Fwd: bcache.patch

Fwd: bcache.patch

[Marcus Sorensen]

---------- Forwarded message ----------
From: Marcus Sorensen <shadowsor-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date: Wed, Nov 9, 2011 at 8:49 PM
Subject: bcache.patch
To: kent.overstreet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org


Was looking for bcache.patch and it's not obvious where I should be
able to find these for various kernel versions. I did suck down the
3.1 kernel tree in the git repo, but I've put some work into my
existing kernel tree and would like to use that.

 Also, various sources seem to state that it relies on filesystem
UUID, I'm assuming this means that I cannot use it to speed up a block
device that I'm exporting elsewhere (e.g. a logical volume exported as
iscsi or Fibre Channel target, that the client has
partitioned/formatted), because locally I should not be aware of or
using the filesystem, but perhaps I could use it if I were exporting
flat files as volumes to be used elsewhere (cached against the local
fs on which the flat files reside). Is this correct?

Thanks,
Marcus Sorensen

Re: bcache.patch

[Kent Overstreet]

I don't intend to keep bcache.patch up to date anymore - I need update
the wiki...

Just do a git merge though, it'd be way easier than dealing with the
patch anyways.

The docs that refer to the fs UUID are way out of date - the backing
device superblock removed the need for that. It should be able to do
exactly what you want.

On Wed, Nov 9, 2011 at 7:53 PM, Marcus Sorensen <shadowsor-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> ---------- Forwarded message ----------
> From: Marcus Sorensen <shadowsor-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> Date: Wed, Nov 9, 2011 at 8:49 PM
> Subject: bcache.patch
> To: kent.overstreet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
>
>
> Was looking for bcache.patch and it's not obvious where I should be
> able to find these for various kernel versions. I did suck down the
> 3.1 kernel tree in the git repo, but I've put some work into my
> existing kernel tree and would like to use that.
>
>  Also, various sources seem to state that it relies on filesystem
> UUID, I'm assuming this means that I cannot use it to speed up a block
> device that I'm exporting elsewhere (e.g. a logical volume exported as
> iscsi or Fibre Channel target, that the client has
> partitioned/formatted), because locally I should not be aware of or
> using the filesystem, but perhaps I could use it if I were exporting
> flat files as volumes to be used elsewhere (cached against the local
> fs on which the flat files reside). Is this correct?
>
> Thanks,
> Marcus Sorensen
> --
> To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>