[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <caf816a4-3ab0-d07e-7559-661999f10f0f@virtuozzo.com>
Date: Tue, 18 Dec 2018 11:12:24 +0300
From: Kirill Tkhai <ktkhai@...tuozzo.com>
To: Matthew Wilcox <willy@...radead.org>
Cc: akpm@...ux-foundation.org, yamada.masahiro@...ionext.com,
const@...eLinux.com, joeypabalinas@...il.com,
arend.vanspriel@...adcom.com, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] scripts/tags: Add more declarations
On 18.12.2018 00:09, Matthew Wilcox wrote:
> On Mon, Dec 17, 2018 at 03:41:46PM +0300, Kirill Tkhai wrote:
>> - '/\<DEFINE_\(MUTEX\|SEMAPHORE\|SPINLOCK\)(\([[:alnum:]_]*\)/\2/v/'
>> + '/\<DEFINE_\(RT_MUTEX\|MUTEX\|SEMAPHORE\|SPINLOCK\)(\([[:alnum:]_]*\)/\2/v/'
>> '/\<DEFINE_\(RAW_SPINLOCK\|RWLOCK\|SEQLOCK\)(\([[:alnum:]_]*\)/\2/v/'
>
> I don't know anything about regexes this complex or tags, but couldn't this
> be '/\<DEFINE_\([[:upper:]_]\)(\([[:alnum:]_]*\)/\2/v/' to catch all
> DEFINE_THINGIES?
I think, we can, but before this it's needed to change arguments order
in macroses like DEFINE_PER_CPU(), and to check, whether tags won't be
polluted something not needed (such wrong defines will be generated
anyway, but if there will be not too many, than it's OK).
Powered by blists - more mailing lists