Re: [PATCH] remove memory.limit v.s. memsw.limit comparison.

[KAMEZAWA Hiroyuki]

On Thu, 4 Jun 2009 16:31:10 +0800
Balbir Singh <balbir@xxxxxxxxxxxxxxxxxx> wrote:

> * KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> [2009-06-04 14:10:43]:
> 
> > From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
> > 
> > Removes memory.limit < memsw.limit at setting limit check completely.
> > 
> > The limitation "memory.limit <= memsw.limit" was added just because
> > it seems sane ...if memory.limit > memsw.limit, only memsw.limit works.
> > 
> > But To implement this limitation, we needed to use private mutex and make
> > the code a bit complated.
> > As Nishimura pointed out, in real world, there are people who only want
> > to use memsw.limit.
> > 
> > Then, this patch removes the check. user-land library or middleware can check
> > this in userland easily if this really concerns.
> > 
> > And this is a good change to charge-and-reclaim.
> > 
> > Now, memory.limit is always checked before memsw.limit
> > and it may do swap-out. But, if memory.limit == memsw.limit, swap-out is
> > finally no help and hits memsw.limit again. So, let's allow the condition
> > memory.limit > memsw.limit. Then we can skip unnecesary swap-out.
> > 
> > Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
> 
> We can't change behaviour this way without breaking userspace scripts,
> API and code. What does it mean for people already using these
> features? Does it break their workflow?
> 

Hopefully no breaks to current users's workflow.
Because this just has influences to "error path" like below

 echo 200M > memory.memsw.limit
 echo 300M > memory.limit
 => ERROR

If the user program made in sane way, above case will never happens because
they set memsw.limit to be greater than memory.limit and above is treated as error.


Thanks,
-Kame


--
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/