ edited/arch/ppc/mm/Makefile 2004-08-02 12:08:32 -07:00
@@ -2,10 +2,6 @@
# Makefile for the linux ppc-specific parts of the memory manager.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-
obj-y := fault.o init.o mem_pieces.o \
mmu_context.o pgtable.o
--- 1.33/arch/ppc/platforms/Makefile 2004-07-29 07:39:02 -07:00
+++ edited/arch/ppc/platforms/Makefile 2004-08-05 11:11:06 -07:00
@@ -2,13 +2,6 @@
# Makefile for the linux kernel.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-ifdef CONFIG_40x
-EXTRA_AFLAGS := -Wa,-m405
-endif
-
# Extra CFLAGS so we don't have to do relative includes
CFLAGS_pmac_setup.o += -Iarch/$(ARCH)/mm
--- 1.31/arch/ppc/syslib/Makefile 2004-08-02 01:00:42 -07:00
+++ edited/arch/ppc/syslib/Makefile 2004-08-05 11:11:14 -07:00
@@ -2,16 +2,6 @@
# Makefile for the linux kernel.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-ifdef CONFIG_4xx
-EXTRA_AFLAGS := -Wa,-m405
-endif
-ifdef CONFIG_E500
-EXTRA_AFLAGS := -Wa,-me500
-endif
-
CFLAGS_prom_init.o += -fPIC
CFLAGS_btext.o += -fPIC
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 18:14 ` Tom Rini
@ 2004-08-05 19:04 ` Sam Ravnborg
2004-08-05 21:03 ` Matt Porter
` (2 subsequent siblings)
3 siblings, 0 replies; 27+ messages in thread
From: Sam Ravnborg @ 2004-08-05 19:04 UTC (permalink / raw)
To: Tom Rini
Cc: Olaf Hering, Giuliano Pochini, kumar.gala, tnt, linuxppc-dev,
linux-kernel, akpm
On Thu, Aug 05, 2004 at 11:14:25AM -0700, Tom Rini wrote:
>
> I mistook AFLAGS for being always invoked with gas, which is not the
> case.
Correct, AFLAGS is kbuild internal. AFLAGS is used only for .S -> .o
and for .S -> .s
>Lets do the following:
Yup - looks better with cpu-as-
Sam
^ permalink raw reply [flat|nested] 27+ messages in thread* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 18:14 ` Tom Rini
2004-08-05 19:04 ` Sam Ravnborg
@ 2004-08-05 21:03 ` Matt Porter
2004-08-06 15:32 ` Olaf Hering
[not found] ` <hhacx8hirb.fsf@alsvidh.mathematik.uni-muenchen.de>
3 siblings, 0 replies; 27+ messages in thread
From: Matt Porter @ 2004-08-05 21:03 UTC (permalink / raw)
To: Tom Rini
Cc: Olaf Hering, Giuliano Pochini, kumar.gala, tnt, linuxppc-dev,
linux-kernel, akpm
On Thu, Aug 05, 2004 at 11:14:25AM -0700, Tom Rini wrote:
>
> On Thu, Aug 05, 2004 at 08:00:25PM +0200, Olaf Hering wrote:
> > On Thu, Aug 05, Tom Rini wrote:
> >
> > > On Thu, Aug 05, 2004 at 04:12:57PM +0200, Olaf Hering wrote:
> > > > On Fri, Jul 30, Tom Rini wrote:
> > > >
> > > > >
> > > > > +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
> > > >
> > > > this should be -Wa,-mppc64bridge for some reasons.
> > >
> > > That, er, doesn't make sense. The assembler needs -Wa,?
> >
> > This makes g5 32bit happy. aflags- is used with 'gcc $options', not for as
> > I'm not sure if the other aflags- should stay.
>
> I mistook AFLAGS for being always invoked with gas, which is not the
> case. Lets do the following:
>
> Signed-off-by: Tom Rini <trini@kernel.crashing.org>
This fixed the build problems (unrecognized -m405) with a PPC440 native
toolchain.
-Matt
^ permalink raw reply [flat|nested] 27+ messages in thread* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-08-05 18:14 ` Tom Rini
2004-08-05 19:04 ` Sam Ravnborg
2004-08-05 21:03 ` Matt Porter
@ 2004-08-06 15:32 ` Olaf Hering
[not found] ` <hhacx8hirb.fsf@alsvidh.mathematik.uni-muenchen.de>
3 siblings, 0 replies; 27+ messages in thread
From: Olaf Hering @ 2004-08-06 15:32 UTC (permalink / raw)
To: Tom Rini
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Thu, Aug 05, Tom Rini wrote:
> On Thu, Aug 05, 2004 at 08:00:25PM +0200, Olaf Hering wrote:
> > On Thu, Aug 05, Tom Rini wrote:
> >
> > > On Thu, Aug 05, 2004 at 04:12:57PM +0200, Olaf Hering wrote:
> > > > On Fri, Jul 30, Tom Rini wrote:
> > > >
> > > > >
> > > > > +aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
> > > >
> > > > this should be -Wa,-mppc64bridge for some reasons.
> > >
> > > That, er, doesn't make sense. The assembler needs -Wa,?
> >
> > This makes g5 32bit happy. aflags- is used with 'gcc $options', not for as
> > I'm not sure if the other aflags- should stay.
>
> I mistook AFLAGS for being always invoked with gas, which is not the
> case. Lets do the following:
>
> Signed-off-by: Tom Rini <trini@kernel.crashing.org>
and this one to enable altivec, because ppc64bridge disables altivec
again.
g5 32bit has CONFIG_POWER4=y and CONFIG_PPC64BRIDGE=y
diff -purNX /tmp/kernel_exclude.txt linux-2.6.8-rc3.trini/arch/ppc/Makefile linux-2.6.8-rc3.trini.fixed/arch/ppc/Makefile
--- linux-2.6.8-rc3.trini/arch/ppc/Makefile 2004-08-06 15:21:00.668653640 +0000
+++ linux-2.6.8-rc3.trini.fixed/arch/ppc/Makefile 2004-08-06 15:25:17.039320563 +0000
@@ -35,9 +35,9 @@ endif
cpu-as-$(CONFIG_4xx) += -Wa,-m405
cpu-as-$(CONFIG_6xx) += -Wa,-maltivec
-cpu-as-$(CONFIG_POWER4) += -Wa,-maltivec
cpu-as-$(CONFIG_E500) += -Wa,-me500
cpu-as-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
+cpu-as-$(CONFIG_POWER4) += -Wa,-maltivec
AFLAGS += $(cpu-as-y)
CFLAGS += $(cpu-as-y)
--
USB is for mice, FireWire is for men!
sUse lINUX ag, nÜRNBERG
^ permalink raw reply [flat|nested] 27+ messages in thread[parent not found: <hhacx8hirb.fsf@alsvidh.mathematik.uni-muenchen.de>]
* [PATCH] ppc32: Fix building of certain CPU types (Was: Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14)
[not found] ` <hhacx8hirb.fsf@alsvidh.mathematik.uni-muenchen.de>
@ 2004-08-06 15:38 ` Tom Rini
0 siblings, 0 replies; 27+ messages in thread
From: Tom Rini @ 2004-08-06 15:38 UTC (permalink / raw)
To: Linus Torvalds
Cc: Olaf Hering, Giuliano Pochini, kumar.gala, tnt, linuxppc-dev,
linux-kernel, akpm, Jens Schmalzing
On Fri, Aug 06, 2004 at 08:40:08AM +0200, Jens Schmalzing wrote:
> Hi,
>
> Tom Rini writes:
>
> > I mistook AFLAGS for being always invoked with gas, which is not the
> > case. Lets do the following:
>
> > +cpu-as-$(CONFIG_POWER4) += -Wa,-maltivec
> > +cpu-as-$(CONFIG_E500) += -Wa,-me500
> > +cpu-as-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
>
> These three lines should be the other way round for a successful build
> on a G5.
Done.
The following corrects how we pass CPU flags to gas. Previous, AFLAGS
was incorrectly assumed to be pased directly to $(AS), which is not the
case.
Concept ack'd by Sam Ravnborg.
Signed-off-by: Tom Rini <trini@kernel.crashing.org>
--- 1.58/arch/ppc/Makefile 2004-07-31 12:16:29 -07:00
+++ edited/arch/ppc/Makefile 2004-08-06 08:35:07 -07:00
@@ -22,28 +22,25 @@
LDFLAGS_vmlinux := -Ttext $(KERNELLOAD) -Bstatic
CPPFLAGS += -Iarch/$(ARCH)
-aflags-y += -Iarch/$(ARCH)
-cflags-y += -Iarch/$(ARCH) -msoft-float -pipe \
+AFLAGS += -Iarch/$(ARCH)
+CFLAGS += -Iarch/$(ARCH) -msoft-float -pipe \
-ffixed-r2 -Wno-uninitialized -mmultiple
CPP = $(CC) -E $(CFLAGS)
CHECK := $(CHECK) -D__powerpc__=1
ifndef CONFIG_E500
-cflags-y += -mstring
+CFLAGS += -mstring
endif
-aflags-$(CONFIG_4xx) += -m405
-cflags-$(CONFIG_4xx) += -Wa,-m405
-aflags-$(CONFIG_6xx) += -maltivec
-cflags-$(CONFIG_6xx) += -Wa,-maltivec
-aflags-$(CONFIG_E500) += -me500
-cflags-$(CONFIG_E500) += -Wa,-me500
-aflags-$(CONFIG_PPC64BRIDGE) += -mppc64bridge
-cflags-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
+cpu-as-$(CONFIG_PPC64BRIDGE) += -Wa,-mppc64bridge
+cpu-as-$(CONFIG_4xx) += -Wa,-m405
+cpu-as-$(CONFIG_6xx) += -Wa,-maltivec
+cpu-as-$(CONFIG_POWER4) += -Wa,-maltivec
+cpu-as-$(CONFIG_E500) += -Wa,-me500
-AFLAGS += $(aflags-y)
-CFLAGS += $(cflags-y)
+AFLAGS += $(cpu-as-y)
+CFLAGS += $(cpu-as-y)
head-y := arch/ppc/kernel/head.o
head-$(CONFIG_8xx) := arch/ppc/kernel/head_8xx.o
--- 1.47/arch/ppc/kernel/Makefile 2004-06-17 23:41:08 -07:00
+++ edited/arch/ppc/kernel/Makefile 2004-08-05 11:10:48 -07:00
@@ -2,16 +2,6 @@
# Makefile for the linux kernel.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-ifdef CONFIG_4xx
-EXTRA_AFLAGS := -Wa,-m405
-endif
-ifdef CONFIG_E500
-EXTRA_AFLAGS := -Wa,-me500
-endif
-
extra-$(CONFIG_PPC_STD_MMU) := head.o
extra-$(CONFIG_40x) := head_4xx.o
extra-$(CONFIG_44x) := head_44x.o
--- 1.17/arch/ppc/mm/Makefile 2004-06-17 23:41:08 -07:00
+++ edited/arch/ppc/mm/Makefile 2004-08-02 12:08:32 -07:00
@@ -2,10 +2,6 @@
# Makefile for the linux ppc-specific parts of the memory manager.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-
obj-y := fault.o init.o mem_pieces.o \
mmu_context.o pgtable.o
--- 1.33/arch/ppc/platforms/Makefile 2004-07-29 07:39:02 -07:00
+++ edited/arch/ppc/platforms/Makefile 2004-08-05 11:11:06 -07:00
@@ -2,13 +2,6 @@
# Makefile for the linux kernel.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-ifdef CONFIG_40x
-EXTRA_AFLAGS := -Wa,-m405
-endif
-
# Extra CFLAGS so we don't have to do relative includes
CFLAGS_pmac_setup.o += -Iarch/$(ARCH)/mm
--- 1.31/arch/ppc/syslib/Makefile 2004-08-02 01:00:42 -07:00
+++ edited/arch/ppc/syslib/Makefile 2004-08-05 11:11:14 -07:00
@@ -2,16 +2,6 @@
# Makefile for the linux kernel.
#
-ifdef CONFIG_PPC64BRIDGE
-EXTRA_AFLAGS := -Wa,-mppc64bridge
-endif
-ifdef CONFIG_4xx
-EXTRA_AFLAGS := -Wa,-m405
-endif
-ifdef CONFIG_E500
-EXTRA_AFLAGS := -Wa,-me500
-endif
-
CFLAGS_prom_init.o += -fPIC
CFLAGS_btext.o += -fPIC
--
Tom Rini
http://gate.crashing.org/~trini/
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c
2004-07-30 19:07 ` Tom Rini
2004-07-30 20:48 ` Giuliano Pochini
@ 2004-08-02 7:45 ` Olaf Hering
1 sibling, 0 replies; 27+ messages in thread
From: Olaf Hering @ 2004-08-02 7:45 UTC (permalink / raw)
To: Tom Rini
Cc: Giuliano Pochini, kumar.gala, tnt, linuxppc-dev, linux-kernel,
akpm
On Fri, Jul 30, Tom Rini wrote:
> On Fri, Jul 30, 2004 at 08:59:01PM +0200, Giuliano Pochini wrote:
>
> > On Thu, 29 Jul 2004 07:43:47 -0700
> > Tom Rini <trini@kernel.crashing.org> wrote:
> > > > I had no time to do a lot of testing, but it seems that binutils 2.15 +
> > > > gcc 3.3.3 is a bad one too. I didn't try to compile the kernel (which
> > > > may also break), but at least I couldn't compile gcc 3.4.1 with the
> > > > above combination. It seems that as doesn't get the -mxxx parameter
> > > > required to compile altivec stuff. Hacking the Makefile to make it
> > > > pass -Wa,-m7455 helped a little, but it eventually failed in another
> > > > weird way. I hadn't time to investigate further, sorry.
> > >
> > > Stock gcc-3.3.3 or from the hammer branch ?
> >
> > Stock.
>
> That is interesting. Olaf, is gcc-3.3.x + binutils-2.15 one of the
> combinations you've got in your matrix of toolchains?
yes, it worked ok for kernel builds. gcc-3_3-branch + binutils 2.15
--
USB is for mice, FireWire is for men!
sUse lINUX ag, nÜRNBERG
^ permalink raw reply [flat|nested] 27+ messages in thread
end of thread, other threads:[~2004-08-06 15:46 UTC | newest]
Thread overview: 27+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-07-28 15:46 [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 check Tom Rini
2004-07-28 22:00 ` Andrew Morton
2004-07-28 22:07 ` Tom Rini
2004-07-28 22:51 ` Andrew Morton
2004-07-29 15:04 ` Sylvain Munaut
2004-07-29 8:05 ` [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c Giuliano Pochini
2004-07-29 14:43 ` Tom Rini
2004-07-30 18:59 ` Giuliano Pochini
2004-07-30 19:07 ` Tom Rini
2004-07-30 20:48 ` Giuliano Pochini
2004-07-30 21:03 ` Tom Rini
2004-08-04 10:22 ` Wolfram Quester
2004-08-04 10:41 ` Guido Guenther
2004-08-04 12:37 ` Guido Guenther
2004-08-05 14:12 ` Olaf Hering
2004-08-05 16:54 ` Tom Rini
2004-08-05 17:00 ` Olaf Hering
2004-08-05 17:20 ` Tom Rini
2004-08-05 17:39 ` Olaf Hering
2004-08-05 18:03 ` Tom Rini
2004-08-05 18:00 ` Olaf Hering
2004-08-05 18:14 ` Tom Rini
2004-08-05 19:04 ` Sam Ravnborg
2004-08-05 21:03 ` Matt Porter
2004-08-06 15:32 ` Olaf Hering
[not found] ` <hhacx8hirb.fsf@alsvidh.mathematik.uni-muenchen.de>
2004-08-06 15:38 ` [PATCH] ppc32: Fix building of certain CPU types (Was: Re: [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14) Tom Rini
2004-08-02 7:45 ` [PATCH][PPC32] Makefile cleanups and gcc-3.4+binutils-2.14 c Olaf Hering
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox