[riscv-port-jdk11u:riscv-port] RFR: 8276799: Implementation of JEP 422: Linux/RISC-V Port [v10]

Fei Yang fyang at openjdk.org
Wed Feb 7 09:08:38 UTC 2024


On Tue, 6 Feb 2024 09:31:04 GMT, kuaiwei <duke at openjdk.org> wrote:

>> Hi, should remove  issue 8243155, 8286301 here? it looks like there is no related implementation.
>
>> Hi, should remove issue 8243155, 8286301 here? it looks like there is no related implementation.
> 
> 8243155 is ported to RISCV, https://github.com/openjdk/riscv-port-jdk11u/pull/3/commits/99ca43f1e7e74f161b40466f49fc61aa734d334d
> 
> 8286301 will be removed, Thanks.

> > @kuaiwei @zhengxiaolinX : Have we finished linking all the related issues? I only see the following issues linked with this PR for now. Is that accurate? Thanks.
> > ```
> > [JDK-8276799](https://bugs.openjdk.org/browse/JDK-8276799): Implementation of JEP 422: Linux/RISC-V Port (Sub-task - P2)
> > [JDK-8202976](https://bugs.openjdk.org/browse/JDK-8202976): Add C1 lea patching support for x86 (Enhancement - P4)
> > [JDK-8243155](https://bugs.openjdk.org/browse/JDK-8243155): AArch64: Add support for SqrtVF (Enhancement - P4)
> > [JDK-8283737](https://bugs.openjdk.org/browse/JDK-8283737): riscv: MacroAssembler::stop() should emit fixed-length instruction sequence (Bug - P4)
> > [JDK-8285437](https://bugs.openjdk.org/browse/JDK-8285437): riscv: Fix MachNode size mismatch for MacroAssembler::verify_oops* (Bug - P4)
> > [JDK-8287418](https://bugs.openjdk.org/browse/JDK-8287418): riscv: Fix correctness issue of MacroAssembler::movptr (Bug - P3)
> > [JDK-8293100](https://bugs.openjdk.org/browse/JDK-8293100): RISC-V: Need to save and restore callee-saved FloatRegisters in StubGenerator::generate_call_stub (Bug - P3)
> > [JDK-8295926](https://bugs.openjdk.org/browse/JDK-8295926): RISC-V: C1: Fix LIRGenerator::do_LibmIntrinsic (Bug - P2)
> > ```
> 
> Yep. Based on the history of the commits, I'd consider the current issue list as okay. Most of the work this patch does is rolling back the backend from JDK19 -> JDK11. So mostly it is reverting patches.

I see some issues are missing here, like the ones previously mentioned in comment: https://github.com/openjdk/riscv-port-jdk11u/pull/3#issuecomment-1715452818.
Also I don't understand why we have the following two issues in the list?

[JDK-8202976](https://bugs.openjdk.org/browse/JDK-8202976): Add C1 lea patching support for x86 (Enhancement - P4)
[JDK-8243155](https://bugs.openjdk.org/browse/JDK-8243155): AArch64: Add support for SqrtVF (Enhancement - P4)


> (BTW the description of this PR is a little erroneous, it's not `initial patch based on openjdk/jdk17u-dev#1427`, but instead, based on the JDK19 initial load [openjdk/jdk at 5905b02](https://github.com/openjdk/jdk/commit/5905b02c0e2643ae8d097562f181953f6c88fc89)).

@kuaiwei : Could you please update the PR description? Thanks.

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

PR Comment: https://git.openjdk.org/riscv-port-jdk11u/pull/3#issuecomment-1931574487


More information about the riscv-port-dev mailing list