Re: [PATCH liburing 3/4] tests: add tests for zerocopy send and notifications

[Pavel Begunkov <asml.silence@gmail.com>]

On 7/25/22 13:08, Ammar Faizi wrote:
> On 7/25/22 6:28 PM, Pavel Begunkov wrote:
>> Don't see any reason for that
> 
> Not that important, just for easy finding. Especially when the number of tests
> increase. And yes, it always increases from time to time.
> 
>> especially since it's not sorted.
> 
> It was, but since that skip-cqe.c exists, it's no longer :p
> OK, OK, that's trivial, never mind. Let's move on.
> 
>>> New test should use the provided exit code protocol. This should have
>>> been "return T_EXIT_SKIP;"
>>
>> Oh, I already hate those rules, sounds like they were specifically
>> honed to make patching harder.
> 
> Lol, how damn hard is it to use it.

Not hard to use, I agree, but it's rather yet another thing that
I need to keep in mind and then it unavoidably gets forgotten and
makes to spend extra 10 minutes to fix/retest/resend/etc., which
is annoying.

Testing is not the most exciting part, frameworks are usually
trying to simplify writing tests even if there is a learning
curve. Implicit rules make it worse. I wouldn't have been
complaining if the compiler failed the build or at least
runtests.sh warned about it.


>> By the way, while we're at it, what is T_EXIT_ERROR? Why it's not used anywhere
>> and how it's different from T_EXIT_FAIL?
> 
> [ Adding Eli to the participants. ]
> 
> Ummm... yeah. I am curious about it too now. I just took a look at commit:
> 
>     ed430fbeb33367 ("tests: migrate some tests to use enum-based exit codes").
> 
> Eli said:
> 
>      From: Eli Schwartz <eschwartz93@gmail.com>
>      Date: Mon, 27 Jun 2022 14:39:05 -0400
>      Subject: [PATCH] tests: migrate some tests to use enum-based exit codes
> 
>      For maintainability and clarity, eschew the use of integer literals in
>      reporting test statuses. Instead, use a helper enum which contains
>      various values from the GNU exitcode protocol. Returning 0 or 1 is
>      obvious, and in the previous commit the ability to read "skip" (77) was
>      implemented. The final exit status is 99, which indicates some kind of
>      error in running the test itself.
> 
>      A partial migration of existing pass/fail values in test sources is
>      included.
> 
>      Signed-off-by: Eli Schwartz <eschwartz93@gmail.com>
> 
> 
> That T_EXIT_ERROR is 99 here. Not sure when to use it in liburing test. Eli?
> 
> [ Just for reference in case you (Eli) want to see the full message:
> 
>    https://lore.kernel.org/io-uring/c89d373f-bc0d-dccf-630f-763e8e1a0fe5@gmail.com/  ]
> 

-- 
Pavel Begunkov