Re: Create printk@xxxxxxxxxxxxxxx

[Willy Tarreau]

On Fri, Oct 09, 2026 at 06:34:27PM +0100, David Laight wrote:
> On Fri, 9 Oct 2026 17:00:53 +0200
> Willy Tarreau <w@xxxxxx> wrote:
> 
> > On Fri, Oct 09, 2026 at 03:26:27PM +0100, David Laight wrote:
> > > On Fri, 9 Oct 2026 15:27:07 +0200
> > > Petr Mladek <pmladek@xxxxxxxx> wrote:
> > >   
> > > > Address    : printk@xxxxxxxxxxxxxxx
> > > > Description: Linux printk and kernel string formatting development
> > > > Owners     : pmladek@xxxxxxxx, john.ogness@xxxxxxxxxxxxx, andriy.shevchenko@xxxxxxxxxxxxxxx, rostedt@xxxxxxxxxxx
> > > > Allow HTML : N
> > > > Archives   : Y
> > > > 
> > > > Reasons for the list and additional info:
> > > > 
> > > > Honestly, the main motivation is to allow sending Sashiko review
> > > > for changes in PRINTK and VSPRINTF subsystems as a reply to
> > > > the mailing list.  
> > > 
> > > There is also the smaller/faster printf code in 'nolibc'.
> > > Might be worth mentioning that as well.  
> > 
> > Better not mix the two, nolibc is purely user land code and doesn't
> > have the same constraints.
> 
> The kernel version is actually horrid, rewriting based (loosely) on the nolibc
> version might be an improvement.
> But I'm not sure how I could do it without it mostly being 'one big patch'.
> The nolibc version is much more of an embedded system version and a usual libc one.
> Things like not calling a function to add every character to the output help!

There's obviously nothing wrong with taking inspiration or code there, but
I meant we should not pollute printk() discussions with nolibc's constraints
and use cases.

Willy