Re: [PATCH bpf-next v2 2/4] selftests/bpf: test_progs: remove global fail/success counts

From: Stanislav Fomichev <sdf@fomichev.me>
To: Daniel Borkmann <daniel@iogearbox.net>
Cc: Stanislav Fomichev <sdf@google.com>,
	netdev@vger.kernel.org, bpf@vger.kernel.org, davem@davemloft.net,
	ast@kernel.org, Andrii Nakryiko <andriin@fb.com>
Subject: Re: [PATCH bpf-next v2 2/4] selftests/bpf: test_progs: remove global fail/success counts
Date: Wed, 21 Aug 2019 10:11:49 -0700	[thread overview]
Message-ID: <20190821171149.GA1717@mini-arch> (raw)
In-Reply-To: <5248b967-2887-2205-3e59-fc067e2ada33@iogearbox.net>

On 08/21, Daniel Borkmann wrote:
> On 8/19/19 9:17 PM, Stanislav Fomichev wrote:
> > Now that we have a global per-test/per-environment state, there
> > is no longer need to have global fail/success counters (and there
> > is no need to save/get the diff before/after the test).
> 
> Thanks for the improvements, just a small comment below, otherwise LGTM.
> 
> > Introduce QCHECK macro (suggested by Andrii) and covert existing tests
> > to it. QCHECK uses new test__fail() to record the failure.
> > 
> > Cc: Andrii Nakryiko <andriin@fb.com>
> > Signed-off-by: Stanislav Fomichev <sdf@google.com>
> [...]
> > @@ -96,17 +93,25 @@ extern struct ipv6_packet pkt_v6;
> >   #define _CHECK(condition, tag, duration, format...) ({			\
> >   	int __ret = !!(condition);					\
> >   	if (__ret) {							\
> > -		error_cnt++;						\
> > +		test__fail();						\
> >   		printf("%s:FAIL:%s ", __func__, tag);			\
> >   		printf(format);						\
> >   	} else {							\
> > -		pass_cnt++;						\
> >   		printf("%s:PASS:%s %d nsec\n",				\
> >   		       __func__, tag, duration);			\
> >   	}								\
> >   	__ret;								\
> >   })
> > +#define QCHECK(condition) ({						\
> > +	int __ret = !!(condition);					\
> > +	if (__ret) {							\
> > +		test__fail();						\
> > +		printf("%s:FAIL:%d ", __func__, __LINE__);		\
> > +	}								\
> > +	__ret;								\
> > +})
> 
> I know it's just a tiny nit but the name QCHECK() really doesn't tell me anything
> if I don't see its definition. Even just a CHECK_FAIL() might be 'better' and
> more aligned with the CHECK() and CHECK_ATTR() we have, at least I don't think
> many would automatically derive 'quiet' from the Q prefix [0].
CHECK_FAIL sounds good, will respin! Thanks!

>   [0] https://lore.kernel.org/bpf/CAEf4BzbUGiUZBWkTWe2=LfhkXYhQGndN9gR6VTZwfV3eytstUw@mail.gmail.com/
> 
> >   #define CHECK(condition, tag, format...) \
> >   	_CHECK(condition, tag, duration, format)
> >   #define CHECK_ATTR(condition, tag, format...) \
> > 
>