Miroslav reported that the livepatch self-tests were failing,
specifically a case in which the consistency model ensures that we do
not patch a current executing function, "TEST: busy target module".

Recent renovations to stack_trace_save_tsk_reliable() left it returning
only an -ERRNO success indication in some configuration combinations:

  klp_check_stack()
    ret = stack_trace_save_tsk_reliable()
      #ifdef CONFIG_ARCH_STACKWALK && CONFIG_HAVE_RELIABLE_STACKTRACE
        stack_trace_save_tsk_reliable()
          ret = arch_stack_walk_reliable()
            return 0
            return -EINVAL
          ...
          return ret;
    ...
    if (ret < 0)
      /* stack_trace_save_tsk_reliable error */
    nr_entries = ret;                               << 0

Previously (and currently for !CONFIG_ARCH_STACKWALK &&
CONFIG_HAVE_RELIABLE_STACKTRACE) stack_trace_save_tsk_reliable()
returned the number of entries that it consumed in the passed storage
array.

In the case of the above config and trace, be sure to return the
stacktrace_cookie.len on stack_trace_save_tsk_reliable() success.

Fixes: 25e39e32b0a3f ("livepatch: Simplify stack trace retrieval")
Reported-by: Miroslav Benes <mbe...@suse.cz>
Signed-off-by: Joe Lawrence <joe.lawre...@redhat.com>
---
 kernel/stacktrace.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/stacktrace.c b/kernel/stacktrace.c
index 27bafc1e271e..90d3e0bf0302 100644
--- a/kernel/stacktrace.c
+++ b/kernel/stacktrace.c
@@ -206,7 +206,7 @@ int stack_trace_save_tsk_reliable(struct task_struct *tsk, 
unsigned long *store,
 
        ret = arch_stack_walk_reliable(consume_entry, &c, tsk);
        put_task_stack(tsk);
-       return ret;
+       return ret ? ret : c.len;
 }
 #endif
 
-- 
2.20.1

Reply via email to