Skip to content

Commit

Permalink
Merge pull request #606 from albertxu216/develop
Browse files Browse the repository at this point in the history 
cs_delay&libbpf_sar合进cpuwatcher，sar新增Idle和Thread功能，可通过传参分别实现不同功能。
  • Loading branch information
@chenamy2017
chenamy2017 authored Dec 1, 2023
2 parents b9f82ff + a6fbdc9 commit 6820581
Show file tree
Hide file tree
Showing 14 changed files with 666 additions and 794 deletions.
18 changes: 9 additions & 9 deletions .github/workflows/libbpf.yml → .github/workflows/cpu_watcher.yml
View file
Original file line number Diff line number Diff line change
@@ -1,18 +1,18 @@
name: libbpf
name: cpu_watcher

on:
push:
branches:
- "*"
paths:
- 'eBPF_Supermarket/CPU_Subsystem/cpu_watcher/cs_delay/libbpf_cs_delay/**'
- '.github/workflows/libbpf.yml'
- 'eBPF_Supermarket/CPU_Subsystem/cpu_watcher/**'
- '.github/workflows/cpu_watcher.yml'
pull_request:
branches:
- "*"
paths:
- 'eBPF_Supermarket/CPU_Subsystem/cpu_watcher/cs_delay/libbpf_cs_delay/**'
- '.github/workflows/libbpf.yml'
- 'eBPF_Supermarket/CPU_Subsystem/cpu_watcher/**'
- '.github/workflows/cpu_watcher.yml'

jobs:
libbpf-project-build-and-test:
Expand All @@ -26,8 +26,8 @@ jobs:
sudo apt install libbpf-dev clang llvm libelf-dev libpcap-dev gcc-multilib build-essential
git submodule update --init --recursive
- name: Run cs_delay
- name: Run cpu_watcher
run: |
cd eBPF_Supermarket/CPU_Subsystem/cpu_watcher/cs_delay/libbpf_cs_delay
make cs_delay
sudo ./cs_delay
cd eBPF_Supermarket/CPU_Subsystem/cpu_watcher/
make cpu_watcher
sudo ./cpu_watcher
10 changes: 5 additions & 5 deletions .github/workflows/system_cpu.yml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -104,7 +104,7 @@ jobs:
uname -r
./run.sh
libbpf_sar-project-build-and-test:
cpu_watcher-project-build-and-test:
runs-on: ubuntu-22.04
steps:
- uses: actions/checkout@v3
Expand All @@ -115,8 +115,8 @@ jobs:
sudo apt install libbpf-dev clang llvm libelf-dev libpcap-dev gcc-multilib build-essential
git submodule update --init --recursive
- name: Run libbpf_sar
- name: Run cpu_watcher
run: |
cd eBPF_Supermarket/CPU_Subsystem/cpu_watcher/libbpf_sar
make libbpf_sar
sudo ./libbpf_sar -t 1
cd eBPF_Supermarket/CPU_Subsystem/cpu_watcher
make cpu_watcher
sudo ./cpu_watcher
0 ..._watcher/cs_delay/BCC_cs_delay/cs_delay.c → ...ket/CPU_Subsystem/BCC_cs_delay/cs_delay.c
View file
File renamed without changes.
0 ...watcher/cs_delay/BCC_cs_delay/cs_delay.py → ...et/CPU_Subsystem/BCC_cs_delay/cs_delay.py
View file
File renamed without changes.
10 changes: 5 additions & 5 deletions ...watcher/cs_delay/libbpf_cs_delay/Makefile → ...market/CPU_Subsystem/cpu_watcher/Makefile
View file
Original file line number Diff line number Diff line change
Expand Up @@ -19,12 +19,12 @@
# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause)
OUTPUT := .output
CLANG ?= clang
LIBBPF_SRC := $(abspath ../../../libbpf/src)
BPFTOOL_SRC := $(abspath ../../../bpftool/src)
LIBBPF_SRC := $(abspath ../libbpf/src)
BPFTOOL_SRC := $(abspath ../bpftool/src)
LIBBPF_OBJ := $(abspath $(OUTPUT)/libbpf.a)
BPFTOOL_OUTPUT ?= $(abspath $(OUTPUT)/bpftool)
BPFTOOL ?= $(BPFTOOL_OUTPUT)/bootstrap/bpftool
LIBBLAZESYM_SRC := $(abspath ../../../blazesym/)
LIBBLAZESYM_SRC := $(abspath ../blazesym/)
LIBBLAZESYM_INC := $(abspath $(LIBBLAZESYM_SRC)/include)
LIBBLAZESYM_OBJ := $(abspath $(OUTPUT)/libblazesym.a)
ARCH ?= $(shell uname -m | sed 's/x86_64/x86/' \
Expand All @@ -34,15 +34,15 @@ ARCH ?= $(shell uname -m | sed 's/x86_64/x86/' \
| sed 's/mips.*/mips/' \
| sed 's/riscv64/riscv/' \
| sed 's/loongarch64/loongarch/')
VMLINUX := ../../../vmlinux/$(ARCH)/vmlinux.h
VMLINUX := ../vmlinux/$(ARCH)/vmlinux.h
# Use our own libbpf API headers and Linux UAPI headers distributed with
# libbpf to avoid dependency on system-wide headers, which could be missing or
# outdated
INCLUDES := -I$(OUTPUT) -I../../../libbpf/include/uapi -I$(dir $(VMLINUX)) -I$(LIBBLAZESYM_INC)
CFLAGS := -g -Wall
ALL_LDFLAGS := $(LDFLAGS) $(EXTRA_LDFLAGS)

APPS = cs_delay
APPS = cpu_watcher

CARGO ?= $(shell which cargo)
ifeq ($(strip $(CARGO)),)
Expand Down
183 changes: 145 additions & 38 deletions ...m/cpu_watcher/libbpf_sar/libbpf_sar.bpf.c → ...U_Subsystem/cpu_watcher/cpu_watcher.bpf.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,13 +20,29 @@
#include <bpf/bpf_helpers.h>
#include <bpf/bpf_core_read.h>
#include <bpf/bpf_tracing.h>
#include "libbpf_sar.h"
#include "cpu_watcher.h"

char LICENSE[] SEC("license") = "Dual BSD/GPL";

const volatile long long unsigned int forks_addr = 0;

// 计数表格，第0项为所统计fork数，第1项为进程切换数,第2项为运行队列长度
#define PF_IDLE 0x00000002 /* I am an IDLE thread */
#define PF_KTHREAD 0x00200000 /* I am a kernel thread */

/*----------------------------------------------*/
/* cs_delay相关maps */
/*----------------------------------------------*/
struct {
__uint(type, BPF_MAP_TYPE_ARRAY);
__uint(max_entries, 1);
__type(key, int);
__type(value, u64);
} start SEC(".maps");//记录时间戳；

/*----------------------------------------------*/
/* sar相关maps */
/*----------------------------------------------*/
// 计数表格，第0项为所统计fork数，第1项为进程切换数,
struct {
__uint(type, BPF_MAP_TYPE_ARRAY);//基于数组的映射
__uint(max_entries, 3);//countMap 可以存储最多 3 对键值对
Expand All @@ -37,28 +53,17 @@ struct {
// 记录开始的时间
struct {
__uint(type, BPF_MAP_TYPE_ARRAY);
__uint(max_entries, 1);
__type(key, int);
__uint(max_entries, 4096);
__type(key, pid_t);
__type(value, u64);
} start SEC(".maps");//记录时间戳；
} procStartTime SEC(".maps");//记录时间戳；

//环形缓冲区；
struct {
__uint(type, BPF_MAP_TYPE_RINGBUF);
__uint(max_entries, 256 * 1024);
} rb SEC(".maps");

//cswch_args结构体
struct cswch_args {
u64 pad;
char prev_comm[16];
pid_t prev_pid;
int prev_prio;
long prev_state;
char next_comm[16];
pid_t next_pid;
int next_prio;
};

// 储存运行队列rq的全局变量
struct {
Expand All @@ -70,17 +75,12 @@ __type(value, struct rq);


struct {
__uint(type, BPF_MAP_TYPE_PERCPU_ARRAY);
__uint(max_entries, 1);
__type(key, u32);
__type(value, int);
__uint(type, BPF_MAP_TYPE_PERCPU_ARRAY);
__uint(max_entries, 1);
__type(key, u32);
__type(value, int);
} runqlen SEC(".maps");//多CPU数组

struct __softirq_info {
u64 pad;
u32 vec;
};

struct {
__uint(type, BPF_MAP_TYPE_PERCPU_HASH);
__uint(max_entries, 4096);
Expand Down Expand Up @@ -111,33 +111,94 @@ __type(key, u32);
__type(value, u64);
} irq_Last_time SEC(".maps");

struct __irq_info {
u64 pad;
u32 irq;
};
// 储存cpu进入空闲的起始时间
struct {
__uint(type, BPF_MAP_TYPE_ARRAY);
__uint(max_entries, 128);
__type(key, u32);
__type(value, u64);
} idleStart SEC(".maps");

// 储存cpu进入空闲的持续时间
struct {
__uint(type, BPF_MAP_TYPE_ARRAY);
__uint(max_entries, 1);
__type(key, u32);
__type(value, u64);
} idleLastTime SEC(".maps");

struct {
__uint(type, BPF_MAP_TYPE_ARRAY);
__uint(max_entries, 1);
__type(key, u32);
__type(value, u64);
} kt_LastTime SEC(".maps");

/*----------------------------------------------*/
/* cs_delay跟踪函数 */
/*----------------------------------------------*/

SEC("kprobe/schedule")
int BPF_KPROBE(schedule)
{
u64 t1;
t1 = bpf_ktime_get_ns()/1000; //bpf_ktime_get_ns返回自系统启动以来所经过的时间(以纳秒为单位)。不包括系统挂起的时间。
int key =0;
bpf_map_update_elem(&start,&key,&t1,BPF_ANY);
return 0;
}

SEC("kretprobe/schedule")
int BPF_KRETPROBE(schedule_exit)
{
u64 t2 = bpf_ktime_get_ns()/1000;
u64 t1,delay;
int key = 0;
u64 *val = bpf_map_lookup_elem(&start,&key);
if (val != 0)
{
t1 = *val;
delay = t2 - t1;
bpf_map_delete_elem(&start, &key);
}else{
return 0;
}

struct event *e;
e = bpf_ringbuf_reserve(&rb, sizeof(*e), 0);
if (!e) return 0;

e->t1=t1;//开始时间
e->t2=t2;//结束时间
e->delay=delay;//时间间隔

/* 成功地将其提交到用户空间进行后期处理 */
bpf_ringbuf_submit(e, 0);

return 0;
}

/*----------------------------------------------*/
/* sar跟踪函数 */
/*----------------------------------------------*/
// 统计fork数
SEC("kprobe/finish_task_switch.isra.0")
int kprobe__finish_task_switch(struct pt_regs *ctx)
{
u32 key = 0;
u64 val, *valp = NULL;
unsigned long total_forks;

if(forks_addr !=0){
valp = (u64 *)forks_addr;
bpf_probe_read_kernel(&total_forks, sizeof(unsigned long), valp);
key = 1;
val = total_forks;
bpf_map_update_elem(&countMap,&key,&val,BPF_ANY);
}

return 0;
}



//获取进程切换数;
SEC("tracepoint/sched/sched_switch")//静态挂载点
int trace_sched_switch2(struct cswch_args *info) {
Expand All @@ -146,7 +207,7 @@ int trace_sched_switch2(struct cswch_args *info) {

// 只有当上一个进程和下一个进程不相同时才执行以下操作，相同则代表是同一个进程
if (prev != next) {
u32 key = 1;
u32 key = 0;
u64 *valp, delta, cur;
struct task_struct *ts;

Expand All @@ -155,7 +216,7 @@ int trace_sched_switch2(struct cswch_args *info) {
u64 time = bpf_ktime_get_ns()/1000;//获取当前时间，ms；

// Step1: 记录next进程的起始时间
bpf_map_update_elem(&start,&pid,&time,BPF_ANY);//上传当前时间到start map中
bpf_map_update_elem(&procStartTime,&pid,&time,BPF_ANY);//上传当前时间到start map中
//procStartTime.update(&pid, &time);//python

// Step2: Syscall时间处理
Expand All @@ -178,6 +239,25 @@ int trace_sched_switch2(struct cswch_args *info) {
return 0;
}

SEC("kprobe/finish_task_switch.isra.0")
int BPF_KPROBE(finish_task_switch,struct task_struct *prev){
pid_t pid=BPF_CORE_READ(prev,pid);
u64 *val, time = bpf_ktime_get_ns();
u64 delta;
// Step1: 记录内核进程（非IDLE）运行时间
if ((BPF_CORE_READ(prev,flags) & PF_KTHREAD) && pid!= 0) {
val = bpf_map_lookup_elem(&procStartTime, &pid);
if (val) {
u32 key = 0;
delta = time - *val*1000;
val = bpf_map_lookup_elem(&kt_LastTime, &key);
if (val) *val += delta;
else bpf_map_update_elem(&kt_LastTime, &key, &delta, BPF_ANY);
}
}
return 0;
}

/*
SEC("kprobe/finish_task_switch")//动态挂载点
int trace_sched_switch(struct cswch_args *info) {
Expand All @@ -195,7 +275,7 @@ int trace_sched_switch(struct cswch_args *info) {
u64 time = bpf_ktime_get_ns()/1000;//获取当前时间，ms；
// Step1: 记录next进程的起始时间
bpf_map_update_elem(&start,&pid,&time,BPF_ANY);//上传当前时间到start map中
bpf_map_update_elem(&procStartTime,&pid,&time,BPF_ANY);//上传当前时间到start map中
//procStartTime.update(&pid, &time);//python
// Step2: Syscall时间处理
Expand Down Expand Up @@ -223,7 +303,7 @@ int trace_sched_switch(struct cswch_args *info) {
//统计运行队列长度
SEC("kprobe/update_rq_clock")
int kprobe_update_rq_clock(struct pt_regs *ctx){
u32 key = 2;
u32 key = 0;
u32 rqkey = 0;
struct rq *p_rq = 0;
p_rq = (struct rq *)bpf_map_lookup_elem(&rq_map, &rqkey);
Expand All @@ -234,7 +314,7 @@ int kprobe_update_rq_clock(struct pt_regs *ctx){
bpf_probe_read_kernel(p_rq, sizeof(struct rq), (void *)PT_REGS_PARM1(ctx));
//使用bpf_probe_read_kernel函数将内核空间中的数据复制到p_rq所指向的内存区域中，以便后续对该数据进行访问和操作。
u64 val = p_rq->nr_running;
bpf_map_update_elem(&countMap,&key,&val,BPF_ANY);
bpf_map_update_elem(&runqlen,&key,&val,BPF_ANY);
return 0;
}

Expand Down Expand Up @@ -313,3 +393,30 @@ int trace_irq_handler_exit(struct __irq_info *info) {
}


//tracepoint:power_cpu_idle 表征了CPU进入IDLE的状态，比较准确
SEC("tracepoint/power/cpu_idle")
int trace_cpu_idle(struct idleStruct *pIDLE) {
u64 delta, time = bpf_ktime_get_ns();
u32 key = pIDLE->cpu_id;
// 按cpuid记录空闲的开始，这十分重要，因为IDLE-0进程可同时运行在两个核上

if (pIDLE->state == -1) {
// 结束idle
u64 *valp = bpf_map_lookup_elem(&idleStart,&key);
if (valp && *valp != 0) {
/*找到初始idle时间*/
delta = time - *valp;//持续空闲时间=当前时间-进入空闲时间；
key = 0;
valp = bpf_map_lookup_elem(&idleLastTime,&key);
if (valp) *valp += delta;//找到持续空闲时间,持续空闲时间更新;
else bpf_map_update_elem(&idleLastTime,&key,&delta,BPF_ANY);//初次记录持续空闲时间;
}
} else {
// 开始idle
u64 val = time;
bpf_map_update_elem(&idleStart,&key,&time,BPF_ANY);
}
return 0;
}


Loading

0 comments on commit 6820581

Please sign in to comment.