Fix: Missing _mm_cvtsi128_si64x in Clang

Author: ashvardanian

include/stringzilla/stringzilla.h

@@ -4022,7 +4022,7 @@ SZ_PUBLIC sz_u64_t sz_checksum_avx2(sz_cptr_t text, sz_size_t length) {
         __m128i low_xmm = _mm256_castsi256_si128(sums_vec.ymm);
         __m128i high_xmm = _mm256_extracti128_si256(sums_vec.ymm, 1);
         __m128i sums_xmm = _mm_add_epi64(low_xmm, high_xmm);
-        sz_u64_t low = (sz_u64_t)_mm_cvtsi128_si64x(sums_xmm);
+        sz_u64_t low = (sz_u64_t)_mm_cvtsi128_si64(sums_xmm);
         sz_u64_t high = (sz_u64_t)_mm_extract_epi64(sums_xmm, 1);
         sz_u64_t result = low + high;
         if (length) result += sz_checksum_serial(text, length);
@@ -4073,7 +4073,7 @@ SZ_PUBLIC sz_u64_t sz_checksum_avx2(sz_cptr_t text, sz_size_t length) {
         __m128i low_xmm = _mm256_castsi256_si128(sums_vec.ymm);
         __m128i high_xmm = _mm256_extracti128_si256(sums_vec.ymm, 1);
         __m128i sums_xmm = _mm_add_epi64(low_xmm, high_xmm);
-        sz_u64_t low = (sz_u64_t)_mm_cvtsi128_si64x(sums_xmm);
+        sz_u64_t low = (sz_u64_t)_mm_cvtsi128_si64(sums_xmm);
         sz_u64_t high = (sz_u64_t)_mm_extract_epi64(sums_xmm, 1);
         result += low + high;
         return result;
@@ -5306,7 +5306,7 @@ SZ_PUBLIC sz_u64_t sz_checksum_avx512(sz_cptr_t text, sz_size_t length) {
         __mmask16 mask = _sz_u16_mask_until(length);
         text_vec.xmms[0] = _mm_maskz_loadu_epi8(mask, text);
         sums_vec.xmms[0] = _mm_sad_epu8(text_vec.xmms[0], _mm_setzero_si128());
-        sz_u64_t low = (sz_u64_t)_mm_cvtsi128_si64x(sums_vec.xmms[0]);
+        sz_u64_t low = (sz_u64_t)_mm_cvtsi128_si64(sums_vec.xmms[0]);
         sz_u64_t high = (sz_u64_t)_mm_extract_epi64(sums_vec.xmms[0], 1);
         return low + high;
     }
@@ -5318,7 +5318,7 @@ SZ_PUBLIC sz_u64_t sz_checksum_avx512(sz_cptr_t text, sz_size_t length) {
         __m128i low_xmm = _mm256_castsi256_si128(sums_vec.ymms[0]);
         __m128i high_xmm = _mm256_extracti128_si256(sums_vec.ymms[0], 1);
         __m128i sums_xmm = _mm_add_epi64(low_xmm, high_xmm);
-        sz_u64_t low = (sz_u64_t)_mm_cvtsi128_si64x(sums_xmm);
+        sz_u64_t low = (sz_u64_t)_mm_cvtsi128_si64(sums_xmm);
         sz_u64_t high = (sz_u64_t)_mm_extract_epi64(sums_xmm, 1);
         return low + high;
     }
@@ -6092,7 +6092,7 @@ SZ_PUBLIC sz_bool_t sz_equal_neon(sz_cptr_t a, sz_cptr_t b, sz_size_t length) {
     return sz_true_k;
 }
 
-SZ_PUBLIC sz_u64_t sz_checksum_neon(sz_cptr_t const *text, size_t length) {
+SZ_PUBLIC sz_u64_t sz_checksum_neon(sz_cptr_t text, sz_size_t length) {
     uint64x2_t sum_vec = vdupq_n_u64(0);
 
     // Process 16 bytes (128 bits) at a time

scripts/bench_token.cpp

@@ -4,6 +4,8 @@
  *
  *  This file is the sibling of `bench_sort.cpp`, `bench_search.cpp` and `bench_similarity.cpp`.
  */
+#include <numeric> // `std::accumulate`
+
 #include <bench.hpp>
 #include <test.hpp> // `random_string`
 
@@ -189,7 +191,7 @@ void bench(strings_type &&strings) {
 
 void bench_on_input_data(int argc, char const **argv) {
     dataset_t dataset = prepare_benchmark_environment(argc, argv);
-
+#if 0
     std::printf("Benchmarking on the entire dataset:\n");
     bench_unary_functions(dataset.tokens, random_generation_functions(100));
     bench_unary_functions(dataset.tokens, random_generation_functions(20));
@@ -211,14 +213,14 @@ void bench_on_input_data(int argc, char const **argv) {
     bench_unary_functions<std::vector<std::string_view>>({dataset.text}, fingerprinting_functions(128, 4 * 1024));
     bench_unary_functions<std::vector<std::string_view>>({dataset.text}, fingerprinting_functions(128, 64 * 1024));
     bench_unary_functions<std::vector<std::string_view>>({dataset.text}, fingerprinting_functions(128, 1024 * 1024));
-
+#endif
     // Baseline benchmarks for real words, coming in all lengths
-    std::printf("Benchmarking on entire dataset:\n");
-    bench<std::vector<std::string_view>>({dataset.text});
-    std::printf("Benchmarking on real lines:\n");
-    bench(dataset.lines);
     std::printf("Benchmarking on real words:\n");
     bench(dataset.tokens);
+    std::printf("Benchmarking on real lines:\n");
+    bench(dataset.lines);
+    std::printf("Benchmarking on entire dataset:\n");
+    bench<std::vector<std::string_view>>({dataset.text});
 
     // Run benchmarks on tokens of different length
     for (std::size_t token_length : {1, 2, 3, 4, 5, 6, 7, 8, 16, 32}) {

scripts/test.cpp

@@ -1546,12 +1546,10 @@ int main(int argc, char const **argv) {
     std::printf("- Uses NEON: %s \n", SZ_USE_ARM_NEON ? "yes" : "no");
     std::printf("- Uses SVE: %s \n", SZ_USE_ARM_SVE ? "yes" : "no");
 
-#if 0
     // Basic utilities
     test_arithmetical_utilities();
     test_memory_utilities();
     test_replacements();
-#endif
 
 // Compatibility with STL
 #if SZ_DETECT_CPP_17 && __cpp_lib_string_view