<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<html><head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
    <meta name="generator" content="Osso Notes">
    <title></title></head>
<body>
<p>Venkatraman S wrote:
<br>&gt; On Wed, Jan 6, 2010 at 6:46 PM, Adrian Hunter &lt;<a href="mailto:adrian.hunter@nokia.com">adrian.hunter@nokia.com</a>&gt;
<br>&gt; wrote:
<br>&gt; &gt; Venkatraman S wrote:
<br>&gt; &gt; &gt;
<br>&gt; &gt; &gt; On Tue, Dec 29, 2009 at 3:18 AM, Tony Lindgren &lt;<a href="mailto:tony@atomide.com">tony@atomide.com</a>&gt;
<br>&gt; wrote:
<br>&gt; &gt; &gt; &gt;
<br>&gt; &gt; &gt; &gt; * Venkatraman S &lt;<a href="mailto:svenkatr@ti.com">svenkatr@ti.com</a>&gt; [091211 07:01]:
<br>&gt; &gt; &gt; &gt; &gt;
<br>&gt; &gt; &gt; &gt; &gt; Here is the most updated version of the patch (thanks to Russell's
<br>&gt; &gt; &gt; &gt; &gt; review). This patch is applicable to OMAP4xxx as well as OMAP3630
<br>&gt; &gt; &gt; &gt; &gt; Reference to previous posts
<br>&gt; &gt; &gt; &gt; &gt; v1&nbsp; <a href="http://marc.info/?l=linux-omap&amp;m=125012097403050&amp;w=2">http://marc.info/?l=linux-omap&amp;m=125012097403050&amp;w=2</a>
<br>&gt; &gt; &gt; &gt; &gt; v2&nbsp; <a href="http://marc.info/?l=linux-omap&amp;m=125137152606644&amp;w=2">http://marc.info/?l=linux-omap&amp;m=125137152606644&amp;w=2</a>
<br>&gt; &gt; &gt; &gt; &gt; v3&nbsp; <a href="http://patchwork.kernel.org/patch/45408/">http://patchwork.kernel.org/patch/45408/</a>
<br>&gt; &gt; &gt; &gt;
<br>&gt; &gt; &gt; &gt; Do you have a patch for drivers/mmc/host/omap_hsmmc.c to
<br>&gt; &gt; &gt; &gt; use this feature? Or some other driver?
<br>&gt; &gt; &gt; &gt;
<br>&gt; &gt; &gt; &gt; Regards,
<br>&gt; &gt; &gt; &gt;
<br>&gt; &gt; &gt; &gt; Tony
<br>&gt; &gt; &gt;
<br>&gt; &gt; &gt; I am about to start working on omap_hsmmc to use the descriptor load
<br>&gt; &gt; &gt; feature. If the DMA changes are acceptable, I can post the driver
<br>&gt; &gt; &gt; patch as well.
<br>&gt; &gt;
<br>&gt; &gt; I presume this is about performance.&nbsp; How does it compare to chained
<br>&gt; DMA?
<br>&gt; &gt; We have a patch for omap_hsmmc for chained DMA that we are still
<br>&gt; testing.
<br>&gt; &gt;
<br>&gt;&nbsp; The main difference would be the number of logical channels used.
<br>&gt; With chaining, I assume you'd request for (or the API internally
<br>&gt; reserves) as many logical channels as there are segments. Here a
<br>&gt; single logical channel would do.
<br>&gt;
<br>
<br>How does the performance compare? Which is faster? Does descriptor autoloading reduce the number of interrupts? </p>
</body>
</html>