<?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 -assume contiguous_assumed_shape and contiguous keyword not working in Software Archive</title>
    <link>https://community.intel.com/t5/Software-Archive/assume-contiguous-assumed-shape-and-contiguous-keyword-not/m-p/1126506#M77657</link>
    <description>&lt;P&gt;Hi,&lt;/P&gt;

&lt;P&gt;I am trying to vectorize the following loop. The loop is defined inside a subroutine using shape arrays. I am using -assume contiguous_assumed_shape as well as contiguous keyword but the compiler still gives me 'strid is unknown to compiler' in the optimization report. Any suggestions are highly appreciated.&lt;/P&gt;

&lt;P&gt;Compiler Options: -qopenmp -O3 -align array64byte -assume contiguous_assumed_shape -xMIC-AVX512 -qopt-report=5 -g&lt;/P&gt;

&lt;P&gt;Fortran Compiler version:&amp;nbsp;ifort (IFORT) 18.0.0 20170811&lt;/P&gt;

&lt;PRE class="brush:fortran;"&gt;real(KIND=8), dimension(:,:), CONTIGUOUS   :: a, b, c
real(KIND=8), dimension(:,:), CONTIGUOUS :: rr


 integer :: nz, nvar, startindx, endindx

do j = startindx, endindx
      do k = 2, nz
         rr(k,j) = rr(k,j) - b(k,j)*rr(k-1,j)
      enddo
      rr(nz,j) = a(nz,j)*rr(nz,j)
enddo
&lt;/PRE&gt;

&lt;P&gt;Optimization Report:&lt;/P&gt;

&lt;P&gt;LOOP BEGIN at ../src/test.f90(22,7)&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15335: loop was not vectorized: vectorization possible but seems inefficient. Use vector always directive or -vec-threshold0 to override&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15329: vectorization support: non-unit strided store was emulated for the variable &amp;lt;rr(k,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(24,10) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;rr(k,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(24,20) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;b(k,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(24,30) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;rr(k-1,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(24,39) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15329: vectorization support: non-unit strided store was emulated for the variable &amp;lt;rr(nz,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(26,7) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;a(nz,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(26,18) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;rr(nz,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(26,28) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15305: vectorization support: vector length 2&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15309: vectorization support: normalized vectorization overhead 0.129&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15452: unmasked strided loads: 5&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15453: unmasked strided stores: 2&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15475: --- begin vector cost summary ---&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15476: scalar cost: 15&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15477: vector cost: 15.500&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15478: estimated potential speedup: 0.950&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15488: --- end vector cost summary ---&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&amp;nbsp; LOOP BEGIN at ../src/test.f90(23,7)&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp; remark #15344: loop was not vectorized: vector dependence prevents vectorization&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp; remark #15346: vector dependence: assumed FLOW dependence between rr(k,j) (24:10) and rr(k-1,j) (24:10)&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp; remark #25439: unrolled with remainder by 2 &amp;nbsp;&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; LOOP END&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&amp;nbsp; LOOP BEGIN at ../src/test.f90(23,7)&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; &amp;lt;Remainder&amp;gt;&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; LOOP END&lt;BR /&gt;
	LOOP END&lt;/P&gt;

&lt;P&gt;&lt;STRONG&gt;Najeeb Ahmad&lt;/STRONG&gt;&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
    <pubDate>Mon, 22 Jan 2018 15:05:13 GMT</pubDate>
    <dc:creator>Najeeb_A_</dc:creator>
    <dc:date>2018-01-22T15:05:13Z</dc:date>
    <item>
      <title>-assume contiguous_assumed_shape and contiguous keyword not working</title>
      <link>https://community.intel.com/t5/Software-Archive/assume-contiguous-assumed-shape-and-contiguous-keyword-not/m-p/1126506#M77657</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;

&lt;P&gt;I am trying to vectorize the following loop. The loop is defined inside a subroutine using shape arrays. I am using -assume contiguous_assumed_shape as well as contiguous keyword but the compiler still gives me 'strid is unknown to compiler' in the optimization report. Any suggestions are highly appreciated.&lt;/P&gt;

&lt;P&gt;Compiler Options: -qopenmp -O3 -align array64byte -assume contiguous_assumed_shape -xMIC-AVX512 -qopt-report=5 -g&lt;/P&gt;

&lt;P&gt;Fortran Compiler version:&amp;nbsp;ifort (IFORT) 18.0.0 20170811&lt;/P&gt;

&lt;PRE class="brush:fortran;"&gt;real(KIND=8), dimension(:,:), CONTIGUOUS   :: a, b, c
real(KIND=8), dimension(:,:), CONTIGUOUS :: rr


 integer :: nz, nvar, startindx, endindx

do j = startindx, endindx
      do k = 2, nz
         rr(k,j) = rr(k,j) - b(k,j)*rr(k-1,j)
      enddo
      rr(nz,j) = a(nz,j)*rr(nz,j)
enddo
&lt;/PRE&gt;

&lt;P&gt;Optimization Report:&lt;/P&gt;

&lt;P&gt;LOOP BEGIN at ../src/test.f90(22,7)&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15335: loop was not vectorized: vectorization possible but seems inefficient. Use vector always directive or -vec-threshold0 to override&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15329: vectorization support: non-unit strided store was emulated for the variable &amp;lt;rr(k,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(24,10) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;rr(k,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(24,20) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;b(k,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(24,30) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;rr(k-1,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(24,39) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15329: vectorization support: non-unit strided store was emulated for the variable &amp;lt;rr(nz,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(26,7) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;a(nz,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(26,18) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15328: vectorization support: non-unit strided load was emulated for the variable &amp;lt;rr(nz,j)&amp;gt;, stride is unknown to compiler&amp;nbsp;&amp;nbsp; [ ../src/test.f90(26,28) ]&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15305: vectorization support: vector length 2&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15309: vectorization support: normalized vectorization overhead 0.129&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15452: unmasked strided loads: 5&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15453: unmasked strided stores: 2&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15475: --- begin vector cost summary ---&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15476: scalar cost: 15&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15477: vector cost: 15.500&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15478: estimated potential speedup: 0.950&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; remark #15488: --- end vector cost summary ---&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&amp;nbsp; LOOP BEGIN at ../src/test.f90(23,7)&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp; remark #15344: loop was not vectorized: vector dependence prevents vectorization&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp; remark #15346: vector dependence: assumed FLOW dependence between rr(k,j) (24:10) and rr(k-1,j) (24:10)&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp; remark #25439: unrolled with remainder by 2 &amp;nbsp;&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; LOOP END&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&amp;nbsp; LOOP BEGIN at ../src/test.f90(23,7)&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; &amp;lt;Remainder&amp;gt;&lt;BR /&gt;
	&amp;nbsp;&amp;nbsp; LOOP END&lt;BR /&gt;
	LOOP END&lt;/P&gt;

&lt;P&gt;&lt;STRONG&gt;Najeeb Ahmad&lt;/STRONG&gt;&lt;/P&gt;

&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Mon, 22 Jan 2018 15:05:13 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Archive/assume-contiguous-assumed-shape-and-contiguous-keyword-not/m-p/1126506#M77657</guid>
      <dc:creator>Najeeb_A_</dc:creator>
      <dc:date>2018-01-22T15:05:13Z</dc:date>
    </item>
    <item>
      <title>Anybody home?</title>
      <link>https://community.intel.com/t5/Software-Archive/assume-contiguous-assumed-shape-and-contiguous-keyword-not/m-p/1126507#M77658</link>
      <description>&lt;P&gt;Anybody home?&lt;/P&gt;</description>
      <pubDate>Tue, 13 Feb 2018 17:15:41 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Archive/assume-contiguous-assumed-shape-and-contiguous-keyword-not/m-p/1126507#M77658</guid>
      <dc:creator>Najeeb_A_</dc:creator>
      <dc:date>2018-02-13T17:15:41Z</dc:date>
    </item>
  </channel>
</rss>

