RFR: Remove Leyden experimental workflow
Remove the Leyden experimental workflow: `-XX:CacheDataStore` switch and friends. We should be using the standard JEP 483/514 workflow for AOT cache. ------------- Commit messages: - Removed native code for handling Leyden experimental workflow - Remove leyden workflow from tests Changes: https://git.openjdk.org/leyden/pull/83/files Webrev: https://webrevs.openjdk.org/?repo=leyden&pr=83&range=00 Stats: 669 lines in 25 files changed: 22 ins; 596 del; 51 mod Patch: https://git.openjdk.org/leyden/pull/83.diff Fetch: git fetch https://git.openjdk.org/leyden.git pull/83/head:pull/83 PR: https://git.openjdk.org/leyden/pull/83
On Fri, 27 Jun 2025 17:28:30 GMT, Ioi Lam <iklam@openjdk.org> wrote:
Remove the Leyden experimental workflow: `-XX:CacheDataStore` switch and friends.
We should be using the standard JEP 483/514 workflow for AOT cache.
Marked as reviewed by asmehra (Committer). Looks good. @iklam I wonder if should move LeydenGCFlags to mainstream. It is not testing anything specific to leyden repo. ------------- PR Review: https://git.openjdk.org/leyden/pull/83#pullrequestreview-2967731510 PR Comment: https://git.openjdk.org/leyden/pull/83#issuecomment-3014173858
On Fri, 27 Jun 2025 19:32:45 GMT, Ashutosh Mehra <asmehra@openjdk.org> wrote:
Looks good. @iklam I wonder if should move LeydenGCFlags to mainstream. It is not testing anything specific to leyden repo.
If we decide to move we should rename it to use `AOT` instead of `Leyden` to avoid confusion. I think we should replace all instances of `leyden` with `AOT`. ------------- PR Comment: https://git.openjdk.org/leyden/pull/83#issuecomment-3014201786
On Fri, 27 Jun 2025 19:46:55 GMT, Vladimir Kozlov <kvn@openjdk.org> wrote:
Looks good. @iklam I wonder if should move LeydenGCFlags to mainstream. It is not testing anything specific to leyden repo.
If we decide to move we should rename it to use `AOT` instead of `Leyden` to avoid confusion. I think we should replace all instances of `leyden` with `AOT`.
OK, I will rename Leyden to AOT in a follow-up changeset. I will to keep this one simple. ------------- PR Comment: https://git.openjdk.org/leyden/pull/83#issuecomment-3014212236
On Fri, 27 Jun 2025 17:28:30 GMT, Ioi Lam <iklam@openjdk.org> wrote:
Remove the Leyden experimental workflow: `-XX:CacheDataStore` switch and friends.
We should be using the standard JEP 483/514 workflow for AOT cache.
Few comments otherwise it is good. Please wait until mainline is merged. Please enable GHA testing src/hotspot/share/cds/cdsConfig.cpp line 395:
393: if (FLAG_IS_DEFAULT(AOTCache) && AOTStubCaching) { 394: log_debug(aot,codecache,init)("AOTCache is not specified - AOTStubCaching is ignored"); 395: }
Correct removal. This was incorrect merge from mainline. src/hotspot/share/cds/cdsConfig.cpp line 467:
465: } 466: if (FLAG_IS_DEFAULT(AOTCache) && AOTStubCaching) { 467: log_debug(aot,codecache,init)("AOTCache is not specified - AOTStubCaching is ignored");
You can remove these checks. There is similar check in `AOTCodeCache::initialize()` for all types of AOT code. ------------- PR Review: https://git.openjdk.org/leyden/pull/83#pullrequestreview-2967705431 PR Comment: https://git.openjdk.org/leyden/pull/83#issuecomment-3014180472 PR Review Comment: https://git.openjdk.org/leyden/pull/83#discussion_r2172719582 PR Review Comment: https://git.openjdk.org/leyden/pull/83#discussion_r2172729434
On Fri, 27 Jun 2025 19:31:23 GMT, Vladimir Kozlov <kvn@openjdk.org> wrote:
Remove the Leyden experimental workflow: `-XX:CacheDataStore` switch and friends.
We should be using the standard JEP 483/514 workflow for AOT cache.
src/hotspot/share/cds/cdsConfig.cpp line 467:
465: } 466: if (FLAG_IS_DEFAULT(AOTCache) && AOTStubCaching) { 467: log_debug(aot,codecache,init)("AOTCache is not specified - AOTStubCaching is ignored");
You can remove these checks. There is similar check in `AOTCodeCache::initialize()` for all types of AOT code.
Can you remove them in your merge? ------------- PR Review Comment: https://git.openjdk.org/leyden/pull/83#discussion_r2172771276
On Fri, 27 Jun 2025 19:52:37 GMT, Ioi Lam <iklam@openjdk.org> wrote:
src/hotspot/share/cds/cdsConfig.cpp line 467:
465: } 466: if (FLAG_IS_DEFAULT(AOTCache) && AOTStubCaching) { 467: log_debug(aot,codecache,init)("AOTCache is not specified - AOTStubCaching is ignored");
You can remove these checks. There is similar check in `AOTCodeCache::initialize()` for all types of AOT code.
Can you remove them in your merge?
Yes, I will do. ------------- PR Review Comment: https://git.openjdk.org/leyden/pull/83#discussion_r2172821157
On Fri, 27 Jun 2025 17:28:30 GMT, Ioi Lam <iklam@openjdk.org> wrote:
Remove the Leyden experimental workflow: `-XX:CacheDataStore` switch and friends.
We should be using the standard JEP 483/514 workflow for AOT cache.
This pull request has now been integrated. Changeset: 03c66c7a Author: Ioi Lam <iklam@openjdk.org> URL: https://git.openjdk.org/leyden/commit/03c66c7a9b905e254342ea7d07216966a5800e... Stats: 669 lines in 25 files changed: 22 ins; 596 del; 51 mod Remove Leyden experimental workflow Reviewed-by: asmehra ------------- PR: https://git.openjdk.org/leyden/pull/83
participants (3)
-
Ashutosh Mehra
-
Ioi Lam
-
Vladimir Kozlov