<?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>topic DPAA RMan as a Kernel-space driver in T-Series</title>
    <link>https://community.nxp.com/t5/T-Series/DPAA-RMan-as-a-Kernel-space-driver/m-p/1160744#M3856</link>
    <description>&lt;P&gt;Hey,&amp;nbsp;&lt;/P&gt;&lt;P&gt;&lt;SPAN&gt;&amp;nbsp;I am working with Freescale t4241 processor and need to implement a RMan kernel-space driver to realize doorbell/portwrite/mailbox transactions as the USDPAA RMan does.&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;I found this documentation &lt;A href="https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fcommunity.nxp.com%2Fpwmxy87654%2Fattachments%2Fpwmxy87654%2Fp-series%2540tkb%2F75%2F2%2FDedicated%2520Channel%2520and%2520Pool%2520Channel%2520Used%2520in%2520Linux%2520Kernel%2520and%2520USDPAA.pdf&amp;amp;data=02%7C01%7C%7C18289f94e244493fa9fd08d85e7daf7b%7C612607c95af74e7f8976faf1ae77be60%7C0%7C0%7C637363241193008538&amp;amp;sdata=IrWS77lNeGBUSvp70Ibr0zzLxP5k8ujpXFj7tjbT4SI%3D&amp;amp;reserved=0" target="_blank"&gt;https://community.nxp.com/pwmxy87654/attachments/pwmxy87654/p-series%40tkb/75/2/Dedicated%20Channel%20and%20Pool%20Channel%20Used%20in%20Linux%20Kernel%20and%20USDPAA.pdf&lt;/A&gt;&lt;/P&gt;&lt;P&gt;Questions regarding it:&lt;/P&gt;&lt;OL&gt;&lt;LI&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;Is it a must to implement like dpaa_eth.c? I.e., init and allocate bman_buffer and FQs for each CPU? Or is it just a design choice?&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/LI&gt;&lt;LI&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;If it is just a design choice, would the implementation like USDPAA applications FRA be possible? I.e., init and allocate bman_buffer and FQs not for any particular CPU.&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/LI&gt;&lt;/OL&gt;&lt;P&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;Questions regarding sending doorbells, the implementation follows USDPAA RMan design. Receiving doorbell works properly, but failed to send doorbells.&amp;nbsp;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;Below are the kernel messages:&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.373723] RMan: get a bman buffer bpid(5) phy-addr(1654ba000),vitraddr(00000000d056b301)&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.376683] dbell sent: 0x6666 to 0&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.380180] sends to device(0) a msg using 10 dest(0x0) other_attr(0x0) fq(0x3500)&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.380183] opt-&amp;gt;tx_fqid 0x3500&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.380187] send: fqid 0x3500 &amp;lt;-- 0x1654ba000 (0)&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;FONT color="#FF0000"&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.380204] pt-rman ffe1e0000.rman: Gets rman outbound transaction error: 0x1&lt;/EM&gt;&lt;/FONT&gt;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;The RMan driver got an error interrupt.&lt;/P&gt;&lt;P&gt;The interrupt status read from the MMEDR registers is 0x1, i.e., OTE (Outbound transaction error).&lt;/P&gt;&lt;P&gt;The DPAARM describes OTE:&lt;/P&gt;&lt;UL&gt;&lt;LI&gt;“A read access to memory has caused a violation. MMECFDR0–3 contain the frame descriptor that caused the memory transaction error. Write 1 to clear.”&lt;/LI&gt;&lt;/UL&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;My questions are regarding this error are:&lt;/P&gt;&lt;OL&gt;&lt;LI&gt;Does it mean that the memory space (buffer) that the FQ is stored that cannot be accessed?&lt;/LI&gt;&lt;LI&gt;Is it because the buffer allocation for the FQ is not valid?&lt;/LI&gt;&lt;LI&gt;Or is it resulted from lacking any initialization or settings on the SRIO port?&lt;/LI&gt;&lt;LI&gt;Any way to diagnose this error? And get more debug information regarding this error?&lt;/LI&gt;&lt;/OL&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thanks in advance,&lt;/P&gt;&lt;P&gt;Shihua&lt;/P&gt;</description>
    <pubDate>Tue, 29 Sep 2020 13:44:11 GMT</pubDate>
    <dc:creator>shihua_huang</dc:creator>
    <dc:date>2020-09-29T13:44:11Z</dc:date>
    <item>
      <title>DPAA RMan as a Kernel-space driver</title>
      <link>https://community.nxp.com/t5/T-Series/DPAA-RMan-as-a-Kernel-space-driver/m-p/1160744#M3856</link>
      <description>&lt;P&gt;Hey,&amp;nbsp;&lt;/P&gt;&lt;P&gt;&lt;SPAN&gt;&amp;nbsp;I am working with Freescale t4241 processor and need to implement a RMan kernel-space driver to realize doorbell/portwrite/mailbox transactions as the USDPAA RMan does.&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;I found this documentation &lt;A href="https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fcommunity.nxp.com%2Fpwmxy87654%2Fattachments%2Fpwmxy87654%2Fp-series%2540tkb%2F75%2F2%2FDedicated%2520Channel%2520and%2520Pool%2520Channel%2520Used%2520in%2520Linux%2520Kernel%2520and%2520USDPAA.pdf&amp;amp;data=02%7C01%7C%7C18289f94e244493fa9fd08d85e7daf7b%7C612607c95af74e7f8976faf1ae77be60%7C0%7C0%7C637363241193008538&amp;amp;sdata=IrWS77lNeGBUSvp70Ibr0zzLxP5k8ujpXFj7tjbT4SI%3D&amp;amp;reserved=0" target="_blank"&gt;https://community.nxp.com/pwmxy87654/attachments/pwmxy87654/p-series%40tkb/75/2/Dedicated%20Channel%20and%20Pool%20Channel%20Used%20in%20Linux%20Kernel%20and%20USDPAA.pdf&lt;/A&gt;&lt;/P&gt;&lt;P&gt;Questions regarding it:&lt;/P&gt;&lt;OL&gt;&lt;LI&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;Is it a must to implement like dpaa_eth.c? I.e., init and allocate bman_buffer and FQs for each CPU? Or is it just a design choice?&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/LI&gt;&lt;LI&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;If it is just a design choice, would the implementation like USDPAA applications FRA be possible? I.e., init and allocate bman_buffer and FQs not for any particular CPU.&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/LI&gt;&lt;/OL&gt;&lt;P&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;&lt;SPAN&gt;Questions regarding sending doorbells, the implementation follows USDPAA RMan design. Receiving doorbell works properly, but failed to send doorbells.&amp;nbsp;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/SPAN&gt;Below are the kernel messages:&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.373723] RMan: get a bman buffer bpid(5) phy-addr(1654ba000),vitraddr(00000000d056b301)&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.376683] dbell sent: 0x6666 to 0&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.380180] sends to device(0) a msg using 10 dest(0x0) other_attr(0x0) fq(0x3500)&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.380183] opt-&amp;gt;tx_fqid 0x3500&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.380187] send: fqid 0x3500 &amp;lt;-- 0x1654ba000 (0)&lt;/EM&gt;&lt;/P&gt;&lt;P&gt;&lt;FONT color="#FF0000"&gt;&lt;EM&gt;[&amp;nbsp;&amp;nbsp; 27.380204] pt-rman ffe1e0000.rman: Gets rman outbound transaction error: 0x1&lt;/EM&gt;&lt;/FONT&gt;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;The RMan driver got an error interrupt.&lt;/P&gt;&lt;P&gt;The interrupt status read from the MMEDR registers is 0x1, i.e., OTE (Outbound transaction error).&lt;/P&gt;&lt;P&gt;The DPAARM describes OTE:&lt;/P&gt;&lt;UL&gt;&lt;LI&gt;“A read access to memory has caused a violation. MMECFDR0–3 contain the frame descriptor that caused the memory transaction error. Write 1 to clear.”&lt;/LI&gt;&lt;/UL&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;My questions are regarding this error are:&lt;/P&gt;&lt;OL&gt;&lt;LI&gt;Does it mean that the memory space (buffer) that the FQ is stored that cannot be accessed?&lt;/LI&gt;&lt;LI&gt;Is it because the buffer allocation for the FQ is not valid?&lt;/LI&gt;&lt;LI&gt;Or is it resulted from lacking any initialization or settings on the SRIO port?&lt;/LI&gt;&lt;LI&gt;Any way to diagnose this error? And get more debug information regarding this error?&lt;/LI&gt;&lt;/OL&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thanks in advance,&lt;/P&gt;&lt;P&gt;Shihua&lt;/P&gt;</description>
      <pubDate>Tue, 29 Sep 2020 13:44:11 GMT</pubDate>
      <guid>https://community.nxp.com/t5/T-Series/DPAA-RMan-as-a-Kernel-space-driver/m-p/1160744#M3856</guid>
      <dc:creator>shihua_huang</dc:creator>
      <dc:date>2020-09-29T13:44:11Z</dc:date>
    </item>
  </channel>
</rss>

