Date: Sat, 14 Sep 2002 14:08:20 +0200 (CEST) From: Dan Lukes <dan@obluda.cz> To: FreeBSD-gnats-submit@FreeBSD.org Subject: kern/42770: MUTEX_DEBUG broke CURRENT kernel Message-ID: <200209141208.g8EC8KA6000443@xkulesh.vol.cz>
next in thread | raw e-mail | index | archive | help
>Number: 42770 >Category: kern >Synopsis: MUTEX_DEBUG broke CURRENT kernel >Confidential: no >Severity: serious >Priority: medium >Responsible: freebsd-bugs >State: open >Quarter: >Keywords: >Date-Required: >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Sat Sep 14 05:10:02 PDT 2002 >Closed-Date: >Last-Modified: >Originator: Dan Lukes >Release: FreeBSD 5.0-CURRENT i386 >Organization: >Environment: sys/kern/syscalls.master,v 1.121 2002/08/19 20:01:31 sys/kern/kern_prot.c,v 1.164 2002/08/01 17:47:55 sys/kern/kern_proc.c,v 1.147 2002/09/11 08:13:53 sys/kern/kern_mutex.c,v 1.107 2002/09/11 08:13:53 sys/vm/vm_glue.c,v 1.155 2002/09/11 08:13:55 sys/vm/vm_map.c,v 1.266 2002/08/24 17:52:08 >Description: MUTEX_DEBUG cause "mutex Giant not owned at vm_map.c:383" (it's triggered during mounting root). The content of stack: setsid->enterpgrp->mtx_init-> ->mtx_validate->kernacc->vm_map_check_protection->_vm_map_lock_read->_mtx_assert syscall setsid is marked as MPS_SAFE, so it doesn't obtain Giant lock, but _vm_map_lock_read required it. >How-To-Repeat: Compile kernel with MUTEX_DEBUG >Fix: workaround - doesn't compile kernel with MUTEX_DEBUG I can't supply the fix, but see the note in mtx_validate (kern_mutex.c) - it seems to be valid not only on alphas. >Release-Note: >Audit-Trail: >Unformatted: To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200209141208.g8EC8KA6000443>