Re: [PATCH v3 00/12] mmc: use nonblock mmc requests to minimize latency

2011-05-12 Thread Jaehoon Chung
Hi Philip.. I have something question..did you test SDMA and ADMA? Actually i didn't test [patch v3] but i tested patch v2 before Per sent. I got that SDMA is faster than ADMA. (using bounce buffer and sdhci) How do you think about that? I want to know your thought. And if you can share the resu

Re: [PATCH v3 00/12] mmc: use nonblock mmc requests to minimize latency

2011-05-09 Thread Philip Rakity
On May 9, 2011, at 5:34 AM, Per Forlin wrote: > On 9 May 2011 04:05, Philip Rakity wrote: >> >> Hi Per, >> >> We noticed on some of our systems if we ADMA or SDMA and a bounce buffer it >> is significantly faster then SDMA. >> > I have not done work with ADMA or SDMA. Where should I look to

Re: [PATCH v3 00/12] mmc: use nonblock mmc requests to minimize latency

2011-05-09 Thread Per Forlin
On 9 May 2011 04:05, Philip Rakity wrote: > > Hi Per, > > We noticed on some of our systems if we ADMA or SDMA and a bounce buffer it > is significantly faster then SDMA. > I have not done work with ADMA or SDMA. Where should I look to read more about it? Are these the right places. DMA iop-dma.c

Re: [PATCH v3 00/12] mmc: use nonblock mmc requests to minimize latency

2011-05-08 Thread Philip Rakity
Hi Per, We noticed on some of our systems if we ADMA or SDMA and a bounce buffer it is significantly faster then SDMA. I believe ADMA will do large transfers. Another data point. Philip On May 7, 2011, at 12:14 PM, Per Forlin wrote: > How significant is the cache maintenance over head? > It

[PATCH v3 00/12] mmc: use nonblock mmc requests to minimize latency

2011-05-07 Thread Per Forlin
How significant is the cache maintenance over head? It depends, the eMMC are much faster now compared to a few years ago and cache maintenance cost more due to multiple cache levels and speculative cache pre-fetch. In relation the cost for handling the caches have increased and is now a bottle neck