On Fri, 18 Jul 2014, Sasha Levin wrote:

On 07/17/2014 12:12 PM, Sasha Levin wrote:

quoted

On 07/17/2014 12:10 PM, Vlastimil Babka wrote:

quoted

quoted

On 07/15/2014 12:28 PM, Hugh Dickins wrote:

quoted

quoted

In the end I decided that we had better look at it as two problems,
the trinity faulting starvation, and the indefinite punching loop,
so 1/2 and 2/2 present both solutions: belt and braces.

I tested that with my reproducer and it was OK, but as I already said, it's not trinity so I didn't observe the new problems in the first place.

I've started seeing a new hang in the lru code, but I'm not sure if
it's related to this patch or not (the locks are the same ones, but
the location is very different).

I'm looking into that.

Hi Hugh,

The new hang I'm seeing is much simpler to analyse (compared to shmem_fallocate) and
doesn't seem to be related. I'll send a separate mail and Cc you just in case, but
I don't think that this patchset has anything to do with it.

Thanks for testing and following up, Sasha.  I agree, that one is
unrelated.  I've just sent a suggestion in your lru_add_drain_all thread.

Otherwise, I've been unable to reproduce the shmem_fallocate hang.

Great.  Andrew, I think we can say that it's now safe to send
1/2 shmem: fix faulting into a hole, not taking i_mutex
2/2 shmem: fix splicing from a hole while it's punched
on to Linus whenever suits you.

(You have some other patches in the mainline-later section of the
mmotm/series file: they're okay too, but not in doubt as these two were.)

Thanks,
Hugh

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>