devxy/build-cran-binaries
Watch
1
0
Fork
You've already forked build-cran-binaries
 0
Code Issues 11 Pull requests Releases Activity

test(verify): report uneven coverage instead of failing on it

Browse source 
Same mistake as the source-fallback share. The packages a non-primary
minor lacks are ABI-risky ones built under the slot's primary minor,
which a client on another minor cannot safely load: their absence is
correct, not a defect. Failing on it blocked resolute and alpine324 on
both arches while they regressed nobody.

Parity gates the claim we make about coverage; MAX_REGRESSIONS gates
whether routing is safe to enable. Only the second belongs in the exit
code.
This commit is contained in:
Patrick Schratz 2026-08-31 09:59:55 +00:00
commit ed131e19fa
No known key found for this signature in database
GPG key ID: 62050D5BC68AB6DC
1 changed files with 9 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

14
scripts/verify-r-minor-routing.sh
View file

@ -43,10 +43,13 @@ MINORS=${MINORS:-"4.4 4.5 4.6"}
EXCLUDED_MINORS=${EXCLUDED_MINORS:-"4.3"}
# How many Path: targets to HEAD-check per slot/minor. 0 disables.
SAMPLE=${SAMPLE:-5}
# Largest package-count shortfall a non-primary minor may have against the best
# minor on the same slot before coverage counts as uneven. A slot built under
# one R minor carries fewer per-minor binaries for the others; until that gap
# closes, "full coverage for ABI-sensitive packages" is not a claim we can make.
# Package-count shortfall against the best minor on the same slot, above which
# coverage is reported as uneven. Reported, not failed on: the packages a
# non-primary minor lacks are ABI-risky ones built under the primary minor,
# which a client on another minor cannot safely load anyway, so their absence
# is correct. This gates the *claim* ("full coverage for ABI-sensitive
# packages"), not whether routing is safe to enable - MAX_REGRESSIONS does
# that.
PARITY_TOLERANCE=${PARITY_TOLERANCE:-25}
# Source fallbacks are reported, not failed on. Since bincraft learned to keep
# a matching-minor generic binary out of a fallback's shadow, a remaining
@ -299,7 +302,8 @@ for arch in $ARCHES; do
[ "$gap" -gt "$PARITY_TOLERANCE" ] && uneven="$uneven R$minor:-$gap"
done
if [ -n "$uneven" ]; then
bad "$slot coverage uneven across minors (vs best $best):$uneven"
printf ' note: %s coverage uneven across minors (vs best %s):%s\n' \
"$slot" "$best" "$uneven"
else
ok "$slot coverage parity across minors (best $best, all within $PARITY_TOLERANCE)"
fi