Am 01.11.2009 18:52 schrieb Randy Dunlap:
> On Sat, 31 Oct 2009 13:11:29 +0100 Tilman Schmidt wrote:
> 
>> Am 31.10.2009 11:17 schrieb Nigel Cunningham:
>>> I've only looked at this one, but something caught my eye...
>>>
>>> André Goddard Rosa wrote:
>>>> diff --git a/arch/blackfin/kernel/cplb-mpu/cplbinit.c
>>>> b/arch/blackfin/kernel/cplb-mpu/cplbinit.c
>>>> index f7b9cdc..b52c1f8 100644
>>>> --- a/arch/blackfin/kernel/cplb-mpu/cplbinit.c
>>>> +++ b/arch/blackfin/kernel/cplb-mpu/cplbinit.c
>>>> @@ -38,7 +38,7 @@ void __init generate_cplb_tables_cpu(unsigned int cpu)
>>>>
>>>>  #ifdef CONFIG_BFIN_EXTMEM_DCACHEABLE
>>>>  	d_cache = CPLB_L1_CHBL;
>>>> -#ifdef CONFIG_BFIN_EXTMEM_WRITETROUGH
>>>> +#ifdef CONFIG_BFIN_EXTMEM_WRITETHROUGH
>>>>  	d_cache |= CPLB_L1_AOW | CPLB_WT;
>>>>  #endif
>>>>  #endif
>>> "Surely that can't be the only time?", I thought. But it is - all of the
>>> other #ifdefs are spelled correctly, so you're probably fixing a
>>> hard-to-find bug here! Good catch!
>>>
>>> Makes me wonder if anyone's written a tool to find things like this -
>>> #ifdefs that will never be defined...
>> I seem to remember someone (Joe Perches?) did, and regularly posted
>> results to LKML in the past, but at some point switched to publishing
>> them on a web page instead.
> 
> It was Robert PJ Day.

Indeed it was, and kudos to him for doing it.
Please forgive my middle-aged brain for the misattribution.

> See http://www.crashcourse.ca/wiki/index.php/Kernel_cleanup

And sure enough, CONFIG_BFIN_EXTMEM_WRITETROUGH is listed there on the
"Badref CONFIG variables" page.

It seems that page would merit a bit of advertising. :-)

-- 
Tilman Schmidt                    E-Mail: tilman@imap.cc
Bonn, Germany
Diese Nachricht besteht zu 100% aus wiederverwerteten Bits.
Ungeöffnet mindestens haltbar bis: (siehe Rückseite)