CloudMarc opened a new pull request, #5364: URL: https://github.com/apache/solr/pull/5364
## What Documents `-XX:-UseBASE64Intrinsics` as an opt-in, commented-out `SOLR_OPTS` setting in `solr.in.sh`/`solr.in.cmd`, mirroring the existing pattern this repo already uses for CVE-2021-44228 (the `log4j2.formatMsgNoLookups` line a few lines above it). ## Why [CVE-2026-114938](https://openjdk.org/groups/vulnerability/advisories/2026-11-17) ("Java4Shell"), disclosed 2026-11-17, is a critical heap-overflow RCE in HotSpot's AVX-512 VBMI Base64 decoder intrinsic. It affects OpenJDK 17.0.6+, 21, 25 and 27 (not 8 or 11) on x86-64 hosts with AVX-512 VBMI (Intel Ice Lake+ Xeon, AMD Zen 4+ EPYC/Ryzen). The disclosing researchers (Calif) demonstrated **post-auth RCE against Apache Solr** specifically. Both `BasicAuthPlugin` and `PKIAuthenticationPlugin` call `Base64.getDecoder().decode()` directly on request-supplied input, which is real exposed surface regardless of exactly which code path the researchers used. As of this PR, no JDK vendor has a GA build with the fix for the 17/21/25 lines — I checked the Adoptium API directly and Temurin only has early-access/beta builds (e.g. `25.0.5.1-beta+1-ea`); Temurin 27 is the one exception already GA. So there's no dependency/base-image bump that resolves this today, and operators need a flag-based mitigation in the meantime. `-XX:-UseBASE64Intrinsics` is a generic HotSpot product flag (`globals.hpp`, checked in the JDK source), honored identically across x86, aarch64, ppc and riscv builds — so turning it off is a safe no-op on platforms this CVE doesn't affect, not just an x86-64-specific workaround. ## Scope Deliberately minimal and non-invasive: a commented-out example a user must opt into, exactly like the existing Log4Shell line, not a default-on behavior change. Didn't touch `bin/solr` itself or add any runtime detection/warning — open to feedback on whether this project wants to go further (e.g. a startup-time check that warns when running on a vulnerable JDK+CPU combination). ## Notes for reviewers - I couldn't find an existing SOLR JIRA ticket for this CVE (searched `project=SOLR AND text~"Java4Shell"`, no hits), so this PR doesn't carry a `SOLR-NNNNN` prefix. Happy to file one if a committer wants it tracked that way — didn't want to guess at scope/priority on the project's own tracker. - Opening as a **draft** for discussion rather than asserting this is the final word on how the project wants to handle a JDK-level CVE with no code fix available. - No changelog fragment included yet; `dev-docs/changelog.adoc` suggests running `./gradlew writeChangelogPr` after the PR exists when there's no JIRA — happy to add that once this has a PR number and there's agreement on the approach. Co-Authored-By: Claude Sonnet 5 <[email protected]> -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
