Re: Reverting the kernel 32/64bit compensation code in autofs userspace?
|[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]|
On 07.05.2012 04:57, Ian Kent wrote: > On Sun, 2012-05-06 at 22:47 +0400, Michael Tokarev wrote: >> Now when kernel has more or less nice interface which works >> with all variants of userspace to date, maybe it's time to >> revert the commit which enables different struct size for >> kernels > 3.3, especially since that commit was initially >> wrong (since it does not take into account any stable trees >> and possible backports to other previous versions)? > > We will need to use something similar to the patch I already sent you as > kernel releases don't get unreleased so I will need to cater for it. Why cater when kernel does not care anymore? You only may want to consider versions between 3.3 and 3.3.4 (iirc), but that is not worth the ugliness I think. That patch for autofs userspace is not needed anymore, since unpatched autofs works fine with all previous/older kernels, and since current versions of kernel, the kernel will work with either size of struct, so that fix in userspace isn't needed again. So you can just remove and forget about it, and remove quite some very strange code piece... ;) I'm not saying the code does not work, I'm saying that it will work regardless of the userspace fix, and since the fix is ugly it might be a good idea to just revert it. And you wont be able to compensate for all backports anyway -- some distro may want to backport this change to some older kernel (e.g. 2.6.32 - why not for, eg, redhat?) in order to let both systemd and autofs work there on mixed 32/64bit environment, -- will you check /etc/redhat.release and other files too to determine which struct size to use? :) Current kernel does not care, and without the last userspace fix the code is at least a bit understandable and it works either way, that's why I think just reverting it is a good idea. But it is not my call obviously. [kernel version] > Because mount.nfs changed to use the kernel to parse nfs mount options > based on kernel version. The kernel RPC code works very differently to > the user space RPC code, since it must wait if a host doesn't respond > where that isn't so important for user space, which leads to long > timeout delays. That's the way it should be done, of course, but for an > interactive application like automount everyone complains about it. > Basically, automount must RPC ping hosts at some point in the mount > trigger to mitigate this. Aha, I see. So nfs "mounter" also needs to know kernel version to pass mount info correctly. I've seen this in other places already, even busybox mount.nfs has kernel version check. Thank you for the explanation. So indeed, the kernel version gathering code should stay, no problem with that. >> And btw, is there some public autofs userspace git/whatever repository >> where one can see current state of the package? > > git clone git://git.kernel.org/pub/scm/linux/storage/autofs/autofs Hm. I found this on http://git.kernel.org/ , but somehow I thought it is some old clone, since it has last modified 4 weeks ago, and I thought all this 32/64 issue was later than that... Ohh, time is flying too fast... :) Thank you, I'll use that. I asked because apparently I will be one of the maintainers of autofs userspace in Debian, and today I'm sponsoring first upload of this package after long inaction of a previous maintainer. See http://bugs.debian.org/671701 for the current discussion if you're interested. /mjt -- To unsubscribe from this list: send the line "unsubscribe autofs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html