From owner-freebsd-questions Thu Jun 8 06:16:01 1995 Return-Path: questions-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id GAA11554 for questions-outgoing; Thu, 8 Jun 1995 06:16:01 -0700 Received: from hda.com (hda.com [199.232.40.182]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id GAA11548 ; Thu, 8 Jun 1995 06:15:57 -0700 Received: (dufault@localhost) by hda.com (8.6.11/8.3) id JAA04674; Thu, 8 Jun 1995 09:13:50 -0400 From: Peter Dufault Message-Id: <199506081313.JAA04674@hda.com> Subject: pthreads (was Re: DCE in BSD4.4-Lite) To: ruchw@bmd.saic.com (Weston Ruch) Date: Thu, 8 Jun 1995 09:13:49 -0400 (EDT) Cc: questions@FreeBSD.org, bde@FreeBSD.org, proven@mit.edu, jkh@FreeBSD.org Reply-To: hackers@FreeBSD.org In-Reply-To: <199506081213.IAA07898@bmd.saic.com> from "Weston Ruch" at Jun 8, 95 08:13:10 am X-Mailer: ELM [version 2.4 PL24] Content-Type: text Content-Length: 3465 Sender: questions-owner@FreeBSD.org Precedence: bulk Weston Ruch writes: > I would like to experiment with the DCE shareware available > at gatekeeper.dec.com. However, the DCE shareware > requires POSIX 4.0 threads. Have POSIX 4.0 threads been > implemented in FreeBSD? Chris Provenzano's beta pthread package based on POSIX1003.4a Draft 8 with a few minor patches compiles and almost runs through its tests on FreeBSD 2.05B. The test that fails is the floating point preemption test, which dies with a: > pid 4510 (test_preemption_) exited with masked floating point exceptions 0x41 I'll bet bde can suggest a fix for this; I recall him commenting on floating point in signal handlers. I coincidentally had decided to check pthreads out this morning to see if it could go on the experimental section of the CD-ROM. Jordan: Do you want this like this on the CD-ROM? It is certainly ready for hacker use. The patches I applied to sipb.mit.edu:/pub/pthreads/pthreads-1_60_beta2.tar.gz were: *** config/configure.dist Mon Apr 3 22:40:57 1995 --- config/configure Thu Jun 8 07:46:14 1995 *************** *** 902,907 **** --- 902,913 ---- romp-*-bsd*) name=romp-bsd ;; + i386-*-freebsd2.*) + name=i386-freebsd-2.0 + sysincludes=freebsd-2.0 + syscall=i386-freebsd2.0 + makefile_frag=netbsd1.0.mk + ;; mips-*-ultrix*) name=r2000-ultrix-4.2 sysincludes=ultrix-4.2 *** machdep/freebsd-2.0/__unistd.h.dist Tue Feb 28 20:27:37 1995 --- machdep/freebsd-2.0/__unistd.h Thu Jun 8 09:03:19 1995 *************** *** 59,65 **** /* too-long path components generate errors */ #define _POSIX_NO_TRUNC 1 /* may disable terminal special characters */ ! #define _POSIX_VDISABLE ((unsigned char)'\377') /* configurable pathname variables */ #define _PC_LINK_MAX 1 --- 59,65 ---- /* too-long path components generate errors */ #define _POSIX_NO_TRUNC 1 /* may disable terminal special characters */ ! #define _POSIX_VDISABLE 0xff /* configurable pathname variables */ #define _PC_LINK_MAX 1 *** tests/Makefile.dist Thu Jun 8 07:50:57 1995 --- tests/Makefile Thu Jun 8 08:46:54 1995 *************** *** 20,27 **** LIBS = -static ../obj/libpthread.a -lm # This list used to include test_select, but that test doesn't terminate. TESTS = test_create test_pthread_join test_switch test_sleep test_readdir \ ! test_fork test_execve test_preemption test_preemption_float \ test_sock_1 test_stdio_1 test_pthread_mutex \ test_pthread_cond_timedwait test_netdb test_pw test_cwd # This list used to include p_bench_semaphore, but the semaphore support isn't --- 20,31 ---- LIBS = -static ../obj/libpthread.a -lm # This list used to include test_select, but that test doesn't terminate. + #TESTS = test_create test_pthread_join test_switch test_sleep test_readdir \ + # test_fork test_execve test_preemption test_preemption_float \ + # test_sock_1 test_stdio_1 test_pthread_mutex \ + # test_pthread_cond_timedwait test_netdb test_pw test_cwd TESTS = test_create test_pthread_join test_switch test_sleep test_readdir \ ! test_fork test_execve test_preemption \ test_sock_1 test_stdio_1 test_pthread_mutex \ test_pthread_cond_timedwait test_netdb test_pw test_cwd # This list used to include p_bench_semaphore, but the semaphore support isn't -- Peter Dufault Real Time Machine Control and Simulation HD Associates, Inc. Voice: 508 433 6936 dufault@hda.com Fax: 508 433 5267