8373794: Move nmethod header from CodeCache #28866

mikabl-arm · 2025-12-17T11:34:46Z

nmethod objects in the CodeCache have the following layout:

| CodeBlob header | NMethod header | Constants | MainCode | StubCode | Data/oops |

Although mutable and immutable metadata have already been moved out of the code cache by JDK-8343789 and JDK-8331087 respectively, the embedded nmethod header fields still occupy ~160 B (with the CodeBlob header adding another 64 B). In JDK25 the total header footprint is 224 B. This space is reserved inside executable memory, which decreases overall executable code density.

This patch relocates the nmethod header to a C-heap-allocated structure and keeps only 8-byte pointer to that header in the CodeCache. The resulting layout is:

| CodeBlob header | Ptr to NMethodHeader | Constants | MainCode | StubCode | Data/oops |

This change reduces the size of the CodeCache-resident header from 224 B to 72 B (64 B CodeBlob header + 8 B pointer), achieving roughly a 3x reduction in header footprint.

This change follows the direction established by JDK-7072317, JDK-8331087 and JDK-8343789.

Testing

The patch has passed tier1-3 and hotspot_all tests on AArch64 and x86_64.

Progress

Change must be properly reviewed (1 review required, with at least 1 Reviewer)
Change must not contain extraneous whitespace
Commit message must refer to an issue

Issue

JDK-8373794: Move nmethod header from CodeCache (Enhancement - P4)

Contributors

Mikhail Ablakatov <mablakatov@openjdk.org>

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/28866/head:pull/28866
$ git checkout pull/28866

Update a local copy of the PR:
$ git checkout pull/28866
$ git pull https://git.openjdk.org/jdk.git pull/28866/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 28866

View PR using the GUI difftool:
$ git pr show -t 28866

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/28866.diff

Using Webrev

Link to Webrev Comment

nmethod objects in the CodeCache have the following layout: | CodeBlob header | NMethod header | Constants | MainCode | StubCode | Data/oops | Although mutable and immutable metadata have already been moved out of the code cache by JDK-8343789 and JDK-8331087 respectively, the embedded `nmethod` header fields still occupy ~160 B (with the `CodeBlob` header adding another 64 B). In JDK25 the total header footprint is 224 B. This space is reserved inside executable memory, which decreases overall executable code density. This patch relocates the `nmethod` header to a C-heap-allocated structure and keeps only 8-byte pointer to that header in the CodeCache. The resulting layout is: | CodeBlob header | Ptr to NMethodHeader | Constants | MainCode | StubCode | Data/oops | This change reduces the size of the CodeCache-resident header from 224 B to 72 B (64 B `CodeBlob` header + 8 B pointer), achieving roughly a 3x reduction in header footprint. This change follows the direction established by JDK-7072317, JDK-8331087 and JDK-8343789.

bridgekeeper · 2025-12-17T11:36:55Z

👋 Welcome back mablakatov! A progress list of the required criteria for merging this PR into master will be added to the body of your pull request. There are additional pull request commands available for use with this pull request.

openjdk · 2025-12-17T11:37:01Z

❗ This change is not yet ready to be integrated.
See the Progress checklist in the description for automated requirements.

openjdk · 2025-12-17T11:37:49Z

@mikabl-arm The following labels will be automatically applied to this pull request:

hotspot
serviceability

When this pull request is ready to be reviewed, an "RFR" email will be sent to the corresponding mailing lists. If you would like to change these labels, use the /label pull request command.

mikabl-arm · 2025-12-17T11:38:46Z

/author set bkilambi

mikabl-arm · 2025-12-17T11:39:13Z

/contributor add mablakatov

openjdk · 2025-12-17T11:39:24Z

@mikabl-arm Only Committers are allowed to issue the author command.

openjdk · 2025-12-17T11:39:38Z

@mikabl-arm
Contributor Mikhail Ablakatov <mablakatov@openjdk.org> successfully added.

mlbridge · 2025-12-17T11:43:25Z

Webrevs

00: Full (bbb3711d)

fg1417 · 2025-12-17T11:47:00Z

/author set bkilambi

openjdk · 2025-12-17T11:48:55Z

@fg1417 Only the pull request author (@mikabl-arm) is allowed to issue the author command.

plummercj

The SA changes look ok, but you'll need to do some testing by hand. The existing SA tests don't do a good job of testing compiler support. Specifically, you should run the clhsdb "livenmethods" and "dumpcodecache" commands and make sure the output looks reasonable. Note you may see an issue with "dumpcodecache" getting stuck after producing some output. You can ignore that. I just filed JDK-8373933 for it.

plummercj · 2025-12-17T17:14:29Z

src/jdk.hotspot.agent/share/classes/sun/jvm/hotspot/code/NMethod.java

 exceptionOffsetField = type.getCIntegerField("_exception_offset");
 deoptHandlerEntryOffsetField = type.getCIntegerField("_deopt_handler_entry_offset");
 origPCOffsetField = type.getCIntegerField("_orig_pc_offset");
 stubOffsetField = type.getCIntegerField("_stub_offset");
 scopesPCsOffsetField = type.getCIntegerField("_scopes_pcs_offset");
 scopesDataOffsetField = type.getCIntegerField("_scopes_data_offset");
+
+ sun.jvm.hotspot.types.Field f = type.getField("_speculations_offset", false, false);
+ if (f != null) {


Why is there a need for a null check here?

_speculations_offset is present only when JVM is compiled with JVMCI enabled, see https://github.com/openjdk/jdk/pull/28866/files/bbb3711df942ec48b0b0a6eeb287f5364afa9098#diff-ccbdfe031dfc301041cac88f76f8eb2dcacff1b9f81c50073de5c6eaeb8b8223R256 . type.getCIntegerField("_speculations_offset") will throw an exception if the field is missing.

Though, if the SA can't be used at all with JVMCI disabled, this check can be omitted as redundant. Please let me know if that's the case.

Though, if the SA can't be used at all with JVMCI disabled, this check can be omitted as redundant. Please let me know if that's the case.

I know of no reason why SA won't work with JVMCI disabled, but I'm also not sure if we test that so there could be bugs.

dean-long · 2025-12-18T02:23:43Z

Wouldn't it be better to move the whole CodeBlob header out of the code cache? Instead of nmethod having a _hdr pointer, CodeBlob would be in malloc space and have a _code pointer into the CodeCache.

mikabl-arm · 2025-12-18T17:15:35Z

Wouldn't it be better to move the whole CodeBlob header out of the code cache? Instead of nmethod having a _hdr pointer, CodeBlob would be in malloc space and have a _code pointer into the CodeCache.

It might be possible to move both CodeBlob and nmethod data to the C-heap. We'd still need something like an _hdr pointer, since JVM resolves a CodeBlob from a method/stub entry point using CodeBlob* CodeCache::find_blob(void* start). So if we move every member field to the C-heap, we store a pointer to that data right before a method/stub entry point in the CodeCache and that's it. A _code pointer pointing from the C-heap into the CodeCache may not be necessary at all.

That said, this would require moving data for other CodeBlob subclasses (AdapterBlob, ExceptionBlob, etc.) to the C-heap, which would significantly broaden the scope of this patch.

@Bhavana-Kilambi , what do you thing? Is this something you've considered?

openjdk bot added serviceability serviceability-dev@openjdk.org hotspot hotspot-dev@openjdk.org labels Dec 17, 2025

openjdk bot added the rfr Pull request is ready for review label Dec 17, 2025

plummercj reviewed Dec 17, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

8373794: Move nmethod header from CodeCache #28866

8373794: Move nmethod header from CodeCache #28866

mikabl-arm commented Dec 17, 2025 •

edited by openjdk bot

Loading

bridgekeeper bot commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

mikabl-arm commented Dec 17, 2025

mikabl-arm commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

mlbridge bot commented Dec 17, 2025

fg1417 commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

plummercj left a comment

plummercj Dec 17, 2025

mikabl-arm Dec 18, 2025

plummercj Dec 18, 2025

dean-long commented Dec 18, 2025

mikabl-arm commented Dec 18, 2025

Labels

5 participants

8373794: Move nmethod header from CodeCache #28866

Are you sure you want to change the base?

8373794: Move nmethod header from CodeCache #28866

Conversation

mikabl-arm commented Dec 17, 2025 • edited by openjdk bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Testing

Progress

Issue

Contributors

Reviewing

bridgekeeper bot commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

mikabl-arm commented Dec 17, 2025

mikabl-arm commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

mlbridge bot commented Dec 17, 2025

Webrevs

fg1417 commented Dec 17, 2025

openjdk bot commented Dec 17, 2025

plummercj left a comment

Choose a reason for hiding this comment

plummercj Dec 17, 2025

Choose a reason for hiding this comment

mikabl-arm Dec 18, 2025

Choose a reason for hiding this comment

plummercj Dec 18, 2025

Choose a reason for hiding this comment

dean-long commented Dec 18, 2025

mikabl-arm commented Dec 18, 2025

Labels

5 participants

mikabl-arm commented Dec 17, 2025 •

edited by openjdk bot

Loading