fix(bpf): cleanup initialising structs and nested ifs (sustainable-co…

…mputing-io#1444)

This commit changes all declarae and then initialise code of form below

```
struct foobar x;
x.a = a;
x.b = b;
```

to
```
struct foobar x = { .a = a, .b = b, };
```
The difference in the code produced is as follows

```
;  new_pid_key.pid = cur_pid;                             | pid_time_t new_pid_key = {.pid = cur_pid };
  0: 63 6a fc ff 00 00 00 00	*(u32 *)(r10 - 0x4) = r6  |  0:	79 a1 68 ff 00 00 00 00	r1 = *(u64 *)(r10 - 0x98)
  1: bf a2 00 00 00 00 00 00	r2 = r10                  |  1:	63 1a fc ff 00 00 00 00	*(u32 *)(r10 - 0x4) = r1
  2: 07 02 00 00 fc ff ff ff	r2 += -0x4                |  2:	bf a2 00 00 00 00 00 00	r2 = r10
  3: bf a3 00 00 00 00 00 00	r3 = r10                  |  3:	07 02 00 00 f0 ff ff ff	r2 += -0x10
  4: 07 03 00 00 70 ff ff ff	r3 += -0x90

```

Modifies calc_delta to remove unnecessary pointer deferencing of second
arg - `val`

Signed-off-by: Sunil Thaha <[email protected]>
(cherry picked from commit ac43e26)

bpfassets/libbpf/src/kepler.bpf.c

@@ -96,11 +96,10 @@ int counter_sched_switch = 0;
 static inline u64 get_on_cpu_time(u32 cur_pid, u32 prev_pid, u64 cur_ts)
 {
  u64 cpu_time = 0;
- u64 *prev_ts;
- pid_time_t prev_pid_key, new_pid_key = {};
+ pid_time_t prev_pid_key = { .pid = prev_pid };
+ pid_time_t new_pid_key = { .pid = cur_pid };
 
- prev_pid_key.pid = prev_pid;
- prev_ts = bpf_map_lookup_elem(&pid_time, &prev_pid_key);
+ u64 *prev_ts = bpf_map_lookup_elem(&pid_time, &prev_pid_key);
  if (prev_ts) {
  // Probably a clock issue where the recorded on-CPU event had a
  // timestamp later than the recorded off-CPU event, or vice versa.
@@ -109,19 +108,18 @@ static inline u64 get_on_cpu_time(u32 cur_pid, u32 prev_pid, u64 cur_ts)
  bpf_map_delete_elem(&pid_time, &prev_pid_key);
  }
  }
- new_pid_key.pid = cur_pid;
+
  bpf_map_update_elem(&pid_time, &new_pid_key, &cur_ts, BPF_NOEXIST);
  return cpu_time;
 }
 
-static inline u64 calc_delta(u64 *prev_val, u64 *val)
+static inline u64 calc_delta(u64 *prev_val, u64 val)
 {
  u64 delta = 0;
-
- if (prev_val) {
- if (*val > *prev_val)
- delta = *val - *prev_val;
+ if (prev_val && val > *prev_val) {
+ delta = val - *prev_val;
  }
+
  return delta;
 }
 
@@ -138,7 +136,7 @@ static inline u64 get_on_cpu_cycles(u32 *cpu_id)
 
  val = c.counter;
  prev_val = bpf_map_lookup_elem(&cpu_cycles, cpu_id);
- delta = calc_delta(prev_val, &val);
+ delta = calc_delta(prev_val, val);
  bpf_map_update_elem(&cpu_cycles, cpu_id, &val, BPF_ANY);
 
  return delta;
@@ -152,12 +150,12 @@ static inline u64 get_on_cpu_instr(u32 *cpu_id)
 
  error = bpf_perf_event_read_value(
  &cpu_instructions_event_reader, *cpu_id, &c, sizeof(c));
- if (error) {
+ if (error)
  return 0;
- }
+
  val = c.counter;
  prev_val = bpf_map_lookup_elem(&cpu_instructions, cpu_id);
- delta = calc_delta(prev_val, &val);
+ delta = calc_delta(prev_val, val);
  bpf_map_update_elem(&cpu_instructions, cpu_id, &val, BPF_ANY);
 
  return delta;
@@ -176,7 +174,7 @@ static inline u64 get_on_cpu_cache_miss(u32 *cpu_id)
  }
  val = c.counter;
  prev_val = bpf_map_lookup_elem(&cache_miss, cpu_id);
- delta = calc_delta(prev_val, &val);
+ delta = calc_delta(prev_val, val);
  bpf_map_update_elem(&cache_miss, cpu_id, &val, BPF_ANY);
 
  return delta;
@@ -241,10 +239,11 @@ int kepler_sched_switch_trace(struct sched_switch_info *ctx)
  // create new process metrics
  cur_pid_metrics = bpf_map_lookup_elem(&processes, &cur_pid);
  if (!cur_pid_metrics) {
- process_metrics_t new_process = {};
- new_process.pid = cur_pid;
- new_process.tgid = tgid;
- new_process.cgroup_id = cgroup_id;
+ process_metrics_t new_process = {
+ .pid = cur_pid,
+ .tgid = tgid,
+ .cgroup_id = cgroup_id,
+ };
  bpf_get_current_comm(&new_process.comm, sizeof(new_process.comm));
  bpf_map_update_elem(
  &processes, &cur_pid, &new_process, BPF_NOEXIST);