lkml.org 
[lkml]   [2013]   [Jun]   [6]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] TTY:vt: convert remain take_over_console's users to do_take_over_console
On Thu, Jun 06, 2013 at 01:47:46PM +0200, Pavel Machek wrote:
> On Thu 2013-06-06 09:23:13, Wang YanQing wrote:
> > On Tue, Jun 04, 2013 at 10:13:18PM +0200, Pavel Machek wrote:
> > > On Tue 2013-05-21 13:15:12, Wang YanQing wrote:
> > > > Impact:
> > > > 1:convert all remain take_over_console to do_take_over_console
> > >
> > > This is step backwards.
> >
> > What is step backwards? do_take_over_console appear MUCH MUCH later in kernel
> > than take_over_console, do_take_over_console is the new API, I can't understand
> > what is step backwards.
>
> "do_*" is internal api, "*" is external api. You sprinkle internal api
> all over the place.
>

internal vs external? No, they only have one difference, callee vs caller hold
the console lock.

> > > > --- a/arch/alpha/kernel/console.c
> > > > +++ b/arch/alpha/kernel/console.c
> > > > @@ -61,7 +61,9 @@ locate_and_init_vga(void *(*sel_func)(void *, void *))
> > > >
> > > > /* Set the VGA hose and init the new console. */
> > > > pci_vga_hose = hose;
> > > > - take_over_console(&vga_con, 0, MAX_NR_CONSOLES-1, 1);
> > > > + console_lock();
> > > > + do_take_over_console(&vga_con, 0, MAX_NR_CONSOLES-1, 1);
> > > > + console_unlock();
> > > > }
> > >
> > > Original was better.
> >
> > Except reduce some console_lock/unlock scatter scattered in kernel, I
> > can't see the "BETTER", and it is not a BIG problem for the benefit to
> > unify the API.
>
> You replaced "calling exported function" with "calling
> internal-sounding function and adding locking too".

do_take_over_console is also exported by:

EXPORT_SYMBOL_GPL(do_take_over_console)


\
 
 \ /
  Last update: 2013-06-06 17:01    [W:0.054 / U:5.060 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site