RFR: 8241074: JDK-8240363 broke the build on AArch64

Erik Österlund erik.osterlund at oracle.com
Tue Mar 17 11:27:01 UTC 2020


Interesting. Perhaps there were cyclic includes that affected me 
differently due to not building
shenandoah... or something. Not sure exactly why it didn't build for me. 
But the include should be there
nevertheless and now it works for everyone.

/Erik

On 2020-03-17 10:46, Andrew Haley wrote:
> On 3/16/20 4:04 PM, Erik Österlund wrote:
>> My recent refactoring from JDK-8240363 broke the build on AArch64. I
>> asked Andrew to sanity check it before pushing,
>> but I think he might have had precompiled headers enabled.
> I didn't use precompiled headers. This is shared code, and it uses
> MacroAssembler, so it surely needs "asm/macroAssembler.hpp" to be
> included.
>
> On all targets, including AArch64, macroAssembler.hpp seems to get
> pulled in as a transitive dependency. I don't know why that didn't
> happen on your AArch64 cross-build.
>



More information about the hotspot-compiler-dev mailing list