Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 21 Jul 2023 15:06:47 +0200
From:   Thomas =?utf-8?Q?Wei=C3=9Fschuh?= <linux@weissschuh.net>
To:     Zhangjin Wu <falcon@tinylab.org>
Cc:     linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org,
        w@1wt.eu
Subject: Re: [PATCH RFC 7/7] selftests/nolibc: proof of concept for TAP output
Message-ID: <1a895dac-f37f-4569-944b-c311a94e3c93@t-8ch.de>
References: <20230719-nolibc-ktap-tmp-v1-7-930bd0c52ff1@weissschuh.net>
 <20230720152940.7623-1-falcon@tinylab.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230720152940.7623-1-falcon@tinylab.org>
Precedence: bulk

On 2023-07-20 23:29:40+0800, Zhangjin Wu wrote:
> The result looks very similar and the change seems not that big, thanks.
> 
> I have a generic question: do we need to compile nolibc-test.c
> independently or at least let users easily compile nolibc-test.c in the
> other places no just in kernel source code, for example, the other libcs
> may want to download and compile it directly.
> 
> The functions used in this change seems not many, is it able to provide
> our clones for them or only provide the clones when we compile them
> out-of-kernel.
> 
> for example:
> 
>     #ifdef NOLIBC_TEST_IN_KERNEL
>     /* -DNOLIBC_TEST_IN_KERNEL from Makefile, for future compatibility */
>     #include "../kselftest.h"
>     #else
>     // our clones of the used functions, for standalone usage
>     #endif

I'd rather just disable the functionality instead. Willy also asked to
stick to the current formatting by default and enable KTAP with a
switch.
So everything will be behind generic wrappers anyways, so we can drop
the KTAP functionality from those wrappers easiy.

Thomas