<div dir="ltr">I am able to vectorize it with the following code;<div><br></div><div><div>#include <stdio.h></div><div>#define N 100351</div><div> </div><div>// This function computes 2D-5 point Jacobi stencil </div><div>void stencil(int a[][N], int b[][N])</div><div>{</div><div> int i, j, k;</div><div> for (k = 0; k < N; k++) {</div><div> for (i = 1; i <= N-2; i++)</div><div> for (j = 1; j <= N-2; j++)</div><div> <span style="white-space:pre"> </span> b[i][j] = 0.25 * (a[i][j] + a[i-1][j] + a[i+1][j] + a[i][j-1] + a[i][j+1]);</div><div><br></div><div>for (i = 1; i <= N-2; i++)</div><div><span style="white-space:pre"> </span>for (j = 1; j <= N-2; j++)</div><div><span style="white-space:pre"> </span> a[i][j] = b[i][j];</div><div> </div><div>}</div><div>}</div></div><div><br></div><div>I removed restrict over here.</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Sun, Jul 2, 2017 at 3:11 AM, hameeza ahmed <span dir="ltr"><<a href="mailto:hahmed2305@gmail.com" target="_blank">hahmed2305@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">further i modified the code to the following;<div><br></div><div><span class=""><div>#include <stdio.h></div><div>#define N 100351</div><div> </div><div>// This function computes 2D-5 point Jacobi stencil </div></span><div>void stencil(int a[restrict][N], int b[restrict][N])</div><div>{</div><div> int i, j, k;</div><div> for (k = 0; k < N; k++) {</div><span class=""><div> for (i = 1; i <= N-2; i++)</div><div> for (j = 1; j <= N-2; j++)</div></span><div> <span style="white-space:pre-wrap"> </span> b[i][j] = 0.25 * (a[i][j] + a[i-1][j] + a[i+1][j] + a[i][j-1] + a[i][j+1]);</div><span class=""><div><br></div><div>for (i = 1; i <= N-2; i++)</div><div><span style="white-space:pre-wrap"> </span>for (j = 1; j <= N-2; j++)</div></span><div><span style="white-space:pre-wrap"> </span> a[i][j] = b[i][j];</div><div> </div><div>}</div><div>}</div><div><br></div><div> but still no vectorization in IR. Also, when I set vector width explicitly to 64, it gives the following error:</div></div><div><br></div><div><div>remark: <unknown>:0:0: loop not vectorized: call instruction cannot be vectorized</div><div>remark: <unknown>:0:0: loop not vectorized: value that could not be identified as reduction is used outside the loop</div></div><div><br></div><div>I need serious help on this. Please guide me.</div></div><div class="HOEnZb"><div class="h5"><div class="gmail_extra"><br><div class="gmail_quote">On Sun, Jul 2, 2017 at 1:54 AM, hameeza ahmed <span dir="ltr"><<a href="mailto:hahmed2305@gmail.com" target="_blank">hahmed2305@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="auto">Does it happen due to loop carried dependence? if yes what is the solution to vectorize such codes?<div dir="auto"><br></div><div dir="auto"><br></div><div dir="auto">please reply. i m waiting.</div></div><div class="m_111627299988929908HOEnZb"><div class="m_111627299988929908h5"><div class="gmail_extra"><br><div class="gmail_quote">On Jul 1, 2017 12:30 PM, "hameeza ahmed" <<a href="mailto:hahmed2305@gmail.com" target="_blank">hahmed2305@gmail.com</a>> wrote:<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">I even tried polly but still my llvm IR does not contain vector instructions. i used the following command;<div><br></div><div><div>clang -S -emit-llvm stencil.c -march=knl -O3 -mllvm -polly -mllvm -polly-vectorizer=stripmine -o stencil_poly.ll</div></div><div><br></div><div>Please specify what is wrong with my code?</div><div><br></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Sat, Jul 1, 2017 at 4:08 PM, hameeza ahmed <span dir="ltr"><<a href="mailto:hahmed2305@gmail.com" target="_blank">hahmed2305@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Hello,<div><br></div><div>I am trying to vectorize following stencil code;</div><div><br></div><div><div>#include <stdio.h></div><div>#define N 100351</div><div> </div><div>// This function computes 2D-5 point Jacobi stencil </div><div>void stencil(int a[restrict][N])</div><div>{</div><div> int i, j, k;</div><div> for (k = 0; k < 100; k++)</div><div> { for (i = 1; i <= N-2; i++)</div><div> { for (j = 1; j <= N-2; j++)</div><div> <span style="white-space:pre-wrap"> </span>{ a[i][j] = 0.25 * (a[i][j] + a[i-1][j] + a[i+1][j] + a[i][j-1] + a[i][j+1]);</div><div> }</div><div>}</div><div>}}</div></div><div><br></div><div>I have used the following commands</div><div><br></div><div>clang -S -emit-llvm stencil.c -march=knl -O3 -mllvm -disable-llvm-optzns -o stencil.ll<br></div><div><br></div><div>opt -S -O3 stencil.ll -o stencil_o3.ll<br></div><div><br></div><div>llc -x86-asm-syntax=intel stencil_o3.ll -o stencil.s<br></div><div><br></div><div>But the code is not vectorized. It still uses the scalar instructions;</div><div><br></div><div>Please correct me.</div><div><br></div><div>Thank You</div><div><br></div><div><br></div><div><br></div><div><br></div></div>
</blockquote></div><br></div>
</blockquote></div></div>
</div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>