Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Thu, 9 Apr 2020 12:32:19 -0700
From:   Jakub Kicinski <kuba@kernel.org>
To:     Keita Suzuki <keitasuzuki.park@sslab.ics.keio.ac.jp>
Cc:     takafumi.kubota1012@sslab.ics.keio.ac.jp, davem@davemloft.net,
        oss-drivers@netronome.com (open list:NETRONOME ETHERNET DRIVERS),
        netdev@vger.kernel.org (open list:NETWORKING DRIVERS),
        linux-kernel@vger.kernel.org (open list)
Subject: Re: [PATCH] nfp: Fix memory leak in nfp_resource_acquire()
Message-ID: <20200409123203.1b5f6534@kicinski-fedora-PC1C0HJN>
In-Reply-To: <20200409174113.28635-1-keitasuzuki.park@sslab.ics.keio.ac.jp>
References: <20200409150210.15488-1-keitasuzuki.park@sslab.ics.keio.ac.jp>
        <20200409174113.28635-1-keitasuzuki.park@sslab.ics.keio.ac.jp>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu,  9 Apr 2020 17:41:11 +0000 Keita Suzuki wrote:
> This patch fixes a memory leak in nfp_resource_acquire(). res->mutex is
> alllocated in nfp_resource_try_acquire(). However, when
> msleep_interruptible() or time_is_before_eq_jiffies() fails, it falls
> into err_fails path where res is freed, but res->mutex is not.
> 
> Fix this by changing call to free to nfp_resource_release().

I don't see a leak here. Maybe you could rephrase the description to
make things clearer?

AFAICS nfp_resource_try_acquire() calls nfp_cpp_mutex_free(res->mutex)
if it's about to return an error. We can only hit msleep or time check
if it returned an error.