[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87jzn6lnou.fsf@cloudflare.com>
Date: Wed, 14 Feb 2024 22:46:46 +0100
From: Jakub Sitnicki <jakub@...udflare.com>
To: Jakub Kicinski <kuba@...nel.org>
Cc: shuah@...nel.org, keescook@...omium.org,
linux-kselftest@...r.kernel.org, netdev@...r.kernel.org
Subject: Re: [PATCH net-next 3/4] selftests: kselftest_harness: support
using xfail
On Wed, Feb 14, 2024 at 08:40 PM +01, Jakub Sitnicki wrote:
[...]
> On second thought, if I can suggest a follow up change so this:
>
> ok 17 # XFAIL SCTP doesn't support IP_BIND_ADDRESS_NO_PORT
>
> ... becomes this
>
> ok 17 ip_local_port_range.ip4_stcp.late_bind # XFAIL SCTP doesn't support IP_BIND_ADDRESS_NO_PORT
>
> You see, we parse test results if they are in TAP format. Lack of test
> name for xfail'ed and skip'ed tests makes it difficult to report in CI
> which subtest was it. Happy to contribute it, once this series gets
> applied.
Should have said "harder", not "difficult". That was an overstatement.
Test name can be extracted from diagnostic lines preceeding the status.
# RUN ip_local_port_range.ip4_stcp.late_bind ...
# XFAIL SCTP doesn't support IP_BIND_ADDRESS_NO_PORT
# OK ip_local_port_range.ip4_stcp.late_bind
ok 17 ip_local_port_range.ip4_stcp.late_bind # XFAIL SCTP doesn't support IP_BIND_ADDRESS_NO_PORT
It just makes the TAP parser easier if the test name is included on the
status line. That would be the motivation here. Let me know what you
think.
[...]
Powered by blists - more mailing lists