Some architectures modify syscall numbers to indicate ABI used
and as a result syscall number as returned by syscall_get_nr
does not correspond to the syscall number used inside the kernel.
Allow an arch to provide a separate implementation for ftrace
that returns the 'real' syscall number from the regs provided.
This will be used by x86 in a follow-up change.
Signed-off-by: Marcin Nowakowski <marcin.nowakow...@imgtec.com>
Cc: Steven Rostedt <rost...@goodmis.org>
Cc: Ingo Molnar <mi...@redhat.com>
---
kernel/trace/trace_syscalls.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c
index 10c058f..24a26d3 100644
--- a/kernel/trace/trace_syscalls.c
+++ b/kernel/trace/trace_syscalls.c
@@ -78,7 +78,11 @@ static inline bool trace_is_compat_syscall(struct pt_regs
*regs)
static inline int
trace_get_syscall_nr(struct task_struct *task, struct pt_regs *regs)
{
+#ifdef arch_trace_syscall_get_nr
+ return arch_trace_syscall_get_nr(task, regs);
+#else
return syscall_get_nr(task, regs);
+#endif
}
--
2.7.4
