Received-SPF: pass (google.com: domain of linux-kernel+bounces-151947-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Message-ID: <5e5cde3e-f3ad-4a9b-bc02-1c473affdcb1@wanadoo.fr>
Date: Fri, 19 Apr 2024 22:38:15 +0200
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH] seq_file: Optimize seq_puts()
To: David Laight <David.Laight@ACULAB.COM>,
 'Al Viro' <viro@zeniv.linux.org.uk>
Cc: Christian Brauner <brauner@kernel.org>, Jan Kara <jack@suse.cz>,
 "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
 "kernel-janitors@vger.kernel.org" <kernel-janitors@vger.kernel.org>,
 "linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>
References: <5c4f7ad7b88f5026940efa9c8be36a58755ec1b3.1704374916.git.christophe.jaillet@wanadoo.fr>
 <4b1a4cc5-e057-4944-be69-d25f28645256@wanadoo.fr>
 <20240415210035.GW2118490@ZenIV>
 <ba306b2a1b5743bab79b3ebb04ece4df@AcuMS.aculab.com>
Content-Language: en-MW
From: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
In-Reply-To: <ba306b2a1b5743bab79b3ebb04ece4df@AcuMS.aculab.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit

Le 16/04/2024 à 22:56, David Laight a écrit :
> From: Al Viro
>> Sent: 15 April 2024 22:01
> ...
>> No need to make it a macro, actually.  And I would suggest going
>> a bit further:
>>
>> static inline void seq_puts(struct seq_file *m, const char *s)
> 
> That probably needs to be 'always_inline'.
> 
>> {
>> 	if (!__builtin_constant_p(*s))
>> 		__seq_puts(m, s);
>> 	else if (s[0] && !s[1])
>> 		seq_putc(m, s[0]);
>> 	else
>> 		seq_write(m, s, __builtin_strlen(s));
>> }
> 
> You missed seq_puts(m, "");
> 
> I did wonder about checking sizeof(s) <= 2 in the #define version.

git grep seq_puts.*\"[^\\].\" | wc -l
77

What would you do in this case?
2 seq_putc() in order to save a memcpy(..., 2), that's it?

It would also slightly change the behaviour, as only the 1st char could 
be added. Actually, it is all or nothing.

Don't know what is the best.

Any thought?

CJ


> That would pick up the cases where a separator is changed/added
> in a loop.
> 
> Could you do:
> 	size_t len = __builtin_strlen(s);
> 	if (!__builtin_constant_p(len))
> 		__seq_puts(m, s);
> 	else switch (len){
> 	case 0: break;
> 	case 1: seq_putc(m, s[0]);
> 	default: seq_write(m, s, len);
> 	}
> 
> 	David
> 
> -
> Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
> Registration No: 1397386 (Wales)
> 
> 
>