linux-kernel - Re: [PATCH 3/4] uprobes/tracing: Kill the pointless local_save_flags/preempt

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20130409145851.GB22913@redhat.com>
Date:	Tue, 9 Apr 2013 16:58:51 +0200
From:	Oleg Nesterov <oleg@...hat.com>
To:	Steven Rostedt <rostedt@...dmis.org>
Cc:	Ananth N Mavinakayanahalli <ananth@...ibm.com>,
	Srikar Dronamraju <srikar@...ux.vnet.ibm.com>,
	Anton Arapov <anton@...hat.com>,
	Frederic Weisbecker <fweisbec@...il.com>,
	Ingo Molnar <mingo@...hat.com>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 3/4] uprobes/tracing: Kill the pointless
	local_save_flags/preempt_count calls

On 04/08, Steven Rostedt wrote:
>
> On Fri, 2013-03-29 at 19:15 +0100, Oleg Nesterov wrote:
> > --- a/kernel/trace/trace_uprobe.c
> > +++ b/kernel/trace/trace_uprobe.c
> > @@ -492,17 +492,13 @@ static int uprobe_trace_func(struct trace_uprobe *tu, struct pt_regs *regs)
> >  	struct ring_buffer_event *event;
> >  	struct ring_buffer *buffer;
> >  	u8 *data;
> > -	int size, i, pc;
> > -	unsigned long irq_flags;
> > +	int size, i;
> >  	struct ftrace_event_call *call = &tu->call;
> >
> > -	local_save_flags(irq_flags);
> > -	pc = preempt_count();
>
> How about instead, just change the above two and have:
>
> 	/* uprobes are never called with preemption disabled */
> 	pc = 0;
> 	irq_flags = 0;
>
> and leave the rest the same. This will help in future reviewers of the
> code to not have to look up what that "0, 0" is for, and then wonder if
> it should be that way. gcc should optimize it to be exactly the same as
> this patch.

Hmm, just to remind which arguments trace_current_buffer_*() has?

Personally I disagree. And, for example, ftrace_syscall_enter/exit just
use 0,0 for the same reason.

So please tell me if you really want the dummy variables/arguments, in
this case I'll change this code even if I do not like it.

Oleg.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/