Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Tue, 23 Jun 2020 14:33:11 -0700 (PDT)
Message-Id: <20200623.143311.995885759487352025.davem@davemloft.net>
To:     likaige@loongson.cn
Cc:     benve@cisco.com, _govind@gmx.com, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org, lixuefeng@loongson.cn,
        yangtiezhu@loongson.cn
Subject: Re: [PATCH RESEND] net/cisco: Fix a sleep-in-atomic-context bug in
 enic_init_affinity_hint()
From:   David Miller <davem@davemloft.net>
In-Reply-To: <1592899989-22049-1-git-send-email-likaige@loongson.cn>
References: <1592899989-22049-1-git-send-email-likaige@loongson.cn>
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

From: Kaige Li <likaige@loongson.cn>
Date: Tue, 23 Jun 2020 16:13:09 +0800

> The kernel module may sleep with holding a spinlock.
> 
> The function call paths (from bottom to top) are:
> 
> [FUNC] zalloc_cpumask_var(GFP_KERNEL)
> drivers/net/ethernet/cisco/enic/enic_main.c, 125: zalloc_cpumask_var in enic_init_affinity_hint
> drivers/net/ethernet/cisco/enic/enic_main.c, 1918: enic_init_affinity_hint in enic_open
> drivers/net/ethernet/cisco/enic/enic_main.c, 2348: enic_open in enic_reset
> drivers/net/ethernet/cisco/enic/enic_main.c, 2341: spin_lock in enic_reset
> 
> To fix this bug, GFP_KERNEL is replaced with GFP_ATOMIC.
> 
> Signed-off-by: Kaige Li <likaige@loongson.cn>

Just grepping around for GFP_KERNEL usage in atomic contexts I guess
is fine.

But you really have to look at the bigger picture.

Calling a NIC driver open function from a context holding a spinlock
is very much the real problem, so many operations have to sleep and
in face that ->ndo_open() method is defined as being allowed to sleep
and that's why the core networking never invokes it with spinlocks
held.