ggml : simplify the SIMD code (#324)

* ggml : simplify the SIMD code

* ggml : generic reduce for all register sizes + comments

ggml.h

@@ -724,6 +724,7 @@ enum ggml_opt_result ggml_opt(
 int ggml_cpu_has_avx(void);
 int ggml_cpu_has_avx2(void);
 int ggml_cpu_has_avx512(void);
+int ggml_cpu_has_fma(void);
 int ggml_cpu_has_neon(void);
 int ggml_cpu_has_arm_fma(void);
 int ggml_cpu_has_f16c(void);

whisper.cpp

@@ -2555,6 +2555,7 @@ const char * whisper_print_system_info(void) {
     s += "AVX = "       + std::to_string(ggml_cpu_has_avx())       + " | ";
     s += "AVX2 = "      + std::to_string(ggml_cpu_has_avx2())      + " | ";
     s += "AVX512 = "    + std::to_string(ggml_cpu_has_avx512())    + " | ";
+    s += "FMA = "       + std::to_string(ggml_cpu_has_fma())       + " | ";
     s += "NEON = "      + std::to_string(ggml_cpu_has_neon())      + " | ";
     s += "ARM_FMA = "   + std::to_string(ggml_cpu_has_arm_fma())   + " | ";
     s += "F16C = "      + std::to_string(ggml_cpu_has_f16c())      + " | ";