Re: [PATCH] linux-2.417 devfs 64bit portablility issue

["Carsten Otte" <COTTE@de.ibm.com>]

Hi Richard!

>BTW: please don't send attachments. Send patches inline instead.
Sorry for sending the patch as attachment, but Notes messes
up whitespace so the patch would'nt apply if I include it directly.

>Sorry, but I find your approach grotesque. Apart from basic warts such
>as not declaring code+data as __init, the approach of populating the
>bitfield from yet another list doesn't appeal to me. I'd much rather
>see an approach which preserved the initialisation using bitmasks.
I do not think this patch is very nice either & it does not work at
all (the initialisation of the array is only called in error case).
I find the overall thing for registering/deregistering devices &
allocating majors very inconsistent.
devfs_alloc_major and devfs_register_*dev do hold the information
about which majors are allocated in two different places without
knowing about each other (bdops field and this private bitfield).
A good solution would be if *dev_register would never return a
major being statically allocated when called with major 0. If this is the
case, I do not see what alloc_major and dealloc_major are useful for.

mit freundlichem Gruß / with kind regards
Carsten Otte

IBM Deutschland Entwicklung GmbH
Linux for eServer development - device driver team
Phone: +49/07031/16-4076
IBM internal phone: *120-4076
--
We are Linux.
Resistance indicates that you're missing the point!