[bpf-next] bpf: change syscall_nr type to int in struct syscall_tp_t

Commit Message

Artem Savkov Oct. 13, 2023, 5:42 a.m. UTC

linux-rt-devel tree contains a patch (b1773eac3f29c ("sched: Add support
for lazy preemption")) that adds an extra member to struct trace_entry.
This causes the offset of args field in struct trace_event_raw_sys_enter
be different from the one in struct syscall_trace_enter:

struct trace_event_raw_sys_enter {
        struct trace_entry         ent;                  /*     0    12 */

        /* XXX last struct has 3 bytes of padding */
        /* XXX 4 bytes hole, try to pack */

        long int                   id;                   /*    16     8 */
        long unsigned int          args[6];              /*    24    48 */
        /* --- cacheline 1 boundary (64 bytes) was 8 bytes ago --- */
        char                       __data[];             /*    72     0 */

        /* size: 72, cachelines: 2, members: 4 */
        /* sum members: 68, holes: 1, sum holes: 4 */
        /* paddings: 1, sum paddings: 3 */
        /* last cacheline: 8 bytes */
};

struct syscall_trace_enter {
        struct trace_entry         ent;                  /*     0    12 */

        /* XXX last struct has 3 bytes of padding */

        int                        nr;                   /*    12     4 */
        long unsigned int          args[];               /*    16     0 */

        /* size: 16, cachelines: 1, members: 3 */
        /* paddings: 1, sum paddings: 3 */
        /* last cacheline: 16 bytes */
};

This, in turn, causes perf_event_set_bpf_prog() fail while running bpf
test_profiler testcase because max_ctx_offset is calculated based on the
former struct, while off on the latter:

  10488         if (is_tracepoint || is_syscall_tp) {
  10489                 int off = trace_event_get_offsets(event->tp_event);
  10490
  10491                 if (prog->aux->max_ctx_offset > off)
  10492                         return -EACCES;
  10493         }

What bpf program is actually getting is a pointer to struct
syscall_tp_t, defined in kernel/trace/trace_syscalls.c. This patch fixes
the problem by aligning struct syscall_tp_t with with struct
syscall_trace_(enter|exit) and changing the tests to use these structs
to dereference context.

Signed-off-by: Artem Savkov <[e-mail]>
Acked-by: Steven Rostedt (Google) <[e-mail]>

---
 kernel/trace/trace_syscalls.c                    | 4 ++--
 tools/testing/selftests/bpf/progs/profiler.inc.h | 2 +-
 tools/testing/selftests/bpf/progs/test_vmlinux.c | 4 ++--
 3 files changed, 5 insertions(+), 5 deletions(-)

Comments

[e-mail] Oct. 13, 2023, 7:50 p.m. UTC | #1

Hello:

This patch was applied to bpf/bpf-next.git (master)
by Andrii Nakryiko <[e-mail]>:

On Fri, 13 Oct 2023 07:42:19 +0200 you wrote:
> linux-rt-devel tree contains a patch (b1773eac3f29c ("sched: Add support
> for lazy preemption")) that adds an extra member to struct trace_entry.
> This causes the offset of args field in struct trace_event_raw_sys_enter
> be different from the one in struct syscall_trace_enter:
> 
> struct trace_event_raw_sys_enter {
>         struct trace_entry         ent;                  /*     0    12 */
> 
> [...]

Here is the summary with links:
  - [bpf-next] bpf: change syscall_nr type to int in struct syscall_tp_t
    https://git.kernel.org/bpf/bpf-next/c/ba8ea72388a1

You are awesome, thank you!

Patch

diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c
index de753403cdafb..9c581d6da843a 100644
--- a/kernel/trace/trace_syscalls.c
+++ b/kernel/trace/trace_syscalls.c
@@ -556,7 +556,7 @@  static int perf_call_bpf_enter(struct trace_event_call *call, struct pt_regs *re
 {
 	struct syscall_tp_t {
 		struct trace_entry ent;
-		unsigned long syscall_nr;
+		int syscall_nr;
 		unsigned long args[SYSCALL_DEFINE_MAXARGS];
 	} __aligned(8) param;
 	int i;
@@ -661,7 +661,7 @@  static int perf_call_bpf_exit(struct trace_event_call *call, struct pt_regs *reg
 {
 	struct syscall_tp_t {
 		struct trace_entry ent;
-		unsigned long syscall_nr;
+		int syscall_nr;
 		unsigned long ret;
 	} __aligned(8) param;
 
diff --git a/tools/testing/selftests/bpf/progs/profiler.inc.h b/tools/testing/selftests/bpf/progs/profiler.inc.h
index f799d87e87002..897061930cb76 100644
--- a/tools/testing/selftests/bpf/progs/profiler.inc.h
+++ b/tools/testing/selftests/bpf/progs/profiler.inc.h
@@ -609,7 +609,7 @@  ssize_t BPF_KPROBE(kprobe__proc_sys_write,
 }
 
 SEC("tracepoint/syscalls/sys_enter_kill")
-int tracepoint__syscalls__sys_enter_kill(struct trace_event_raw_sys_enter* ctx)
+int tracepoint__syscalls__sys_enter_kill(struct syscall_trace_enter* ctx)
 {
 	struct bpf_func_stats_ctx stats_ctx;
 
diff --git a/tools/testing/selftests/bpf/progs/test_vmlinux.c b/tools/testing/selftests/bpf/progs/test_vmlinux.c
index 4b8e37f7fd06c..78b23934d9f8f 100644
--- a/tools/testing/selftests/bpf/progs/test_vmlinux.c
+++ b/tools/testing/selftests/bpf/progs/test_vmlinux.c
@@ -16,12 +16,12 @@  bool kprobe_called = false;
 bool fentry_called = false;
 
 SEC("tp/syscalls/sys_enter_nanosleep")
-int handle__tp(struct trace_event_raw_sys_enter *args)
+int handle__tp(struct syscall_trace_enter *args)
 {
 	struct __kernel_timespec *ts;
 	long tv_nsec;
 
-	if (args->id != __NR_nanosleep)
+	if (args->nr != __NR_nanosleep)
 		return 0;
 
 	ts = (void *)args->args[0];