Mailing-List: contact zsh-workers-help@zsh.org; run by ezmlm Precedence: bulk X-No-Archive: yes List-Id: Zsh Workers List List-Post: List-Help: X-Spam-Checker-Version: SpamAssassin 3.4.1 (2015-04-28) on f.primenet.com.au X-Spam-Level: X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00 autolearn=ham autolearn_force=no version=3.4.1 Date: Thu, 17 Mar 2016 20:09:48 +0000 From: Daniel Shahaf To: Peter Stephenson Cc: zsh-workers@zsh.org Subject: Re: segfault in completion for configure Message-ID: <20160317200948.GA32439@tarsus.local2> References: <20160311143202.4008e29b@pwslap01u.europe.root.pri> <160311150056.ZM30997@torch.brasslantern.com> <20160312031116.GC28459@zira.vinc17.org> <160312082029.ZM2340@torch.brasslantern.com> <160312093420.ZM14020@torch.brasslantern.com> <20160313215831.GA23404@zira.vinc17.org> <160314194323.ZM6887@torch.brasslantern.com> <20160317152435.GC11303@cventin.lip.ens-lyon.fr> <160317111515.ZM16697@torch.brasslantern.com> <20160317182406.2cb566ed@pwslap01u.europe.root.pri> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <20160317182406.2cb566ed@pwslap01u.europe.root.pri> User-Agent: Mutt/1.5.23 (2014-03-12) X-Seq: zsh-workers 38176 Peter Stephenson wrote on Thu, Mar 17, 2016 at 18:24:06 +0000: > Or is that if we know we've been int he signal handler at that point > we already know the place where we needed to queue signals... We could convert the pattern matching globals to macros that execute code whenever the globals are accessed: diff --git a/Src/pattern.c b/Src/pattern.c index 72c7d97..a4fb24f 100644 --- a/Src/pattern.c +++ b/Src/pattern.c @@ -303,6 +303,12 @@ static long patalloc; /* size allocated for same */ static int patflags; /* flags passed down to patcompile */ static int patglobflags; /* globbing flags & approx */ +#define patflags (f(), patflags) +void f(void) +{ + /* ... */ +} + /* * Increment pointer to metafied multibyte string. */ Now, f() can assert() some invariant, or check if SIGWINCH had been delivered (and if so, drop into a gdb breakpoint), etc... I'm proposing this just as a means of debugging for this one problem, not as a permanent change. Cheers, Daniel