[jdk17u-dev] RFR: 8299570: [JVMCI] Insufficient error handling when CodeBuffer is exhausted

Aleksey Shipilev shade at openjdk.org
Wed May 3 09:32:51 UTC 2023


On Fri, 28 Apr 2023 11:03:36 GMT, Goetz Lindenmaier <goetz at openjdk.org> wrote:

> I backport this for parity with 17.0.8-oracle.
> 
> I resolved three files due to context and NULL/nullptr.

Questions! :)

src/hotspot/cpu/aarch64/nativeInst_aarch64.cpp line 532:

> 530: // Generate a trampoline for a branch to dest.  If there's no need for a
> 531: // trampoline, simply patch the call directly to dest.
> 532: void NativeCall::trampoline_jump(CodeBuffer &cbuf, address dest, JVMCI_TRAPS) {

In the original changeset, this hunk has the `#if INCLUDE_JVMCI` addition. Plus `#endif` later.

src/hotspot/cpu/aarch64/nativeInst_aarch64.cpp line 534:

> 532: void NativeCall::trampoline_jump(CodeBuffer &cbuf, address dest, JVMCI_TRAPS) {
> 533:   MacroAssembler a(&cbuf);
> 534:   

Yeah, original changeset does not have a newline here?

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

Changes requested by shade (Reviewer).

PR Review: https://git.openjdk.org/jdk17u-dev/pull/1308#pullrequestreview-1405783813
PR Review Comment: https://git.openjdk.org/jdk17u-dev/pull/1308#discussion_r1180306394
PR Review Comment: https://git.openjdk.org/jdk17u-dev/pull/1308#discussion_r1180306805


More information about the jdk-updates-dev mailing list