writable contexts for bpf raw tracepoints
Subsystems (12)
v5 cover writable contexts for bpf raw tracepoints
This adds an opt-in interface for tracepoints to expose a writable context to BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE programs that are attached, while supporting read-only access from existing BPF_PROG_TYPE_RAW_TRACEPOINT programs, as well as from non-BPF-based tracepoints.
The initial motivation is to support tracing that can be observed from the remote end of an NBD socket, e.g. by adding flags to the struct nbd_request header. Earlier attempts included adding an NBD-specific tracepoint fd, but in code review, I was recommended to implement it more generically -- as a result, this patchset is far simpler than my initial try.
v4->v5:
* rebased onto bpf-next/master and fixed merge conflicts
* "tools: sync bpf.h" also syncs comments that have previously changed in bpf-next
v3->v4:
* fixed a silly copy/paste typo in include/trace/events/bpf_test_run.h
(_TRACE_NBD_H -> _TRACE_BPF_TEST_RUN_H)
* fixed incorrect/misleading wording in patch 1's commit message, since the pointer cannot be directly dereferenced in a BPF_PROG_TYPE_RAW_TRACEPOINT
* cleaned up the error message wording if the prog_tests fail
* Addressed feedback from Yonghong
* reject non-pointer-sized accesses to the buffer pointer
* use sizeof(struct nbd_request) as one-byte-past-the-end in raw_tp_writable_reject_nbd_invalid.c
* use BPF_MOV64_IMM instead of BPF_LD_IMM64
v2->v3:
* Andrew addressed Josef's comments:
* C-style commenting in nbd.c
* Collapsed identical events into a single DECLARE_EVENT_CLASS. This saves about 2kB of kernel text
v1->v2:
* add selftests
* sync tools/include/uapi/linux/bpf.h
* reject variable offset into the buffer
* add string representation of PTR_TO_TP_BUFFER to reg_type_str
Andrew Hall (1):
nbd: add tracepoints for send/receive timing
Matt Mullins (4):
bpf: add writable context for raw tracepoints
nbd: trace sending nbd requests
tools: sync bpf.h
selftests: bpf: test writable buffers in raw tps
MAINTAINERS | 1 + drivers/block/nbd.c | 13 +++ include/linux/bpf.h | 2 + include/linux/bpf_types.h | 1 + include/linux/tracepoint-defs.h | 1 + include/trace/bpf_probe.h | 27 ++++- include/trace/events/bpf_test_run.h | 50 ++++++++ include/trace/events/nbd.h | 107 ++++++++++++++++++ include/uapi/linux/bpf.h | 1 + kernel/bpf/syscall.c | 8 +- kernel/bpf/verifier.c | 31 +++++ kernel/trace/bpf_trace.c | 24 ++++ net/bpf/test_run.c | 4 + tools/include/uapi/linux/bpf.h | 10 +- tools/lib/bpf/libbpf.c | 1 + tools/lib/bpf/libbpf_probes.c | 1 + .../raw_tp_writable_reject_nbd_invalid.c | 42 +++++++ .../bpf/prog_tests/raw_tp_writable_test_run.c | 80 +++++++++++++ .../selftests/bpf/verifier/raw_tp_writable.c | 34 ++++++ 19 files changed, 433 insertions(+), 5 deletions(-) create mode 100644 include/trace/events/bpf_test_run.h create mode 100644 include/trace/events/nbd.h create mode 100644 tools/testing/selftests/bpf/prog_tests/raw_tp_writable_reject_nbd_invalid.c create mode 100644 tools/testing/selftests/bpf/prog_tests/raw_tp_writable_test_run.c create mode 100644 tools/testing/selftests/bpf/verifier/raw_tp_writable.c
Signature · 1 line
2.17.1
PATCH v5 1/5 bpf: add writable context for raw tracepoints
This is an opt-in interface that allows a tracepoint to provide a safe buffer that can be written from a BPF_PROG_TYPE_RAW_TRACEPOINT program. The size of the buffer must be a compile-time constant, and is checked before allowing a BPF program to attach to a tracepoint that uses this feature.
The pointer to this buffer will be the first argument of tracepoints that opt in; the pointer is valid and can be bpf_probe_read() by both BPF_PROG_TYPE_RAW_TRACEPOINT and BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE programs that attach to such a tracepoint, but the buffer to which it points may only be written by the latter.
--- include/linux/bpf.h | 2 ++ include/linux/bpf_types.h | 1 + include/linux/tracepoint-defs.h | 1 + include/trace/bpf_probe.h | 27 +++++++++++++++++++++++++-- include/uapi/linux/bpf.h | 1 + kernel/bpf/syscall.c | 8 ++++++-- kernel/bpf/verifier.c | 31 +++++++++++++++++++++++++++++++ kernel/trace/bpf_trace.c | 24 ++++++++++++++++++++++++ 8 files changed, 91 insertions(+), 4 deletions(-)
include/linux/bpf.h +2 -0
@@ -272,6 +272,7 @@ enum bpf_reg_type {PTR_TO_SOCK_COMMON_OR_NULL ,/* reg points to sock_common or NULL */ PTR_TO_TCP_SOCK ,/* reg points to struct tcp_sock */ PTR_TO_TCP_SOCK_OR_NULL ,/* reg points to struct tcp_sock or NULL */ + PTR_TO_TP_BUFFER, /* reg points to a writable raw tp's buffer */ };/* The information passed from prog-specific *_is_valid_access
@@ -361,6 +362,7 @@ struct bpf_prog_aux {u32 used_map_cnt;u32 max_ctx_offset;u32 max_pkt_offset; +u32 max_tp_access;u32 stack_depth;u32 id;u32 func_cnt;/* used by non-func prog as the number of func progs */
include/linux/bpf_types.h +1 -0
@@ -25,6 +25,7 @@ BPF_PROG_TYPE(BPF_PROG_TYPE_KPROBE, kprobe) BPF_PROG_TYPE(BPF_PROG_TYPE_TRACEPOINT, tracepoint) BPF_PROG_TYPE(BPF_PROG_TYPE_PERF_EVENT, perf_event) BPF_PROG_TYPE(BPF_PROG_TYPE_RAW_TRACEPOINT, raw_tracepoint) +BPF_PROG_TYPE(BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, raw_tracepoint_writable)#endif #ifdef CONFIG_CGROUP_BPF BPF_PROG_TYPE (BPF_PROG_TYPE_CGROUP_DEVICE , cg_dev)
include/linux/tracepoint-defs.h +1 -0
@@ -45,6 +45,7 @@ struct bpf_raw_event_map {struct tracepoint *tp;void *bpf_func;u32 num_args; + u32 writable_size; }__aligned(32) ;#endif
include/trace/bpf_probe.h +25 -2
@@ -69,8 +69,7 @@ __bpf_trace_##call(void *__data, proto) \* to make sure that if the tracepoint handling changes, the * bpf probe will fail to compile unless it too is updated. */ -#undef DEFINE_EVENT -#define DEFINE_EVENT (template, call, proto, args) \ +#define __DEFINE_EVENT (template, call, proto, args, size) \ static inline void bpf_test_probe_##call(void) \ { \ check_trace_callback_type_##call(__bpf_trace_##template); \
@@ -81,12 +80,36 @@ __bpf_trace_tp_map_##call = { \ .tp = &__tracepoint_##call, \ .bpf_func = (void *)__bpf_trace_##template, \ .num_args = COUNT_ARGS(args), \ + .writable_size = size, \ }; +#define FIRST (x, ...) x + +#undef DEFINE_EVENT_WRITABLE +#define DEFINE_EVENT_WRITABLE (template, call, proto, args, size) \ +static inline void bpf_test_buffer_##call(void) \ +{ \ +/* BUILD_BUG_ON() is ignored if the code is completely eliminated, but \ +* BUILD_BUG_ON_ZERO() uses a different mechanism that is not \ +* dead-code-eliminated. \ +*/ \ +FIRST (proto); \ + (void )BUILD_BUG_ON_ZERO (size !=sizeof (*FIRST (args))); \ +} \ +__DEFINE_EVENT(template, call, PARAMS(proto), PARAMS(args), size) + +#undef DEFINE_EVENT +#define DEFINE_EVENT (template, call, proto, args) \ + __DEFINE_EVENT(template, call, PARAMS(proto), PARAMS(args), 0)#undef DEFINE_EVENT_PRINT#define DEFINE_EVENT_PRINT (template, name, proto, args, print) \ DEFINE_EVENT(template, name, PARAMS(proto), PARAMS(args))#include TRACE_INCLUDE (TRACE_INCLUDE_FILE ) + +#undef DEFINE_EVENT_WRITABLE +#undef __DEFINE_EVENT +#undef FIRST +#endif /* CONFIG_BPF_EVENTS */
include/uapi/linux/bpf.h +1 -0
@@ -168,6 +168,7 @@ enum bpf_prog_type {BPF_PROG_TYPE_SK_REUSEPORT ,BPF_PROG_TYPE_FLOW_DISSECTOR ,BPF_PROG_TYPE_CGROUP_SYSCTL , + BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, }; enum bpf_attach_type {
kernel/bpf/syscall.c +6 -2
@@ -1789,12 +1789,16 @@ static int bpf_raw_tracepoint_open(const union bpf_attr *attr) } raw_tp->btp = btp; - prog =bpf_prog_get_type (attr->raw_tracepoint.prog_fd, -BPF_PROG_TYPE_RAW_TRACEPOINT ); + prog =bpf_prog_get (attr->raw_tracepoint.prog_fd);if (IS_ERR (prog)) { err =PTR_ERR (prog); goto out_free_tp; } +if (prog->type !=BPF_PROG_TYPE_RAW_TRACEPOINT && + prog->type !=BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE ) { + err = -EINVAL ; + goto out_put_prog; + } err =bpf_probe_register (raw_tp->btp, prog);if (err)
kernel/bpf/verifier.c +31 -0
@@ -405,6 +405,7 @@ static const char * const reg_type_str[] = { [PTR_TO_SOCK_COMMON_OR_NULL] = "sock_common_or_null", [PTR_TO_TCP_SOCK] = "tcp_sock", [PTR_TO_TCP_SOCK_OR_NULL] = "tcp_sock_or_null", + [PTR_TO_TP_BUFFER] = "tp_buffer", };static char slot_type_char[] = {
@@ -1993,6 +1994,32 @@ static int check_ctx_reg(struct bpf_verifier_env *env,return 0 ; } +static int check_tp_buffer_access (struct bpf_verifier_env *env, +const struct bpf_reg_state *reg, +int regno,int off,int size) +{ +if (off <0 ) { +verbose (env, +"R%d invalid tracepoint buffer access: off=%d, size=%d" , + regno, off, size); +return -EACCES ; + } +if (!tnum_is_const (reg->var_off) || reg->var_off.value) { +char tn_buf[48 ]; + +tnum_strn (tn_buf,sizeof (tn_buf), reg->var_off); +verbose (env, +"R%d invalid variable buffer offset: off=%d, var_off=%s" , + regno, off, tn_buf); +return -EACCES ; + } +if (off + size > env->prog->aux->max_tp_access) + env->prog->aux->max_tp_access = off + size; + +return 0 ; +} + +/* truncate register to smaller size (in bytes) * must be called with size < BPF_REG_SIZE */
@@ -2137,6 +2164,10 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn err =check_sock_access (env, insn_idx, regno, off, size, t);if (!err && value_regno >=0 )mark_reg_unknown (env, regs, value_regno); + }else if (reg->type ==PTR_TO_TP_BUFFER ) { + err =check_tp_buffer_access (env, reg, regno, off, size); +if (!err && t ==BPF_READ && value_regno >=0 ) +mark_reg_unknown (env, regs, value_regno); }else {verbose (env,"R%d invalid mem access '%s'\n" , regno, reg_type_str[reg->type]);
kernel/trace/bpf_trace.c +24 -0
@@ -915,6 +915,27 @@ const struct bpf_verifier_ops raw_tracepoint_verifier_ops = {const struct bpf_prog_ops raw_tracepoint_prog_ops = { }; +static bool raw_tp_writable_prog_is_valid_access (int off,int size, +enum bpf_access_type type, +const struct bpf_prog *prog, +struct bpf_insn_access_aux *info) +{ +if (off ==0 ) { +if (size !=sizeof (u64) || type !=BPF_READ ) +return false; + info->reg_type =PTR_TO_TP_BUFFER ; + } +return raw_tp_prog_is_valid_access (off, size, type, prog, info); +} + +const struct bpf_verifier_ops raw_tracepoint_writable_verifier_ops = { + .get_func_proto = raw_tp_prog_func_proto, + .is_valid_access = raw_tp_writable_prog_is_valid_access, +}; + +const struct bpf_prog_ops raw_tracepoint_writable_prog_ops = { +}; +static bool pe_prog_is_valid_access (int off,int size,enum bpf_access_type type,const struct bpf_prog *prog,struct bpf_insn_access_aux *info)
@@ -1204,6 +1225,9 @@ static int __bpf_probe_register(struct bpf_raw_event_map *btp, struct bpf_prog *if (prog->aux->max_ctx_offset > btp->num_args *sizeof (u64))return -EINVAL ; +if (prog->aux->max_tp_access > btp->writable_size) +return -EINVAL ; +return tracepoint_probe_register (tp, (void *)btp->bpf_func, prog); }
Signature · 1 line
2.17.1
PATCH v5 2/5 nbd: trace sending nbd requests
In reply to Matt MullinsThis adds a tracepoint that can both observe the nbd request being sent to the server, as well as modify that request , e.g., setting a flag in the request that will cause the server to collect detailed tracing data.
The struct request * being handled is included to permit correlation with the block tracepoints.
--- MAINTAINERS | 1 + drivers/block/nbd.c | 5 ++++ include/trace/events/nbd.h | 56 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 62 insertions(+) create mode 100644 include/trace/events/nbd.h
MAINTAINERS +1 -0
@@ -10741,6 +10741,7 @@ L: [email protected] L: [email protected] F: Documentation/blockdev/nbd.txt F: drivers/block/nbd.c +F: include/trace/events/nbd.h F: include/uapi/linux/nbd.h NETWORK DROP MONITOR
drivers/block/nbd.c +5 -0
@@ -44,6 +44,9 @@#include <linux/nbd-netlink.h> #include <net/genetlink.h> +#define CREATE_TRACE_POINTS +#include <trace/events/nbd.h> +static DEFINE_IDR (nbd_index_idr);static DEFINE_MUTEX (nbd_index_mutex);static int nbd_total_devices =0 ;
@@ -526,6 +529,8 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index) handle =nbd_cmd_handle (cmd);memcpy (request.handle, &handle,sizeof (handle)); +trace_nbd_send_request (&request, nbd->index,blk_mq_rq_from_pdu (cmd)); +dev_dbg (nbd_to_dev (nbd),"request %p: sending control (%s@%llu,%uB)\n" , req,nbdcmd_to_ascii (type), (unsigned long long )blk_rq_pos (req) <<9 ,blk_rq_bytes (req));
include/trace/events/nbd.h new file +56 -0
56 lines not on this page · raw message
Signature · 1 line
2.17.1
PATCH v5 3/5 nbd: add tracepoints for send/receive timing
In reply to Matt MullinsFrom: Andrew Hall <[email protected]>
This adds four tracepoints to nbd, enabling separate tracing of payload and header sending/receipt.
In the send path for headers that have already been sent, we also explicitly initialize the handle so it can be referenced by the later tracepoint.
--- drivers/block/nbd.c | 8 ++++++ include/trace/events/nbd.h | 51 ++++++++++++++++++++++++++++++++++++++ 2 files changed, 59 insertions(+)
drivers/block/nbd.c +8 -0
@@ -513,6 +513,10 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)if (sent) {if (sent >=sizeof (request)) { skip = sent -sizeof (request); + +/* initialize handle for tracing purposes */ + handle =nbd_cmd_handle (cmd); + goto send_pages; }iov_iter_advance (&from, sent);
@@ -536,6 +540,7 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index) (unsigned long long)blk_rq_pos(req) << 9, blk_rq_bytes(req)); result =sock_xmit (nbd, index,1 , &from, (type ==NBD_CMD_WRITE ) ?MSG_MORE :0 , &sent); +trace_nbd_header_sent (req, handle);if (result <=0 ) {if (was_interrupted (result)) {/* If we havne't sent anything we can just return BUSY,
@@ -608,6 +613,7 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index) bio = next; } out: +trace_nbd_payload_sent (req, handle); nsock->pending =NULL ; nsock->sent =0 ;return 0 ;
@@ -655,6 +661,7 @@ static struct nbd_cmd *nbd_read_stat(struct nbd_device *nbd, int index) tag, req);return ERR_PTR (-ENOENT ); } +trace_nbd_header_received (req, handle); cmd =blk_mq_rq_to_pdu (req);mutex_lock (&cmd->lock);
@@ -708,6 +715,7 @@ static struct nbd_cmd *nbd_read_stat(struct nbd_device *nbd, int index) } } out: +trace_nbd_payload_received (req, handle);mutex_unlock (&cmd->lock);return ret ?ERR_PTR (ret) : cmd; }
include/trace/events/nbd.h +51 -0
57 lines not on this page · raw message
Signature · 1 line
2.17.1
PATCH v5 4/5 tools: sync bpf.h
In reply to Matt MullinsThis adds BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, and fixes up the
error: enumeration value ‘BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE’ not handled in switch [-Werror=switch-enum]
build errors it would otherwise cause in libbpf.
--- tools/include/uapi/linux/bpf.h | 10 +++++++++- tools/lib/bpf/libbpf.c | 1 + tools/lib/bpf/libbpf_probes.c | 1 + 3 files changed, 11 insertions(+), 1 deletion(-)
tools/include/uapi/linux/bpf.h +9 -1
@@ -168,6 +168,7 @@ enum bpf_prog_type {BPF_PROG_TYPE_SK_REUSEPORT ,BPF_PROG_TYPE_FLOW_DISSECTOR ,BPF_PROG_TYPE_CGROUP_SYSCTL , + BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, }; enum bpf_attach_type {
@@ -1737,12 +1738,19 @@ union bpf_attr {* error if an eBPF program tries to set a callback that is not * supported in the current kernel. * -* The supported callback values that *argval* can combine are: +* *argval* is a flag array which can combine these flags: * * * **BPF_SOCK_OPS_RTO_CB_FLAG** (retransmission time out) * * **BPF_SOCK_OPS_RETRANS_CB_FLAG** (retransmission) * * **BPF_SOCK_OPS_STATE_CB_FLAG** (TCP state change) * +* Therefore, this function can be used to clear a callback flag by +* setting the appropriate bit to zero. e.g. to disable the RTO +* callback: +* +* **bpf_sock_ops_cb_flags_set(bpf_sock,** +* **bpf_sock->bpf_sock_ops_cb_flags & ~BPF_SOCK_OPS_RTO_CB_FLAG)** +* * Here are some examples of where one could call such eBPF * program: *
tools/lib/bpf/libbpf.c +1 -0
@@ -2136,6 +2136,7 @@ static bool bpf_prog_type__needs_kver(enum bpf_prog_type type)case BPF_PROG_TYPE_UNSPEC :case BPF_PROG_TYPE_TRACEPOINT :case BPF_PROG_TYPE_RAW_TRACEPOINT : +case BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE :case BPF_PROG_TYPE_PERF_EVENT :case BPF_PROG_TYPE_CGROUP_SYSCTL :return false;
tools/lib/bpf/libbpf_probes.c +1 -0
@@ -93,6 +93,7 @@ probe_load(enum bpf_prog_type prog_type, const struct bpf_insn *insns,case BPF_PROG_TYPE_CGROUP_DEVICE :case BPF_PROG_TYPE_SK_MSG :case BPF_PROG_TYPE_RAW_TRACEPOINT : +case BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE :case BPF_PROG_TYPE_LWT_SEG6LOCAL :case BPF_PROG_TYPE_LIRC_MODE2 :case BPF_PROG_TYPE_SK_REUSEPORT :
Signature · 1 line
2.17.1
PATCH v5 5/5 selftests: bpf: test writable buffers in raw tps
In reply to Matt MullinsThis tests that:
* a BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE cannot be attached if it uses either:
* a variable offset to the tracepoint buffer, or
* an offset beyond the size of the tracepoint buffer
* a tracer can modify the buffer provided when attached to a writable tracepoint in bpf_prog_test_run
--- include/trace/events/bpf_test_run.h | 50 ++++++++++++ net/bpf/test_run.c | 4 + .../raw_tp_writable_reject_nbd_invalid.c | 42 ++++++++++ .../bpf/prog_tests/raw_tp_writable_test_run.c | 80 +++++++++++++++++++ .../selftests/bpf/verifier/raw_tp_writable.c | 34 ++++++++ 5 files changed, 210 insertions(+) create mode 100644 include/trace/events/bpf_test_run.h create mode 100644 tools/testing/selftests/bpf/prog_tests/raw_tp_writable_reject_nbd_invalid.c create mode 100644 tools/testing/selftests/bpf/prog_tests/raw_tp_writable_test_run.c create mode 100644 tools/testing/selftests/bpf/verifier/raw_tp_writable.c
include/trace/events/bpf_test_run.h new file +50 -0
@@ -0,0 +1,50 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM bpf_test_run + +#if !defined(_TRACE_BPF_TEST_RUN_H) || defined(TRACE_HEADER_MULTI_READ ) +#define _TRACE_BPF_TEST_RUN_H + +#include <linux/tracepoint.h> + +DECLARE_EVENT_CLASS (bpf_test_finish, + +TP_PROTO (int *err), + +TP_ARGS (err), + +TP_STRUCT__entry ( +__field (int, err) + ), + +TP_fast_assign ( + __entry->err = *err; + ), + +TP_printk ("bpf_test_finish with err=%d" , __entry->err) +); + +#ifdef DEFINE_EVENT_WRITABLE +#undef BPF_TEST_RUN_DEFINE_EVENT +#define BPF_TEST_RUN_DEFINE_EVENT (template, call, proto, args, size) \ + DEFINE_EVENT_WRITABLE(template, call, PARAMS(proto), \ + PARAMS(args), size) +#else +#undef BPF_TEST_RUN_DEFINE_EVENT +#define BPF_TEST_RUN_DEFINE_EVENT (template, call, proto, args, size) \ + DEFINE_EVENT(template, call, PARAMS(proto), PARAMS(args)) +#endif + +BPF_TEST_RUN_DEFINE_EVENT (bpf_test_finish, bpf_test_finish, + +TP_PROTO (int *err), + +TP_ARGS (err), + +sizeof (int ) +); + +#endif + +/* This part must be outside protection */ +#include <trace/define_trace.h>
net/bpf/test_run.c +4 -0
@@ -13,6 +13,9 @@#include <net/sock.h> #include <net/tcp.h> +#define CREATE_TRACE_POINTS +#include <trace/events/bpf_test_run.h> +static int bpf_test_run (struct bpf_prog *prog,void *ctx,u32 repeat,u32 *retval,u32 *time) {
@@ -100,6 +103,7 @@ static int bpf_test_finish(const union bpf_attr *kattr,if (err != -ENOSPC ) err =0 ; out: +trace_bpf_test_finish (&err);return err; }
tools/testing/selftests/bpf/prog_tests/raw_tp_writable_reject_nbd_invalid.c new file +42 -0
42 lines not on this page · raw message
tools/testing/selftests/bpf/prog_tests/raw_tp_writable_test_run.c new file +80 -0
80 lines not on this page · raw message
tools/testing/selftests/bpf/verifier/raw_tp_writable.c new file +34 -0
@@ -0,0 +1,34 @@ +{ + "raw_tracepoint_writable: reject variable offset", + .insns = { +/* r6 is our tp buffer */ +BPF_LDX_MEM (BPF_DW ,BPF_REG_6 ,BPF_REG_1 ,0 ), + +BPF_LD_MAP_FD (BPF_REG_1 ,0 ), +/* move the key (== 0) to r10-8 */ +BPF_MOV32_IMM (BPF_REG_0 ,0 ), +BPF_MOV64_REG (BPF_REG_2 ,BPF_REG_10 ), +BPF_ALU64_IMM (BPF_ADD ,BPF_REG_2 ,-8 ), +BPF_STX_MEM (BPF_DW ,BPF_REG_2 ,BPF_REG_0 ,0 ), +/* lookup in the map */ +BPF_RAW_INSN (BPF_JMP |BPF_CALL ,0 ,0 ,0 , + BPF_FUNC_map_lookup_elem), + +/* exit clean if null */ +BPF_JMP_IMM (BPF_JNE ,BPF_REG_0 ,0 ,1 ), +BPF_EXIT_INSN (), + +/* shift the buffer pointer to a variable location */ +BPF_LDX_MEM (BPF_W ,BPF_REG_0 ,BPF_REG_0 ,0 ), +BPF_ALU64_REG (BPF_ADD ,BPF_REG_6 ,BPF_REG_0 ), +/* clobber whatever's there */ +BPF_MOV64_IMM (BPF_REG_7 ,4242 ), +BPF_STX_MEM (BPF_DW ,BPF_REG_6 ,BPF_REG_7 ,0 ), + +BPF_MOV64_IMM (BPF_REG_0 ,0 ), + BPF_EXIT_INSN(), + }, + .fixup_map_hash_8b = { 1, }, + .prog_type =BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE , + .errstr = "R6 invalid variable buffer offset: off=0, var_off=(0x0; 0xffffffff)", +},
Signature · 1 line
2.17.1
Alexei Starovoitov
AppliedQuoted 12 lines · Matt Mullins
This adds an opt-in interface for tracepoints to expose a writable context to
BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLEprograms that are attached, while supporting read-only access from existingBPF_PROG_TYPE_RAW_TRACEPOINTprograms, as well as from non-BPF-based tracepoints.The initial motivation is to support tracing that can be observed from the remote end of an NBD socket, e.g. by adding flags to the struct
nbd_requestheader. Earlier attempts included adding an NBD-specific tracepoint fd, but in code review, I was recommended to implement it more generically -- as a result, this patchset is far simpler than my initial try.
Quoted 4 lines
v4->v5:
* rebased onto bpf-next/master and fixed merge conflicts
* "tools: syncbpf.h" also syncs comments that have previously changed in bpf-next
Applied. Thanks!