* linux-next: build warning after merge of the kbuild tree
@ 2011-05-02 2:18 Stephen Rothwell
2011-05-02 2:24 ` Dave Jones
0 siblings, 1 reply; 15+ messages in thread
From: Stephen Rothwell @ 2011-05-02 2:18 UTC (permalink / raw)
To: Michal Marek; +Cc: linux-next, linux-kernel, Dave Jones, Sam Ravnborg
[-- Attachment #1: Type: text/plain, Size: 483 bytes --]
Hi Michal,
After merging the kbuild tree, today's linux-next build (x86_64
allmodconfig) produced this warning:
At top level:
cc1: warning: unrecognized command line option "-Wno-unused-but-set-variable"
Presumably caused by commit af0e5d565d2f ("kbuild: Disable
-Wunused-but-set-variable for gcc 4.6.0") though I don't see why. I am
using gcc version 4.4.5.
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 2:18 linux-next: build warning after merge of the kbuild tree Stephen Rothwell
@ 2011-05-02 2:24 ` Dave Jones
2011-05-02 3:45 ` Stephen Rothwell
0 siblings, 1 reply; 15+ messages in thread
From: Dave Jones @ 2011-05-02 2:24 UTC (permalink / raw)
To: Stephen Rothwell; +Cc: Michal Marek, linux-next, linux-kernel, Sam Ravnborg
On Mon, May 02, 2011 at 12:18:17PM +1000, Stephen Rothwell wrote:
> Hi Michal,
>
> After merging the kbuild tree, today's linux-next build (x86_64
> allmodconfig) produced this warning:
>
> At top level:
> cc1: warning: unrecognized command line option "-Wno-unused-but-set-variable"
>
> Presumably caused by commit af0e5d565d2f ("kbuild: Disable
> -Wunused-but-set-variable for gcc 4.6.0") though I don't see why. I am
> using gcc version 4.4.5.
Which version of the patch got merged ? This should be guarded by cc-option checks.
Dave
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 2:24 ` Dave Jones
@ 2011-05-02 3:45 ` Stephen Rothwell
2011-05-02 4:02 ` Dave Jones
2011-05-02 4:24 ` Stephen Rothwell
0 siblings, 2 replies; 15+ messages in thread
From: Stephen Rothwell @ 2011-05-02 3:45 UTC (permalink / raw)
To: Dave Jones; +Cc: Michal Marek, linux-next, linux-kernel, Sam Ravnborg
Hi Dave,
On Sun, 1 May 2011 22:24:18 -0400 Dave Jones <davej@redhat.com> wrote:
>
> Which version of the patch got merged ? This should be guarded by cc-option checks.
It looks right to me (that's why I wrote "though I don't see why") but
empirically we have a problem.
This is what is in -next today:
From: Dave Jones <davej@redhat.com>
Date: Thu, 21 Apr 2011 17:28:13 -0400
Subject: [PATCH] kbuild: Disable -Wunused-but-set-variable for gcc 4.6.0
Disable the new -Wunused-but-set-variable that was added in gcc 4.6.0
It produces more false positives than useful warnings.
This can still be enabled using W=1
Signed-off-by: Dave Jones <davej@redhat.com>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Tested-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Michal Marek <mmarek@suse.cz>
---
Makefile | 4 ++++
scripts/Makefile.build | 1 +
2 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/Makefile b/Makefile
index e509cc7..4527dc2 100644
--- a/Makefile
+++ b/Makefile
@@ -567,6 +567,10 @@ ifndef CONFIG_CC_STACKPROTECTOR
KBUILD_CFLAGS += $(call cc-option, -fno-stack-protector)
endif
+# This warning generated too much noise in a regular build.
+# Use make W=1 to enable this warning (see scripts/Makefile.build)
+KBUILD_CFLAGS += $(call cc-option, -Wno-unused-but-set-variable)
+
ifdef CONFIG_FRAME_POINTER
KBUILD_CFLAGS += -fno-omit-frame-pointer -fno-optimize-sibling-calls
else
diff --git a/scripts/Makefile.build b/scripts/Makefile.build
index 9c0c481..9fb19c0 100644
--- a/scripts/Makefile.build
+++ b/scripts/Makefile.build
@@ -66,6 +66,7 @@ warning-1 += -Wmissing-format-attribute
warning-1 += -Wmissing-prototypes
warning-1 += -Wold-style-definition
warning-1 += $(call cc-option, -Wmissing-include-dirs)
+warning-1 += $(call cc-option, -Wunused-but-set-variable)
warning-2 := -Waggregate-return
warning-2 += -Wcast-align
--
1.7.4.4
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
^ permalink raw reply related [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 3:45 ` Stephen Rothwell
@ 2011-05-02 4:02 ` Dave Jones
2011-05-02 4:44 ` Stephen Rothwell
2011-05-02 4:24 ` Stephen Rothwell
1 sibling, 1 reply; 15+ messages in thread
From: Dave Jones @ 2011-05-02 4:02 UTC (permalink / raw)
To: Stephen Rothwell; +Cc: Michal Marek, linux-next, linux-kernel, Sam Ravnborg
On Mon, May 02, 2011 at 01:45:24PM +1000, Stephen Rothwell wrote:
> diff --git a/scripts/Makefile.build b/scripts/Makefile.build
> index 9c0c481..9fb19c0 100644
> --- a/scripts/Makefile.build
> +++ b/scripts/Makefile.build
> @@ -66,6 +66,7 @@ warning-1 += -Wmissing-format-attribute
> warning-1 += -Wmissing-prototypes
> warning-1 += -Wold-style-definition
> warning-1 += $(call cc-option, -Wmissing-include-dirs)
> +warning-1 += $(call cc-option, -Wunused-but-set-variable)
Subtly different from what I wrote.. the version I wrote did this ..
+KBUILD_EXTRA_WARNINGS += $(call cc-option, -Wunused-but-set-variable)
I just tested with the gcc 4.4.5 on master.kernel.org, with a tip of Linus'
tree with my original diff, and it didn't complain at all.
What does gcc -v say ?
gcc version 4.4.5 20101112 (Red Hat 4.4.5-2) (GCC) seems to be fine.
(note the above hunk isn't relevant unless using W=1 builds anyway,
so I'm at a loss as to what broke it..)
Dave
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 3:45 ` Stephen Rothwell
2011-05-02 4:02 ` Dave Jones
@ 2011-05-02 4:24 ` Stephen Rothwell
2011-05-02 4:36 ` Dave Jones
2011-05-02 4:53 ` Stephen Rothwell
1 sibling, 2 replies; 15+ messages in thread
From: Stephen Rothwell @ 2011-05-02 4:24 UTC (permalink / raw)
To: Dave Jones; +Cc: Michal Marek, linux-next, linux-kernel, Sam Ravnborg
[-- Attachment #1: Type: text/plain, Size: 645 bytes --]
Hi Dave,
On Mon, 2 May 2011 13:45:24 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> On Sun, 1 May 2011 22:24:18 -0400 Dave Jones <davej@redhat.com> wrote:
> >
> > Which version of the patch got merged ? This should be guarded by cc-option checks.
>
> It looks right to me (that's why I wrote "though I don't see why") but
> empirically we have a problem.
I do not get it for every compile fo a file, just a few times during a
whole kernel build. Some config/arches do not produce any (like sparc32
defconfig).
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 4:24 ` Stephen Rothwell
@ 2011-05-02 4:36 ` Dave Jones
2011-05-02 5:57 ` Stephen Rothwell
2011-05-02 4:53 ` Stephen Rothwell
1 sibling, 1 reply; 15+ messages in thread
From: Dave Jones @ 2011-05-02 4:36 UTC (permalink / raw)
To: Stephen Rothwell; +Cc: Michal Marek, linux-next, linux-kernel, Sam Ravnborg
On Mon, May 02, 2011 at 02:24:54PM +1000, Stephen Rothwell wrote:
> Hi Dave,
>
> On Mon, 2 May 2011 13:45:24 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > On Sun, 1 May 2011 22:24:18 -0400 Dave Jones <davej@redhat.com> wrote:
> > >
> > > Which version of the patch got merged ? This should be guarded by cc-option checks.
> >
> > It looks right to me (that's why I wrote "though I don't see why") but
> > empirically we have a problem.
>
> I do not get it for every compile fo a file, just a few times during a
> whole kernel build. Some config/arches do not produce any (like sparc32
> defconfig).
That's interesting. That suggests that something isn't being built with KBUILD_CFLAGS.
I did a full defconfig build on x86-64 successfully (allmodconfig is churning away).
Can you point me at something that doesn't build ?
Dave
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 4:02 ` Dave Jones
@ 2011-05-02 4:44 ` Stephen Rothwell
0 siblings, 0 replies; 15+ messages in thread
From: Stephen Rothwell @ 2011-05-02 4:44 UTC (permalink / raw)
To: Dave Jones; +Cc: Michal Marek, linux-next, linux-kernel, Sam Ravnborg
[-- Attachment #1: Type: text/plain, Size: 1656 bytes --]
Hi Dave,
On Mon, 2 May 2011 00:02:16 -0400 Dave Jones <davej@redhat.com> wrote:
>
> Subtly different from what I wrote.. the version I wrote did this ..
>
> +KBUILD_EXTRA_WARNINGS += $(call cc-option, -Wunused-but-set-variable)
Due to other changes in the kbuild tree, I suspect. However, as you say,
this part is not relevant to the warning I am getting.
> I just tested with the gcc 4.4.5 on master.kernel.org, with a tip of Linus'
> tree with my original diff, and it didn't complain at all.
Also see my other email. For a i386 defconfig build, I get this warning
only twice, for a sparc64 defconfig I get 4 and for a sparc32 defconfig,
none:
At top level:
cc1: warning: unrecognized command line option "-Wno-unused-but-set-variable"
> What does gcc -v say ?
> gcc version 4.4.5 20101112 (Red Hat 4.4.5-2) (GCC) seems to be fine.
$ /opt/cross/gcc-4.4.5-nolibc/x86_64-linux/bin/x86_64-linux-gcc -v
Using built-in specs.
Target: x86_64-linux
Configured with: /home/tony/buildall/src/gcc/configure --target=x86_64-linux --enable-targets=all --prefix=/home/tony/buildall/cross.x86_64/gcc-4.4.5-nolibc/x86_64-linux/ --enable-languages=c --with-newlib --without-headers --with-system-libunwind --disable-nls --disable-threads --disable-shared --disable-libmudflap --disable-libssp --disable-libgomp --disable-decimal-float --enable-checking=release --with-mpfr=/home/tony/buildall/src/sys-x86_64 --with-gmp=/home/tony/buildall/src/sys-x86_64 --disable-bootstrap
Thread model: single
gcc version 4.4.5 (GCC)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 4:24 ` Stephen Rothwell
2011-05-02 4:36 ` Dave Jones
@ 2011-05-02 4:53 ` Stephen Rothwell
2011-05-02 11:13 ` Michal Marek
1 sibling, 1 reply; 15+ messages in thread
From: Stephen Rothwell @ 2011-05-02 4:53 UTC (permalink / raw)
To: Dave Jones; +Cc: Michal Marek, linux-next, linux-kernel, Sam Ravnborg
[-- Attachment #1: Type: text/plain, Size: 2495 bytes --]
Hi Dave,
hmmm,
On my build machine:
$ /opt/cross/gcc-4.4.5-nolibc/x86_64-linux/bin/x86_64-linux-gcc -Wno-unused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
0
$ /opt/cross/gcc-4.4.5-nolibc/x86_64-linux/bin/x86_64-linux-gcc -Wunused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
cc1: error: unrecognized command line option "-Wunused-but-set-variable"
1
$ gcc -Wno-unused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
0
$ gcc -Wunused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
cc1: error: unrecognized command line option "-Wunused-but-set-variable"
1
$ gcc -v
Using built-in specs.
Target: x86_64-linux-gnu
Configured with: ../src/configure -v --with-pkgversion='Debian 4.4.5-8' --with-bugurl=file:///usr/share/doc/gcc-4.4/README.Bugs --enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr --program-suffix=-4.4 --enable-shared --enable-multiarch --enable-linker-build-id --with-system-zlib --libexecdir=/usr/lib --without-included-gettext --enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.4 --libdir=/usr/lib --enable-nls --enable-clocale=gnu --enable-libstdcxx-debug --enable-objc-gc --with-arch-32=i586 --with-tune=generic --enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu --target=x86_64-linux-gnu
Thread model: posix
gcc version 4.4.5 (Debian 4.4.5-8)
On master.kernel.org:
$ gcc -Wno-unused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
0
$ gcc -Wunused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
0
$ gcc -v
Using built-in specs.
Target: x86_64-redhat-linux
Configured with: ../configure --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-bugurl=http://bugzilla.redhat.com/bugzilla --enable-bootstrap --enable-shared --enable-threads=posix --enable-checking=release --with-system-zlib --enable-__cxa_atexit --disable-libunwind-exceptions --enable-gnu-unique-object --enable-languages=c,c++,objc,obj-c++,java,fortran,ada --enable-java-awt=gtk --disable-dssi --with-java-home=/usr/lib/jvm/java-1.5.0-gcj-1.5.0.0/jre --enable-libgcj-multifile --enable-java-maintainer-mode --with-ecj-jar=/usr/share/java/eclipse-ecj.jar --disable-libjava-multilib --with-ppl --with-cloog --with-tune=generic --with-arch_32=i686 --build=x86_64-redhat-linux
Thread model: posix
gcc version 4.4.5 20101112 (Red Hat 4.4.5-2) (GCC)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 4:36 ` Dave Jones
@ 2011-05-02 5:57 ` Stephen Rothwell
0 siblings, 0 replies; 15+ messages in thread
From: Stephen Rothwell @ 2011-05-02 5:57 UTC (permalink / raw)
To: Dave Jones; +Cc: Michal Marek, linux-next, linux-kernel, Sam Ravnborg
[-- Attachment #1: Type: text/plain, Size: 2477 bytes --]
Hi Dave,
On Mon, 2 May 2011 00:36:43 -0400 Dave Jones <davej@redhat.com> wrote:
>
> That's interesting. That suggests that something isn't being built with KBUILD_CFLAGS.
> I did a full defconfig build on x86-64 successfully (allmodconfig is churning away).
> Can you point me at something that doesn't build ?
I have just done an i386 defconfig build with V=1. Every compilation has
that flag passed to it, but only one complains:
i386-linux-gcc -Wp,-MD,kernel/.futex.o.d -nostdinc -isystem /opt/cross/gcc-4.4.5-nolibc/i386-linux/bin/../lib/gcc/i386-linux/4.4.5/include -I/scratch/sfr/next/arch/x86/include -Iarch/x86/include/generated -Iinclude -I/scratch/sfr/next/include -include include/generated/autoconf.h -I/scratch/sfr/next/kernel -Ikernel -D__KERNEL__ -Wall -Wundef -Wstrict-prototypes -Wno-trigraphs -fno-strict-aliasing -fno-common -Werror-implicit-function-declaration -Wno-format-security -fno-delete-null-pointer-checks -Os -m32 -msoft-float -mregparm=3 -freg-struct-return -mpreferred-stack-boundary=2 -march=i686 -mtune=generic -Wa,-mtune=generic32 -ffreestanding -DCONFIG_AS_CFI=1 -DCONFIG_AS_CFI_SIGNAL_FRAME=1 -DCONFIG_AS_CFI_SECTIONS=1 -pipe -Wno-sign-compare -fno-asynchronous-unwind-tables -mno-sse -mno-mmx -mno-sse2 -mno-3dnow -Wframe-larger-than=2048 -fno-stack-protector -Wno-unused-but-set-variable -fno-omit-frame-pointer -fno-optimize-sibling-calls -Wdeclaration-after-statement -Wno-pointer-sign -fno-strict-overflow -fconserve-stack -D"KBUILD_STR(s)=#s" -D"KBUILD_BASENAME=KBUILD_STR(futex)" -D"KBUILD_MODNAME=KBUILD_STR(futex)" -c -o kernel/futex.o /scratch/sfr/next/kernel/futex.c
/scratch/sfr/next/kernel/futex.c: In function 'fixup_pi_state_owner':
/scratch/sfr/next/kernel/futex.c:1549: warning: 'curval' may be used uninitialized in this function
/scratch/sfr/next/kernel/futex.c: In function 'futex_lock_pi_atomic':
/scratch/sfr/next/kernel/futex.c:678: warning: 'curval' may be used uninitialized in this function
/scratch/sfr/next/kernel/futex.c: In function 'handle_futex_death':
/scratch/sfr/next/kernel/futex.c:2454: warning: 'nval' may be used uninitialized in this function
At top level:
cc1: warning: unrecognized command line option "-Wno-unused-but-set-variable"
I removed kernel/futex.o and recompiled to verify. futex.c doesn't seem
to be built any differently to anything else.
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 4:53 ` Stephen Rothwell
@ 2011-05-02 11:13 ` Michal Marek
2011-05-02 12:13 ` [PATCH] kbuild: Fix passing -Wno-* options to gcc 4.4+ Michal Marek
2011-05-02 15:17 ` linux-next: build warning after merge of the kbuild tree Valdis.Kletnieks
0 siblings, 2 replies; 15+ messages in thread
From: Michal Marek @ 2011-05-02 11:13 UTC (permalink / raw)
To: Stephen Rothwell; +Cc: Dave Jones, linux-next, linux-kernel, Sam Ravnborg
On Mon, May 02, 2011 at 02:53:38PM +1000, Stephen Rothwell wrote:
> Hi Dave,
>
> hmmm,
>
> On my build machine:
> $ /opt/cross/gcc-4.4.5-nolibc/x86_64-linux/bin/x86_64-linux-gcc -Wno-unused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
> 0
> $ /opt/cross/gcc-4.4.5-nolibc/x86_64-linux/bin/x86_64-linux-gcc -Wunused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
> cc1: error: unrecognized command line option "-Wunused-but-set-variable"
> 1
> $ gcc -Wno-unused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
> 0
> $ gcc -Wunused-but-set-variable -c -xc /dev/null -o /tmp/xx.o; echo $?
> cc1: error: unrecognized command line option "-Wunused-but-set-variable"
> 1
I reproduced this myself with 4.5.1 and 4.6.0 and was about to file a
bug at gcc.gnu.org, but I learned that this is a feature :-(, see here:
http://gcc.gnu.org/PR28322 . It treats invalid -Wno-* options as
warnings and it only issues them when there is another warning or error
in the source.
I have a patch below, but after learning the above, I'm going to rewrite
it to some more generic cc-disable-warning fuction, because it can hit
us anytime again.
Michal
Subject: [PATCH] kbuild: Fix passing -Wno-unused-but-set-variable to some gcc versions
Some versions of gcc will happily accept -Wno-<anything> in the
cc-option test and complain later when compiling some random file that
-Wno-unused-but-set-variable is not recognized. Work around this by
testing if gcc supports the -Wunused-but-set-variable option insteaed.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Michal Marek <mmarek@suse.cz>
diff --git a/Makefile b/Makefile
index 4527dc2..db6b6df 100644
--- a/Makefile
+++ b/Makefile
@@ -569,7 +569,12 @@ endif
# This warning generated too much noise in a regular build.
# Use make W=1 to enable this warning (see scripts/Makefile.build)
-KBUILD_CFLAGS += $(call cc-option, -Wno-unused-but-set-variable)
+# We test for support of the -Wunused-but-set-variable option, because
+# some version of gcc will happily accept any -Wno-* switch and complain
+# later during build.
+ifeq ($(call cc-option-yn, -Wunused-but-set-variable), y)
+KBUILD_CFLAGS += -Wno-unused-but-set-variable
+endif
ifdef CONFIG_FRAME_POINTER
KBUILD_CFLAGS += -fno-omit-frame-pointer -fno-optimize-sibling-calls
^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH] kbuild: Fix passing -Wno-* options to gcc 4.4+
2011-05-02 11:13 ` Michal Marek
@ 2011-05-02 12:13 ` Michal Marek
2011-05-03 0:44 ` Stephen Rothwell
2011-05-02 15:17 ` linux-next: build warning after merge of the kbuild tree Valdis.Kletnieks
1 sibling, 1 reply; 15+ messages in thread
From: Michal Marek @ 2011-05-02 12:13 UTC (permalink / raw)
To: sfr; +Cc: davej, linux-next, linux-kernel, sam
Starting with 4.4, gcc will happily accept -Wno-<anything> in the
cc-option test and complain later when compiling a file that has some
other warning. This rather unexpected behavior is intentional as per
http://gcc.gnu.org/PR28322, so work around it by testing for support of
the opposite option (without the no-). Introduce a new Makefile function
cc-disable-warning that does this and update two uses of cc-option in
the toplevel Makefile.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Michal Marek <mmarek@suse.cz>
---
Documentation/kbuild/makefiles.txt | 12 ++++++++++++
Makefile | 4 ++--
scripts/Kbuild.include | 5 +++++
3 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/Documentation/kbuild/makefiles.txt b/Documentation/kbuild/makefiles.txt
index 835b64a..47435e5 100644
--- a/Documentation/kbuild/makefiles.txt
+++ b/Documentation/kbuild/makefiles.txt
@@ -501,6 +501,18 @@ more details, with real examples.
gcc >= 3.00. For gcc < 3.00, -malign-functions=4 is used.
Note: cc-option-align uses KBUILD_CFLAGS for $(CC) options
+ cc-disable-warning
+ cc-disable-warning checks if gcc supports a given warning and returns
+ the commandline switch to disable it. This special function is needed,
+ because gcc 4.4 and later accept any unknown -Wno-* option and only
+ warn about it if there is another warning in the source file.
+
+ Example:
+ KBUILD_CFLAGS += $(call cc-disable-warning, unused-but-set-variable)
+
+ In the above example, -Wno-unused-but-set-variable will be added to
+ KBUILD_CFLAGS only if gcc really accepts it.
+
cc-version
cc-version returns a numerical version of the $(CC) compiler version.
The format is <major><minor> where both are two digits. So for example
diff --git a/Makefile b/Makefile
index 4527dc2..f919209 100644
--- a/Makefile
+++ b/Makefile
@@ -569,7 +569,7 @@ endif
# This warning generated too much noise in a regular build.
# Use make W=1 to enable this warning (see scripts/Makefile.build)
-KBUILD_CFLAGS += $(call cc-option, -Wno-unused-but-set-variable)
+KBUILD_CFLAGS += $(call cc-disable-warning, unused-but-set-variable)
ifdef CONFIG_FRAME_POINTER
KBUILD_CFLAGS += -fno-omit-frame-pointer -fno-optimize-sibling-calls
@@ -616,7 +616,7 @@ CHECKFLAGS += $(NOSTDINC_FLAGS)
KBUILD_CFLAGS += $(call cc-option,-Wdeclaration-after-statement,)
# disable pointer signed / unsigned warnings in gcc 4.0
-KBUILD_CFLAGS += $(call cc-option,-Wno-pointer-sign,)
+KBUILD_CFLAGS += $(call cc-disable-warning, pointer-sign)
# disable invalid "can't wrap" optimizations for signed / pointers
KBUILD_CFLAGS += $(call cc-option,-fno-strict-overflow)
diff --git a/scripts/Kbuild.include b/scripts/Kbuild.include
index fae2d8d..c034dd7 100644
--- a/scripts/Kbuild.include
+++ b/scripts/Kbuild.include
@@ -118,6 +118,11 @@ cc-option-yn = $(call try-run,\
cc-option-align = $(subst -functions=0,,\
$(call cc-option,-falign-functions=0,-malign-functions=0))
+# cc-disable-warning
+# Usage: cflags-y += $(call cc-disable-warning,unused-but-set-variable)
+cc-disable-warning = $(call try-run,\
+ $(CC) $(KBUILD_CPPFLAGS) $(KBUILD_CFLAGS) -W$(strip $(1)) -c -xc /dev/null -o "$$TMP",-Wno-$(strip $(1)))
+
# cc-version
# Usage gcc-ver := $(call cc-version)
cc-version = $(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-version.sh $(CC))
--
1.7.4.1
^ permalink raw reply related [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 11:13 ` Michal Marek
2011-05-02 12:13 ` [PATCH] kbuild: Fix passing -Wno-* options to gcc 4.4+ Michal Marek
@ 2011-05-02 15:17 ` Valdis.Kletnieks
2011-05-02 15:31 ` Michal Marek
1 sibling, 1 reply; 15+ messages in thread
From: Valdis.Kletnieks @ 2011-05-02 15:17 UTC (permalink / raw)
To: Michal Marek
Cc: Stephen Rothwell, Dave Jones, linux-next, linux-kernel, Sam Ravnborg
[-- Attachment #1: Type: text/plain, Size: 460 bytes --]
On Mon, 02 May 2011 13:13:37 +0200, Michal Marek said:
> +# We test for support of the -Wunused-but-set-variable option, because
> +# some version of gcc will happily accept any -Wno-* switch and complain
> +# later during build.
> +ifeq ($(call cc-option-yn, -Wunused-but-set-variable), y)
> +KBUILD_CFLAGS += -Wno-unused-but-set-variable
> +endif
Do we need to do this more generically for all the other -Wno-foo
options under the various W={123} options?
[-- Attachment #2: Type: application/pgp-signature, Size: 227 bytes --]
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: linux-next: build warning after merge of the kbuild tree
2011-05-02 15:17 ` linux-next: build warning after merge of the kbuild tree Valdis.Kletnieks
@ 2011-05-02 15:31 ` Michal Marek
0 siblings, 0 replies; 15+ messages in thread
From: Michal Marek @ 2011-05-02 15:31 UTC (permalink / raw)
To: Valdis.Kletnieks
Cc: Stephen Rothwell, Dave Jones, linux-next, linux-kernel, Sam Ravnborg
On 2.5.2011 17:17, Valdis.Kletnieks@vt.edu wrote:
> On Mon, 02 May 2011 13:13:37 +0200, Michal Marek said:
>
>> +# We test for support of the -Wunused-but-set-variable option, because
>> +# some version of gcc will happily accept any -Wno-* switch and complain
>> +# later during build.
>> +ifeq ($(call cc-option-yn, -Wunused-but-set-variable), y)
>> +KBUILD_CFLAGS += -Wno-unused-but-set-variable
>> +endif
>
> Do we need to do this more generically for all the other -Wno-foo
> options under the various W={123} options?
I only see -Wno-unused-parameter in scripts/Makefile.build and that
should be supported by any gcc version. Also please have a look at
http://marc.info/?l=linux-kernel&m=130433840412201&w=2, which is an
updated version.
Michal
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH] kbuild: Fix passing -Wno-* options to gcc 4.4+
2011-05-02 12:13 ` [PATCH] kbuild: Fix passing -Wno-* options to gcc 4.4+ Michal Marek
@ 2011-05-03 0:44 ` Stephen Rothwell
2011-05-03 8:52 ` Michal Marek
0 siblings, 1 reply; 15+ messages in thread
From: Stephen Rothwell @ 2011-05-03 0:44 UTC (permalink / raw)
To: Michal Marek; +Cc: davej, linux-next, linux-kernel, sam
[-- Attachment #1: Type: text/plain, Size: 862 bytes --]
HI Michal,
On Mon, 2 May 2011 14:13:01 +0200 Michal Marek <mmarek@suse.cz> wrote:
>
> Starting with 4.4, gcc will happily accept -Wno-<anything> in the
> cc-option test and complain later when compiling a file that has some
> other warning. This rather unexpected behavior is intentional as per
> http://gcc.gnu.org/PR28322, so work around it by testing for support of
> the opposite option (without the no-). Introduce a new Makefile function
> cc-disable-warning that does this and update two uses of cc-option in
> the toplevel Makefile.
>
> Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
> Signed-off-by: Michal Marek <mmarek@suse.cz>
This fixed things for me, thanks.
Tested-by: Stephen Rothwell <sfr@canb.auug.org.au>
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH] kbuild: Fix passing -Wno-* options to gcc 4.4+
2011-05-03 0:44 ` Stephen Rothwell
@ 2011-05-03 8:52 ` Michal Marek
0 siblings, 0 replies; 15+ messages in thread
From: Michal Marek @ 2011-05-03 8:52 UTC (permalink / raw)
To: Stephen Rothwell; +Cc: davej, linux-next, linux-kernel, sam
On 3.5.2011 02:44, Stephen Rothwell wrote:
> HI Michal,
>
> On Mon, 2 May 2011 14:13:01 +0200 Michal Marek<mmarek@suse.cz> wrote:
>>
>> Starting with 4.4, gcc will happily accept -Wno-<anything> in the
>> cc-option test and complain later when compiling a file that has some
>> other warning. This rather unexpected behavior is intentional as per
>> http://gcc.gnu.org/PR28322, so work around it by testing for support of
>> the opposite option (without the no-). Introduce a new Makefile function
>> cc-disable-warning that does this and update two uses of cc-option in
>> the toplevel Makefile.
>>
>> Reported-by: Stephen Rothwell<sfr@canb.auug.org.au>
>> Signed-off-by: Michal Marek<mmarek@suse.cz>
>
> This fixed things for me, thanks.
>
> Tested-by: Stephen Rothwell<sfr@canb.auug.org.au>
Thanks for spotting the warnings and for testing. I pushed this to
kbuild-2.6.git#kbuild.
Michal
^ permalink raw reply [flat|nested] 15+ messages in thread
end of thread, other threads:[~2011-05-03 8:52 UTC | newest]
Thread overview: 15+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2011-05-02 2:18 linux-next: build warning after merge of the kbuild tree Stephen Rothwell
2011-05-02 2:24 ` Dave Jones
2011-05-02 3:45 ` Stephen Rothwell
2011-05-02 4:02 ` Dave Jones
2011-05-02 4:44 ` Stephen Rothwell
2011-05-02 4:24 ` Stephen Rothwell
2011-05-02 4:36 ` Dave Jones
2011-05-02 5:57 ` Stephen Rothwell
2011-05-02 4:53 ` Stephen Rothwell
2011-05-02 11:13 ` Michal Marek
2011-05-02 12:13 ` [PATCH] kbuild: Fix passing -Wno-* options to gcc 4.4+ Michal Marek
2011-05-03 0:44 ` Stephen Rothwell
2011-05-03 8:52 ` Michal Marek
2011-05-02 15:17 ` linux-next: build warning after merge of the kbuild tree Valdis.Kletnieks
2011-05-02 15:31 ` Michal Marek
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).