Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:4 as permitted sender) client-ip=2620:137:e000::3:4;
Date:   Mon, 9 Oct 2023 08:17:50 -0700
From:   Jakub Kicinski <kuba@kernel.org>
To:     Chengfeng Ye <dg573847474@gmail.com>
Cc:     3chas3@gmail.com, davem@davemloft.net, horms@kernel.org,
        linux-atm-general@lists.sourceforge.net, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 1/2] atm: solos-pci: Fix potential deadlock on
 &cli_queue_lock
Message-ID: <20231009081750.2073013d@kernel.org>
In-Reply-To: <CAAo+4rUE=+9Kp8CvMH3w15dJotkX03h=5YMV+hu-YSobkwj1NA@mail.gmail.com>
References: <20231005074858.65082-1-dg573847474@gmail.com>
        <20231006162835.79484017@kernel.org>
        <CAAo+4rUE=+9Kp8CvMH3w15dJotkX03h=5YMV+hu-YSobkwj1NA@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Sat, 7 Oct 2023 23:58:36 +0800 Chengfeng Ye wrote:
> > and irqsave here. I think you're right that it's just softirq (== bh)
> > that may deadlock, so no need to take the irqsave() version in process
> > context.  
> 
> Yes, spin_lock_bh() is enough.
> 
> I just found spin_lock_irqsave() is more frequently used in this file, so I
> also used spin_lock_irqsave() here for uniformity consideration at that time.
> 
> Should I send a new patch series to change this to spin_lock_bh()? That's
> better for performance consideration.

Yes, performance is one reason and another is that the code will 
be easier to understand if the locking matches the requirements.