Re: [PATCH] hfsplus: fix FS driver name in printks

[Vyacheslav Dubeyko <slava@dubeyko.com>]

On Tue, 2013-01-29 at 09:22 +0000, Mike Fleetwood wrote:
> On Tue, Jan 29, 2013 at 09:49:25AM +0400, Vyacheslav Dubeyko wrote:
> > On Mon, 2013-01-28 at 20:23 +0000, Mike Fleetwood wrote:
> > > Correct the name of the hfsplus FS driver as used in printk calls.
> > > "hfs:" -> "hfsplus:".
> > > 
> > > Signed-off-by: Mike Fleetwood <mike.fleetwood@googlemail.com>
> > > ---
> > > 
> > > Hi,
> > > 
> > > Is there a current reason why the hfsplus FS driver uses "hfs:" almost
> > > exclusively rather than "hfsplus:" as its name in printk calls?
> > 
> > There are as minimum two reason for leaving "hfs:" prefix in peace: (1)
> > historical - it is like code style of "old" library; (2) the prefix
> > "hfs:" is shorter - so, it gives opportunity to make more descriptive
> > comments by means of one line under 80 symbols kernel code style
> > requirement.
> > 
> > By the way, did you check your patch by scripts/checkpatch.pl script?
> > 
> > Moreover, there are hfsplus driver's patches in linux-next that uses
> > "hfs:" prefix.
> > 
> > I doubt that this patch can improve hfsplus driver quality. It looks
> > like changes in many places without changing anything in essence.
> > 
> > With the best regards,
> > Vyacheslav Dubeyko.  
> > 
> 
> In terms of line length I was applying the exception in the CodingStyle
> which says "never break user-visible strings such as printk messages,
> because that breaks the ability to grep for them." to allow lines be
> longer than 80 characters.
> 
> I did use checkpatch.pl.  It reported this for every printk:
>     WARNING: Prefer netdev_err(netdev, ... then dev_err(dev, ... then pr_err(...  to printk(KERN_ERR ...

So, if we begin to modify error messages in the hfsplus driver then,
maybe, it makes sense to exchange the printk() on pr_err() and to
prepare define for "hfsplus:" prefix? Then, we make checkpatch.pl
completely happy.

> After seeing that ext2/3/4, btrfs and xfs use printk and not any of
> those functions I followed the majority.  It also reported a couple of:
>     WARNING: line over 80 characters
> I was applying the above exception.
> 
> Absoutely this patch doesn't fix any faults and is not for 3.8.0-rc* but
> for linux next.  I just though that it would be useful for users to be
> told the name of the FS driver generating the message rather than a
> different one.
> 

Anyway, I think that changing "hfs:" on "hfsplus:" is not important.

> Would an equlivant patch be accepted for linux-next?
> How do I send it to linux-next?
> 

I mean git://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git
when I am talking about linux-next. So, you can prepare patches are
based on this repository.

With the best regards,
Vyacheslav Dubeyko.

> Thank you Vyacheslav for reviewing my patch,
> Mike
> 
> > 
> > > Assuming not here's a patch to fix.
> > > 
> > > (Any code which may have been copied between hfs and hfsplus has since
> > > diverged significantly).
> > > 
> > > Thanks,
> > > Mike
> > >