On Mon, 25 Jan 2021 17:20:23 -0600 Lijun Pan wrote:
> Ensure that received Command-Response Queue (CRQ) entries are
> properly read in order by the driver. dma_rmb barrier has
> been added before accessing the CRQ descriptor to ensure
> the entire descriptor is read before processing.
>
> Fixes: 032c5e82847a ("Driver for IBM System i/p VNIC protocol")
> Signed-off-by: Lijun Pan <l...@linux.ibm.com>
> ---
> v2: drop dma_wmb according to Jakub's opinion
>
> drivers/net/ethernet/ibm/ibmvnic.c | 8 ++++++++
> 1 file changed, 8 insertions(+)
>
> diff --git a/drivers/net/ethernet/ibm/ibmvnic.c
> b/drivers/net/ethernet/ibm/ibmvnic.c
> index 9778c83150f1..d84369bd5fc9 100644
> --- a/drivers/net/ethernet/ibm/ibmvnic.c
> +++ b/drivers/net/ethernet/ibm/ibmvnic.c
> @@ -5084,6 +5084,14 @@ static void ibmvnic_tasklet(struct tasklet_struct *t)
> while (!done) {
> /* Pull all the valid messages off the CRQ */
> while ((crq = ibmvnic_next_crq(adapter)) != NULL) {
> + /* Ensure that the entire CRQ descriptor queue->msgs
> + * has been loaded before reading its contents.
I still find this sentence confusing, maybe you mean to say stored
instead of loaded?
> + * This barrier makes sure ibmvnic_next_crq()'s
> + * crq->generic.first & IBMVNIC_CRQ_CMD_RSP is loaded
> + * before ibmvnic_handle_crq()'s
> + * switch(gen_crq->first) and switch(gen_crq->cmd).
Yup, that makes perfect sense. It's about ordering of the loads.
> + */
> + dma_rmb();
> ibmvnic_handle_crq(crq, adapter);
> crq->generic.first = 0;
> }
