<?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 SSE intsruction _mm_insert_epi32 in Intel® ISA Extensions</title>
    <link>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780474#M261</link>
    <description>&lt;SPAN style="font-size: x-small;"&gt;&lt;P&gt;if a register has to be filled with data upto an arbitrary byte/word/dword position is it possible? for example: &lt;BR /&gt;&lt;BR /&gt;if cnt = 0 or 1 or 2 or 3. how can the following be done&lt;BR /&gt;&lt;BR /&gt;for (int i=0; i &amp;lt; cnt; i++)&lt;BR /&gt;{&lt;BR /&gt; Register = _mm_insert_epi32(Register, i);&lt;BR /&gt; Register1 = _mm_slli_si128(Register1, i);&lt;BR /&gt;}&lt;BR /&gt;&lt;BR /&gt;Both are being dissallowed since i is required to be a constant value. Is there a workaround? &lt;BR /&gt;suppose at run time only two values are to be filled (cnt = 2).&lt;BR /&gt;&lt;BR /&gt;Thanks &lt;BR /&gt;Deepak&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;/P&gt;&lt;/SPAN&gt;</description>
    <pubDate>Tue, 13 Jul 2010 11:51:07 GMT</pubDate>
    <dc:creator>inteleverywhere</dc:creator>
    <dc:date>2010-07-13T11:51:07Z</dc:date>
    <item>
      <title>SSE intsruction _mm_insert_epi32</title>
      <link>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780474#M261</link>
      <description>&lt;SPAN style="font-size: x-small;"&gt;&lt;P&gt;if a register has to be filled with data upto an arbitrary byte/word/dword position is it possible? for example: &lt;BR /&gt;&lt;BR /&gt;if cnt = 0 or 1 or 2 or 3. how can the following be done&lt;BR /&gt;&lt;BR /&gt;for (int i=0; i &amp;lt; cnt; i++)&lt;BR /&gt;{&lt;BR /&gt; Register = _mm_insert_epi32(Register, i);&lt;BR /&gt; Register1 = _mm_slli_si128(Register1, i);&lt;BR /&gt;}&lt;BR /&gt;&lt;BR /&gt;Both are being dissallowed since i is required to be a constant value. Is there a workaround? &lt;BR /&gt;suppose at run time only two values are to be filled (cnt = 2).&lt;BR /&gt;&lt;BR /&gt;Thanks &lt;BR /&gt;Deepak&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;/P&gt;&lt;/SPAN&gt;</description>
      <pubDate>Tue, 13 Jul 2010 11:51:07 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780474#M261</guid>
      <dc:creator>inteleverywhere</dc:creator>
      <dc:date>2010-07-13T11:51:07Z</dc:date>
    </item>
    <item>
      <title>SSE intsruction _mm_insert_epi32</title>
      <link>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780475#M262</link>
      <description>Use switch() operator</description>
      <pubDate>Tue, 13 Jul 2010 12:18:37 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780475#M262</guid>
      <dc:creator>Arthur_Moroz</dc:creator>
      <dc:date>2010-07-13T12:18:37Z</dc:date>
    </item>
    <item>
      <title>SSE intsruction _mm_insert_epi32</title>
      <link>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780476#M263</link>
      <description>Would that notnegate the optimization? Also is there a way to do a bit shift (not byet shift) on an m128i bit value.&lt;BR /&gt;&lt;BR /&gt;Reg&lt;BR /&gt;Deepak&lt;BR /&gt;</description>
      <pubDate>Wed, 14 Jul 2010 06:50:01 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780476#M263</guid>
      <dc:creator>inteleverywhere</dc:creator>
      <dc:date>2010-07-14T06:50:01Z</dc:date>
    </item>
    <item>
      <title>SSE intsruction _mm_insert_epi32</title>
      <link>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780477#M264</link>
      <description>My dear friend,&lt;BR /&gt;&lt;BR /&gt;1) It is compiler who does optimisation, not you. For compiler switch() is as good optimizable as any other operator.&lt;BR /&gt;&lt;BR /&gt;2) there's no way to bit shift whole register&lt;BR /&gt;</description>
      <pubDate>Wed, 14 Jul 2010 07:16:11 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780477#M264</guid>
      <dc:creator>Arthur_Moroz</dc:creator>
      <dc:date>2010-07-14T07:16:11Z</dc:date>
    </item>
    <item>
      <title>SSE intsruction _mm_insert_epi32</title>
      <link>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780478#M265</link>
      <description>Thank yousir.</description>
      <pubDate>Wed, 14 Jul 2010 12:24:38 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-ISA-Extensions/SSE-intsruction-mm-insert-epi32/m-p/780478#M265</guid>
      <dc:creator>inteleverywhere</dc:creator>
      <dc:date>2010-07-14T12:24:38Z</dc:date>
    </item>
  </channel>
</rss>

