Received-SPF: pass (google.com: domain of linux-kernel+bounces-127242-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Date: Tue, 2 Apr 2024 01:57:50 +0100
From: Al Viro <viro@zeniv.linux.org.uk>
To: Maxim Moskalets <maximmosk4@gmail.com>
Cc: linux-kernel@vger.kernel.org,
	Maxim Moskalets <Maxim.Moskalets@kaspersky.com>
Subject: Re: [PATCH 0/8] kernel: replace seq_puts by seq_putc
Message-ID: <20240402005750.GD538574@ZenIV>
References: <20240326184514.8478-1-Maxim.Moskalets@kaspersky.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20240326184514.8478-1-Maxim.Moskalets@kaspersky.com>
Sender: Al Viro <viro@ftp.linux.org.uk>

On Tue, Mar 26, 2024 at 09:45:14PM +0300, Maxim Moskalets wrote:
> Using seq_putc for single characters is faster and more appropriate
> than seq_puts, since only one character is passed and there is no need
> to use a more powerful and less fast function.

Could we simply do this:

static inline void seq_puts(struct seq_file *m, const char *s)
{
	if (__builtin_constant_p(*s) && s[0] && !s[1])
		seq_putc(m, s[0]);
	else
		__seq_puts(m, s);
}

IIRC, __builtin_constant_p(*s) is true when s is a string literal.
Works for recent gcc and clang...