[patch] Fix non-PCH builds

Maurizio Cimadamore maurizio.cimadamore at oracle.com
Sat Nov 3 02:30:13 UTC 2018


Yep

Maurizio

On 01/11/2018 21:48, Vladimir Ivanov wrote:
> Looks good.
>
> Maurizio, I assume the fix should go into foreign and then it'll be 
> automatically propagated into foreign+vector, right?
>
> Best regards,
> Vladimir Ivanov
>
> On 01/11/2018 13:32, Aleksey Shipilev wrote:
>> Hi,
>>
>> Not sure what is the workflow for Panama patches. This trivial fix is 
>> needed to build foreign+vector
>> branch with --disable-precompiled-headers. Please push where you see 
>> fit?
>>
>> ---- 8< --------- --------- --------- --------- --------- --------- 
>> --------- ---------
>>
>> diff -r 0260ce01533b src/hotspot/share/prims/universalNativeInvoker.cpp
>> --- a/src/hotspot/share/prims/universalNativeInvoker.cpp    Wed Oct 
>> 31 20:59:05 2018 +0100
>> +++ b/src/hotspot/share/prims/universalNativeInvoker.cpp    Thu Nov 
>> 01 21:30:03 2018 +0100
>> @@ -22,9 +22,12 @@
>>    */
>>
>>   #include "precompiled.hpp"
>> +#include "code/codeBlob.hpp"
>>   #include "prims/universalNativeInvoker.hpp"
>>   #include "runtime/interfaceSupport.inline.hpp"
>>   #include "runtime/jniHandles.inline.hpp"
>> +#include "runtime/stubCodeGenerator.hpp"
>> +#include "prims/methodHandles.hpp"
>>
>>   BufferBlob* UniversalNativeInvoker::_invoke_native_blob = NULL;
>>
>> ---- 8< --------- --------- --------- --------- --------- --------- 
>> --------- ---------
>>
>> Unfortunately, there is a dependency between stubCodeGenerator.hpp 
>> and methodHandles.hpp that does
>> not allow these headers to go alphabetically.
>>
>> Thanks,
>> -Aleksey
>>


More information about the panama-dev mailing list