From mboxrd@z Thu Jan 1 00:00:00 1970 From: Greg KH Date: Fri, 20 Feb 2004 01:54:34 +0000 Subject: Re: HOWTO use udev to manage /dev Message-Id: <20040220015433.GC3134@kroah.com> List-Id: References: <20040219185932.GA10527@kroah.com> <20040219191636.GC10527@kroah.com> <20040219230749.GA15848@kroah.com> <20040219235602.GI15848@kroah.com> In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable To: =?iso-8859-1?B?RnLpZOlyaWMgTC4gVy4=?= Meunier <1@pervalidus.net> Cc: linux-kernel@vger.kernel.org, linux-hotplug-devel@lists.sourceforge.net On Thu, Feb 19, 2004 at 09:51:52PM -0300, Fr=E9d=E9ric L. W. Meunier wrote: > On Thu, 19 Feb 2004, Greg KH wrote: >=20 > > So if you take out the line about starting udevd, does it > > work for you? >=20 > No. >=20 > > How about changing the #!/bin/bash to #!/bin/sash in the > > first line for the start_udev script? >=20 > I didn't have it, but compiled and changed. Yes, it works. >=20 > > What distro is this? >=20 > Slackware, with a cute rc.S. /bin/bash was also recompiled, shared: >=20 > $ ldd /bin/bash > libreadline.so.4 =3D> /usr/lib/libreadline.so.4 (0x4001c000) > libhistory.so.4 =3D> /usr/lib/libhistory.so.4 (0x40049000) > libncurses.so.5 =3D> /lib/libncurses.so.5 (0x40050000) > libdl.so.2 =3D> /lib/libdl.so.2 (0x4008f000) > libc.so.6 =3D> /lib/libc.so.6 (0x40092000) > /lib/ld-linux.so.2 =3D> /lib/ld-linux.so.2 (0x40000000) >=20 > Maybe the problem ? Does yours differ ? Mine does differ, but it is dynamic: $ ldd /bin/bash linux-gate.so.1 =3D> (0xffffe000) libtermcap.so.2 =3D> /lib/libtermcap.so.2 (0x4d5b5000) libdl.so.2 =3D> /lib/libdl.so.2 (0x4d3b4000) libc.so.6 =3D> /lib/tls/libc.so.6 (0x4d254000) /lib/ld-linux.so.2 =3D> /lib/ld-linux.so.2 (0x4d238000) > bash from Slackware: >=20 > libtermcap.so.2 =3D> /lib/libtermcap.so.2 (0x4001c000) > libdl.so.2 =3D> /lib/libdl.so.2 (0x4005c000) > libc.so.6 =3D> /lib/libc.so.6 (0x4005f000) > /lib/ld-linux.so.2 =3D> /lib/ld-linux.so.2 (0x40000000) >=20 > OK, I'll later boot with it and see if it works. If it does, > I'll run strace with the other. How about using sash? That is statically linked. > > Can you run strace on the start_udev script after boot to see who is > > needing access to /dev/null? >=20 > I forgot to run it, but noticed there was a /dev/null, but a > text file (0644). And I didn't create it anywhere. That sounds like some program is trying to write to it. Hm, there is a patch in the Red Hat version of udev that basically makes udev do the start_udev logic, in the .c file because they do not have a shell in their initrd. If you can dig it out of there, that might be a solution for you to use. Other than that, how about running strace on start_udev when your rc.S script calls it? That might help out. thanks, greg k-h ------------------------------------------------------- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356&alloc_id438&op=CCk _______________________________________________ Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net Linux-hotplug-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel