simde: port simde to Elbrus

Author: makise-homura

libobs/util/simde/simde-arch.h

@@ -181,6 +181,49 @@
 #define SIMDE_ARCH_H8300
 #endif
 
+/* Elbrus (8S, 8SV and successors)
+   <https://en.wikipedia.org/wiki/Elbrus-8S> */
+#if defined(__e2k__)
+#define SIMDE_ARCH_E2K
+#endif
+
+/* SIMD ISA extensions for Elbrus (equivalent to X86 ones) */
+#if defined(SIMDE_ARCH_E2K)
+#if defined(__MMX__)
+#define SIMDE_ARCH_X86_MMX 1
+#endif
+#if defined(__SSE__)
+#define SIMDE_ARCH_X86_SSE 1
+#endif
+#if defined(__SSE2__)
+#define SIMDE_ARCH_X86_SSE2 1
+#endif
+#if defined(__SSE3__)
+#define SIMDE_ARCH_X86_SSE3 1
+#endif
+#if defined(__SSSE3__)
+#define SIMDE_ARCH_X86_SSSE3 1
+#endif
+#if defined(__SSE4_1__)
+#define SIMDE_ARCH_X86_SSE4_1 1
+#endif
+#if defined(__SSE4_2__)
+#define SIMDE_ARCH_X86_SSE4_2 1
+#endif
+#if defined(__AVX__)
+#define SIMDE_ARCH_X86_AVX 1
+#endif
+#if defined(__AVX2__)
+#define SIMDE_ARCH_X86_AVX2 1
+#endif
+#if defined(__FMA__)
+#define SIMDE_ARCH_X86_FMA 1
+#endif
+#if defined(__PCLMUL__)
+#define SIMDE_ARCH_X86_PCLMUL 1
+#endif
+#endif
+
 /* HP/PA / PA-RISC
    <https://en.wikipedia.org/wiki/PA-RISC> */
 #if defined(__PA8000__) || defined(__HPPA20__) || defined(__RISC2_0__) || \

libobs/util/simde/simde-common.h

@@ -324,6 +324,11 @@ HEDLEY_DIAGNOSTIC_POP
 #define SIMDE_VECTORIZE_REDUCTION(r) HEDLEY_PRAGMA(omp simd reduction(r))
 #endif
 #define SIMDE_VECTORIZE_ALIGNED(a) HEDLEY_PRAGMA(omp simd aligned(a))
+#elif defined(__e2k__)
+#define SIMDE_VECTORIZE
+#define SIMDE_VECTORIZE_SAFELEN(l)
+#define SIMDE_VECTORIZE_REDUCTION(r)
+#define SIMDE_VECTORIZE_ALIGNED(a)
 #elif defined(SIMDE_ENABLE_CILKPLUS)
 #define SIMDE_VECTORIZE HEDLEY_PRAGMA(simd)
 #define SIMDE_VECTORIZE_SAFELEN(l) HEDLEY_PRAGMA(simd vectorlength(l))

libobs/util/simde/x86/sse2.h

@@ -5862,7 +5862,7 @@ simde__m128i simde_mm_slli_epi16(simde__m128i a, const int imm8)
 	simde__m128i_private r_, a_ = simde__m128i_to_private(a);
 
 #if defined(SIMDE_VECTOR_SUBSCRIPT_SCALAR)
-	r_.i16 = a_.i16 << (imm8 & 0xff);
+	r_.i16 = a_.i16 << (uint8_t)(imm8 & 0xff);
 #else
 	const int s =
 		(imm8 >
@@ -5982,7 +5982,7 @@ simde__m128i simde_mm_slli_epi64(simde__m128i a, const int imm8)
 	simde__m128i_private r_, a_ = simde__m128i_to_private(a);
 
 #if defined(SIMDE_VECTOR_SUBSCRIPT_SCALAR)
-	r_.i64 = a_.i64 << imm8;
+	r_.i64 = a_.i64 << (uint8_t)imm8;
 #else
 	SIMDE_VECTORIZE
 	for (size_t i = 0; i < (sizeof(r_.i64) / sizeof(r_.i64[0])); i++) {
@@ -6028,7 +6028,7 @@ simde__m128i simde_mm_srli_epi16(simde__m128i a, const int imm8)
 	simde__m128i_private r_, a_ = simde__m128i_to_private(a);
 
 #if defined(SIMDE_VECTOR_SUBSCRIPT_SCALAR)
-	r_.u16 = a_.u16 >> imm8;
+	r_.u16 = a_.u16 >> (uint8_t)imm8;
 #else
 	SIMDE_VECTORIZE
 	for (size_t i = 0; i < (sizeof(r_.i16) / sizeof(r_.i16[0])); i++) {
@@ -6082,7 +6082,7 @@ simde__m128i simde_mm_srli_epi32(simde__m128i a, const int imm8)
 	simde__m128i_private r_, a_ = simde__m128i_to_private(a);
 
 #if defined(SIMDE_VECTOR_SUBSCRIPT_SCALAR)
-	r_.u32 = a_.u32 >> (imm8 & 0xff);
+	r_.u32 = a_.u32 >> (uint8_t)(imm8 & 0xff);
 #else
 	SIMDE_VECTORIZE
 	for (size_t i = 0; i < (sizeof(r_.i32) / sizeof(r_.i32[0])); i++) {
@@ -6148,7 +6148,7 @@ simde__m128i simde_mm_srli_epi64(simde__m128i a, const int imm8)
 	r_.neon_u64 = vshlq_u64(a_.neon_u64, vdupq_n_s64(-imm8));
 #else
 #if defined(SIMDE_VECTOR_SUBSCRIPT_SCALAR) && !defined(SIMDE_BUG_GCC_94488)
-	r_.u64 = a_.u64 >> imm8;
+	r_.u64 = a_.u64 >> (uint8_t)imm8;
 #else
 	SIMDE_VECTORIZE
 	for (size_t i = 0; i < (sizeof(r_.i64) / sizeof(r_.i64[0])); i++) {