linux-kernel - Re: [PATCH 2/6] device: Add #define dev_fmt similar to #define pr

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Sat, 7 Jul 2018 10:35:28 +0200
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     Corey Minyard <minyard@....org>
Cc:     Joe Perches <joe@...ches.com>, x86@...nel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 2/6] device: Add #define dev_fmt similar to #define pr_fmt

On Fri, Jul 06, 2018 at 03:50:16PM -0500, Corey Minyard wrote:
> On 07/06/2018 10:50 AM, Greg Kroah-Hartman wrote:
> > On Fri, Jul 06, 2018 at 08:41:34AM -0700, Joe Perches wrote:
> > > On Fri, 2018-07-06 at 17:30 +0200, Greg Kroah-Hartman wrote:
> > > > On Wed, May 09, 2018 at 08:15:46AM -0700, Joe Perches wrote:
> > > > > Add a prefixing macro to dev_<level> uses similar to the pr_fmt
> > > > > prefixing macro used in pr_<level> calls.
> > > > > 
> > > > > This can help avoid some string duplication in dev_<level> uses.
> > > > > 
> > > > > The default, like pr_fmt, is an empty #define dev_fmt(fmt) fmt
> > > > > 
> > > > > Rename the existing dev_<level> functions to _dev_<level> and
> > > > > introduce #define dev_<level> _dev_<level> macros that use the
> > > > > new #define dev_fmt
> > > > > 
> > > > > Miscellanea:
> > > > > 
> > > > > o Consistently use #defines with fmt, ... and ##__VA_ARGS__
> > > > > o Remove unnecessary externs
> > > > SHouldn't these be separate patches please?
> > > Multiple patches touching the same lines are unnecessary work.
> > > 
> > > So I don't think so as it's just touching bits that need change
> > > anyway.
> > > 
> > > > > Signed-off-by: Joe Perches <joe@...ches.com>
> > > > > ---
> > > > >   drivers/base/core.c    |  12 +++---
> > > > >   include/linux/device.h | 103 ++++++++++++++++++++++++++++---------------------
> > > > >   2 files changed, 64 insertions(+), 51 deletions(-)
> > > > Ok this seems like a lot of churn for no real apparent gain.  What is
> > > > all of this getting us?
> > > > 
> > > > What is the benifit, you have more code now,
> > > > why is that good?
> > > IPMI and a few other subsystems prefix all their output
> > > dev_<level> with specific duplicated content.
> > That's crazy and strange as the whole idea was that dev_* would uniquely
> > identify the driver/device such that "prefixes" would not be needed.
> > 
> > Ugh.
> > 
> > > This centralizes those prefixes just like pr_fmt.
> > > 
> > > The IPMI changes are already in mainline and now
> > > need this mechanism to maintain their output content.
> > > 
> > > IPMI and a few other subsystems prefix all their output
> > > dev_<level> with specific duplicated content.
> > > 
> > > This centralizes those prefixes just like pr_fmt.
> > > 
> > > commit a2d70dfdda6f ("ipmi: msghandler: Add and use pr_fmt and dev_fmt,
> > > remove PFX")
> > Ok, I'll take this.  I don't like it, as drivers should really not need
> > this at all.  But ipmi is known for doing "odd" things anyway...
> 
> I've been trying to make things less odd as time has gone by.  And I'm happy
> to pull out this change.  I've been slowly working on the logging in
> the ipmi driver to clean this kind of stuff up.

Cleaning up the logging would be great to see happen.  Until then, we
can leave these patches in.

thanks,

greg k-h