Re: Re: [PATCH v6 0/7] introduce memory hinting API for external process

From: SeongJae Park <sjpark@amazon.com>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: Minchan Kim <minchan@kernel.org>,
	LKML <linux-kernel@vger.kernel.org>,
	linux-mm <linux-mm@kvack.org>, <linux-api@vger.kernel.org>,
	<oleksandr@redhat.com>, Suren Baghdasaryan <surenb@google.com>,
	Tim Murray <timmurray@google.com>,
	Daniel Colascione <dancol@google.com>,
	Sandeep Patil <sspatil@google.com>,
	Sonny Rao <sonnyrao@google.com>,
	Brian Geffon <bgeffon@google.com>, Michal Hocko <mhocko@suse.com>,
	Johannes Weiner <hannes@cmpxchg.org>,
	Shakeel Butt <shakeelb@google.com>,
	John Dias <joaodias@google.com>,
	Joel Fernandes <joel@joelfernandes.org>, <sj38.park@gmail.com>,
	<alexander.h.duyck@linux.intel.com>, Jann Horn <jannh@google.com>
Subject: Re: Re: [PATCH v6 0/7] introduce memory hinting API for external process
Date: Thu, 20 Feb 2020 10:16:31 +0100	[thread overview]
Message-ID: <20200220091631.31949-1-sjpark@amazon.com> (raw)
In-Reply-To: <20200219120123.07dda51c29006a892059ccde@linux-foundation.org> (raw)

On Wed, 19 Feb 2020 12:01:23 -0800 Andrew Morton <akpm@linux-foundation.org> wrote:

> On Tue, 18 Feb 2020 17:44:26 -0800 Minchan Kim <minchan@kernel.org> wrote:
> 
> > Now, we have MADV_PAGEOUT and MADV_COLD as madvise hinting API. With that,
> > application could give hints to kernel what memory range are preferred to be
> > reclaimed. However, in some platform(e.g., Android), the information
> > required to make the hinting decision is not known to the app.
> > Instead, it is known to a centralized userspace daemon(e.g., ActivityManagerService),
> > and that daemon must be able to initiate reclaim on its own without any app
> > involvement.
> > 
> 
> This patchset doesn't seem to be getting a lot of interest from other
> potential users?  It seems very specialized.  Are there or will there
> ever be any users of this apart from one Android daemon?
> 
> Also, it doesn't terribly hard for ActivityManagerService to tell
> another process "now run madvise with these arguments".  Please explain
> why this is not practical in ActivityManagerService and also within
> other potential users of this syscall.

I personally have interest in and hope successful development/merge of this
patchset.

The interested usecases of 'madvise_process()' for me is optimizations of
general memory-intensive workloads having dynamic data access patterns on
hierarchical memory systems (e.g., multi-tier memory or fast storage based swap
devices).  In more detail, I'm already using a part of this patchset for my RFC
patchset implementing Data Access Monitoring-based Operation Schemes[1].  For
my specific case, I don't need new system call but only target task argument,
though.

Once in a past, before joining my current company, I tried using 'madvise()' to
optimize some scientific HPC programs.  The improvement results were clear, but
optimizing each of the workloads was challenging and time-consuming.  I believe
this new systemcall to be very helpful for such cases, either.

[1] https://lore.kernel.org/linux-mm/20200218085309.18346-1-sjpark@amazon.com/

Thanks,
SeongJae Park