Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 27 Jul 2011 22:59:31 -0400
From:      Attilio Rao <attilio@freebsd.org>
To:        Nathan Whitehorn <nwhitehorn@freebsd.org>
Cc:        svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org, Sergey Kandaurov <pluknet@freebsd.org>, Andreas Tobler <andreast@freebsd.org>
Subject:   Re: svn commit: r224216 - in head/sys: ia64/ia64 mips/mips powerpc/aim sparc64/sparc64
Message-ID:  <CAJ-FndBjU1XmM-pJPTJy8On_18sSyLp_FeTWE8cATYhz4gtVdg@mail.gmail.com>
In-Reply-To: <CAJ-FndBS=WS6_Uk5p1pw%2BQYureRGyFW2NDF%2B0HvB6B9fELt0dw@mail.gmail.com>
References:  <201107191241.p6JCfvqw072816@svn.freebsd.org> <4E2C9012.602@freebsd.org> <4E2DB8AC.3000703@FreeBSD.org> <4E2DBCBF.2030402@freebsd.org> <CAJ-FndBS=WS6_Uk5p1pw%2BQYureRGyFW2NDF%2B0HvB6B9fELt0dw@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
I think that the following patch may better reflect the definition of size_t:
http://www.freebsd.org/~attilio/sintrcnt-fixup32.diff

Do you think you can test it out?

Thanks,
Attilio

2011/7/25 Attilio Rao <attilio@freebsd.org>:
> Guys, I'm a bit out of that for a couple of days, flying to another
> continent, I'm not entirely sure the patch is right because I think we
> need an unified approach for specifying a 32-bits value, but you can
> check-in as a showstopper for the moment, I will make a better one in
> a couple of days.
>
> Attilio
>
> 2011/7/25 Nathan Whitehorn <nwhitehorn@freebsd.org>:
>> On 07/25/11 13:40, Andreas Tobler wrote:
>>>
>>> On 24.07.11 23:35, Nathan Whitehorn wrote:
>>>>
>>>> On 07/19/11 07:41, Attilio Rao wrote:
>>>>>
>>>>> Author: attilio
>>>>> Date: Tue Jul 19 12:41:57 2011
>>>>> New Revision: 224216
>>>>> URL: http://svn.freebsd.org/changeset/base/224216
>>>>>
>>>>> Log:
>>>>> On 64 bit architectures size_t is 8 bytes, thus it should use an 8 bytes
>>>>> storage.
>>>>> Fix the sintrcnt/sintrnames specification.
>>>>>
>>>>> No MFC is previewed for this patch.
>>>>>
>>>>
>>>> These also need to be .long on 32-bit PowerPC, otherwise this change
>>>> triggers kernel panics and faults in sysctl. Please change it as quickly
>>>> as possible.
>>>
>>> Like the below?
>>>
>>> I'll test and I could commit if ok. (Build and approval)
>>>
>>> Gruss,
>>> Andreas
>>>
>>> Index: locore32.S
>>> ===================================================================
>>> --- locore32.S (revision 224326)
>>> +++ locore32.S (working copy)
>>> @@ -91,13 +91,13 @@
>>> GLOBAL(intrnames)
>>> .space INTRCNT_COUNT * (MAXCOMLEN + 1) * 2
>>> GLOBAL(sintrnames)
>>> - .word INTRCNT_COUNT * (MAXCOMLEN + 1) * 2
>>> + .long INTRCNT_COUNT * (MAXCOMLEN + 1) * 2
>>>
>>> .align 4
>>> GLOBAL(intrcnt)
>>> .space INTRCNT_COUNT * 4 * 2
>>> GLOBAL(sintrcnt)
>>> - .word INTRCNT_COUNT * 4 * 2
>>> + .long INTRCNT_COUNT * 4 * 2
>>>
>>> .text
>>> .globl btext
>>
>> Exactly like that.
>> -Nathan
>>
>
>
>
> --
> Peace can only be achieved by understanding - A. Einstein
>



-- 
Peace can only be achieved by understanding - A. Einstein



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-FndBjU1XmM-pJPTJy8On_18sSyLp_FeTWE8cATYhz4gtVdg>