Re: [PATCH 01/43] signal: Add block_sigmask() for adding sigmask to current->blocked

[Oleg Nesterov <oleg@redhat.com>]

On 08/22, Matt Fleming wrote:
>
> (Adding linux-arch to Cc so arch maintainers will hopefully see this)

Yes, thanks.

So far I applied only this patch to

	git://git.kernel.org/pub/scm/linux/kernel/git/oleg/misc.git ptrace

But I am going to take the whole series unless I have the nack from
maintainers. Everything looks correct.

IOW, I am going to wait a bit to collect the ACKs from the maintainers
in case (I hope ;) they want to review these changes.

OK?

Oleg.

> On Fri, 2011-08-19 at 17:46 +0100, Matt Fleming wrote:
> > From: Matt Fleming <matt.fleming@intel.com>
> > 
> > This patch abstracts the code sequence for adding a signal handler's
> > sa_mask to current->blocked because the sequence is identical for all
> > architectures. Furthermore, in the past some architectures actually
> > got this code wrong, so introduce a wrapper that all architectures can
> > use.
> > 
> > Cc: Oleg Nesterov <oleg@redhat.com>
> > Cc: Thomas Gleixner <tglx@linutronix.de>
> > Cc: Ingo Molnar <mingo@elte.hu>
> > Cc: H. Peter Anvin <hpa@zytor.com>
> > Cc: Tejun Heo <tj@kernel.org>
> > Cc: Andrew Morton <akpm@linux-foundation.org>
> > Signed-off-by: Matt Fleming <matt.fleming@intel.com>
> > ---
> >  arch/x86/kernel/signal.c |    6 +-----
> >  include/linux/signal.h   |    1 +
> >  kernel/signal.c          |   21 +++++++++++++++++++++
> >  3 files changed, 23 insertions(+), 5 deletions(-)
> > 
> > diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
> > index 54ddaeb2..46a01bd 100644
> > --- a/arch/x86/kernel/signal.c
> > +++ b/arch/x86/kernel/signal.c
> > @@ -682,7 +682,6 @@ static int
> >  handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
> >  		struct pt_regs *regs)
> >  {
> > -	sigset_t blocked;
> >  	int ret;
> >  
> >  	/* Are we from a system call? */
> > @@ -733,10 +732,7 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
> >  	 */
> >  	regs->flags &= ~X86_EFLAGS_TF;
> >  
> > -	sigorsets(&blocked, &current->blocked, &ka->sa.sa_mask);
> > -	if (!(ka->sa.sa_flags & SA_NODEFER))
> > -		sigaddset(&blocked, sig);
> > -	set_current_blocked(&blocked);
> > +	block_sigmask(ka, sig);
> >  
> >  	tracehook_signal_handler(sig, info, ka, regs,
> >  				 test_thread_flag(TIF_SINGLESTEP));
> > diff --git a/include/linux/signal.h b/include/linux/signal.h
> > index a822300..7987ce74 100644
> > --- a/include/linux/signal.h
> > +++ b/include/linux/signal.h
> > @@ -254,6 +254,7 @@ extern void set_current_blocked(const sigset_t *);
> >  extern int show_unhandled_signals;
> >  
> >  extern int get_signal_to_deliver(siginfo_t *info, struct k_sigaction *return_ka, struct pt_regs *regs, void *cookie);
> > +extern void block_sigmask(struct k_sigaction *ka, int signr);
> >  extern void exit_signals(struct task_struct *tsk);
> >  
> >  extern struct kmem_cache *sighand_cachep;
> > diff --git a/kernel/signal.c b/kernel/signal.c
> > index 291c970..7a08164 100644
> > --- a/kernel/signal.c
> > +++ b/kernel/signal.c
> > @@ -2314,6 +2314,27 @@ relock:
> >  	return signr;
> >  }
> >  
> > +/**
> > + * block_sigmask - add @ka's signal mask to current->blocked
> > + * @ka: action for @signr
> > + * @signr: signal that has been successfully delivered
> > + *
> > + * This function should be called when a signal has succesfully been
> > + * delivered. It adds the mask of signals for @ka to current->blocked
> > + * so that they are blocked during the execution of the signal
> > + * handler. In addition, @signr will be blocked unless %SA_NODEFER is
> > + * set in @ka->sa.sa_flags.
> > + */
> > +void block_sigmask(struct k_sigaction *ka, int signr)
> > +{
> > +	sigset_t blocked;
> > +
> > +	sigorsets(&blocked, &current->blocked, &ka->sa.sa_mask);
> > +	if (!(ka->sa.sa_flags & SA_NODEFER))
> > +		sigaddset(&blocked, signr);
> > +	set_current_blocked(&blocked);
> > +}
> > +
> >  /*
> >   * It could be that complete_signal() picked us to notify about the
> >   * group-wide signal. Other threads should be notified now to take
> 
> 
>