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;
MIME-Version: 1.0
In-Reply-To: <20200325191103.GA6495@infradead.org>
References: <20200325002847.2140-1-nbowler@draconx.ca> <20200325191103.GA6495@infradead.org>
From:   Nick Bowler <nbowler@draconx.ca>
Date:   Wed, 25 Mar 2020 16:15:44 -0400
Message-ID: <CADyTPExA+Ng=QR33ZMA9qffSOEMrMVDsgU3G2nLvA9Zn3DYD+w@mail.gmail.com>
Subject: Re: [PATCH] nvme: Fix NVME_IOCTL_ADMIN_CMD compat address handling.
To:     Christoph Hellwig <hch@infradead.org>
Cc:     linux-nvme@lists.infradead.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 25/03/2020, Christoph Hellwig <hch@infradead.org> wrote:
> A couple of comments:
>
> No need for the "." the end of the subject line.
>
> I also think you should just talk of the nvme_user_cmd function,
> as that also is used for the NVME_IOCTL_IO_CMD ioctl.  Also there now
> is a nvme_user_cmd64 variant that needs the same fix, can you also
> include that?

Fair enough.  I can make the same change there... but I don't know how
to actually test the nvme_user_cmd64 function because I cannot find any
users of the NVME_IOCTL_ADMIN64_CMD or NVME_IOCTL_IO64_CMD ioctls.

>> +	if (in_compat_syscall()) {
>> +		/*
>> +		 * On real 32-bit kernels this implementation ignores the
>> +		 * upper bits of address fields so we must replicate that
>> +		 * behaviour in the compat case.
>
> s/real //g please, there are no fake 32-vit kernels :)

OK.

I shall prepare a v2 patch then with this feedback addressed.

Thanks,
  Nick