From mboxrd@z Thu Jan 1 00:00:00 1970 From: Yegor Yefremov Date: Wed, 03 Aug 2011 15:41:18 +0200 Subject: [Buildroot] [PATCH 1/3] bump udev and clean up In-Reply-To: References: <4E298D06.4050107@visionsystems.de> <4E298E70.50200@visionsystems.de> <871uxe10hq.fsf@macbook.be.48ers.dk> <4E2E67C8.7090208@visionsystems.de> <871uxcuqg0.fsf@macbook.be.48ers.dk> <87sjpst92p.fsf@macbook.be.48ers.dk> Message-ID: <4E394FFE.1020706@visionsystems.de> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Hi Diego, please use http://en.wikipedia.org/wiki/Top-posting#Bottom-posting. Am 03.08.2011 14:37, schrieb Diego Iastrubni: > With the update of 173... lets bring this up again... > > I am having problems in my setup: I see the /dev/ directory quite > empty. I found out that you removed the call to udevstart in S10udev, > calling it manually fixed the "emptiness problem" and /dev/ is filled > up as I expect. > > Is the problem caused by my side (I am using tmpfs, and not udevfs, > and my kernel is *way* old). What is the equivalent of udevstart in > udev? I've looked at what Debian maintainers do: they use udevadm trigger --action=add http://linux.die.net/man/8/udevadm With the newer kernel you don't have such a problem as empty /dev/, because in BR devtmpfs will be used to fill that directory before mounting rootfs. So just starting udevd is sufficient. > On Wed, Jul 27, 2011 at 3:53 PM, Diego Iastrubni wrote: >> On Wed, Jul 27, 2011 at 1:19 PM, Diego Iastrubni wrote: >>> You are probably right, qouting from >>> http://www.spinics.net/lists/hotplug/msg04998.html : >>> >>> It's generally not a good idea to build new low-level tools against >>> outdated kernels or run new low-level tools against them. Things can >>> easily go wrong. We usually only test the other way around: new >>> kernels on old userspace. That should work fine for quite a while. >>> >>> So, this patch demands: >>> >>> 1) newer toolchains >>> 2) kernel 2.6.33 and above. >>> >>> (2) is bad, since 2.6.32 is still maintained. For (1), I have a patch >>> in the works, I will test and report. >> >> See patch attached. Just put it inside packages/udev and rebuild. >> Works for me, and should not break other things, at least I hope. >> > _______________________________________________ > buildroot mailing list > buildroot at busybox.net > http://lists.busybox.net/mailman/listinfo/buildroot >