From owner-freebsd-threads@FreeBSD.ORG Sun Jun 5 21:09:23 2005 Return-Path: X-Original-To: freebsd-threads@freebsd.org Delivered-To: freebsd-threads@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 27DDE16A41C for ; Sun, 5 Jun 2005 21:09:23 +0000 (GMT) (envelope-from lewissaunders@myrealbox.com) Received: from smtp-send.myrealbox.com (smtp-send.myrealbox.com [192.108.102.143]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5EFD543D5F for ; Sun, 5 Jun 2005 21:09:20 +0000 (GMT) (envelope-from lewissaunders@myrealbox.com) Received: from sweet lewissaunders [81.101.64.94] by smtp-send.myrealbox.com with NetMail SMTP Agent $Revision: 1.5 $ on Linux via secured & encrypted transport (TLS); Sun, 05 Jun 2005 15:09:18 -0600 Message-ID: <000301c56a12$d5bce850$5e406551@sweet> From: "Lewis Saunders" To: Date: Sun, 5 Jun 2005 22:06:59 +0100 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=original Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2180 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 Subject: 5.4-RELEASE fails libpthread tests X-BeenThere: freebsd-threads@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Threading on FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 05 Jun 2005 21:09:23 -0000 Hi, Is it normal for most of the tests in /usr/src/lib/libpthread/test/ to fail? I've just upgraded from 5.3-RELEASE to 5.4-RELEASE, and running make there gives: Test static library: -------------------------------------------------------------------------- Test c_user c_system c_total chng passed/FAILED h_user h_system h_total % chng -------------------------------------------------------------------------- hello_d 0.00 0.00 0.00 passed -------------------------------------------------------------------------- hello_s 0.00 0.01 0.01 passed -------------------------------------------------------------------------- join_leak_d 0.00 0.07 0.07 *** FAILED *** -------------------------------------------------------------------------- mutex_d 0.00 144.76 144.76 *** FAILED *** -------------------------------------------------------------------------- sem_d 0.00 0.00 0.00 *** FAILED *** -------------------------------------------------------------------------- sigsuspend_d 0.00 0.00 0.00 passed -------------------------------------------------------------------------- sigwait_d 0.00 0.00 0.00 *** FAILED *** -------------------------------------------------------------------------- guard_s.pl 0.00 0.29 0.29 *** FAILED *** (30/30 failed) -------------------------------------------------------------------------- propagate_s.pl 0.00 0.07 0.07 *** FAILED *** (1/1 failed) -------------------------------------------------------------------------- Totals 0.00 0.01 0.01 0.00 3 / 9 passed (33.33%) 0.00 0.00 0.00 0.00% -------------------------------------------------------------------------- *** Error code 1 Stop in /usr/src/lib/libpthread/test I'm afraid I never thought to run the tests on 5.3. The guard tests also dump core: Core was generated by `guard_b'. Program terminated with signal 11, Segmentation fault. #0 0x0805c17f in kse_thr_interrupt () (gdb) bt full #0 0x0805c17f in kse_thr_interrupt () No symbol table info available. (gdb) info threads * 4 LWP 100073 0x0805c17f in kse_thr_interrupt () 3 Thread 4 (sleeping) 0x0804a2eb in _thr_sched_switch_unlocked () 2 Thread 3 (LWP 100106) 0x0804a3f8 in kse_sched_single () 1 Thread 1 (runnable) 0x0805eb68 in __vfprintf () Do these fail for everyone because things aren't finished yet or is it just me? I'm also having immense difficulties with the new nVIDIA drivers, which are supposedly thread safe now. They seem to make things die in random places and make a mess of the stack (lots of ?? in backtraces, and ldexp() calling itself which really can't happen), and sometimes prevent gdb from switching threads. They even make a program linked with libc_r crash, which worked fine with the previous driver. Would it be worth updating to -CURRENT? Otherwise I might have to port to Linux which would be rather a pity :) Thanks for any advice, Lewis