1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282
|
// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (c) 2020, 2025, Oracle and/or its affiliates.
*/
#include <linux/bpf.h>
#include <stddef.h>
#include <stdint.h>
#include <bpf/bpf_helpers.h>
#include <dtrace/conf.h>
#include <dtrace/dif_defines.h>
#include <dtrace/faults_defines.h>
#include <dt_bpf_maps.h>
#include <dt_dctx.h>
#include <dt_state.h>
#include "probe_error.h"
#ifndef noinline
# define noinline __attribute__((noinline))
#endif
extern struct bpf_map_def cpuinfo;
extern struct bpf_map_def probes;
extern struct bpf_map_def state;
extern struct bpf_map_def usdt_names;
extern uint64_t BOOTTM;
extern uint64_t NPROBES;
extern uint64_t PC;
extern uint64_t STBSZ;
extern uint64_t STKSIZ;
extern uint64_t STACK_OFF;
extern uint64_t STACK_SKIP;
extern uint64_t TASK_COMM;
extern uint64_t TASK_REAL_PARENT;
extern uint64_t TASK_TGID;
#define error(dctx, fault, illval) \
({ \
dt_probe_error((dctx), (uint64_t)&PC, (fault), (illval)); \
-1; \
})
noinline uint64_t dt_bvar_args(const dt_dctx_t *dctx, uint32_t idx)
{
dt_mstate_t *mst = dctx->mst;
if (idx >= sizeof(mst->argv) / sizeof(mst->argv[0]))
return error(dctx, DTRACEFLT_ILLOP, 0);
return mst->argv[idx];
}
noinline uint64_t dt_bvar_caller(const dt_dctx_t *dctx)
{
uint64_t buf[3] = { 0, };
volatile uint64_t
skip = (uint64_t)(&STACK_SKIP);
if (bpf_get_stack(dctx->ctx, buf, sizeof(buf),
skip & BPF_F_SKIP_FIELD_MASK) < 0)
return 0;
/* If we had to skip any frames, account for the dt_bvar_caller() frame. */
if (skip)
return buf[2];
return buf[1];
}
noinline uint64_t dt_bvar_curcpu(const dt_dctx_t *dctx)
{
uint32_t key = bpf_get_smp_processor_id();
void *val = bpf_map_lookup_elem(&cpuinfo, &key);
if (val == NULL) {
/*
* Typically, we would use 'return error(...);' but
* that confuses the verifier because it returns -1.
* So, instead, we explicitly return 0.
*/
error(dctx, DTRACEFLT_ILLOP, 0);
return 0;
}
return (uint64_t)val;
}
noinline uint64_t dt_bvar_curthread(const dt_dctx_t *dctx)
{
return bpf_get_current_task();
}
noinline uint64_t dt_bvar_epid(const dt_dctx_t *dctx)
{
dt_mstate_t *mst = dctx->mst;
return (((uint64_t)mst->prid) << 32) | mst->stid;
}
noinline uint64_t dt_bvar_errno(const dt_dctx_t *dctx)
{
dt_mstate_t *mst = dctx->mst;
return mst->syscall_errno;
}
noinline uint64_t dt_bvar_execname(const dt_dctx_t *dctx)
{
uint64_t ptr;
/* &(current->comm) */
ptr = bpf_get_current_task();
if (ptr == 0)
return error(dctx, DTRACEFLT_BADADDR, ptr);
return (uint64_t)ptr + (uint64_t)&TASK_COMM;
}
noinline uint64_t dt_bvar_gid(const dt_dctx_t *dctx)
{
return bpf_get_current_uid_gid() >> 32;
}
noinline uint64_t dt_bvar_id(const dt_dctx_t *dctx)
{
dt_mstate_t *mst = dctx->mst;
return mst->prid;
}
noinline uint64_t dt_bvar_pid(const dt_dctx_t *dctx)
{
return bpf_get_current_pid_tgid() >> 32;
}
noinline uint64_t dt_bvar_ppid(const dt_dctx_t *dctx)
{
uint64_t ptr;
int32_t val = -1;
/* Chase pointers val = current->real_parent->tgid. */
ptr = bpf_get_current_task();
if (ptr == 0)
return error(dctx, DTRACEFLT_BADADDR, ptr);
if (bpf_probe_read((void *)&ptr, 8,
(const void *)(ptr + (uint64_t)&TASK_REAL_PARENT)))
return error(dctx, DTRACEFLT_BADADDR, ptr + (uint64_t)&TASK_REAL_PARENT);
if (bpf_probe_read((void *)&val, 4,
(const void *)(ptr + (uint64_t)&TASK_TGID)))
return error(dctx, DTRACEFLT_BADADDR, ptr + (uint64_t)&TASK_TGID);
return (uint64_t)val;
}
noinline uint64_t dt_bvar_probedesc(const dt_dctx_t *dctx, uint32_t idx)
{
dt_mstate_t *mst = dctx->mst;
uint32_t key = mst->prid;
if (key < ((uint64_t)&NPROBES)) {
dt_bpf_probe_t *pinfo;
uint64_t off = 0;
pinfo = bpf_map_lookup_elem(&probes, &key);
if (pinfo == NULL)
return (uint64_t)dctx->strtab;
switch (idx) {
case DIF_VAR_PROBEPROV:
off = pinfo->prv;
break;
case DIF_VAR_PROBEMOD:
off = pinfo->mod;
break;
case DIF_VAR_PROBEFUNC:
off = pinfo->fun;
break;
case DIF_VAR_PROBENAME:
off = pinfo->prb;
}
if (off > (uint64_t)&STBSZ)
return (uint64_t)dctx->strtab;
return (uint64_t)(dctx->strtab + off);
} else {
char *s;
s = bpf_map_lookup_elem(&usdt_names, &key);
if (s == NULL)
return (uint64_t)dctx->strtab;
switch (idx) {
case DIF_VAR_PROBENAME:
s += DTRACE_FUNCNAMELEN;
case DIF_VAR_PROBEFUNC:
s += DTRACE_MODNAMELEN;
case DIF_VAR_PROBEMOD:
s += DTRACE_PROVNAMELEN;
case DIF_VAR_PROBEPROV:
}
return (uint64_t)s;
}
}
noinline uint64_t dt_bvar_stackdepth(const dt_dctx_t *dctx)
{
uint32_t bufsiz = (uint32_t) (uint64_t) (&STKSIZ);
char *buf = dctx->mem + (uint64_t)(&STACK_OFF);
uint64_t retv;
volatile uint64_t
skip = (uint64_t)(&STACK_SKIP);
retv = bpf_get_stack(dctx->ctx, buf, bufsiz,
skip & BPF_F_SKIP_FIELD_MASK);
if (retv < 0)
return error(dctx, DTRACEFLT_BADSTACK, 0 /* FIXME */);
/*
* While linux/bpf.h does not describe the meaning of bpf_get_stack()'s
* return value outside of its sign, it is presumably the length of the
* copied stack.
*
* If retv==bufsiz, presumably the stack is larger than what we
* can retrieve. But it's also possible that the buffer was exactly
* large enough. So, leave it to the user to interpret the result.
*
* If we had to skip any frames, account for the dt_bvar_stackdepth() frame.
*/
if (skip)
return retv / sizeof(uint64_t) - 1;
return retv / sizeof(uint64_t);
}
noinline uint64_t dt_bvar_tid(const dt_dctx_t *dctx)
{
return bpf_get_current_pid_tgid() & 0x00000000ffffffffUL;
}
noinline uint64_t dt_bvar_timestamp(const dt_dctx_t *dctx)
{
dt_mstate_t *mst = dctx->mst;
if (mst->tstamp == 0)
mst->tstamp = bpf_ktime_get_ns();
return mst->tstamp;
}
noinline uint64_t dt_bvar_ucaller(const dt_dctx_t *dctx)
{
uint64_t buf[2] = { 0, };
if (bpf_get_stack(dctx->ctx, buf, sizeof(buf), BPF_F_USER_STACK) < 0)
return 0;
return buf[1];
}
noinline uint64_t dt_bvar_uid(const dt_dctx_t *dctx)
{
return bpf_get_current_uid_gid() & 0x00000000ffffffffUL;
}
noinline uint64_t dt_bvar_ustackdepth(const dt_dctx_t *dctx)
{
uint32_t bufsiz = (uint32_t) (uint64_t) (&STKSIZ);
char *buf = dctx->mem + (uint64_t)(&STACK_OFF);
uint64_t retv;
retv = bpf_get_stack(dctx->ctx, buf, bufsiz, BPF_F_USER_STACK);
if (retv < 0)
return error(dctx, DTRACEFLT_BADSTACK, 0 /* FIXME */);
/* See dt_bvar_stackdepth() above. */
return retv / sizeof(uint64_t);
}
noinline uint64_t dt_bvar_walltimestamp(const dt_dctx_t *dctx)
{
return bpf_ktime_get_ns() + ((uint64_t)&BOOTTM);
}
|