Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
Date:   Mon, 21 Mar 2022 15:21:10 +0000
From:   Matthew Wilcox <willy@infradead.org>
To:     JeffleXu <jefflexu@linux.alibaba.com>
Cc:     dhowells@redhat.com, linux-cachefs@redhat.com, xiang@kernel.org,
        chao@kernel.org, linux-erofs@lists.ozlabs.org,
        torvalds@linux-foundation.org, gregkh@linuxfoundation.org,
        linux-fsdevel@vger.kernel.org, joseph.qi@linux.alibaba.com,
        bo.liu@linux.alibaba.com, tao.peng@linux.alibaba.com,
        gerry@linux.alibaba.com, eguan@linux.alibaba.com,
        linux-kernel@vger.kernel.org, luodaowen.backend@bytedance.com
Subject: Re: [PATCH v5 03/22] cachefiles: introduce on-demand read mode
Message-ID: <YjiX5oXYkmN6WrA3@casper.infradead.org>
References: <20220316131723.111553-1-jefflexu@linux.alibaba.com>
 <20220316131723.111553-4-jefflexu@linux.alibaba.com>
 <YjiAVezd5B9auhcP@casper.infradead.org>
 <6bc551d2-15fc-5d17-c99b-8db588c6b671@linux.alibaba.com>
 <YjiLACenpRV4XTcs@casper.infradead.org>
 <adb957da-8909-06d8-1b2c-b8a293b37930@linux.alibaba.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <adb957da-8909-06d8-1b2c-b8a293b37930@linux.alibaba.com>
Precedence: bulk

On Mon, Mar 21, 2022 at 11:18:05PM +0800, JeffleXu wrote:
> >> Besides, IMHO read-write lock shall be more performance friendly, since
> >> most cases are the read side.
> > 
> > That's almost never true.  rwlocks are usually a bad idea because you
> > still have to bounce the cacheline, so you replace lock contention
> > (which you can see) with cacheline contention (which is harder to
> > measure).  And then you have questions about reader/writer fairness
> > (should new readers queue behind a writer if there's one waiting, or
> > should a steady stream of readers be able to hold a writer off
> > indefinitely?)
> 
> Interesting, I didn't notice it before. Thanks for explaining it.

No problem.  It's hard to notice.

> BTW what I want is just
> 
> ```
> PROCESS 1		PROCESS 2
> =========		=========
> #lock			#lock
> set DEAD state		if (not DEAD)
> flush xarray		   enqueue into xarray
> #unlock			#unlock
> ```
> 
> I think it is a generic paradigm. So it seems that the spinlock inside
> xarray is already adequate for this job?

Absolutely; just use xa_lock() to protect both setting & testing the
flag.