deploy: e06e54c

core_simd/simd/index.html

@@ -19,7 +19,7 @@ <h4 id="operations-use-the-best-instructions-available"><a class="doc-anchor" hr
 Instead, they map to a reasonable implementation of the operation for the target.</p>
 <p>Consistency between targets is not compromised to use faster or fewer instructions.
 In some cases, <code>std::arch</code> will provide a faster function that has slightly different behavior than the <code>std::simd</code> equivalent.
-For example, <a href="https://doc.rust-lang.org/nightly/core/core_arch/x86/sse/fn._mm_min_ps.html" title="fn core::core_arch::x86::sse::_mm_min_ps"><code>_mm_min_ps</code></a><sup id="fnref1"><a href="#fn1">1</a></sup> can be slightly faster than <a href="num/trait.SimdFloat.html#tymethod.simd_min" title="method core_simd::simd::num::SimdFloat::simd_min"><code>SimdFloat::simd_min</code></a>, but does not conform to the IEEE standard also used by <a href="https://doc.rust-lang.org/nightly/core/primitive.f32.html#method.min" title="method f32::min"><code>f32::min</code></a>.
+For example, <code>_mm_min_ps</code><sup id="fnref1"><a href="#fn1">1</a></sup> can be slightly faster than <a href="num/trait.SimdFloat.html#tymethod.simd_min" title="method core_simd::simd::num::SimdFloat::simd_min"><code>SimdFloat::simd_min</code></a>, but does not conform to the IEEE standard also used by <a href="https://doc.rust-lang.org/nightly/core/primitive.f32.html#method.min" title="method f32::min"><code>f32::min</code></a>.
 When necessary, <a href="struct.Simd.html" title="struct core_simd::simd::Simd"><code>Simd&lt;T, N&gt;</code></a> can be converted to the types provided by <code>std::arch</code> to make use of target-specific functions.</p>
 <p>Many targets simply don’t have SIMD, or don’t support SIMD for a particular element type.
 In those cases, regular scalar operations are generated instead.</p>

src/core_simd/swizzle_dyn.rs.html

@@ -230,7 +230,7 @@ <h2>Files</h2></div></nav><div class="sidebar-resizer"></div>
                 </span><span class="number">16 </span>=&gt; transize(vqtbl1q_u8, <span class="self">self</span>, idxs),
                 <span class="attr">#[cfg(all(target_feature = <span class="string">"avx2"</span>, not(target_feature = <span class="string">"avx512vbmi"</span>)))]
                 </span><span class="number">32 </span>=&gt; transize_raw(avx2_pshufb, <span class="self">self</span>, idxs),
-                <span class="attr">#[cfg(target_feature = <span class="string">"avx512vl,avx512vbmi"</span>)]
+                <span class="attr">#[cfg(all(target_feature = <span class="string">"avx512vl"</span>, target_feature = <span class="string">"avx512vbmi"</span>))]
                 </span><span class="number">32 </span>=&gt; transize(x86::_mm256_permutexvar_epi8, <span class="self">self</span>, idxs),
                 <span class="comment">// Notable absence: avx512bw shuffle
                 // If avx512bw is available, odds of avx512vbmi are good