• Home
  • Features
  • Pricing
  • Docs
  • Announcements
  • Sign In

SRI-CSL / yices2 / 25536263059
67%
master: 71%

Build:
Build:
LAST BUILD BRANCH: fix-bv64-interval-abstraction-ub
DEFAULT BRANCH: master
Ran 08 May 2026 04:21AM UTC
Jobs 1
Files 486
Run time 1min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst

08 May 2026 04:12AM UTC coverage: 67.087% (+0.008%) from 67.079%
25536263059

push

github

disteph
mcsat: gate preprocessor diag prints on YICES_PP_TRACE; trim windows CI

Two fixes / aids for the windows-latest|release|--enable-thread-safety
--enable-mcsat mcsat_tuples failure investigation:

1. Gate every preprocessor.c TEMP DIAG stderr print behind an env var
   YICES_PP_TRACE. The existing prints were unconditional, so they
   leaked into the regression-test stderr stream and broke ~21 mcsat
   regression tests (`make test`) that diff stderr against expected
   output. With the gate, the prints are silent unless the env var is
   set, so:

     - `make test` regressions are clean again on every CI config
     - the prints are still available when needed: set
       YICES_PP_TRACE=1 (any non-empty / non-"0" value) before running
       the failing binary.

   Implementation: a tiny static helper `pp_diag_enabled()` reads the
   env var once (cached in a static int -- benign one-time race under
   --enable-thread-safety), and a `PP_DIAG(...)` macro wraps
   fprintf(stderr,...)+fflush(stderr). All call sites switched in
   build_value_from_flat, preprocessor_build_tuple_model, and
   preprocessor_build_model.

   Verified locally:
     mcsat_tuples                          -> 0   "[pp_build|build_value" lines
     YICES_PP_TRACE=1 mcsat_tuples         -> 631 "[pp_build|build_value" lines

2. Trim Windows CI to only the failing config and only the API tests
   so each iteration takes ~10 minutes instead of an hour:

     - Matrix: mode=[release], config-opt=[--enable-thread-safety
       --enable-mcsat] (was 2x3=6 jobs, now 1).
     - Test step: just `make check-api` with YICES_PP_TRACE=1
       exported, so mcsat_tuples emits the full preprocessor trace.
       The `make test` regression sweep is dropped for now.

   Both changes are TEMP and clearly commented; the original matrix
   and test step are restored at the end of this debugging session.

43 of 49 new or added lines in 1 file covered. (87.76%)

76 existing lines in 1 file now uncovered.

84692 of 126242 relevant lines covered (67.09%)

1633053.41 hits per line

Uncovered Changes

Lines Coverage ∆ File
6
81.64
0.18% src/mcsat/preprocessor.c

Coverage Regressions

Lines Coverage ∆ File
76
81.64
0.18% src/mcsat/preprocessor.c
Jobs
ID Job ID Ran Files Coverage
1 25536263059.1 08 May 2026 04:21AM UTC 486
67.09
GitHub Action Run
Source Files on build 25536263059
  • Tree
  • List 486
  • Changed 1
  • Source Changed 0
  • Coverage Changed 1
Coverage ∆ File Lines Relevant Covered Missed Hits/Line
  • Back to Repo
  • f7f89fbd on github
  • Prev Build on mcsat-tuples-support (#25530974729)
  • Next Build on mcsat-tuples-support (#25538560463)
STATUS · Troubleshooting · Open an Issue · Sales · Support · CAREERS · ENTERPRISE · START FREE · SCHEDULE DEMO
ANNOUNCEMENTS · TWITTER · TOS & SLA · Supported CI Services · What's a CI service? · Automated Testing

© 2026 Coveralls, Inc