linux-kernel - Re: [PATCH 0/7] uprobes: single-step fixes

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: <20120908170601.GA19311@redhat.com>
Date:	Sat, 8 Sep 2012 19:06:01 +0200
From:	Oleg Nesterov <oleg@...hat.com>
To:	Ingo Molnar <mingo@...e.hu>, Peter Zijlstra <peterz@...radead.org>,
	Srikar Dronamraju <srikar@...ux.vnet.ibm.com>
Cc:	Ananth N Mavinakayanahalli <ananth@...ibm.com>,
	Anton Arapov <anton@...hat.com>,
	"H. Peter Anvin" <hpa@...or.com>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Roland McGrath <roland@...k.frob.com>,
	Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH 0/7] uprobes: single-step fixes

On 09/03, Oleg Nesterov wrote:
>
> Sebastian, I changed your patches a bit:
>
> 	1/7:
>
> 		- Change the subject and update the changelog. In particular,
> 		  s/utrace/uprobes/. I am wondering where this typo came from ;)

Hmm. I just noticed this patch is buggy. arch_uprobe_disable_step(&uprobe->arch)
is not safe after put_uprobe().

Srikar, I fixed this in my tree with the following change,

	--- kernel/events/uprobes.c~	2012-09-02 16:52:54.000000000 +0200
	+++ kernel/events/uprobes.c	2012-09-08 18:56:44.000000000 +0200
	@@ -1536,10 +1536,10 @@ static void handle_singlestep(struct upr
		else
			WARN_ON_ONCE(1);
	 
	+	arch_uprobe_disable_step(&uprobe->arch);
		put_uprobe(uprobe);
		utask->active_uprobe = NULL;
		utask->state = UTASK_RUNNING;
	-	arch_uprobe_disable_step(&uprobe->arch);
		xol_free_insn_slot(current);
	 
		spin_lock_irq(&current->sighand->siglock);

I hope your ack is still valid.

And this also allows us to rely on utask->state in disable_step(), see
the new 8/7 I'll send in a minute. I was going to fix this later, but
I just realized that "disable if trapped" is more buggy than I thought.

Assuming that you are agree with 6 and 8. I'd prefer the new one as a
separate change, but if you prefer to join them please let me know.

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/