On Mon, Feb 22, 2016 at 4:55 PM, Or Gerlitz <gerlitz.or@xxxxxxxxx> wrote:
> On Mon, Feb 22, 2016 at 3:49 PM, Christoph Hellwig <hch@xxxxxx> wrote:
>
>> @@ -36,9 +36,7 @@
>> -#define ISER_RECV_DATA_SEG_LEN 8192
>
> Sagi,
>
> Can you please explain WW (When & Why) we changed this from 128 bytes
> to 8192 bytes?!! do we use 8000++ bytes for every RX descriptor or
> this is nowadays only for the login phase? what's the trick?
Oh, I see now that this is isert and not iser... if this descriptor is
used for every IO and set is such hard-coded way, that is terribly
inefficient, NM for now I guess, as this patchset doesn't introduce
this,
>> @@ -92,7 +90,7 @@ enum iser_conn_state {
>> struct iser_rx_desc {
>> struct iser_ctrl iser_header;
>> struct iscsi_hdr iscsi_header;
>> - char data[ISER_RECV_DATA_SEG_LEN];
>> + char data[ISCSI_DEF_MAX_RECV_SEG_LEN];
>> u64 dma_addr;
>> struct ib_sge rx_sg;
>> char pad[ISER_RX_PAD_SIZE];
BTW that pad is probably just c&p from the initiator, has nothing to do here
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
