_LP64, int32_t and intptr_t

[Michael Franz]

Ben,

What I do for this is, delete the test_gamma from the bsd makefile.
That way it never runs the test.  This is the same solution the
IcedTea project does.  There is no need to copy over any dylibs.

Michael



On 12/16/08, Ben Evans <benjamin.john.evans at googlemail.com> wrote:
> Hi Greg,
>
> On Tue, Dec 16, 2008 at 4:59 PM, Greg Lewis <glewis at eyesbeyond.com> wrote:
>
>> On Tue, Dec 16, 2008 at 11:05:31AM -0500, Michael Franz wrote:
>> > I have not built the latest repo, I was working on one from 13/14.
>> > There may have been merges done after I did my pull.
>> >
>> > These errors are in Java code and are unrelated to the patches.
>> > Further evidence that the repo has changed since my last build.
>> >
>> > I will not be able to look at this until tonight.  If you can, pull a
>> > different revision and try again.
>>
>> All commits to the repository are sent to this list as far as I can tell.
>> I certainly haven't committed any changes since the merge in late November
>> that introduced the problem in HotSpot.  I have some local hacks
>> (different
>> from Michael's) to try and get around that, but I'm seeing other errors
>> that look like they are due to the version of as on my Mac.
>>
>> We are pretty due to do some more merging, but I'm hoping our friends at
>> Sun (hi Xiaobin!) can maybe come up with an acceptable change they are
>> willing to commit to the main repo to fix the HotSpot compile problems.
>>
>> Ben, it might be worth trying to bootstrap with the version of OpenJDK
>> that is available in binary form on Landon Fuller's pages.  That is what
>> Michael uses.
>
>
> Using the OpenJDK build on Landon's pages enables the patched build to
> complete, at least as far as the self-test error mentioned here:
>
> http://mail.openjdk.java.net/pipermail/bsd-port-dev/2008-September/000044.html
>
> As Karen suggests on that page, I've copied the .dylib's over a copy of the
> existing JDK build, so I think I have a working build now.
>
> I'm still struggling to get the invokedynamic patches applied, but that's
> another story.
>
> Thanks,
>
> Ben
>

-- 
Sent from Gmail for mobile | mobile.google.com