On Mon, Apr 11, 2016 at 6:51 PM, Andy Shevchenko <andy.shevchenko@xxxxxxxxx> wrote: > On Mon, Apr 11, 2016 at 6:33 PM, Bryan O'Donoghue > <pure.logic@xxxxxxxxxxxxxxxxx> wrote: >> On Thu, 2016-04-07 at 23:37 +0300, Andy Shevchenko wrote: >> >> Preface. I tried this on Galileo and it appears to work. I'll do some >> throughput testing to verify but, initially the results are positive :) > > I submitted (and pushed into my branch) a bit changed version (see my v2). > >>> + lpss->dma_maxburst = 8; >> >> Are these dwords ? If those are bytes then the maxburst value looks >> small. In the BSP the max burst is 32 bytes. > > max_burst is in items of given size (here is 32 bytes for memory and 8 > bytes for UART). I took this value from Quark BSP, but I'll be happy > to adjust to more optimal values. +Jarkko. Oy vei, seems we don't set memory side of transfer neither in 8250_dma, nor in spi-pxa2xx-dma. -- With Best Regards, Andy Shevchenko -- To unsubscribe from this list: send the line "unsubscribe dmaengine" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
