From: Michael Jeanson Date: Mon, 5 Nov 2018 16:35:52 +0000 (-0500) Subject: Fix: signal: Distinguish between kernel_siginfo and siginfo (v4.20) X-Git-Tag: v2.11.0-rc3~11 X-Git-Url: https://git.lttng.org./?a=commitdiff_plain;h=5ee84cfdf32ddba083a7f62b9ffd4e91f8b43c1f;p=lttng-modules.git Fix: signal: Distinguish between kernel_siginfo and siginfo (v4.20) See upstream commit : commit ae7795bc6187a15ec51cf258abae656a625f9980 Author: Eric W. Biederman Date: Tue Sep 25 11:27:20 2018 +0200 signal: Distinguish between kernel_siginfo and siginfo Linus recently observed that if we did not worry about the padding member in struct siginfo it is only about 48 bytes, and 48 bytes is much nicer than 128 bytes for allocating on the stack and copying around in the kernel. The obvious thing of only adding the padding when userspace is including siginfo.h won't work as there are sigframe definitions in the kernel that embed struct siginfo. So split siginfo in two; kernel_siginfo and siginfo. Keeping the traditional name for the userspace definition. While the version that is used internally to the kernel and ultimately will not be padded to 128 bytes is called kernel_siginfo. The definition of struct kernel_siginfo I have put in include/signal_types.h A set of buildtime checks has been added to verify the two structures have the same field offsets. To make it easy to verify the change kernel_siginfo retains the same size as siginfo. The reduction in size comes in a following change. Signed-off-by: Michael Jeanson Signed-off-by: Mathieu Desnoyers --- diff --git a/instrumentation/events/lttng-module/signal.h b/instrumentation/events/lttng-module/signal.h index 68045ce6..6c484ba2 100644 --- a/instrumentation/events/lttng-module/signal.h +++ b/instrumentation/events/lttng-module/signal.h @@ -36,21 +36,24 @@ * SEND_SIG_NOINFO means that si_code is SI_USER, and SEND_SIG_PRIV * means that si_code is SI_KERNEL. */ -#if (LINUX_VERSION_CODE < KERNEL_VERSION(3,4,0)) +#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,20,0)) LTTNG_TRACEPOINT_EVENT(signal_generate, - TP_PROTO(int sig, struct siginfo *info, struct task_struct *task), + TP_PROTO(int sig, struct kernel_siginfo *info, struct task_struct *task, + int group, int result), - TP_ARGS(sig, info, task), + TP_ARGS(sig, info, task, group, result), TP_FIELDS( ctf_integer(int, sig, sig) LTTNG_FIELDS_SIGINFO(info) ctf_array_text(char, comm, task->comm, TASK_COMM_LEN) ctf_integer(pid_t, pid, task->pid) + ctf_integer(int, group, group) + ctf_integer(int, result, result) ) ) -#else +#elif (LINUX_VERSION_CODE >= KERNEL_VERSION(3,4,0)) LTTNG_TRACEPOINT_EVENT(signal_generate, TP_PROTO(int sig, struct siginfo *info, struct task_struct *task, @@ -67,6 +70,20 @@ LTTNG_TRACEPOINT_EVENT(signal_generate, ctf_integer(int, result, result) ) ) +#else +LTTNG_TRACEPOINT_EVENT(signal_generate, + + TP_PROTO(int sig, struct siginfo *info, struct task_struct *task), + + TP_ARGS(sig, info, task), + + TP_FIELDS( + ctf_integer(int, sig, sig) + LTTNG_FIELDS_SIGINFO(info) + ctf_array_text(char, comm, task->comm, TASK_COMM_LEN) + ctf_integer(pid_t, pid, task->pid) + ) +) #endif /** @@ -83,6 +100,21 @@ LTTNG_TRACEPOINT_EVENT(signal_generate, * This means, this can show which signals are actually delivered, but * matching generated signals and delivered signals may not be correct. */ +#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,20,0)) +LTTNG_TRACEPOINT_EVENT(signal_deliver, + + TP_PROTO(int sig, struct kernel_siginfo *info, struct k_sigaction *ka), + + TP_ARGS(sig, info, ka), + + TP_FIELDS( + ctf_integer(int, sig, sig) + LTTNG_FIELDS_SIGINFO(info) + ctf_integer(unsigned long, sa_handler, (unsigned long) ka->sa.sa_handler) + ctf_integer(unsigned long, sa_flags, ka->sa.sa_flags) + ) +) +#else LTTNG_TRACEPOINT_EVENT(signal_deliver, TP_PROTO(int sig, struct siginfo *info, struct k_sigaction *ka), @@ -96,6 +128,7 @@ LTTNG_TRACEPOINT_EVENT(signal_deliver, ctf_integer(unsigned long, sa_flags, ka->sa.sa_flags) ) ) +#endif #if (LINUX_VERSION_CODE < KERNEL_VERSION(3,4,0)) LTTNG_TRACEPOINT_EVENT_CLASS(signal_queue_overflow,