[llvm] [llvm] Improve implementation of StringRef::find_last_of and cie (PR #71865)

[via llvm-commits]

https://github.com/serge-sans-paille updated https://github.com/llvm/llvm-project/pull/71865

>From 5ebe6d8729ea7b9fa69f2c1ff39c8ddd76997503 Mon Sep 17 00:00:00 2001
From: serge-sans-paille <sguelton at mozilla.com>
Date: Thu, 9 Nov 2023 20:41:40 +0100
Subject: [PATCH] [llvm] Improve implementation of StringRef::find_last_of for
 the usual case of 2 chars

Almost all usage of StringRef::find_last_of in Clang/LLVM use a Needle
of 2 elements, which can easily be optimized in SSE2. The IPC of the
improved version is significantly better as shown in

        https://godbolt.org/z/h1dsdcMd8

And it does not require an extra structure.
---
 llvm/lib/Support/StringRef.cpp | 33 ++++++++++++++++++++++++++++++++-
 1 file changed, 32 insertions(+), 1 deletion(-)

diff --git a/llvm/lib/Support/StringRef.cpp b/llvm/lib/Support/StringRef.cpp
index feee47ca693b251..c9923cdfdb76754 100644
--- a/llvm/lib/Support/StringRef.cpp
+++ b/llvm/lib/Support/StringRef.cpp
@@ -15,6 +15,10 @@
 #include "llvm/Support/Error.h"
 #include <bitset>
 
+#ifdef __SSE2__
+#include <emmintrin.h>
+#endif
+
 using namespace llvm;
 
 // MSVC emits references to this into the translation units which reference it.
@@ -268,17 +272,44 @@ StringRef::size_type StringRef::find_first_not_of(StringRef Chars,
   return npos;
 }
 
+#ifdef __SSE2__
+
+StringRef::size_type vectorized_find_last_of_specialized(const char *Data,
+                                                         size_t Sz, char C0,
+                                                         char C1) {
+  __m128i Needle0 = _mm_set1_epi8(Chars[0]);
+  __m128i Needle1 = _mm_set1_epi8(Chars[1]);
+  do {
+    Sz = Sz < 16 ? 0 : Sz - 16;
+    __m128i Buffer = _mm_loadu_si128((const __m128i *)(Data + Sz));
+    unsigned Mask = _mm_movemask_epi8(_mm_or_si128(
+        _mm_cmpeq_epi8(Buffer, Needle0), _mm_cmpeq_epi8(Buffer, Needle1)));
+    if (Mask != 0) {
+      return Sz + sizeof(Mask) * CHAR_BIT - llvm::countl_zero(Mask);
+    }
+  } while (Sz);
+  return npos;
+}
+#endif
+
 /// find_last_of - Find the last character in the string that is in \arg C,
 /// or npos if not found.
 ///
 /// Note: O(size() + Chars.size())
 StringRef::size_type StringRef::find_last_of(StringRef Chars,
                                              size_t From) const {
+  size_type Sz = std::min(From, Length);
+
+#ifdef __SSE2__
+  if (Chars.size() == 2) {
+    return vectorized_find_last_of_specialized(Data, Sz, Chars[0], Chars[1]);
+#endif
+
   std::bitset<1 << CHAR_BIT> CharBits;
   for (char C : Chars)
     CharBits.set((unsigned char)C);
 
-  for (size_type i = std::min(From, Length) - 1, e = -1; i != e; --i)
+  for (size_type i = Sz - 1, e = -1; i != e; --i)
     if (CharBits.test((unsigned char)Data[i]))
       return i;
   return npos;