On Mon, 2013-01-21 at 06:22 -0800, Vinod Koul wrote: > On Mon, Jan 21, 2013 at 11:45:51AM +0200, Andy Shevchenko wrote: > > > > + return 0; > > > hmmm, why not use BLOCK_TS value. That way you dont need to look at > > > direction > > > and along with burst can easily calculate residue... > > > > Do you mean to read CTL hi/lo and do > > > > desc->len - ctlhi.block_ts * ctllo.src_tr_width? > > > Yes > > I think it could be not precise when memory-to-peripheral transfer is > > going on. In that case you probably will have src_tr_width like 32 bits, > > meanwhile peripheral may receive only byte stream. > Nope that is not the case. > SAR/DAR is always incremented in src/dstn_tr_width granularity. For example if > you are using MEM to DMA, then SAR will always increment in case of x86 in > 4byte > granularity as we will read bursts not singles. > > Also if check the spec, it says "Once the transfer starts, the read-back > value is the > total number of data items already read from the source peripheral, > regardless of > what is the flow controller" > > So basically you get what is read from buffer in case of MEM->PER and get what > is read from FIFO in case of PER->MEM which IMO gives you better or equal > results > than your calulation.
I will try this. Indeed I don't like usage of direction as well, and your solution seems much clear in that sense. -- Andy Shevchenko <[email protected]> Intel Finland Oy -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
