Re: [Qemu-devel] [PATCH arm-devs v1 04/13] net/cadence_gem: simplify rx buf descriptor walking

[Peter Crosthwaite]

On Mon, Dec 2, 2013 at 10:12 PM, Peter Maydell <address@hidden> wrote:
> On 2 December 2013 07:11, Peter Crosthwaite
> <address@hidden> wrote:
>> There was a replication of the rx descriptor address walking logic.
>> Reorder the flow control to remove. This refactoring also obsoletes
>> the local variables packet_desc_addr and last_desc_addr.
>>
>> Signed-off-by: Peter Crosthwaite <address@hidden>
>> ---
>>
>>  hw/net/cadence_gem.c | 39 ++++++++++-----------------------------
>>  1 file changed, 10 insertions(+), 29 deletions(-)
>>
>> diff --git a/hw/net/cadence_gem.c b/hw/net/cadence_gem.c
>> index 3f30caf..73ac0d8 100644
>> --- a/hw/net/cadence_gem.c
>> +++ b/hw/net/cadence_gem.c
>> @@ -586,7 +586,6 @@ static int gem_mac_address_filter(GemState *s, const 
>> uint8_t *packet)
>>  static ssize_t gem_receive(NetClientState *nc, const uint8_t *buf, size_t 
>> size)
>>  {
>>      unsigned    desc[2];
>> -    hwaddr packet_desc_addr, last_desc_addr;
>>      GemState *s;
>>      unsigned   rxbufsize, bytes_to_copy;
>>      unsigned   rxbuf_offset;
>> @@ -667,17 +666,16 @@ static ssize_t gem_receive(NetClientState *nc, const 
>> uint8_t *buf, size_t size)
>>
>>      DB_PRINT("config bufsize: %d packet size: %ld\n", rxbufsize, size);
>>
>> -    packet_desc_addr = s->rx_desc_addr;
>> -    while (1) {
>> -        DB_PRINT("read descriptor 0x%x\n", (unsigned)packet_desc_addr);
>> +    while (bytes_to_copy) {
>> +        DB_PRINT("read descriptor 0x%x\n", (unsigned)s->rx_desc_addr);
>
> Can we get here with bytes_to_copy == 0 ?  The logic sets size to a minimum
> of 64 bytes but we've already set up bytes_to_copy before that.
>

Should we be able to get here? Wouldn't that means you are trying to
recieve a length 0 packet? Could just convert to do loop to receive a
length 0 packet when gem_recieve is called with size = 0.


Regards,
Peter

> -- PMM
>