peggy9lac

&nbsp; &nbsp; [Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [PATCH] Speed up DVB TS stream delivery from <a href="http://www.chistraightenerbuy.com/ghd-hair-straightener-c-3.html"><strong>GHD IV Hot Pink Styler</strong></a> DMA buffer into dvb-core's buffer Subject: Re: [PATCH] Speed up DVB TS stream delivery from DMA buffer into dvb-core's buffer From: Andreas Oberritter &lt;obi@xxxxxxxxxxx&gt; Date: Sun, 10 Apr 2011 14:28:31 +0200 Cc: Mauro Carvalho Chehab &lt;mchehab@xxxxxxxxxx&gt;, Linux Media Mailing List &lt;linux-media@xxxxxxxxxxxxxxx&gt; In-reply-to: &lt;4D9F63E1.6060808@xxxxxxxxxxx&gt; User-agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.15) Gecko/20110406 Lightning/1.0b2 Thunderbird/3.1.9 On 04/08/2011 09:37 PM, Marko Ristola wrote: &gt; &gt; Here is some statistics without likely and with likely functions. &gt; &gt; It seems that using likely() gives better performance with Phenom I too. I'm not sure whether that's the right conclusion. See below. &gt; % Plain knl % No likely % With likely % &gt; dvb_ringbuffer_write 5,9 62,8 8,7 81,3 5,7 79,2 &gt; dvb_dmx_swfilter_packet 1,2 12,8 0,7 6,5 0,8 11,1 &gt; dvb_dmx_swfilter_204 2,3 24,5 1,3 12,1 0,7 9,7 &gt; &gt; &gt; Here &quot;Plain knl %&quot; is &quot;perf top -d 30&quot; percentage. &gt; 24,5 12,1 and 9,7 are percentages without a patch, with basic patch &gt; and last is with &quot;likely&quot; functions using patch. The varying ratio between dvb_ringbuffer_write and dvb_dmx_swfilter_packet, which haven't been modified by the patch, seems to indicate that the numbers were influenced <a href="http://www.chistraightenerbuy.com/ghd-hair-straightener-c-3.html"><strong>ghd hair straighteners</strong></a> by other activities on the system or by the data used during <a href="http://www.chistraightenerbuy.com/ghd-hair-straightener-c-3.html"><strong>cheap ghd straighteners</strong></a> the tests. 8,7% for dvb_ringbuffer_write in test #2 could also indicate a higher data rate than in #1 and #3. However, adding likely() certainly won't do any harm. Feel free to send an incremental patch. Regards, Andreas -- To unsubscribe from this list: send the line &quot;unsubscribe linux-media&quot; in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html References: [PATCH] Speed up DVB TS stream delivery from DMA buffer into dvb-core's buffer From: Marko Ristola Re: [PATCH] Speed up DVB TS stream delivery from DMA buffer into dvb-core's buffer From: Marko Ristola Prev by Date: Re: [PATCH] Speed up DVB TS stream delivery from DMA buffer into dvb-core's buffer Next by Date: Re: [PATCH 2.6.39] soc_camera: OMAP1: fix missing bytesperline and sizeimage initialization Previous by thread: Re: [PATCH] Speed up DVB TS stream delivery from DMA buffer into dvb-core's buffer Next by thread: Re: [PATCH] Speed up DVB TS stream delivery from DMA buffer into dvb-core's buffer Index(es): Date Thread [Linux&nbsp;Input] &nbsp; &nbsp; [Video&nbsp;for&nbsp;Linux] &nbsp; &nbsp; [Mplayer&nbsp;Users] &nbsp; &nbsp; [Linux&nbsp;USB&nbsp;Devel] &nbsp; &nbsp; [Linux&nbsp;Audio&nbsp;Users] &nbsp; &nbsp; [Photos] &nbsp; &nbsp; [Yosemite&nbsp;Photos] &nbsp; &nbsp; [Linux&nbsp;Kernel] &nbsp; &nbsp; [Linux&nbsp;SCSI] &nbsp; &nbsp; [XFree86] &nbsp; &nbsp; [Devices] &nbsp; &nbsp; [Yosemite&nbsp;Backpacking] &nbsp; &nbsp; [Linux&nbsp;Input] <a href="http://www.htzh.net/view.php?id=20590"><strong>ACER Aspire 5100 Laptop CPU Cooling Fan AB7505UX-EB3 | Super ...</strong></a>