测得的mcycle小于spike中的minstret

Question

我用spike来运行 "riscv-tools/riscv-tests/build/benchmarks"中的测试程序：

$ spike multiply.riscv

并且输出显示：

mcycle = 24096
minstret = 24103

为什么 mcycle 小于 minstret？

难道这意味着尖峰在一个周期内可以运行条以上的指令？
（我试图追踪尖峰代码，但找不到如何计算 mcycle。）

Answer 1

在这种情况下，mcycle 和 minstret 值的打印不是来自 Spike，而是来自测试（基准）。有代码：

https://github.com/ucb-bar/riscv-benchmarks/blob/master/common/syscalls.c

#define NUM_COUNTERS 2
static uintptr_t counters[NUM_COUNTERS];
static char* counter_names[NUM_COUNTERS];

static int handle_stats(int enable)
{
  int i = 0;
#define READ_CTR(name) do { \
    while (i >= NUM_COUNTERS) ; \
    uintptr_t csr = read_csr(name); \
    if (!enable) { csr -= counters[i]; counter_names[i] = #name; } \
    counters[i++] = csr; \
  } while (0)

  READ_CTR(mcycle);
  READ_CTR(minstret);

#undef READ_CTR
  return 0;
}

阅读 mcycle 和 minstret 之间有一些代码。现在您知道读数之间有多少代码了。

在 Spike 中，mcycle 和 minstret 根据定义总是相等的（它们由相同的代码处理）：https://github.com/riscv/riscv-isa-sim/blob/9e012462f53113dc9ed00d7fbb89aeafeb9b89e9/riscv/processor.cc#L347

    case CSR_MINSTRET:
    case CSR_MCYCLE:
      if (xlen == 32)
        state.minstret = (state.minstret >> 32 << 32) | (val & 0xffffffffU);
      else
        state.minstret = val;
      break;

syscalls.c 从 https://github.com/ucb-bar/riscv-benchmarks/blob/master/multiply/bmark.mk 链接到 multiply.riscv 二进制文件 - multiply_riscv_bin = multiply.riscv $(multiply_riscv_bin): ... $(patsubst %.c, %.o, ... syscalls.c ... )

在syscalls.c函数中有_init调用测试的main并打印值，记录在SYS_stats"syscall"和handle_stats.

void _init(int cid, int nc)
{
  init_tls();
  thread_entry(cid, nc);

  // only single-threaded programs should ever get here.
  int ret = main(0, 0);

  char buf[NUM_COUNTERS * 32] __attribute__((aligned(64)));
  char* pbuf = buf;
  for (int i = 0; i < NUM_COUNTERS; i++)
    if (counters[i])
      pbuf += sprintf(pbuf, "%s = %d\n", counter_names[i], counters[i]);
  if (pbuf != buf)
    printstr(buf);

  exit(ret);
}

测得的mcycle小于spike中的minstret

The measured mcycle is less than minstret in Spike

riscv