lkml.org 
[lkml]   [2011]   [Sep]   [27]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH v5 3.1.0-rc4-tip 26/26] uprobes: queue signals while thread is singlestepping.
* Peter Zijlstra <peterz@infradead.org> [2011-09-27 15:03:46]:

> On Tue, 2011-09-20 at 17:35 +0530, Srikar Dronamraju wrote:
> > +#ifdef CONFIG_UPROBES
> > + if (!group && t->utask && t->utask->active_uprobe)
> > + pending = &t->utask->delayed;
> > +#endif
> > +
> > /*
> > * Short-circuit ignored signals and support queuing
> > * exactly one non-rt signal, so that we can get more
> > @@ -1106,6 +1111,11 @@ static int __send_signal(int sig, struct siginfo *info, struct task_struct *t,
> > }
> > }
> >
> > +#ifdef CONFIG_UPROBES
> > + if (!group && t->utask && t->utask->active_uprobe)
> > + return 0;
> > +#endif
> > +
> > out_set:
> > signalfd_notify(t, sig);
> > sigaddset(&pending->signal, sig);
> > @@ -1569,6 +1579,13 @@ int send_sigqueue(struct sigqueue *q, struct task_struct *t, int group)
> > }
> > q->info.si_overrun = 0;
> >
> > +#ifdef CONFIG_UPROBES
> > + if (!group && t->utask && t->utask->active_uprobe) {
> > + pending = &t->utask->delayed;
> > + list_add_tail(&q->list, &pending->list);
> > + goto out;
> > + }
> > +#endif
> > signalfd_notify(t, sig);
> > pending = group ? &t->signal->shared_pending : &t->pending;
> > list_add_tail(&q->list, &pending->list);
> > @@ -2199,7 +2216,10 @@ int get_signal_to_deliver(siginfo_t *info, struct k_sigaction *return_ka,
> > spin_unlock_irq(&sighand->siglock);
> > goto relock;
> > }
> > -
> > +#ifdef CONFIG_UPROBES
> > + if (current->utask && current->utask->active_uprobe)
> > + break;
> > +#endif
>
> That's just crying for something like:
>
> #ifdef CONFIG_UPROBES
> static inline bool uprobe_delay_signal(struct task_struct *p)
> {
> return p->utask && p->utask->active_uprobe;
> }
> #else
> static inline bool uprobe_delay_signal(struct task_struct *p)
> {
> return false;
> }
> #endif
>
> That'll instantly kill the #ifdeffery as well as describe wtf you're
> actually doing.


Okay,

I did a rethink and implemented this patch a little differently using
block_all_signals, unblock_all_signals. This wouldnt need the
#ifdeffery + no changes in kernel/signal.c

Will post the same in the next patchset.

--
Thanks and Regards
Srikar


\
 
 \ /
  Last update: 2011-09-27 15:29    [W:0.105 / U:21.100 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site