Repository navigation
openjdk-21/25/26, ghc-9.6.7 -> gen-15; drop the glibc-bedrock-2.42-sysroot package - #1462
Conversation
…sroot package gen-15 of gominimal/bootstrap builds the amd64 openjdk-21/25/26 and ghc-9.6.7 against the system glibc with the gcc seed: PT_INTERP /usr/lib/ld-linux-x86-64.so.2, no RUNPATH beyond their own $ORIGIN entries, and ghc-cc a plain gcc wrapper that adds that loader. The four pins move to gen-15 on both arches (arm64's objects are the gen-13 bytes under their gen-15 names, ghc-9.6.7's rebuilt from the shared script and identical apart from BUILDINFO) and their amd64 runtime dep becomes glibc. ghc drops the sysroot from its amd64 build deps; its check that nothing shipped names the sysroot stays. Nothing imports glibc-bedrock-2.42-sysroot any more, so the package goes. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> (cherry picked from commit 1151b19)
|
Warning Review limit reached
This review includes 7 billable files and costs up to $1.75.
Reviews can continue after your included limit without a manual trigger. An admin must approve usage-based billing. Or wait 15 minutes for your next included review. View limit details
|
The arm64 JDK objects are the gen-13 bytes re-published under gen-15 (bootstrap generations/gen-15/manifest.json lists each sha256 at its gen-15 url), so the url moves and the sha256 correctly stays. Comments only; no spec hash changes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
norrietaylor
left a comment
There was a problem hiding this comment.
Scope matches the title and body. The diff moves the amd64 tops of openjdk-21/25/26 and ghc-9.6.7 to gen-15 and swaps the sysroot runtime dep for glibc. It moves the arm64 pins to gen-15 and deletes glibc-bedrock-2.42-sysroot. Clobber none found. git log on the touched packages shows #1242 (gen-13) as the last change, and this PR builds on it. Correctness rg finds no remaining importer of glibc-bedrock-2.42-sysroot, so the delete leaves no dangling import. The three # source-hash-guard: allow markers match MARKER in .github/scripts/source_hash_guard.py:85: added line, matching sha256, and a reason. The amd64 "nothing shipped names the bedrock sysroot" check in packages/ghc/build.sh stays in place. The new openjdk header comments call each package "the boot JDK for jdk-NN", but nothing imports openjdk-21/25/26. One stale reference sits outside the diff: packages/ghc/build.sh:65 still says "the gen-13 prebuilt". Not checked I did not verify that the gen-15 amd64 objects really have the stated PT_INTERP and RUNPATH. That needs a fetch and a build. The ghc build.sh guard covers the ghc outputs at build time.
| let libxtst = import "../libxtst/build.ncl" in | ||
| let libxcursor = import "../libxcursor/build.ncl" in | ||
| # openjdk-21: OpenJDK 21.0.2 from gen-13 of gominimal/bootstrap: the JVM ladder | ||
| # openjdk-21: the boot JDK for jdk-21, OpenJDK 21.0.2 from gen-15 of gominimal/bootstrap: the JVM ladder |
There was a problem hiding this comment.
low · The new header calls openjdk-21 "the boot JDK for jdk-21". This is not true. packages/jdk-21/build.ncl downloads the Temurin 21.0.13 binary and imports no openjdk package. The PR body also says "nothing imports openjdk-21/25/26". The same false claim is in packages/openjdk-25/build.ncl:17 ("boot JDK for jdk-25") and packages/openjdk-26/build.ncl:17 ("boot JDK for jdk-27"). The next reader can conclude that a change here rebuilds the jdk-NN packages. Remove the phrase from all three headers, or state that no package consumes it today.
Supersedes #1322 (same change, replayed onto main after #1242 squash-merged; that branch carried #1242's 81 commits and conflicted).
The amd64 openjdk-21/25/26 and ghc-9.6.7 tops move to gen-15 of gominimal/bootstrap, where they were rebuilt against the system glibc with the plain loader:
PT_INTERP /usr/lib/ld-linux-x86-64.so.2, only$ORIGINRUNPATHs,ghc-ccruns the gcc on PATH. They now takeglibcas their runtime dep like arm64 does, so theglibc-bedrock-2.42-sysrootpackage has no consumers and is deleted;ghcdrops it as a build dep and its build.sh keeps the "nothing shipped names the sysroot" check. The arm64 pins move to the same objects under the gen-15 prefix (ghc-9.6.7 arm64 is byte-identical to gen-14 except BUILDINFO).Moves 10 packages (
mip dump --packages, amd64, measured 2026-11-12): openjdk-21, openjdk-25, openjdk-26, ghc-9.6.7, ghc, cabal, stack, haskell-language-server, tamarin-prover and dev-language-servers. jdk-21/25/27 do not move: on main they are Temurin binaries (#1137), and nothing imports openjdk-21/25/26. gen-13 objects pinned elsewhere are unchanged.The source-hash guard flags the arm64 openjdk pins because their url moves while the sha256 stays. That is correct here: gen-15 re-publishes the gen-13 arm64 objects unchanged (bootstrap
generations/gen-15/manifest.jsonlists each sha256 at its gen-15 url), so each carries a# source-hash-guard: allowcomment.After this, nothing in pkgs reads the hex0 → stage0-* → glibc-bedrock-2.42 chain; #1442 removes its last three orphans and the remaining 24 can go in one follow-up PR.
🤖 Generated with Claude Code