Re: [PATCH bpf-next v6 10/10] selftests/bpf: test case for register_bpf_struct_ops().

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 





On 10/26/23 13:31, Eduard Zingerman wrote:
On Sat, 2023-10-21 at 22:03 -0700, thinker.li@xxxxxxxxx wrote:
From: Kui-Feng Lee <thinker.li@xxxxxxxxx>

Create a new struct_ops type called bpf_testmod_ops within the bpf_testmod
module. When a struct_ops object is registered, the bpf_testmod module will
invoke test_2 from the module.

Signed-off-by: Kui-Feng Lee <thinker.li@xxxxxxxxx>

Hello,

Sorry for the late response, was moving through the patch-set very slowly.
Please note that CI currently fails for this series [0], reported error is:

testing_helpers.c:13:10: fatal error: 'rcu_tasks_trace_gp.skel.h' file not found
    13 | #include "rcu_tasks_trace_gp.skel.h"
       |          ^~~~~~~~~~~~~~~~~~~~~~~~~~~

I get the same error when try to run tests locally (after full clean).
On the other hand it looks like `kern_sync_rcu_tasks_trace` changes
are not really necessary, when I undo these changes but keep changes in:

- .../selftests/bpf/bpf_testmod/bpf_testmod.c
- .../selftests/bpf/bpf_testmod/bpf_testmod.h
- .../bpf/prog_tests/test_struct_ops_module.c
- .../selftests/bpf/progs/struct_ops_module.c

struct_ops_module/regular_load test still passes.

Regarding assertion:

+	ASSERT_EQ(skel->bss->test_2_result, 7, "test_2_result");

Could you please leave a comment explaining why the value is 7?
I don't understand what invokes 'test_2' but changing it to 8
forces test to fail, so something does call 'test_2' :)

Also, when running test_maps I get the following error:

libbpf: bpf_map_create_opts has non-zero extra bytes
map_create_opts(317):FAIL:bpf_map_create() error:Invalid argument (name=hash_of_maps)

According to what Andrii Nakryiko said,
once [1] is landed, this error should be fixed.

[1] https://lore.kernel.org/all/20231029011509.2479232-1-andrii@xxxxxxxxxx/



[0] https://patchwork.kernel.org/project/netdevbpf/patch/20231022050335.2579051-11-thinker.li@xxxxxxxxx/
     (look for 'Logs for x86_64-gcc / build / build for x86_64 with gcc ')

[...]





[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux