linux-3.16.1

[bdubbs@…]

New point version

[bdubbs@…]

Now linux-3.15.8

[ken@…]

now linux-3.16.0

[Armin]

Consider changing

    ()  path to uevent helper [CONFIG_UEVENT_HELPER_PATH]

To

    [ ] Support for uevent helper [CONFIG_UEVENT_HELPER]

On the kernel config page (for 3.16).

[ken@…]

That change on the kernel config page looks appropriate (I queried this when it was introduced, didn't get a satisfactory explanation, found that going with the default 'Y' even on a recent LFS (non-systemd) build didn't cause me any grief) - but I've just tried selecting it on LFS-7.5 and there was a little noise from the uevent part of the bootscripts. That box is now building current svn, will see if the noise is still there when it boots.

[ken@…]

The noise is still there:

 S10udev line 48: /proc/sys/kernel/hotplug: No such file or directory

which comes from

      # Udev handles uevents itself, so we don't need to have
      # the kernel call out to any binary in response to them
      echo > /proc/sys/kernel/hotplug

I had not noticed this message until I turned off CONFIG_UEVENT_HELPER. Clearly, it does not cause a problem, but it looks messy and likely to worry people. Should we keep that old line of code, and wrap it in a test to see if the file exists, or should we drop it ? e.g.

      test -f /proc/sys/kernel/hotplug && echo > /proc/sys/kernel/hotplug

If we do that, I suppose the comment should be updated to reflect this was a legacy feature.

[bdubbs@…]

I'll just remove that line when I update the kernel. I plan on waiting about a week to do that betting on 3.16.1 by then. I just don't want to do two updates before -rc1.

[bdubbs@…]

Fixed at revision 10690.

Fixed hotplug earlier at revision 10689.