Rongsong Shen
d26ff8870d
The buglog was one 27K-line JSON file whose fields covered only symptom,
root cause and fix, as free text: 2% of entries named the code version a
bug was found on, 12% the build profile, 36% a way to reproduce, and 58% of
the file was OpenWolf hook auto-detections ("Missing guard clause",
"Function not marked async" on .md files).
- .claude/skills/buglog/bug-record.md: the record standard. Eight
sections (Summary as the title, Description, Environment, Reproduction;
Root cause, Fix, Verification, References once fixed), each with the
content it must contain -- e.g. Environment needs the OS, the libc the
failing binary loads (from its ELF interpreter: glibc 2.42 in the nix
shell vs 2.40 on the host), toolchain, build profile and commit.
- .wolf/bugs/<id>.md: 500 records migrated from buglog.json; every field
verified to round-trip. They carry legacy: true and an explicit "not
recorded" where the old log had no such field; component derived from
file paths, severity from keywords (else unknown), fix_commits from
hashes in status_note/fix that git resolves (91 of 463 fixed).
- .wolf/buglog-archive.json: the 710 hook auto-detections, verbatim.
- .wolf/buglog.json is now an index generated from the records, keeping
the fields OpenWolf's hooks and `openwolf bug search` read.
- scripts/buglog.py rewritten: new (template + Environment collected from
the machine and --binary), check, status (refuses a status the record
doesn't support; fixed needs Root cause, Fix, Verification and
--commit hashes git resolves), note, recur, set, show, list, triage,
env, index. 29 unittest cases in scripts/tests/test_buglog.py.
- scripts/git-hooks/pre-commit runs `check --staged`: incomplete staged
records, a legacy record hand-changed to fixed without the fix-time
sections, a new record claiming legacy: true, or a stale index refuse
the commit.
- scripts/gitea-sync.py reads the records: issue body = the record's
sections plus linked fix commits (origin remote, GITEA_COMMIT_REPO to
override), and a body hash so an issue's text is refreshed when the
record changes.
- bug-1866 rewritten to the full standard as the worked example.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
39 lines
3.9 KiB
Markdown
39 lines
3.9 KiB
Markdown
---
|
|
id: "bug-1605"
|
|
status: "cannot-reproduce"
|
|
status_updated: "2026-09-24"
|
|
status_note: "transient build failure, passed on immediate retry; likely bug-1604 collision"
|
|
component: "lib"
|
|
severity: "unknown"
|
|
timestamp: "2026-08-20T10:27:00.000Z"
|
|
last_seen: "2026-09-19T09:00:00Z"
|
|
occurrences: 1
|
|
files: ["src/lib/controller.capnp.c (generated)", "src/controller/config_repl.c (consumer)"]
|
|
tags: ["build", "capnp", "codegen", "concurrent-build-collision", "controller", "needs-followup", "ninja", "pre-existing", "race", "transient"]
|
|
related: ["bug-1439", "bug-1493", "bug-1604"]
|
|
repro_attempts: "1 retry, clean"
|
|
legacy: true
|
|
---
|
|
|
|
# bug-1605: controller ASAN build: two failures in the same ninja run -- (1) controller/config_repl.c:1921:28…
|
|
|
|
## Description
|
|
controller ASAN build: two failures in the same ninja run -- (1) controller/config_repl.c:1921:28: error: 'RequestReply_data_probeconfigtagreplymsg' undeclared (first use in this function); did you mean 'RequestReply_data_upgradeconfigreplymsg' -- and (2) lib/controller.capnp.c:25013:1: fatal error: opening dependency file lib/libxce.a.p/meson-generated_.._controller.capnp.c.o.d: No such file or directory, compilation terminated; ninja: build stopped: subcommand failed.
|
|
|
|
## Environment
|
|
_Not recorded (migrated from buglog.json)._
|
|
|
|
## Reproduction
|
|
- Attempts: 1 retry, clean
|
|
|
|
## Root cause
|
|
Matches the capnp custom_target's documented missing-dependency-edge race (see comment block at top of src/tools/xce-build.sh, 'retry once before debugging'): config_repl.c was compiled against a stale/partially-regenerated controller.capnp.h that lacked the 'probeconfigtagreplymsg' enumerant even though src/schema/controller.capnp:596 defines it (probeconfigtagreplymsg @20: ProbeConfigTagReply), and lib/controller.capnp.c's own compile then hit an I/O race writing its .d dependency file, most likely because the generator custom_target re-ran/rewrote lib/ mid-build without ninja having a correct dependency edge ordering the two. Not caused by the concurrent main.c/testsuite/test.c argp_program_version diff (those files are unrelated to config_repl.c/capnp). CORRECTION (same session, minutes later): discovered via `ps aux` a THIRD live instance of a concurrent `nix develop ./src --command bash src/tools/xce-build.sh controller asan` process actively doing rm -rf xce.build + fresh meson compile in this SAME shared worktree at 10:38, i.e. immediately after this bug's own two build attempts (10:23-10:33) -- and .wolf/memory.md/cerebrum.md already documented an earlier occurrence THIS SAME SESSION of exactly this hazard (bug-1604: two concurrent xce-build.sh runs sharing default BUILD_DIR=xce.build race on rm -rf, one silently wiping/regenerating the other's in-progress build, corrupting or vanishing already-verified binaries). The 'known capnp codegen dependency-edge race' explanation above is therefore NOT confirmed as the true root cause -- a concurrent cross-session xce-build.sh collision (bug-1604) is at least as likely, and possibly the actual cause, of both the stale-header undeclared-enum error and the .d-file open I/O failure (a concurrent rm -rf mid-compile would produce exactly both symptoms). Do not treat this entry as proof the capnp custom_target race is real without ruling out concurrent-session collision first -- check `ps aux | grep xce-build.sh` before attributing a build failure to codegen internals in this worktree.
|
|
|
|
## Fix
|
|
Confirmed transient: re-ran the exact same command (nix develop ./src --command bash src/tools/xce-build.sh controller asan) a second time with zero source changes in between and the full clean build compiled successfully (202/202, zero errors). Not root-caused/fixed in the underlying capnp custom_target dependency graph -- xce-build.sh's own comments already document this as a known, unfixed gap and recommend exactly this retry-once workaround.
|
|
|
|
## Verification
|
|
_Not recorded (migrated from buglog.json)._
|
|
|
|
## References
|
|
- Related: bug-1439, bug-1493, bug-1604
|