Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 23 May 2012 02:43:05 +0000 (UTC)
From:      Benjamin Kaduk <bjk@freebsd.org>
To:        John Baldwin <jhb@freebsd.org>
Cc:        svn-src-head@freebsd.org, Gleb Kurtsou <gleb.kurtsou@gmail.com>, src-committers@freebsd.org, svn-src-all@freebsd.org
Subject:   Re: svn commit: r235601 - head/include/protocols
Message-ID:  <alpine.BSF.2.00.1205230240100.22739@freefall.freebsd.org>
In-Reply-To: <201205211127.34052.jhb@freebsd.org>
References:  <201205181001.q4IA1VED044374@svn.freebsd.org> <201205180941.48076.jhb@freebsd.org> <20120518152436.GA9116@reks> <201205211127.34052.jhb@freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 21 May 2012, John Baldwin wrote:

> On Friday, May 18, 2012 11:24:36 am Gleb Kurtsou wrote:
>> On (18/05/2012 09:41), John Baldwin wrote:
>>> On Friday, May 18, 2012 6:01:31 am Gleb Kurtsou wrote:
>>>> Author: gleb
>>>> Date: Fri May 18 10:01:31 2012
>>>> New Revision: 235601
>>>> URL: http://svn.freebsd.org/changeset/base/235601
>>>>
>>>> Log:
>>>>   Don't use ino_t in dumprestore protocol definition.
>>>>
>>>>   Since ino_t size is about to change to 64-bits, replace ino_t used in
>>>>   dump protocol definition with 32-bit dump_ino_t to preserve backward
>>>>   compatibility.  At some point, it may be necessary to use spare fields
>>>>   in struct in order to fully support 64-bit inode numbers.
>>>>
>>>>   Sponsored by:	Google Summer of Code 2011
>>>
>>> A question about your stat changes: did you expand dev_t to 32 bits for
> the
>>> AFS folks, or did you leave it as 16 bits?
>>
>> dev_t is already 32-bit. Changing it to 64-bit was discussed at some
>> point and from what I recall no decision was made:
>>
>> http://marc.info/?t=129119478700005&r=1&w=2
>>
>> I'm going to commit preparatory changes only for now. Then publish diff
>> for testing. We can still change dev_t to 64-bit if needed. Although I
>> didn't work on it.
>
> Ah, it was 64-bit they asked for.  If it is easy to do so, I'd favor changing
> it since you've already done all the hard work of rolling a new stat
> structure.  I'd rather err on wasting 32-bits for dev_t than having to do all
> this over again.

Hi John,

Thanks for remembering this -- I'm still pretty swamped with other stuff 
and missed it.
It would indeed be nice for us to have a 64-bit dev_t to work with.

-Ben



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.BSF.2.00.1205230240100.22739>