Re: [PATCH] rpmsg: virtio: add endianness conversions

[Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>]

Hi Mathieu,

Thanks for your review.

On Mon, Jun 15, 2020 at 02:47:07PM -0600, Mathieu Poirier wrote:
> On Mon, Jun 08, 2020 at 02:39:32PM +0200, Guennadi Liakhovetski wrote:
> > According to the VirtIO 1.0 spec data, sent over virtual queues must
> > be in little-endian format. Update the RPMsg VirtIO implementation
> > to enforce that.
> > 
> > Suggested-by: Michael S. Tsirkin <mst@redhat.com>
> > Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
> > ---
> >  drivers/rpmsg/virtio_rpmsg_bus.c | 61 ++++++++++++++++++++++------------------
> >  1 file changed, 33 insertions(+), 28 deletions(-)
> > 
> > diff --git a/drivers/rpmsg/virtio_rpmsg_bus.c b/drivers/rpmsg/virtio_rpmsg_bus.c
> > index 07d4f33..b8ff42b 100644
> > --- a/drivers/rpmsg/virtio_rpmsg_bus.c
> > +++ b/drivers/rpmsg/virtio_rpmsg_bus.c
> > @@ -11,6 +11,8 @@
> >  
> >  #define pr_fmt(fmt) "%s: " fmt, __func__
> >  
> > +#include <asm/byteorder.h>
> > +
> >  #include <linux/dma-mapping.h>
> >  #include <linux/idr.h>
> >  #include <linux/jiffies.h>
> > @@ -22,6 +24,7 @@
> >  #include <linux/scatterlist.h>
> >  #include <linux/slab.h>
> >  #include <linux/sched.h>
> > +#include <linux/types.h>
> >  #include <linux/virtio.h>
> >  #include <linux/virtio_ids.h>
> >  #include <linux/virtio_config.h>
> > @@ -84,11 +87,11 @@ struct virtproc_info {
> >   * Every message sent(/received) on the rpmsg bus begins with this header.
> >   */
> >  struct rpmsg_hdr {
> > -	u32 src;
> > -	u32 dst;
> > -	u32 reserved;
> > -	u16 len;
> > -	u16 flags;
> > +	__le32 src;
> > +	__le32 dst;
> > +	__le32 reserved;
> > +	__le16 len;
> > +	__le16 flags;
> >  	u8 data[];
> >  } __packed;
> >  
> > @@ -106,8 +109,8 @@ struct rpmsg_hdr {
> >   */
> >  struct rpmsg_ns_msg {
> >  	char name[RPMSG_NAME_SIZE];
> > -	u32 addr;
> > -	u32 flags;
> > +	__le32 addr;
> > +	__le32 flags;
> >  } __packed;
> >  
> >  /**
> > @@ -335,8 +338,8 @@ static int virtio_rpmsg_announce_create(struct rpmsg_device *rpdev)
> >  		struct rpmsg_ns_msg nsm;
> >  
> >  		strncpy(nsm.name, rpdev->id.name, RPMSG_NAME_SIZE);
> > -		nsm.addr = rpdev->ept->addr;
> > -		nsm.flags = RPMSG_NS_CREATE;
> > +		nsm.addr = cpu_to_le32(rpdev->ept->addr);
> > +		nsm.flags = cpu_to_le32(RPMSG_NS_CREATE);
> 
> So how will this work with existing firmware?  

Is this currently running on any BE machines?

> >  		err = rpmsg_sendto(rpdev->ept, &nsm, sizeof(nsm), RPMSG_NS_ADDR);
> >  		if (err)
> > @@ -359,8 +362,8 @@ static int virtio_rpmsg_announce_destroy(struct rpmsg_device *rpdev)
> >  		struct rpmsg_ns_msg nsm;
> >  
> >  		strncpy(nsm.name, rpdev->id.name, RPMSG_NAME_SIZE);
> > -		nsm.addr = rpdev->ept->addr;
> > -		nsm.flags = RPMSG_NS_DESTROY;
> > +		nsm.addr = cpu_to_le32(rpdev->ept->addr);
> > +		nsm.flags = cpu_to_le32(RPMSG_NS_DESTROY);
> >  
> >  		err = rpmsg_sendto(rpdev->ept, &nsm, sizeof(nsm), RPMSG_NS_ADDR);
> >  		if (err)
> > @@ -612,15 +615,15 @@ static int rpmsg_send_offchannel_raw(struct rpmsg_device *rpdev,
> >  		}
> >  	}
> >  
> > -	msg->len = len;
> > +	msg->len = cpu_to_le16(len);
> >  	msg->flags = 0;
> > -	msg->src = src;
> > -	msg->dst = dst;
> > +	msg->src = cpu_to_le32(src);
> > +	msg->dst = cpu_to_le32(dst);
> >  	msg->reserved = 0;
> >  	memcpy(msg->data, data, len);
> >  
> >  	dev_dbg(dev, "TX From 0x%x, To 0x%x, Len %d, Flags %d, Reserved %d\n",
> > -		msg->src, msg->dst, msg->len, msg->flags, msg->reserved);
> > +		src, dst, len, msg->flags, msg->reserved);
> >  #if defined(CONFIG_DYNAMIC_DEBUG)
> >  	dynamic_hex_dump("rpmsg_virtio TX: ", DUMP_PREFIX_NONE, 16, 1,
> >  			 msg, sizeof(*msg) + msg->len, true);
> > @@ -704,13 +707,15 @@ static int rpmsg_recv_single(struct virtproc_info *vrp, struct device *dev,
> >  {
> >  	struct rpmsg_endpoint *ept;
> >  	struct scatterlist sg;
> > +	unsigned int msg_len = le16_to_cpu(msg->len);
> 
> And here too, along with everything else that follows?
> 
> As Michael suggested using __virtioXY types would allow to use the
> virtioXY_to_cpu() and cpu_to_virtioXY() macros, which is definitely a step in
> the right direction.

Ok, sure, I'll redo this, I just wanted to see if there would be any more comments 
before submitting a new version.

Thanks
Guennadi

> 
> Thanks,
> Mathieu
> 
> >  	int err;
> >  
> >  	dev_dbg(dev, "From: 0x%x, To: 0x%x, Len: %d, Flags: %d, Reserved: %d\n",
> > -		msg->src, msg->dst, msg->len, msg->flags, msg->reserved);
> > +		le32_to_cpu(msg->src), le32_to_cpu(msg->dst), msg_len,
> > +		le16_to_cpu(msg->flags), le32_to_cpu(msg->reserved));
> >  #if defined(CONFIG_DYNAMIC_DEBUG)
> >  	dynamic_hex_dump("rpmsg_virtio RX: ", DUMP_PREFIX_NONE, 16, 1,
> > -			 msg, sizeof(*msg) + msg->len, true);
> > +			 msg, sizeof(*msg) + msg_len, true);
> >  #endif
> >  
> >  	/*
> > @@ -718,15 +723,15 @@ static int rpmsg_recv_single(struct virtproc_info *vrp, struct device *dev,
> >  	 * the reported payload length.
> >  	 */
> >  	if (len > vrp->buf_size ||
> > -	    msg->len > (len - sizeof(struct rpmsg_hdr))) {
> > -		dev_warn(dev, "inbound msg too big: (%d, %d)\n", len, msg->len);
> > +	    msg_len > (len - sizeof(struct rpmsg_hdr))) {
> > +		dev_warn(dev, "inbound msg too big: (%d, %d)\n", len, msg_len);
> >  		return -EINVAL;
> >  	}
> >  
> >  	/* use the dst addr to fetch the callback of the appropriate user */
> >  	mutex_lock(&vrp->endpoints_lock);
> >  
> > -	ept = idr_find(&vrp->endpoints, msg->dst);
> > +	ept = idr_find(&vrp->endpoints, le32_to_cpu(msg->dst));
> >  
> >  	/* let's make sure no one deallocates ept while we use it */
> >  	if (ept)
> > @@ -739,8 +744,8 @@ static int rpmsg_recv_single(struct virtproc_info *vrp, struct device *dev,
> >  		mutex_lock(&ept->cb_lock);
> >  
> >  		if (ept->cb)
> > -			ept->cb(ept->rpdev, msg->data, msg->len, ept->priv,
> > -				msg->src);
> > +			ept->cb(ept->rpdev, msg->data, msg_len, ept->priv,
> > +				le32_to_cpu(msg->src));
> >  
> >  		mutex_unlock(&ept->cb_lock);
> >  
> > @@ -846,15 +851,15 @@ static int rpmsg_ns_cb(struct rpmsg_device *rpdev, void *data, int len,
> >  	/* don't trust the remote processor for null terminating the name */
> >  	msg->name[RPMSG_NAME_SIZE - 1] = '\0';
> >  
> > -	dev_info(dev, "%sing channel %s addr 0x%x\n",
> > -		 msg->flags & RPMSG_NS_DESTROY ? "destroy" : "creat",
> > -		 msg->name, msg->addr);
> > -
> >  	strncpy(chinfo.name, msg->name, sizeof(chinfo.name));
> >  	chinfo.src = RPMSG_ADDR_ANY;
> > -	chinfo.dst = msg->addr;
> > +	chinfo.dst = le32_to_cpu(msg->addr);
> > +
> > +	dev_info(dev, "%sing channel %s addr 0x%x\n",
> > +		 le32_to_cpu(msg->flags) & RPMSG_NS_DESTROY ? "destroy" : "creat",
> > +		 msg->name, chinfo.dst);
> >  
> > -	if (msg->flags & RPMSG_NS_DESTROY) {
> > +	if (le32_to_cpu(msg->flags) & RPMSG_NS_DESTROY) {
> >  		ret = rpmsg_unregister_device(&vrp->vdev->dev, &chinfo);
> >  		if (ret)
> >  			dev_err(dev, "rpmsg_destroy_channel failed: %d\n", ret);
> > -- 
> > 1.9.3
> >