理解trace-cmd工作原理
This is a very useful file for synchronizing user space with events happening in the kernel. Writing strings into this file will be written into the ftrace buffer.
- 采用wpa_supplicant源代码进行分析，其中trace_marker操作位于: src/utils/wpa_debug.c
- int wpa_debug_open_linux_tracing(void)
- void wpa_printf(int level, const char *fmt, ...)
- /sys/kernel/debug/tracing/trace_marker
如下是简单的示例使用代码
int trace_fd = open("trace_marker", WR_ONLY);
void trace_write(const char *fmt, ...)
{
va_list ap;
char buf[256];
int n;
if (trace_fd < 0)
return;
va_start(ap, fmt);
n = vsnprintf(buf, 256, fmt, ap);
va_end(ap);
write(trace_fd, buf, n);
}这里需要需要注意的是上面的王trace_marker中些内容，其实是相当于调用printk，所以后面使用trace-cmd进行event监听的时候是要监听printk的。
- ./wpa_supplicant -c/etc/wpa_supplicant/wpa_supplicant.conf -iwlan0 -Dnl80211,wext -T
- 监听事件：
- 监听所有event：trace-cmd record -e all -P 4338
- 只见听一个printk事件：trace-cmd record -e printk -P 4338
- 只见多个事件：trace-cmd record -e printk -e sys_enter -P 4338
- 上面
trace-cmd record命令会生成trace.dat文件 - 使用
kernelshark查看trace.dat
cat /sys/kernel/debug/tracing/available_events
.config - Linux/arm 4.19.71 Kernel Configuration
→ Kernel hacking → Tracers ────────────────────────────────────────────────────────────────────────────────────
┌──────────────────────────────────────────────── Tracers ─────────────────────────────────────────────────┐
│ Arrow keys navigate the menu. <Enter> selects submenus ---> (or empty submenus ----). Highlighted │
│ letters are hotkeys. Pressing <Y> includes, <N> excludes, <M> modularizes features. Press <Esc><Esc> │
│ to exit, <?> for Help, </> for Search. Legend: [*] built-in [ ] excluded <M> module < > module │
│ capable │
│ ┌──────────────────────────────────────────────────────────────────────────────────────────────────────┐ │
│ │ --- Tracers │ │
│ │ -*- Kernel Function Tracer │ │
│ │ [*] Kernel Function Graph Tracer │ │
│ │ [ ] Enable trace events for preempt and irq disable/enable │ │
│ │ [*] Interrupts-off Latency Tracer │ │
│ │ [*] Scheduling Latency Tracer │ │
│ │ [ ] Tracer to detect hardware latencies (like SMIs) │ │
│ │ [ ] Trace syscalls │ │
│ │ -*- Create a snapshot trace buffer │ │
│ │ -*- Allow snapshot to swap per CPU │ │
│ │ Branch Profiling (No branch profiling) ---> │ │
│ │ [*] Trace max stack │ │
│ │ [*] Support for tracing block IO actions │ │
│ │ [*] Enable kprobes-based dynamic events │ │
│ │ [ ] Enable uprobes-based dynamic events │ │
│ │ [*] enable/disable function tracing dynamically │ │
│ │ [*] Kernel function profiler │ │
│ │ [ ] Perform a startup test on ftrace │ │
│ │ [ ] Add tracepoint that benchmarks tracepoints │ │
│ │ < > Ring buffer benchmark stress tester │ │
│ │ [ ] Ring buffer startup self test │ │
│ │ < > Preempt / IRQ disable delay thread to test latency tracers │ │
│ │ [ ] Show eval mappings for trace events │ │
│ │ [*] Trace gpio events │ │
│ │ │ │
│ │ │ │
│ └──────────────────────────────────────────────────────────────────────────────────────────────────────┘ │
├──────────────────────────────────────────────────────────────────────────────────────────────────────────┤
│ <Select> < Exit > < Help > < Save > < Load > │
└──────────────────────────────────────────────────────────────────────────────────────────────────────────┘