Re: 2.6.10-rc1-mm5: REISER4_LARGE_KEY is still selectable

[Adrian Bunk <bunk@stusta.de>]

On Fri, Nov 12, 2004 at 07:37:59PM +0300, Vladimir Saveliev wrote:

> Hello

Hi Vladimir,

> On Fri, 2004-11-12 at 16:23, Adrian Bunk wrote:
> > On Fri, Nov 12, 2004 at 10:07:59AM +0300, Vladimir Saveliev wrote:
> > > On Thu, 2004-11-11 at 19:50, Adrian Bunk wrote:
> > > > REISER4_LARGE_KEY is still selectable in reiser4-include-reiser4.patch 
> > > > (and we agreed that it shouldn't be).
> > > 
> > > Sorry, concerning this problem - what did we agree about?
> > 
> > depending on the setting of REISER4_LARGE_KEY, there are two binary 
> > incompatible variants of reiser4 (which can't be both supported by one 
> > kernel).
> > 
> > Therefore, REISER4_LARGE_KEY shouldn't be asked but always enabled.
> 
> One may create reiser4 with so called short keys. In current state of
> code enabling LARGE_KEY will make impossible to use that filesystem.

exactly that's the problem.

> So, while reiser4 is not able to distinguish key type on fly we let user
> to look for and undefive REISER4_LARGE_KEY macro directly in source
> code?

My personal preference was to simply remove the REISER4_LARGE_KEY=n 
case, but Hans' suggestion to hide this option in a .h file without 
being visible from the Kconfig files [1] is also OK.

cu
Adrian

[1] http://www.ussg.iu.edu/hypermail/linux/kernel/0410.2/2834.html

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed