<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>LayerscapeのトピックRe: Situation for LS1021A High-Resolution Timers?</title>
    <link>https://community.nxp.com/t5/Layerscape/Situation-for-LS1021A-High-Resolution-Timers/m-p/606655#M1712</link>
    <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;For anyone else interested, turns out the trick is to set the "always-on" property in the device tree node for the CPU's (ARM architecture) timer.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Doing so gives a guarantee to the kernel that the timer will not stop due to power management; I still have to see how well that holds for my platform, YMMV.&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
    <pubDate>Wed, 28 Sep 2016 07:06:31 GMT</pubDate>
    <dc:creator>stefannickl</dc:creator>
    <dc:date>2016-09-28T07:06:31Z</dc:date>
    <item>
      <title>Situation for LS1021A High-Resolution Timers?</title>
      <link>https://community.nxp.com/t5/Layerscape/Situation-for-LS1021A-High-Resolution-Timers/m-p/606652#M1709</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;I'm working with a standard Linux 4.4.19-rt27 kernel on the TQ Systems platform for the LS1021A, and&amp;nbsp; following &lt;A href="https://community.nxp.com/external-link.jspa?url=http%3A%2F%2Felinux.org%2FHigh_Resolution_Timers" rel="nofollow" target="_blank"&gt;http://elinux.org/High_Resolution_Timers&lt;/A&gt;&lt;A&gt;,&lt;/A&gt; I'm not getting high-resolution timers despite having them activated in the kernel config.&lt;/P&gt;&lt;P style="min-height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;I've tried all preemption levels that can be configured from PREEMPT_RT to LAZY, each time cyclictest warns about unavailable high resolution timers and gives results that correspond to the jiffy period.&lt;/P&gt;&lt;P style="min-height: 8pt; padding: 0px;"&gt;&lt;/P&gt;&lt;P&gt;Any advice how this could be solved?&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;I've seen replies to older posts referring to SDK kernels, but according to the FAE the platform work by Freescale/NXP is now pushed to mainline instead (very agreeable move in general), and the BSP by TQ Systems is also based on the standard kernel.&lt;/P&gt;&lt;P style="min-height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Best regards,&lt;/P&gt;&lt;P&gt;Stefan&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Fri, 09 Sep 2016 20:13:25 GMT</pubDate>
      <guid>https://community.nxp.com/t5/Layerscape/Situation-for-LS1021A-High-Resolution-Timers/m-p/606652#M1709</guid>
      <dc:creator>stefannickl</dc:creator>
      <dc:date>2016-09-09T20:13:25Z</dc:date>
    </item>
    <item>
      <title>Re: Situation for LS1021A High-Resolution Timers?</title>
      <link>https://community.nxp.com/t5/Layerscape/Situation-for-LS1021A-High-Resolution-Timers/m-p/606653#M1710</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;&lt;SPAN style="font-size: 10.0pt;"&gt;NXP offers SDK Linux BSP 2.0 for TWR-LS1021a board. This SDK supports high resolution timer. See attachment.&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN style="font-size: 10.0pt;"&gt;Use this SDK if high resolution timer is needed.&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;BR /&gt;Have a great day,&lt;BR /&gt;Pavel Chubakov&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;-----------------------------------------------------------------------------------------------------------------------&lt;BR /&gt;Note: If this post answers your question, please click the Correct Answer button. Thank you!&lt;BR /&gt;-----------------------------------------------------------------------------------------------------------------------&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Mon, 12 Sep 2016 03:27:45 GMT</pubDate>
      <guid>https://community.nxp.com/t5/Layerscape/Situation-for-LS1021A-High-Resolution-Timers/m-p/606653#M1710</guid>
      <dc:creator>Pavel</dc:creator>
      <dc:date>2016-09-12T03:27:45Z</dc:date>
    </item>
    <item>
      <title>Re: Situation for LS1021A High-Resolution Timers?</title>
      <link>https://community.nxp.com/t5/Layerscape/Situation-for-LS1021A-High-Resolution-Timers/m-p/606654#M1711</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;Hello Pavel,&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;thanks for your reply.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;The screenshots you've provided show the kernel compile-time configuration option for high resolution timers, which I have enabled as well.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;However &lt;A href="https://community.nxp.com/external-link.jspa?url=http%3A%2F%2Felinux.org%2FHigh_Resolution_Timers" rel="nofollow noopener noreferrer" target="_blank"&gt;http://elinux.org/High_Resolution_Timers&lt;/A&gt; describes a number of tests for whether high resolution timers are actually available at run time, and those come out negative despite:&lt;/P&gt;&lt;PRE&gt;# zcat /proc/config.gz |egrep "RT_FULL|HIGH_RES" CONFIG_HIGH_RES_TIMERS=y CONFIG_PREEMPT_RT_FULL=y&amp;nbsp; # grep resolution /proc/timer_list&amp;nbsp;&amp;nbsp; .resolution: 10000000 nsecs&amp;nbsp;&amp;nbsp; .resolution: 10000000 nsecs&amp;nbsp;&amp;nbsp; .resolution: 10000000 nsecs [...]&amp;nbsp; # grep event_handler /proc/timer_list&amp;nbsp; event_handler:&amp;nbsp; tick_handle_periodic&amp;nbsp; event_handler:&amp;nbsp; tick_handle_periodic&amp;nbsp; # cyclictest -n -p 80 -i 500 -l 5000 # /dev/cpu_dma_latency set to 0us WARN: High resolution timers not available policy: fifo: loadavg: 0.00 0.00 0.00 2/105 335 T: 0 (&amp;nbsp; 335) P:80 I:500 C:&amp;nbsp;&amp;nbsp; 5000 Min:&amp;nbsp;&amp;nbsp; 1758 Act: 9748 Avg: 9743 Max: 9767&lt;/PRE&gt;&lt;P&gt;This was for HZ=100, latency goes down to 1ms for HZ=1000.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Can you confirm that the SDK kernel actually provides high-res timers at runtime?&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Are there any plans to update the SDK kernel to 4.x over the next year or so?&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Best regards,&lt;/P&gt;&lt;P&gt;Stefan&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Mon, 12 Sep 2016 07:00:03 GMT</pubDate>
      <guid>https://community.nxp.com/t5/Layerscape/Situation-for-LS1021A-High-Resolution-Timers/m-p/606654#M1711</guid>
      <dc:creator>stefannickl</dc:creator>
      <dc:date>2016-09-12T07:00:03Z</dc:date>
    </item>
    <item>
      <title>Re: Situation for LS1021A High-Resolution Timers?</title>
      <link>https://community.nxp.com/t5/Layerscape/Situation-for-LS1021A-High-Resolution-Timers/m-p/606655#M1712</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;For anyone else interested, turns out the trick is to set the "always-on" property in the device tree node for the CPU's (ARM architecture) timer.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Doing so gives a guarantee to the kernel that the timer will not stop due to power management; I still have to see how well that holds for my platform, YMMV.&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Wed, 28 Sep 2016 07:06:31 GMT</pubDate>
      <guid>https://community.nxp.com/t5/Layerscape/Situation-for-LS1021A-High-Resolution-Timers/m-p/606655#M1712</guid>
      <dc:creator>stefannickl</dc:creator>
      <dc:date>2016-09-28T07:06:31Z</dc:date>
    </item>
  </channel>
</rss>

