From owner-freebsd-current Thu Sep 24 13:39:44 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA08823 for freebsd-current-outgoing; Thu, 24 Sep 1998 13:39:44 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from grape.carrier.kiev.ua (grape.carrier.kiev.ua [193.193.193.219]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id NAA08761 for ; Thu, 24 Sep 1998 13:39:24 -0700 (PDT) (envelope-from archer@grape.carrier.kiev.ua) Received: (from archer@localhost) by grape.carrier.kiev.ua (8.9.1/8.9.1) id XAA02983; Thu, 24 Sep 1998 23:39:00 +0300 (EEST) (envelope-from archer) Message-ID: <19980924233859.63322@carrier.kiev.ua> Date: Thu, 24 Sep 1998 23:38:59 +0300 From: Alexander Litvin To: Luoqi Chen Cc: current@FreeBSD.ORG Subject: Strange ld (was: deadlock in vm_fault()) References: <199809242016.QAA27364@lor.watermarkgroup.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.88e In-Reply-To: <199809242016.QAA27364@lor.watermarkgroup.com>; from Luoqi Chen on Thu, Sep 24, 1998 at 04:16:40PM -0400 Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Thu, Sep 24, 1998 at 04:16:40PM -0400, Luoqi Chen wrote: > > Unfortunately, your patch does not solve my problem. May be, it is of > > different source. > > > > Just now I have it (ld) hanging in the same place: > > > > UID PID PPID CPU PRI NI VSZ RSS WCHAN STAT TT TIME COMMAND > > 0 1506 1505 0 28 0 496 384 - R+ vb 4:14.76 /usr/obj/elf/usr/src/tmp/usr/libexec/elf/ld -m elf_i386 -Bstatic -o cat /usr/obj/elf/usr/src/tmp/usr/lib/crt1.o /usr/obj/elf/usr/src/tmp/usr/lib/crti.o /usr/obj/elf/usr/src/tmp/usr/lib/crtbegin.o -L/usr/obj/elf/usr/src/tmp/usr/lib -L/usr/obj/elf/usr/src/tmp/usr/lib -L/usr/libdata/gcc cat.o /usr/obj/elf/usr/src/tmp/usr/lib/libgcc.a -lc /usr/obj/elf/usr/src/tmp/usr/lib/libgcc.a /usr/obj/elf/usr/src/tmp/usr/lib/crtend.o /usr/obj/elf/usr/src/tmp/usr/lib/crtn.o > > > > The strange thing is that I cannot generate a coredump of it with gcore: > > > > root:/usr/obj/elf/usr/src/tmp/usr/libexec/elf:grape:> gcore ld 1506 > > gcore: The executable ld does not belong to process 1506! > > Text segment size (in bytes): executable 65793, process 335872 > > > This doesn't look like a kernel related problem, most likely it is a bug > in ld that caused it to loop forever. To generate a coredump you probably > need to specify the full path of ld in the temporary tool directory. If > I remember correctly, gcore has not been converted to dump ELF format cores. > Try using kill -QUIT instead, or use gdb to attach the process directly. Well, I tried gcore with full path -- it doesn't work either. The problem is really not kernel related -- that ld (in temporary tree) hangs, even if it is just run, without any arguments or input files. I am now in the process of determining how it happened for such 'ld' to be built (I have a buildworld output written to a file). Though it may happen that the fact that _such_ ld was built during -j12 world is kernel related ;) :( > -lq --- I can't understand it. I can't even understand the people who can understand it. -- Queen Juliana of the Netherlands. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message