Re: [PATCH V2 2/3] drivers/staging: zcache: host services and PAMservices

[Nitin Gupta]

On 02/09/2011 06:46 PM, Minchan Kim wrote:
> Hi Nitin,
>
> Sorry for late response.
>
> On Thu, Feb 10, 2011 at 2:36 AM, Nitin Gupta<ngupta@xxxxxxxxxx>  wrote:
> > On 02/09/2011 11:39 AM, Dan Magenheimer wrote:
> > > > From: Minchan Kim [mailto:minchan.kim@xxxxxxxxx]
> > >
> > > > As I read your comment, I can't find the benefit of zram compared to
> > > > frontswap.
> > >
> > > Well, I am biased, but I agree that frontswap is a better technical
> > > solution than zram. ;-)  But "dynamic-ity" is very important to
> > > me and may be less important to others.
> >
> >
> > I agree that frontswap is better than zram when considering swap as the use
> > case - no bio overhead, dynamic resizing. However, zram being a *generic*
> > block-device has some unique cases too like hosting files on /tmp, various
> > caches under /var or any place where a compressed in-memory block device can
> > help.
>
> Yes. I mentioned that benefit but I am not sure the reason is enough.
> What I had in mind long time ago is that zram's functionality into brd.
> So someone who want to compress contents could use it with some mount
> option to enable compression.
> By such way, many ramdisk user can turn it on easily.
> If many user begin using the brd, we can see many report about
> performance then solve brd performance s as well as zcache world-wide
> usage.
>
> Hmm,  the idea is too late?
>
> > So, frontswap and zram have overlapping use case of swap but are not the
> > same.
>
> If we can insert zram's functionality into brd, maybe there is no
> reason to coexist.
>
>
> >

I thought about this before starting with zram development but thought 
adding compression (and in future, defrag) and use of custom allocator 
is just too much of a hassle and thus dropped the idea. If someone is 
anyhow interested in merging brd and zram I would be glad to help but I 
still think that is simply not worth the effort.

Thanks,
Nitin
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/