<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:clearspace="http://www.jivesoftware.com/xmlns/clearspace/rss" xmlns:dc="http://purl.org/dc/elements/1.1/" version="2.0">
  <channel>
    <title>Intel Communities: Message List</title>
    <link>http://communities.intel.com/index.jspa?view=discussions</link>
    <description>Most recent forum messages</description>
    <language>en</language>
    <pubDate>Fri, 19 Apr 2013 12:15:06 GMT</pubDate>
    <generator>Jive SBS 5.0.2.0  (http://jivesoftware.com/products/clearspace/)</generator>
    <dc:date>2013-04-19T12:15:06Z</dc:date>
    <dc:language>en</dc:language>
    <item>
      <title>Re: Memory consistency issue using RCCE</title>
      <link>http://communities.intel.com/message/190939?tstart=0#190939</link>
      <description>&lt;!-- [DocumentBodyStart:ade542ea-9bf3-4d9c-bea7-cbacd507600c] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;Hi Nicolas,&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;WCB = Write combine buffer.&lt;/p&gt;&lt;p&gt;If you looking to RCCE_put.c code, you will find the RCCE_put() not using &lt;span style="font-family: intel-neo-sans-1, intel-neo-sans-2, tahoma, helvetica, sans-serif; background-color: #ffffff;"&gt;RCCE_fool_write_combine_buffer&lt;/span&gt; to flush the data, while RCCE_put_char() (which used by &lt;span style="font-family: intel-neo-sans-1, intel-neo-sans-2, tahoma, helvetica, sans-serif; background-color: #ffffff;"&gt;RCCE_flag_write()&lt;/span&gt;) flush the data by using &lt;span style="font-family: intel-neo-sans-1, intel-neo-sans-2, tahoma, helvetica, sans-serif; background-color: #ffffff;"&gt;RCCE_fool_write_combine_buffer&lt;/span&gt; to make sure the data already written in MPB.&lt;/p&gt;&lt;p&gt;Other way, I think to implement those function based on no-cache mode. &lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Regards&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Hayder&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:ade542ea-9bf3-4d9c-bea7-cbacd507600c] --&gt;</description>
      <pubDate>Fri, 19 Apr 2013 12:15:06 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/190939?tstart=0#190939</guid>
      <dc:date>2013-04-19T12:15:06Z</dc:date>
      <clearspace:dateToText>1 month, 1 day ago</clearspace:dateToText>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
    <item>
      <title>Re: Memory consistency issue using RCCE</title>
      <link>http://communities.intel.com/message/190931?tstart=0#190931</link>
      <description>&lt;!-- [DocumentBodyStart:4a64ad0f-b66c-4c73-a421-f044a22c7c4a] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;Hi, &lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;I have already got like this issue before in my work. I got impression, the problem is the conflicts between &lt;/p&gt;&lt;p&gt;RCCE_flag_write() and RCCE_put(), because there is no flush for WCB in RCCE_put() to ensure the data gets flushed.&lt;/p&gt;&lt;p&gt;My suggestion is to add this:&lt;/p&gt;&lt;p&gt;*(int *)RCCE_fool_write_combine_buffer = 1;&lt;/p&gt;&lt;p&gt;after each RCCU_put() to make sure the WCB gets flushed, or implement your application based on non-cache mode.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;I hope this help.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Hayder&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:4a64ad0f-b66c-4c73-a421-f044a22c7c4a] --&gt;</description>
      <pubDate>Fri, 19 Apr 2013 08:27:53 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/190931?tstart=0#190931</guid>
      <dc:date>2013-04-19T08:27:53Z</dc:date>
      <clearspace:dateToText>1 month, 1 day ago</clearspace:dateToText>
      <clearspace:replyCount>2</clearspace:replyCount>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
    <item>
      <title>Write Combine Buffer status</title>
      <link>http://communities.intel.com/message/184255?tstart=0#184255</link>
      <description>&lt;!-- [DocumentBodyStart:7dfcf1d4-8926-4099-88ce-df315bd18d20] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;Hi, &lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Is there a way to know the write combine buffer when is full or flushing?&lt;/p&gt;&lt;p&gt;Or, does hardware support register to know status of WCB? &lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Thanks in Advance&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:7dfcf1d4-8926-4099-88ce-df315bd18d20] --&gt;</description>
      <pubDate>Mon, 25 Feb 2013 15:24:11 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/184255?tstart=0#184255</guid>
      <dc:date>2013-02-25T15:24:11Z</dc:date>
      <clearspace:dateToText>2 months, 3 weeks ago</clearspace:dateToText>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
    <item>
      <title>Re: Linked data structures in shared off-chip memory</title>
      <link>http://communities.intel.com/message/168531?tstart=0#168531</link>
      <description>&lt;!-- [DocumentBodyStart:9abdd261-74bc-44fc-9eef-d7bdc3e669e9] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;First of all, Thanks for this nice post.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;&amp;gt;&amp;gt;Allocate a "physical remapping" area (i.e., a set of LUT slots) per core. Map in this range, on all cores, at &lt;span style="font-style: inherit; font-size: 13px; font-family: inherit;"&gt;&lt;strong&gt;the same&lt;/strong&gt;&lt;/span&gt; virtual address. Configure the LUTs to point to the same system memory. Now, you can &lt;span style="font-style: inherit; font-size: 13px; font-family: inherit;"&gt;&lt;strong&gt;exchange virtual addresses freely across cores&lt;/strong&gt;&lt;/span&gt; (as long as they refer to objects in this special shared section).&amp;lt;&amp;lt;&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;I have thought in advance about this idea, but I got impression that's a poor and rather restrictive solution.&lt;/p&gt;&lt;p&gt;Could you give us more details about it? if possible with examples.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Regards&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Hayder&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:9abdd261-74bc-44fc-9eef-d7bdc3e669e9] --&gt;</description>
      <pubDate>Thu, 04 Oct 2012 09:59:15 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/168531?tstart=0#168531</guid>
      <dc:date>2012-10-04T09:59:15Z</dc:date>
      <clearspace:dateToText>7 months, 2 weeks ago</clearspace:dateToText>
      <clearspace:replyCount>1</clearspace:replyCount>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
    <item>
      <title>Re: Linked data structures in shared off-chip memory</title>
      <link>http://communities.intel.com/message/167657?tstart=0#167657</link>
      <description>&lt;!-- [DocumentBodyStart:98252ff8-ebac-4d09-84a8-fe9c1837256c] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;You can't using pointer approach, because the virtual address is not valid for other cores. The SCC is like cluster on chip and it is not follow real shared memory model. That means map shared memory are into every process on the system at different virtual address. So, you need using techniques of cluster programming model to passing data from core to other.&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:98252ff8-ebac-4d09-84a8-fe9c1837256c] --&gt;</description>
      <pubDate>Sat, 22 Sep 2012 22:09:14 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/167657?tstart=0#167657</guid>
      <dc:date>2012-09-22T22:09:14Z</dc:date>
      <clearspace:dateToText>7 months, 4 weeks ago</clearspace:dateToText>
      <clearspace:replyCount>8</clearspace:replyCount>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
    <item>
      <title>Re: A non-cacheable block of memory</title>
      <link>http://communities.intel.com/message/167557?tstart=0#167557</link>
      <description>&lt;!-- [DocumentBodyStart:2282fae3-d5e6-4eeb-a7c4-3ee34ca62826] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;I have already installed the code and really nice tool. I use sccKit 1.4.2.2 and linux kernel 2.6.38.3.&lt;/p&gt;&lt;p&gt;If you want to install, you should use buildroot-2011.05 and following the instruction SMC paper. Please, don't forget to change the address of sys_call_table.&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:2282fae3-d5e6-4eeb-a7c4-3ee34ca62826] --&gt;</description>
      <pubDate>Fri, 21 Sep 2012 15:27:44 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/167557?tstart=0#167557</guid>
      <dc:date>2012-09-21T15:27:44Z</dc:date>
      <clearspace:dateToText>8 months, 1 day ago</clearspace:dateToText>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
    <item>
      <title>Re: A non-cacheable block of memory</title>
      <link>http://communities.intel.com/message/167554?tstart=0#167554</link>
      <description>&lt;!-- [DocumentBodyStart:9eaa7625-70b8-4ae6-a430-81d5a37106d1] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;&lt;span class="GingerNoCheckStart"&gt;&lt;/span&gt;Thanks for your response.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Other way, we could use mprotect mechanism which support by SMC group to set the block of memory as Cacheable or UC.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&lt;span class="GingerNoCheckEnd"&gt;&lt;/span&gt;&amp;nbsp;&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:9eaa7625-70b8-4ae6-a430-81d5a37106d1] --&gt;</description>
      <pubDate>Fri, 21 Sep 2012 13:12:54 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/167554?tstart=0#167554</guid>
      <dc:date>2012-09-21T13:12:54Z</dc:date>
      <clearspace:dateToText>8 months, 1 day ago</clearspace:dateToText>
      <clearspace:replyCount>2</clearspace:replyCount>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
    <item>
      <title>Re: Cross-Compiler Update</title>
      <link>http://communities.intel.com/message/166002?tstart=0#166002</link>
      <description>&lt;!-- [DocumentBodyStart:06e10b06-2bcf-41fb-b635-66a59722e5b7] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;Hi Edward,&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;I don't know exactly why you got at this error, but I think there is mistake when you use this compiler.&lt;/p&gt;&lt;p&gt;If you could posting the procedure of compile your code and the message error.&lt;/p&gt;&lt;p&gt;Any way, I attached picture that contained the procedure of test the compiler and you could follow it and check your code.&lt;/p&gt;&lt;p&gt;1. Add your new compiler to your Path by using:&lt;/p&gt;&lt;p&gt;export PATH=$PATH:/(directory of cross compiler)/build/install/i586-unknown-linux-gnu/bin&lt;/p&gt;&lt;p&gt;2. To be sure the compiler in your Path:&lt;/p&gt;&lt;p&gt;which i586-unknown-linux-gnu-gcc&lt;/p&gt;&lt;p&gt;3. to compile your code use i586-unknown-linux-gnu-gcc or i586-unknown-linux-gnu-g++&lt;/p&gt;&lt;p&gt;4. You could also find in attached file the output of run hello world program.&lt;/p&gt;&lt;p&gt;BTY, the program run in MCPC and SCC.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Regards!&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Hayder&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:06e10b06-2bcf-41fb-b635-66a59722e5b7] --&gt;</description>
      <pubDate>Tue, 04 Sep 2012 06:08:21 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/166002?tstart=0#166002</guid>
      <dc:date>2012-09-04T06:08:21Z</dc:date>
      <clearspace:dateToText>8 months, 2 weeks ago</clearspace:dateToText>
      <clearspace:replyCount>1</clearspace:replyCount>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
    <item>
      <title>Re: Cross-Compiler Update</title>
      <link>http://communities.intel.com/message/165919?tstart=0#165919</link>
      <description>&lt;!-- [DocumentBodyStart:7bca34ea-abe1-4181-a01c-e1d1dea8a023] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;Hi Edward,&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;First of all, you need to install Crosstool-ng, and you could download from this website:&lt;/p&gt;&lt;p&gt;&lt;a class="jive-link-external-small" href="http://crosstool-ng.org/" target="_blank"&gt;http://crosstool-ng.org/&lt;/a&gt;&lt;span&gt;.&lt;/span&gt;&lt;/p&gt;&lt;p&gt;You could tracking those instructions which found in this tutorials to understand how to build cross-compiler:&lt;/p&gt;&lt;p&gt;&lt;a class="jive-link-external-small" href="http://www.aclevername.com/articles/linux-xilinx-tutorial/crosstool-ng-1.3.2.html" target="_blank"&gt;http://www.aclevername.com/articles/linux-xilinx-tutorial/crosstool-ng-1.3.2.html&lt;/a&gt;&lt;/p&gt;&lt;p&gt;&lt;a class="jive-link-external-small" href="http://getglitched.com/?page_id=253" target="_blank"&gt;http://getglitched.com/?page_id=253&lt;/a&gt;&lt;/p&gt;&lt;p&gt;Other way, I attached the configuration file (I got it from Isa&amp;iacute;as and modified it) for crosstool-ng 1.11.3, that will build a 4.4.6 toolchain with C/C++ and fortran. It works and you could just download it and copy the contents of file and place it in the .config file under your crosstool-ng working directory and instruct crosstool-ng to build the compiler by using ./ct-ng build.&lt;/p&gt;&lt;p&gt;BTY, the .config file is already existing in crosstool directory, just you need modified it.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Do not hesitate to ask more questions.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Regards!&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Hayder&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:7bca34ea-abe1-4181-a01c-e1d1dea8a023] --&gt;</description>
      <pubDate>Mon, 03 Sep 2012 06:09:54 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/165919?tstart=0#165919</guid>
      <dc:date>2012-09-03T06:09:54Z</dc:date>
      <clearspace:dateToText>8 months, 2 weeks ago</clearspace:dateToText>
      <clearspace:replyCount>3</clearspace:replyCount>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
    <item>
      <title>Re: Cross-Compiler Update</title>
      <link>http://communities.intel.com/message/162942?tstart=0#162942</link>
      <description>&lt;!-- [DocumentBodyStart:8a02ad21-5ff4-40f5-893a-b5150f2daeb0] --&gt;&lt;div class="jive-rendered-content"&gt;&lt;p&gt;Hi Edward,&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;You have two options to use new version of gcc on SCC:&lt;/p&gt;&lt;p&gt;1) You should be able to build a custom linux that has new version of gcc, or using sccKit 1.4.2.2 which uses buildroot 2011.11 and comes with kernel 3.1.4 that its default cross-gcc seems to be gcc-4.4.6.&lt;/p&gt;&lt;p&gt;2) You could generate a new version of gcc by using crosstool-ng.&lt;/p&gt;&lt;p style="min-height: 8pt; height: 8pt; padding: 0px;"&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Hopefully, this reply help you.&lt;/p&gt;&lt;/div&gt;&lt;!-- [DocumentBodyEnd:8a02ad21-5ff4-40f5-893a-b5150f2daeb0] --&gt;</description>
      <pubDate>Mon, 30 Jul 2012 10:03:30 GMT</pubDate>
      <author>webadmin@intel.com</author>
      <guid>http://communities.intel.com/message/162942?tstart=0#162942</guid>
      <dc:date>2012-07-30T10:03:30Z</dc:date>
      <clearspace:dateToText>9 months, 3 weeks ago</clearspace:dateToText>
      <clearspace:replyCount>5</clearspace:replyCount>
      <clearspace:objectType>0</clearspace:objectType>
    </item>
  </channel>
</rss>

