[git pull] hwclock fixes

[git pull] hwclock fixes

[Sami Kerola]

The following changes since commit 63de90d44f4f05a1f48942148edcf1740ba4a623:

  libmount: cleanup docs (2011-07-23 01:27:39 +0200)

are available in the git repository at:
  https://github.com/kerolasa/lelux-utiliteetit hwclock

Sami Kerola (11):
      hwclock: remove clock-ppc.c
      hwclock: remove misleading information
      hwclock: coding style clean up
      hwclock: include-what-you-use header check
      hwclock: move path definitions to pathnames.h
      hwclock: remove inb & outb functions
      hwclock: fix compiler warnings
      hwclock: remove goto statement
      hwclock: add variable initialization
      hwclock: move long options away from global scope
      hwclock: validate numeric option arguments

 configure.ac           |    2 -
 hwclock/Makefile.am    |    4 +-
 hwclock/README.hwclock |   10 -
 hwclock/clock-ppc.c    |  459 ---------
 hwclock/clock.h        |   20 +-
 hwclock/cmos.c         |  936 +++++++++---------
 hwclock/hwclock.c      | 2654 ++++++++++++++++++++++++------------------------
 hwclock/kd.c           |  239 +++---
 hwclock/rtc.c          |  565 ++++++-----
 include/pathnames.h    |    9 +
 10 files changed, 2223 insertions(+), 2675 deletions(-)
 delete mode 100644 hwclock/clock-ppc.c

-- 
   Sami Kerola
   http://www.iki.fi/kerolasa/

Re: [git pull] hwclock fixes

[Karel Zak]

On Tue, Jul 26, 2011 at 12:42:46AM +0200, Sami Kerola wrote:
> are available in the git repository at:
>   https://github.com/kerolasa/lelux-utiliteetit hwclock
>
>       hwclock: remove clock-ppc.c

OK

>       hwclock: remove misleading information

OK

>       hwclock: coding style clean up

OK

>       hwclock: include-what-you-use header check

 Hmm.. I see something completely different in the patch:

    Showing 5 changed files with 171 additions and 216 deletions. 

 it seems like "use libc error facilities".

>       hwclock: move path definitions to pathnames.h

OK

>       hwclock: remove inb & outb functions

Are you really sure? I think it depends on the header files, for
example it seems GNU Hurd reads information from cmos

http://www.spinics.net/lists/util-linux-ng/msg04179.html

and we already talked about it:

http://www.spinics.net/lists/util-linux-ng/index.html#04673

>       hwclock: fix compiler warnings

This is wrong for __alpha__ code where the variables are used.

Move the comment before the function and the begin of the function to
the #ifndef __alpha_ block, then you can use __unused__ for non-aplha
only.

>       hwclock: remove goto statement

OK

>       hwclock: add variable initialization

Good catch.

>       hwclock: move long options away from global scope

But the enum {} does not have to be in the main().

>       hwclock: validate numeric option arguments

OK

    Karel

-- 
 Karel Zak  <kzak@redhat.com>
 http://karelzak.blogspot.com

Re: [git pull] hwclock fixes

[Sami Kerola]

On Tue, Jul 26, 2011 at 16:22, Karel Zak <kzak@redhat.com> wrote:
> On Tue, Jul 26, 2011 at 12:42:46AM +0200, Sami Kerola wrote:
>> are available in the git repository at:
>> =A0 https://github.com/kerolasa/lelux-utiliteetit hwclock
>> =A0 =A0 =A0 hwclock: include-what-you-use header check
>
> =A0Hmm.. I see something completely different in the patch:
>
> =A0 =A0Showing 5 changed files with 171 additions and 216 deletions.
>
> =A0it seems like "use libc error facilities".

Do'h. One of my rebase fix something etc had lead to accidental
squash. The commit is split to two as it was meant to be.

>> =A0 =A0 =A0 hwclock: remove inb & outb functions
>
> Are you really sure? I think it depends on the header files, for
> example it seems GNU Hurd reads information from cmos
>
> http://www.spinics.net/lists/util-linux-ng/msg04179.html
>
> and we already talked about it:
>
> http://www.spinics.net/lists/util-linux-ng/index.html#04673

I did not study history well enough before proposing this. The commit
is thrown away.

>> =A0 =A0 =A0 hwclock: fix compiler warnings
>
> This is wrong for __alpha__ code where the variables are used.
>
> Move the comment before the function and the begin of the function to
> the #ifndef __alpha_ block, then you can use __unused__ for non-aplha
> only.

I changed the commit to have complete functions in precompiler if
statements. To me it looks most readable way of doing what you seem to
mean.

>> =A0 =A0 =A0 hwclock: move long options away from global scope
>
> But the enum {} does not have to be in the main().

Moved to the beginning of the file.

All is pushed to my repo with force. Let me know if there is something
other that needs nudging.

--=20
=A0=A0 Sami Kerola
=A0=A0 http://www.iki.fi/kerolasa/

Re: [git pull] hwclock fixes

[Karel Zak]

On Tue, Jul 26, 2011 at 06:38:26PM +0200, Sami Kerola wrote:
> All is pushed to my repo with force. Let me know if there is something
> other that needs nudging.

 Merged.

-- 
 Karel Zak  <kzak@redhat.com>
 http://karelzak.blogspot.com