Received: from sss.pgh.pa.us (sss.pgh.pa.us [209.114.132.154]) by mail.postgresql.org (8.11.1/8.11.1) with ESMTP id eBIHeV601896 for ; Mon, 18 Dec 2000 12:40:31 -0500 (EST) (envelope-from tgl@sss.pgh.pa.us) Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1]) by sss.pgh.pa.us (8.11.1/8.11.1) with ESMTP id eBIHeWL20040; Mon, 18 Dec 2000 12:40:32 -0500 (EST) To: Ian Lance Taylor cc: mathijs@ilse.nl, pgsql-hackers@postgresql.org Subject: Re: Re: [SQL] PostgreSQL crashes on me :( In-reply-to: <20001218173340.29569.qmail@daffy.airs.com> References: <20001217215118.D64770@ilse.nl> <21339.977111275@sss.pgh.pa.us> <20001218173340.29569.qmail@daffy.airs.com> Comments: In-reply-to Ian Lance Taylor message dated "18 Dec 2000 09:33:40 -0800" Date: Mon, 18 Dec 2000 12:40:32 -0500 Message-ID: <20037.977161232@sss.pgh.pa.us> From: Tom Lane X-Archive-Number: 200012/770 X-Sequence-Number: 2291 Ian Lance Taylor writes: > Therefore, the scenario you describe is possible, but only if there > happens to be both a timer interrupt and a SIGCHLD signal within a > couple of instructions after the select returns. Right. A small failure window would explain the infrequency of the symptom. > Ideally, signal handlers should not make system calls. However, if > this is impossible, then signal handlers must save and restore errno. I have just committed fixes to do that in current sources (7.1 branch). I doubt we're going to make a 7.0.4 release, but if we do then this ought to be back-patched. regards, tom lane