All of lore.kernel.org
 help / color / mirror / Atom feed
From: Markos Chandras <Markos.Chandras@imgtec.com>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH v3] Disable o32 ABI for MIPS64 architectures
Date: Fri, 28 Mar 2014 10:41:22 +0000	[thread overview]
Message-ID: <533551D2.2010003@imgtec.com> (raw)
In-Reply-To: <5335510C.3090006@imgtec.com>

On 03/28/2014 10:38 AM, Vicente Olivert Riera wrote:
> On 03/28/2014 10:27 AM, Markos Chandras wrote:
>> On 03/28/2014 09:51 AM, Vicente Olivert Riera wrote:
>>> On 03/28/2014 12:02 AM, Joshua Kinard wrote:
>>>> On 03/27/2014 05:56, Vicente Olivert Riera wrote:
>>>>> On 03/26/2014 09:28 PM, Peter Korsgaard wrote:
>>>>>>>>>>> "Vicente" == Vicente Olivert Riera <Vincent.Riera@imgtec.com>
>>>>>>>>>>> writes:
>>>>>>
>>>>>> Hi,
>>>>>>
>>>>>>    >>> +config BR2_MIPS_OABI32
>>>>>>    >>> +    bool
>>>>>>    >>> +    default y        if BR2_mips || BR2_mipsel
>>>>>>    >>> +    default n        if BR2_mips64 || BR2_mips64el
>>>>>>    >>
>>>>>>    >> Did you forget to remove this or is there a reason to keep it?
>>>>>> In the
>>>>>>    >> latter case, please add an explanatory comment.
>>>>>>
>>>>>>    > Why would I want to remove this? The BR2_MIPS_OABI32 symbol is
>>>>>> used by
>>>>>>    > glibc and uclibc packages.
>>>>>>
>>>>>> Arnout is just asking about the 'default n' line. Symbols are 'n' by
>>>>>> default, so the line doesn't do anything as far as I can see.
>>>>>>
>>>>>
>>>>> I have maintained that line because if you remove it and then select
>>>>> any
>>>>> MIPS64 target, the "# BR2_MIPS_OABI32 is not set" is not present on
>>>>> the
>>>>> .config file.
>>>>>
>>>>> Isn't that a problem?
>>>>
>>>> Is the definition of BR2_MIPS_OABI32 missing completely, or do you see
>>>> "BR2_MIPS_OABI32=y"?  If it's missing completely, then it's virtually
>>>> the
>>>> same as it being there as a comment, which will get stripped
>>>> out/ignored by
>>>> the build system.  Kconfig/Kbuild (whatever it is called) only cares
>>>> if the
>>>> symbol is defined to "y".  I believe that is then checked for in
>>>> Makefiles
>>>> and is in a generated config.h file for use as a C #define.
>>>>
>>>> That's how it works in the Linux kernel at least...
>>>
>>> If you remove the "default 'n'" line, and you select a MIPS64 target,
>>> then "BR2_MIPS_OABI32=y" is completely missing in the .config file.
>>>
>>>
>>
>> That's not a problem is it? It's ok if the line is missing. A missing
>> line means "this option is not available". A line starting with "#"
>> means "the option is available but currently disabled"
>
> I want to be sure about that, because there are packages which use the
> BR2_MIPS_OABI32 symbol. So, maybe having that line starting with "#"
> means BR2_MIPS_OABI32=n, and everything works fine, and don't having
> that line at all means that symbol is not defined and the packages which
> use that symbol fail because of that. That's what I want to know.
>

I think not having that line, or having that line but prefixed with "#" 
is the same thing as far as the Kconfig dependencies are concerned.
(pretty much what Joshua said already)

-- 
markos

  reply	other threads:[~2014-03-28 10:41 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-26 14:15 [Buildroot] [PATCH v3] Disable o32 ABI for MIPS64 architectures Vicente Olivert Riera
2014-03-26 17:18 ` Arnout Vandecappelle
2014-03-26 18:03   ` Vicente Olivert Riera
2014-03-26 21:28     ` Peter Korsgaard
2014-03-27  9:56       ` Vicente Olivert Riera
2014-03-28  0:02         ` Joshua Kinard
2014-03-28  9:51           ` Vicente Olivert Riera
2014-03-28 10:27             ` Markos Chandras
2014-03-28 10:38               ` Vicente Olivert Riera
2014-03-28 10:41                 ` Markos Chandras [this message]
2014-03-28 10:52                   ` Joshua Kinard
2014-03-28 11:01                     ` Vicente Olivert Riera
2014-03-28 11:07                 ` Thomas Petazzoni

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=533551D2.2010003@imgtec.com \
    --to=markos.chandras@imgtec.com \
    --cc=buildroot@busybox.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.