Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 4 Nov 2022 12:58:30 +0900
From:   Sergey Senozhatsky <senozhatsky@chromium.org>
To:     Johannes Weiner <hannes@cmpxchg.org>
Cc:     Minchan Kim <minchan@kernel.org>,
        Sergey Senozhatsky <senozhatsky@chromium.org>,
        Nhat Pham <nphamcs@gmail.com>, akpm@linux-foundation.org,
        linux-mm@kvack.org, linux-kernel@vger.kernel.org,
        ngupta@vflare.org, sjenning@redhat.com, ddstreet@ieee.org,
        vitaly.wool@konsulko.com
Subject: Re: [PATCH 2/5] zsmalloc: Consolidate zs_pool's migrate_lock and
 size_class's locks
Message-ID: <Y2SN5tMH8CqYHsYK@google.com>
References: <20221026200613.1031261-1-nphamcs@gmail.com>
 <20221026200613.1031261-3-nphamcs@gmail.com>
 <Y2Hj+H4VzlN/fcmR@google.com>
 <Y2Li412OGB6g8ARA@google.com>
 <Y2PbrOqRMLDsYev0@cmpxchg.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <Y2PbrOqRMLDsYev0@cmpxchg.org>
Precedence: bulk

On (22/11/03 11:18), Johannes Weiner wrote:
> > > I'm not in love with this, to be honest. One big pool lock instead
> > > of 255 per-class locks doesn't look attractive, as one big pool lock
> > > is going to be hammered quite a lot when zram is used, e.g. as a regular
> > > block device with a file system and is under heavy parallel writes/reads.
> 
> TBH the class always struck me as an odd scope to split the lock. Lock
> contention depends on how variable the compression rate is of the
> hottest incoming data, which is unpredictable from a user POV.
> 
> My understanding is that the primary usecase for zram is swapping, and
> the pool lock is the same granularity as the swap locking.

That's what we thought until a couple of merge windows ago we figured
(the hard way) that SUSE uses ZRAM as a normal block device with a real
file-system on it. And they use it often enough to immediately spot the
regression which we landed.

> Do you have a particular one in mind? (I'm thinking journaled ones are
> not of much interest, since their IO tends to be fairly serialized.)
> 
> btrfs?

Probably some parallel fio workloads? Seq, random reads/writes from
numerous workers.

I personally sometimes use ZRAM when I want to compile something and
I care only about the package, I don't need .o for recomplilation or
something, just the final package.