From: H Hartley Sweeten <hartleys@visionengravers.com>
To: Russell King - ARM Linux <linux@arm.linux.org.uk>
Cc: Linux Kernel <linux-kernel@vger.kernel.org>,
ARM Kernel <linux-arm-kernel@lists.infradead.org>,
"mikpe@it.uu.se" <mikpe@it.uu.se>,
"ebiederm@xmission.com" <ebiederm@xmission.com>,
"vapier@gentoo.org" <vapier@gentoo.org>,
"tony.luck@intel.com" <tony.luck@intel.com>
Subject: RE: [PATCH] ARM: quiet sparse noise due to __ARCH_WANT_SYS_RT_SIG(ACTION|SUSPEND)
Date: Tue, 21 Jun 2011 16:08:09 -0500 [thread overview]
Message-ID: <ADE657CA350FB648AAC2C43247A983F001F38227C2CB@AUSP01VMBX24.collaborationhost.net> (raw)
In-Reply-To: <20110621191624.GK23234@n2100.arm.linux.org.uk>
On Tuesday, June 21, 2011 12:16 PM, Russell King wrote:
> On Tue, Jun 21, 2011 at 12:09:45PM -0700, H Hartley Sweeten wrote:
>> ARM defines __ARCH_WANT_SYS_RT_SIG(ACTION|SUSPEND) which
>> produces the following sparse warnings in kernel/signal.c:
>>
>> warning: symbol 'sys_rt_sigaction' was not declared. Should it be static?
>> warning: symbol 'sys_rt_sigsuspend' was not declared. Should it be static?
>>
>> Since ARM doesn't include <asm-generic/syscalls.h>, due to different
>> calling conventions for some system calls, prototype the functions
>> in <asm/unistd.h> to quiet the noise.
>
> NAK.
>
> asm/unistd.h is an exported header. Please don't pollute it with kernel
> internal stuff. Instead, follow the hint in asm-generic and create a new
> header file to describe syscalls called... asm/syscalls.h.
>
> And.. the problem is a lot larger than you mention above:
>
> arch/arm/kernel/signal.c:68:16: warning: symbol 'sys_sigsuspend' was not declared. Should it be static?
> arch/arm/kernel/signal.c:84:1: warning: symbol 'sys_sigaction' was not declared. Should it be static?
> arch/arm/kernel/signal.c:328:16: warning: symbol 'sys_sigreturn' was not declared. Should it be static?
> arch/arm/kernel/signal.c:360:16: warning: symbol 'sys_rt_sigreturn' was not declared. Should it be static?
> arch/arm/kernel/sys_arm.c:34:16: warning: symbol 'sys_fork' was not declared. Should it be static?
> arch/arm/kernel/sys_arm.c:47:16: warning: symbol 'sys_clone' was not declared. Should it be static?
> arch/arm/kernel/sys_arm.c:57:16: warning: symbol 'sys_vfork' was not declared. Should it be static?
> arch/arm/kernel/sys_arm.c:65:16: warning: symbol 'sys_execve' was not declared. Should it be static?
> arch/arm/kernel/sys_arm.c:129:17: warning: symbol 'sys_arm_fadvise64_64' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:144:17: warning: symbol 'sys_oabi_stat64' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:154:17: warning: symbol 'sys_oabi_lstat64' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:164:17: warning: symbol 'sys_oabi_fstat64' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:174:17: warning: symbol 'sys_oabi_fstatat64' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:196:17: warning: symbol 'sys_oabi_fcntl64' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:249:17: warning: symbol 'sys_oabi_epoll_ctl' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:270:17: warning: symbol 'sys_oabi_epoll_wait' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:304:17: warning: symbol 'sys_oabi_semtimedop' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:343:17: warning: symbol 'sys_oabi_semop' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:349:16: warning: symbol 'sys_oabi_ipc' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:367:17: warning: symbol 'sys_oabi_bind' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:377:17: warning: symbol 'sys_oabi_connect' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:387:17: warning: symbol 'sys_oabi_sendto' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:400:17: warning: symbol 'sys_oabi_sendmsg' was not declared. Should it be static?
> arch/arm/kernel/sys_oabi-compat.c:426:17: warning: symbol 'sys_oabi_socketcall' was not declared. Should it be static?
> arch/arm/kernel/traps.c:464:16: warning: symbol 'arm_syscall' was not declared. Should it be static?
>
> So you actually need something like this (pasted, so whitespace
> damaged). I'll pull it out of my low priority queue and queue it up
> for the next merge window.
Yes, the problem is bigger than just sys_rt_sig(action|suspend).
I was also working on fixing the issues above by adding an asm/syscalls.h.
I just wasn't sure how to handle the sys_rt_sig(action|suspend) functions.
Your patch was whitespace damaged so I applied it by hand to test.
It does fix the sparse issues you list above. So FWIW....
Tested-by: H Hartley Sweeten <hsweeten@visionengravers.com>
But, it still does not fix the sparse warnings in kernel/signal.c:
kernel/signal.c:2718:1: warning: symbol 'sys_rt_sigaction' was not declared. Should it be static?
kernel/signal.c:2810:1: warning: symbol 'sys_rt_sigsuspend' was not declared. Should it be static?
FYI, there are a couple other related sparse warnings in arch/arm/kernel, but
these are not exactly "syscalls":
arch/arm/kernel/ptrace.c:795:16: warning: symbol 'syscall_trace' was not declared. Should it be static?
arch/arm/kernel/signal.c:787:1: warning: symbol 'do_notify_resume' was not declared. Should it be static?
arch/arm/kernel/traps.c:343:29: warning: symbol 'do_undefinstr' was not declared. Should it be static?
arch/arm/kernel/traps.c:386:17: warning: symbol 'do_unexp_fiq' was not declared. Should it be static?
arch/arm/kernel/traps.c:398:17: warning: symbol 'bad_mode' was not declared. Should it be static?
arch/arm/kernel/traps.c:661:6: warning: symbol '__bad_xchg' was not declared. Should it be static?
arch/arm/kernel/traps.c:674:1: warning: symbol 'baddataabort' was not declared. Should it be static?
arch/arm/kernel/traps.c:706:6: warning: symbol '__readwrite_bug' was not declared. Should it be static?
arch/arm/kernel/traps.c:728:17: warning: symbol '__div0' was not declared. Should it be static?
arch/arm/kernel/traps.c:735:6: warning: symbol 'abort' was not declared. Should it be static?
Regards,
Hartley
next prev parent reply other threads:[~2011-06-21 21:08 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-06-21 19:09 H Hartley Sweeten
2011-06-21 19:16 ` Russell King - ARM Linux
2011-06-21 21:08 ` H Hartley Sweeten [this message]
2011-06-21 19:31 ` Mike Frysinger
2011-06-21 22:32 ` H Hartley Sweeten
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ADE657CA350FB648AAC2C43247A983F001F38227C2CB@AUSP01VMBX24.collaborationhost.net \
--to=hartleys@visionengravers.com \
--cc=ebiederm@xmission.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@arm.linux.org.uk \
--cc=mikpe@it.uu.se \
--cc=tony.luck@intel.com \
--cc=vapier@gentoo.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®