* Re: [PATCH 2/2] perf Makefile: default WERROR to off.
2014-08-26 6:53 ` [PATCH 2/2] perf Makefile: default WERROR to off Namhyung Kim
@ 2014-08-26 12:11 ` John Spencer
0 siblings, 0 replies; 2+ messages in thread
From: John Spencer @ 2014-08-26 12:11 UTC (permalink / raw)
To: Namhyung Kim; +Cc: linux-perf-users, linux-kernel
Namhyung Kim wrote:
> On Mon, 25 Aug 2014 21:37:08 +0200, John Spencer wrote:
>> From 4a72032828a6784d93f4becf723303a17d723544 Mon Sep 17 00:00:00 2001
>> From: John Spencer <maillist-linux@barfooze.de>
>> Date: Mon, 25 Aug 2014 21:25:43 +0200
>> Subject: [PATCH 2/2] perf Makefile: default WERROR to off.
>>
>> Having WERROR on by default breaks build everywhere the author hasn't
>> tested so far; including musl libc which warns about wrong includes.
>>
>> The unsuspecting user will think there's something broken and only
>> if he's coureageuous enough to grep the Makefile infrastructure find
>> out that he may be able to get the build working with WERROR=0.
>>
>> OTOH with WERROR defaulting to off, anything will work for the user
>> and the maintainer can knowingly use WERROR=1 to do his test builds.
>
> I think it's intentionally turned on to fix any warnings in the first
> place. But yes, it might miss some non-popular systems tho..
yes, and it's sufficient if someone tries to build it with a
newer/different compiler/libc version/whatever to make it fail.
having -Werror on by default is just silly and annoying because 90% of
warnings (at least with -Wall) are utterly unimportant (like unused
variables etc).
there's probably a good reason the kernel itself is not built with -Werror.
(in my head i call people hardcoding -Werror "Werror-nazis"...)
but i agree that *some* warnings should be fatal, for example implicit
function declarations, or pointer-to-int casts, because those can
actually break code.
>
> How many errors/warnings do you see on your system (musl libc?). Any
> chance to post fixes instead?
there are tons of warnings, mostly from mixing kernel headers with
userspace headers (and libc_compat.h is only patched for GLIBC - my
musl-related libc_compat.h patches have been ignored so far, but they're
archived at https://github.com/sabotage-linux/kernel-headers ).
In file included from perf.c:12:0:
util/cache.h:74:15: warning: redundant redeclaration of 'strlcpy'
[-Wredundant-decls]
In file included from util/util.h:53:0,
from builtin.h:4,
from perf.c:9:
/usr/include/string.h:84:8: note: previous declaration of 'strlcpy' was here
In file included from util/util.h:67:0,
from builtin.h:4,
from perf.c:9:
/usr/include/sys/poll.h:1:2: warning: #warning redirecting incorrect
#include <sys/poll.h> to <poll.h> [-Wcpp]
In file included from
/src/build/kernel/linux-3.14.14.org/arch/x86/include/uapi/asm/ioctl.h:1:0,
from
/src/build/kernel/linux-3.14.14.org/include/uapi/linux/ioctl.h:4,
from
/src/build/kernel/linux-3.14.14.org/include/uapi/linux/perf_event.h:18,
from util/../perf.h:156,
from util/cache.h:7,
from perf.c:12:
/src/build/kernel/linux-3.14.14.org/include/uapi/asm-generic/ioctl.h:65:0:
warning: "_IOC" redefined [enabled by default]
In file included from /usr/include/sys/ioctl.h:7:0,
from util/util.h:69,
from builtin.h:4,
from perf.c:9:
/usr/include/bits/ioctl.h:1:0: note: this is the location of the
previous definition
In file included from
/src/build/kernel/linux-3.14.14.org/arch/x86/include/uapi/asm/ioctl.h:1:0,
from
/src/build/kernel/linux-3.14.14.org/include/uapi/linux/ioctl.h:4,
from
/src/build/kernel/linux-3.14.14.org/include/uapi/linux/perf_event.h:18,
from util/../perf.h:156,
from util/cache.h:7,
from perf.c:12:
/src/build/kernel/linux-3.14.14.org/include/uapi/asm-generic/ioctl.h:76:0:
warning: "_IO" redefined [enabled by default]
(same for _IOR, _IOW, _IOWR, _IOC)
these warnings are repeated over an over again as they come from headers.
so basically we have
- util/cache.h:74:15: warning: redundant redeclaration of 'strlcpy'
- uapi/asm-generic/ioctl.h: warning: "_IO*" redefined (userspace and
kernel space defining the same thing)
- util/util.h: incorrect inclusion of sys/poll.h rather than poll.h (see
man 3p poll)
tho it's possible I may have missed something else in the wall of warnings.
Thanks,
--JS
>
> Thanks,
> Namhyung
>
>
>> Signed-off-by: John Spencer <maillist-linux@barfooze.de>
>>
>> ---
>> tools/perf/config/Makefile | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/tools/perf/config/Makefile b/tools/perf/config/Makefile
>> index 1f67aa0..b1d639a 100644
>> --- a/tools/perf/config/Makefile
>> +++ b/tools/perf/config/Makefile
>> @@ -102,7 +102,7 @@ ifeq ($(call get-executable,$(BISON)),)
>> endif
>>
>> # Treat warnings as errors unless directed not to
>> -ifneq ($(WERROR),0)
>> +ifeq ($(WERROR),1)
>> CFLAGS += -Werror
>> endif
>
^ permalink raw reply [flat|nested] 2+ messages in thread