<?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 Hi Eloi, in Intel® MPI Library</title>
    <link>https://community.intel.com/t5/Intel-MPI-Library/Cannot-use-jemalloc-with-IntelMPI/m-p/1039992#M4289</link>
    <description>&lt;P&gt;Hi Eloi,&lt;/P&gt;

&lt;P&gt;Thanks for getting in touch.&amp;nbsp; Yes, Intel MPI does implement its own malloc although it's mostly a wrapper around the OS's malloc call (with some very slight modifications).&lt;/P&gt;

&lt;P&gt;If you're looking to replace the memory management subsystem used within Intel MPI, that's easy enough to do.&amp;nbsp; The &lt;A href="https://software.intel.com/en-us/node/528845"&gt;Unified Memory Management&lt;/A&gt; chapter in our Reference Manual gives you some examples.&amp;nbsp; Give that a try and let me know how it goes.&lt;/P&gt;

&lt;P&gt;Regards,&lt;BR /&gt;
	~Gergana&lt;/P&gt;</description>
    <pubDate>Mon, 09 Mar 2015 21:33:14 GMT</pubDate>
    <dc:creator>Gergana_S_Intel</dc:creator>
    <dc:date>2015-03-09T21:33:14Z</dc:date>
    <item>
      <title>Cannot use jemalloc with IntelMPI</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Cannot-use-jemalloc-with-IntelMPI/m-p/1039990#M4287</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;

&lt;P&gt;I've tried to bench several memory allocators on Linux (64-bit) such as ptmalloc2, tcmalloc and jemalloc with an application linked against IntelMPI (4.1.3.049).&lt;/P&gt;

&lt;P&gt;Launching any application linked with jemalloc will cause the execution to abort with a signal 11. But the same application, when not linked with IntelMPI will work without any issue.&lt;/P&gt;

&lt;P&gt;Is IntelMPI doing its own malloc/free ?&lt;BR /&gt;
	How can this issue be overcome ?&lt;/P&gt;

&lt;P&gt;&lt;SPAN style="font-size: 1em; line-height: 1.5;"&gt;Thanks,&lt;/SPAN&gt;&lt;BR /&gt;
	&lt;SPAN style="font-size: 1em; line-height: 1.5;"&gt;Eloi&lt;/SPAN&gt;&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Sat, 07 Mar 2015 12:02:11 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Cannot-use-jemalloc-with-IntelMPI/m-p/1039990#M4287</guid>
      <dc:creator>Eloi_Gaudry</dc:creator>
      <dc:date>2015-03-07T12:02:11Z</dc:date>
    </item>
    <item>
      <title>Here is the backtrace I got</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Cannot-use-jemalloc-with-IntelMPI/m-p/1039991#M4288</link>
      <description>&lt;P&gt;Here is the backtrace I got within gdb:&lt;/P&gt;

&lt;PRE class="brush:bash;"&gt;Program received signal SIGSEGV, Segmentation fault.
0x00007ffff0c8241c in free (ptr=&amp;lt;error reading variable: Cannot access memory at address 0x7fffff7feff0&amp;gt;) at ../../i_rtc_hook.c:45
45	../../i_rtc_hook.c: No such file or directory.
(gdb) where
#0  0x00007ffff0c8241c in free (ptr=&amp;lt;error reading variable: Cannot access memory at address 0x7fffff7feff0&amp;gt;) at ../../i_rtc_hook.c:45
#1  0x00007ffff0c82480 in free (ptr=0x7fffe7839080) at ../../i_rtc_hook.c:65
#2  0x00007ffff0c82480 in free (ptr=0x7fffe7839080) at ../../i_rtc_hook.c:65
#3  0x00007ffff0c82480 in free (ptr=0x7fffe7839080) at ../../i_rtc_hook.c:65
#4  0x00007ffff0c82480 in free (ptr=0x7fffe7839080) at ../../i_rtc_hook.c:65
#5  0x00007ffff0c82480 in free (ptr=0x7fffe7839080) at ../../i_rtc_hook.c:65
&lt;/PRE&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Mon, 09 Mar 2015 10:11:55 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Cannot-use-jemalloc-with-IntelMPI/m-p/1039991#M4288</guid>
      <dc:creator>Eloi_Gaudry</dc:creator>
      <dc:date>2015-03-09T10:11:55Z</dc:date>
    </item>
    <item>
      <title>Hi Eloi,</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Cannot-use-jemalloc-with-IntelMPI/m-p/1039992#M4289</link>
      <description>&lt;P&gt;Hi Eloi,&lt;/P&gt;

&lt;P&gt;Thanks for getting in touch.&amp;nbsp; Yes, Intel MPI does implement its own malloc although it's mostly a wrapper around the OS's malloc call (with some very slight modifications).&lt;/P&gt;

&lt;P&gt;If you're looking to replace the memory management subsystem used within Intel MPI, that's easy enough to do.&amp;nbsp; The &lt;A href="https://software.intel.com/en-us/node/528845"&gt;Unified Memory Management&lt;/A&gt; chapter in our Reference Manual gives you some examples.&amp;nbsp; Give that a try and let me know how it goes.&lt;/P&gt;

&lt;P&gt;Regards,&lt;BR /&gt;
	~Gergana&lt;/P&gt;</description>
      <pubDate>Mon, 09 Mar 2015 21:33:14 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Cannot-use-jemalloc-with-IntelMPI/m-p/1039992#M4289</guid>
      <dc:creator>Gergana_S_Intel</dc:creator>
      <dc:date>2015-03-09T21:33:14Z</dc:date>
    </item>
    <item>
      <title>For those of you keeping</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Cannot-use-jemalloc-with-IntelMPI/m-p/1039993#M4290</link>
      <description>&lt;P&gt;For those of you keeping track at home, the real issue was in the linking order when using a different memory allocator.&amp;nbsp; Intel MPI already uses the malloc, relloc, free routines.&amp;nbsp; If you're simply trying to replace the malloc library used (as a "drop-in" solution) from default glibc to something else (e.g. jemalloc), make sure that's linked *before* linking with libmpi for your application.&lt;/P&gt;

&lt;P&gt;~Gergana&lt;/P&gt;</description>
      <pubDate>Tue, 10 Mar 2015 21:01:37 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Cannot-use-jemalloc-with-IntelMPI/m-p/1039993#M4290</guid>
      <dc:creator>Gergana_S_Intel</dc:creator>
      <dc:date>2015-03-10T21:01:37Z</dc:date>
    </item>
  </channel>
</rss>

