From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <zsh-workers-return-34618-mason-zsh=primenet.com.au@zsh.org>
Received: (qmail 17800 invoked by alias); 23 Feb 2015 14:15:45 -0000
Mailing-List: contact zsh-workers-help@zsh.org; run by ezmlm
Precedence: bulk
X-No-Archive: yes
List-Id: Zsh Workers List <zsh-workers.zsh.org>
List-Post: <mailto:zsh-workers@zsh.org>
List-Help: <mailto:zsh-workers-help@zsh.org>
X-Seq: 34618
Received: (qmail 16101 invoked from network); 23 Feb 2015 14:15:33 -0000
X-Spam-Checker-Version: SpamAssassin 3.3.2 (2011-06-06) on f.primenet.com.au
X-Spam-Level: 
X-Spam-Status: No, score=-6.9 required=5.0 tests=BAYES_00,RCVD_IN_DNSWL_HI,
	SPF_HELO_PASS autolearn=ham version=3.3.2
From: Kamil Dudka <kdudka@redhat.com>
To: Bart Schaefer <schaefer@brasslantern.com>
Cc: zsh-workers@zsh.org
Subject: Re: PATCH Re: deadlock in free() called from a signal handler
Date: Mon, 23 Feb 2015 15:15:22 +0100
Message-ID: <2067297.ZNlDuaAVYz@kdudka.brq.redhat.com>
User-Agent: KMail/4.14.4 (Linux/3.18.7-200.fc21.x86_64; KDE/4.14.4; x86_64; ; )
In-Reply-To: <150220182812.ZM11805@torch.brasslantern.com>
References: <37490085.zXPQGCoLTl@kdudka.brq.redhat.com> <150219092329.ZM17912@torch.brasslantern.com> <150220182812.ZM11805@torch.brasslantern.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="us-ascii"
X-Scanned-By: MIMEDefang 2.68 on 10.5.11.22

On Friday 20 February 2015 18:28:12 Bart Schaefer wrote:
> On Feb 19,  9:23am, Bart Schaefer wrote:
> }
> } In every case I can remember so far, when this happens it means that we
> } ought to be using the signal queuing macros at a scope outside the call
> } to the malloc library.
> }  [...]
> } So ... per my very first remark above, it's probably worth examining the
> } context in execcmd() to see if signal queuing is appropriate there, or
> } if we just need it in setunderscore().  But it may also make sense to
> } replace that one last malloc() and put queuing in zfree()/zsfree() too.
> 
> (Above edited for silly typo.)
> 
> There doesn't appear to be an appropriate scope for queuing signals in
> execcmd().  However, setunderscore() and all of the functions in text.c
> manipulate global state, so they should be considered non-reentrant.
> 
> The following patch does not yet put signal queuing into zfree(), but
> does wrap it around setunderscore() and the three non-static functions
> in text.c, as well as replace that one remaining malloc() with zalloc().
> Hopefully this doesn't slow down execution noticeably.

Thank you for the patch!  We will give it a try.

Kamil