RFR: 8254362: x86_32 builds fail after JDK-8253180

Vladimir Kozlov kvn at openjdk.java.net
Sun Oct 11 20:28:08 UTC 2020


On Sun, 11 Oct 2020 17:52:05 GMT, Aleksey Shipilev <shade at openjdk.org> wrote:

> `r15_thread` is not available on x86_32. I noticed that JDK-8253180 introduces:
> 
> void C1SafepointPollStub::emit_code(LIR_Assembler* ce) {
> #ifdef _LP64
> ...
>   __ movptr(Address(r15_thread, JavaThread::saved_exception_pc_offset()), rscratch1);
> ...
> #else
>   ShouldNotReachHere();
> #endif /* _LP64 */
> }
> 
> ...and we should do the same in `C2SafepointPollStubTable` to unbreak x86_32.
> 
> Testing:
>  - [x] x86_32 build
>  - [x] x86_32 hotspot:tier1 (lots of unrelated failures that look like JDK-8254125, never a `ShouldNotReachHere`)
>  - [x] x86_64 build
>  - [x] x86_64 hotspot:tier1

Good.

-------------

Marked as reviewed by kvn (Reviewer).

PR: https://git.openjdk.java.net/jdk/pull/593


More information about the hotspot-compiler-dev mailing list