性能记录如何获取目标进程的指令的虚拟内存地址以及性能用于存储的数据结构

Question

I'm reading source code of perf and try to make sense of how perf record get the the virtual memory address of an Instruction of the target process that trigger perf to count(for example the virtual address of call test ,which is assembly code watched by perf report 's annotation). And what data structure does perf use to store the virtual address of the instruction that make perf-event happen and trigger counter?

Answer 1

In order to get the instruction pointers for samples, you must include PERF_SAMPLE_IP in perf_attr.sample_type for the perf_event_open call. This will include the ip field in the struct for PERF_RECORD_SAMPLE events that you can read from the mmap ringbuffer. See the manpage for details.

PERF_RECORD_SAMPLE
    This record indicates a sample.

       struct {
            struct perf_event_header header;
            u64    sample_id;   /* if PERF_SAMPLE_IDENTIFIER */
            u64    ip;          /* if PERF_SAMPLE_IP */
            u32    pid, tid;    /* if PERF_SAMPLE_TID */
            u64    time;        /* if PERF_SAMPLE_TIME */
            ...