Code snippet for ad-hoc profiling with RDTSC/RDTSCP.

tsc.cpp

struct tsc {                                                              
    uint32_t cycles_low, cycles_high;                                     
    uint64_t cycles;                                                      
                                                                          
    tsc() : cycles(0) {}                                                  
                                                                          
    void begin() {                                                        
        asm volatile (                                                    
            "CPUID\n\t"/*serialize*/                                      
            "RDTSC\n\t"/*read the clock*/                                 
            "mov %%edx, %0\n\t"                                           
            "mov %%eax, %1\n\t"                                           
            : "=r" (cycles_high), "=r" (cycles_low)                       
            :: "%rax", "%rbx", "%rcx", "%rdx");                           
    }                                                                     
                                                                          
    void end() {                                                          
        uint32_t cycles_high1, cycles_low1;                               
        asm volatile (                                                    
            "RDTSCP\n\t"/*read the clock*/                                
            "mov %%edx, %0\n\t"                                           
            "mov %%eax, %1\n\t"                                           
            "CPUID\n\t"                                                   
            : "=r" (cycles_high1), "=r" (cycles_low1)                     
            :: "%rax", "%rbx", "%rcx", "%rdx");                           
        cycles += ((uint64_t(cycles_high1) << 32) + uint64_t(cycles_low1))
            - ((uint64_t(cycles_high) << 32) + uint64_t(cycles_low));     
    }                                                                     
};