From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <yuanhan.liu@linux.intel.com>
Received: from mga14.intel.com (mga14.intel.com [192.55.52.115])
 by dpdk.org (Postfix) with ESMTP id 80658C5BA
 for <dev@dpdk.org>; Fri, 29 Jan 2016 09:56:49 +0100 (CET)
Received: from fmsmga001.fm.intel.com ([10.253.24.23])
 by fmsmga103.fm.intel.com with ESMTP; 29 Jan 2016 00:56:48 -0800
X-ExtLoop1: 1
X-IronPort-AV: E=Sophos;i="5.22,363,1449561600"; d="scan'208";a="891726581"
Received: from yliu-dev.sh.intel.com (HELO yliu-dev) ([10.239.66.49])
 by fmsmga001.fm.intel.com with ESMTP; 29 Jan 2016 00:56:36 -0800
Date: Fri, 29 Jan 2016 16:57:23 +0800
From: Yuanhan Liu <yuanhan.liu@linux.intel.com>
To: Tetsuya Mukawa <mukawa@igel.co.jp>
Message-ID: <20160129085723.GV4257@yliu-dev.sh.intel.com>
References: <1453108389-21006-2-git-send-email-mukawa@igel.co.jp>
 <1453374478-30996-6-git-send-email-mukawa@igel.co.jp>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1453374478-30996-6-git-send-email-mukawa@igel.co.jp>
User-Agent: Mutt/1.5.23 (2014-03-12)
Cc: dev@dpdk.org
Subject: Re: [dpdk-dev] [RFC PATCH 5/5] virtio: Extend virtio-net PMD to
 support container environment
X-BeenThere: dev@dpdk.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: patches and discussions about DPDK <dev.dpdk.org>
List-Unsubscribe: <http://dpdk.org/ml/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <http://dpdk.org/ml/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
X-List-Received-Date: Fri, 29 Jan 2016 08:56:50 -0000

On Thu, Jan 21, 2016 at 08:07:58PM +0900, Tetsuya Mukawa wrote:
> +static int
> +virt_read_pci_cfg(struct virtio_hw *hw, void *buf, size_t len, off_t offset)
> +{
> +	qtest_read_pci_cfg(hw, "virtio-net", buf, len, offset);
> +	return 0;
> +}
> +
> +static void *
> +virt_get_mapped_addr(struct virtio_hw *hw, uint8_t bar,
> +		     uint32_t offset, uint32_t length)
> +{
> +	uint64_t base;
> +	uint64_t size;
> +
> +	if (qtest_get_bar_size(hw, "virtio-net", bar, &size) < 0) {
> +		PMD_INIT_LOG(ERR, "invalid bar: %u", bar);
> +		return NULL;
> +	}
> +
> +	if (offset + length < offset) {
> +		PMD_INIT_LOG(ERR, "offset(%u) + lenght(%u) overflows",
> +			offset, length);
> +		return NULL;
> +	}
> +
> +	if (offset + length > size) {
> +		PMD_INIT_LOG(ERR,
> +			"invalid cap: overflows bar space: %u > %"PRIu64,
> +			offset + length, size);
> +		return NULL;
> +	}
> +
> +	if (qtest_get_bar_addr(hw, "virtio-net", bar, &base) < 0) {
> +		PMD_INIT_LOG(ERR, "invalid bar: %u", bar);
> +		return NULL;
> +	}

So, I understood the usage now, and the cfg_ops abstraction doesn't look
good yet necessary to me.  For EAL managed pci device, bar length and
addr are stored at memory_resources[], and for your case, it's from the
qtest. And judging that it's compile time decision, I'd like it to be:

    #ifdef /* RTE_LIBRTE_VIRTIO_HOST_MODE */
    
    static uint32_t
    get_bar_size(...)
    {
    	return qtest_get_bar_size(..);
    }
    
    static uint64-t
    get_bar_addr(...)
    {
    	return qtest_get_bar_addr(..);
    }
    
    ...
    ...
    
    #else
    
    static  uint32_t
    get_bar_size(...)
    {
    	return dev->mem_resource[bar].addr;
    }
    
    ...
    
    }
    #endif


And then you just need do related changes at virtio_read_caps() and
get_cfg_addr(). That'd be much simpler, without introducing duplicate
code and uncessary complex.

What do you think of that?

	--yliu

> +
> +	return (void *)(base + offset);
> +}
> +
> +static const struct virtio_pci_cfg_ops virt_cfg_ops = {
> +	.map			= virt_map_pci_cfg,
> +	.unmap			= virt_unmap_pci_cfg,
> +	.get_mapped_addr	= virt_get_mapped_addr,
> +	.read			= virt_read_pci_cfg,
> +};
> +#endif /* RTE_LIBRTE_VIRTIO_HOST_MODE */