James Bottomley wrote:
>
> bios are supposed to be used by filesystems ... they represent the input
> to block. In principle, it seems nicer that PC requests deal with pages
> as their natural input ... that way we don't have to worry about
> exporting the block interfaces for bio->request construction.
>
This is my point. I'm a filesystem and I want a natural route from
bio to request, only with a paradigm shift as you call it.
For a filesystem the most natural and comfortable vehicle is a bio.
So the interface I would like to have is:
struct request *blk_make_request(struct request_queue *, struct bio *, gfp_t);
And let block layer take care of it's internals. Off course when crafting above
the generic_make_request will use above API so there is absolutely no duplication
of code and block layer is free to change as it feels like, only in a single place.
>> I will try the blk_rq_map_user, I'm sure it will not be pretty, though.
>
> Great, thanks ... we can also adjust block interfaces to help.
>
> James
>
I'm on it. It's not as bad as I thought, and it should work for what I have now.
I'm touching fast-path block layer code though, so we'll have to test it for a long
while.
I have a gut feeling that I will regret it some day, but we'll cross that bridge when
we get there.
Thank you for your help and concern.
Boaz
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
