On 4/28/18 12:06 PM, Alexei Starovoitov wrote:
On Sat, Apr 28, 2018 at 11:17:30AM -0700, Y Song wrote:
On Sat, Apr 28, 2018 at 9:56 AM, Alexei Starovoitov
<alexei.starovoi...@gmail.com> wrote:
On Sat, Apr 28, 2018 at 12:02:04AM -0700, Yonghong Song wrote:
The test attached a raw_tracepoint program to sched/sched_switch.
It tested to get stack for user space, kernel space and user
space with build_id request. It also tested to get user
and kernel stack into the same buffer with back-to-back
bpf_get_stack helper calls.
Whenever the kernel stack is available, the user space
application will check to ensure that the kernel function
for raw_tracepoint ___bpf_prog_run is part of the stack.
Signed-off-by: Yonghong Song <y...@fb.com>
...
+static int get_stack_print_output(void *data, int size)
+{
+ bool good_kern_stack = false, good_user_stack = false;
+ const char *expected_func = "___bpf_prog_run";
so the test works with interpreter only?
I guess that's ok for now, but needs to fixed for
configs with CONFIG_BPF_JIT_ALWAYS_ON=y
I did not test CONFIG_BPF_JIT_ALWAYS_ON=y.
I can have a followup patch for this if the patch set does not need respin.
I was thinking to apply the set and do the fix in the follow up,
but testing it with jit_enable=1 I don't see it's failing,
so something is wrong with the test.
Yes, it is because the return value test
if (CHECK(err < 0, "perf_event_poller", "err %d errno %d\n", err,
...
the "err < 0" is not right as all the return values are nonnegative.
Also get_stack_raw_tp_action() keeps spawning new 'dd' in the background
which is not killed after test stops.
Please fix both issues in respin.
I will fix both and resend the patch.
