RFR (small) 6518907: purge IA64 platform-specific ifdefs in source code

[Vladimir Kozlov]

Hi, Volker

We will wait your review. Have a good Holiday!

Regards,
Vladimir

On Oct 30, 2012, at 7:54 AM, Volker Simonis wrote:

> Hi,
> 
> thank you for remembering us:)
> 
> We at SAP are indeed using this code - at least most of it.
> 
> We don't support the SA at all, so the changes there will not affect
> us (and I suppose they won't affect ZERO as well).
> 
> But most of the other changes are required for IA64 (e.g.
> need_register_stack_bang() in compile.hpp.
> 
> I'm currently out of office and Thursday is public holiday in Germany
> so probably there will not be a lot of folks in the offices at the end
> of this week, but we would really appreciate if you'd give us some
> more time to review this change. Maybe some of the arcane IA64 stuff
> can be really removed, but I think most of it is still required for an
> IA64 port to work.
> 
> Regards,
> Volker
> 
> On Tue, Oct 30, 2012 at 5:48 AM, David Holmes <david.holmes at oracle.com> wrote:
>> Notwithstanding we may have to abandon this if the community is using the
>> IA64 "support" that exists in the shared code ...
>> 
>> Are the SA changes consistent with what Yumin had proposed and is fixing
>> under:
>> 
>> https://jbs.oracle.com/bugs/browse/JDK-8000412
>> 
>> ---
>> 
>> This change in the windows code needs closer examination:
>> 
>> -#ifdef _M_IA64
>> -#define __CPU__ ia64
>> -#elif _M_AMD64
>> -#define __CPU__ amd64
>> -#else
>> -#define __CPU__ i486
>> -#endif
>> 
>> while it appears that __CPU__ is not being used in our source code, it may
>> be that this was needed to affect other headers and/or the compiler. This
>> dates back a long way.
>> 
>> ---
>> 
>> There are no makefile changes to remove ia64 settings. If we are going to do
>> this then we should do it all.
>> 
>> Cheers,
>> David
>> 
>> 
>> On 30/10/2012 4:52 AM, Christian Thalinger wrote:
>>> 
>>> 
>>> On Oct 29, 2012, at 10:55 AM, Vladimir Kozlov<vladimir.kozlov at oracle.com>
>>> wrote:
>>> 
>>>> 
>>>> On Oct 29, 2012, at 10:48 AM, Christian Thalinger wrote:
>>>> 
>>>>> Sure:  http://cr.openjdk.java.net/~twisti/6518907
>>>>> 
>>>>> Btw. the bug synopsis (which is called Summary in JIRA) is actually:
>>>>> 
>>>>> 6518907: too many platform-specific ifdefs in source code
>>>>> 
>>>>> -- Chris
>>>>> 
>>>>> On Oct 29, 2012, at 10:35 AM, Vladimir
>>>>> Kozlov<vladimir.kozlov at oracle.com>  wrote:
>>>>> 
>>>>>> Christian,
>>>>>> 
>>>>>> Can you move web rev  to cr.openjdk so people outside can review it. We
>>>>>> then can ask RH and SAP if they use that code.
>>>>>> 
>>>>>> Thanks,
>>>>>> Vladimir
>>>>>> 
>>>>>> On Oct 29, 2012, at 10:31 AM, Christian Thalinger wrote:
>>>>>> 
>>>>>>> 
>>>>>>> On Oct 29, 2012, at 9:38 AM, Morris Meyer<morris.meyer at oracle.com>
>>>>>>> wrote:
>>>>>>> 
>>>>>>>> Took a pass at purging the platform-specific IA64 refs throughout the
>>>>>>>> source base.
>>>>>>> 
>>>>>>> 
>>>>>>> In general I think this is okay.  The only thing I'm worried is that
>>>>>>> maybe some defines or such may be used when compiling Zero on IA64.
>>>>>>> Although I don't know if anyone is doing that.
>>> 
>>> 
>>> Apparently Debian is doing it:
>>> 
>>> http://packages.debian.org/sid/ia64/openjdk-7-jre-headless
>>> 
>>> -- Chris
>>> 
>>>>>>> 
>>>>>>> -- Chris
>>>>>>> 
>>>>>>>> 
>>>>>>>> Thanks,
>>>>>>>> 
>>>>>>>>        --mm
>>>>>>>> 
>>>>>>>> JBS - https://jbs.oracle.com/bugs/browse/JDK-6518907
>>>>>>>> WEBREV -
>>>>>>>> http://javaweb.us.oracle.com/~mameyer/webrevs/01/JDK-6518907/
>>>>>>> 
>>>>>>> 
>>>>>> 
>>>>> 
>>>> 
>>> 
>>