FAILURE MAP
← Case archive

FA-16531 / Floating-point arithmetic / Open access

Clearing one floating flag clears all accrued flags · case 01

Clearing one floating flag clears all accrued flags.

Verified by executionVariant 1 · 10 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

Clearing one floating flag clears all accrued flags. The faulty expression is remaining=set() if clear else accrued.

THE FAILURE

Clearing one floating flag clears all accrued flags. The faulty expression is remaining=set() if clear else accrued.

Unsuccessful approach: The attempted local correction remaining=accrued & set(clear) still violates the explicit regression fixtures.

Case contract

Stipulated floating exception environment: apply an operation flag set, accrue sticky flags, select enabled trap by priority invalid/divzero/overflow/underflow/inexact, save prior state and clear only requested flags after reporting the trap. Inputs are valid flag names.

Why this case matters

An offline floating representation model isolates a reproducible arithmetic fault.

1 / The failure

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json
import math
import struct
N = 1
observations = []
def solve(sticky, enabled, raised, clear):
    priority=['invalid','divzero','overflow','underflow','inexact']
    prior=set(sticky)
    new=set(raised)
    if 'overflow' in new: new.add('inexact')
    if 'underflow' in new: new.add('inexact')
    accrued=prior|new
    pending=new & set(enabled)
    trap=next((flag for flag in priority if flag in pending),None)
    remaining=set() if clear else accrued
    return [sorted(prior),sorted(accrued),trap,sorted(remaining)]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('sticky survives', solve(["invalid"],[],["inexact"],[]), [["invalid"],["inexact","invalid"],None,["inexact","invalid"]])
check('overflow implies inexact', solve([],[],["overflow"],[]), [[],["inexact","overflow"],None,["inexact","overflow"]])
check('underflow implies inexact', solve([],[],["underflow"],[]), [[],["inexact","underflow"],None,["inexact","underflow"]])
check('priority', solve([],["overflow","invalid"],["overflow","invalid"],[]), [[],["inexact","invalid","overflow"],"invalid",["inexact","invalid","overflow"]])
check('old flag no retrap', solve(["invalid"],["invalid"],["inexact"],[]), [["invalid"],["inexact","invalid"],None,["inexact","invalid"]])
check('clear after trap', solve([],["divzero"],["divzero"],["divzero"]), [[],["divzero"],"divzero",[]])
check('selective clear', solve(["invalid","inexact"],[],[],["inexact"]), [["inexact","invalid"],["inexact","invalid"],None,["invalid"]])
check('masked flag still sticky', solve([],[],["divzero"],[]), [[],["divzero"],None,["divzero"]])
check('five distinct flag sets', solve([],[],["invalid","divzero","inexact","underflow","overflow"][:N],[]), [[],sorted(["invalid","divzero","inexact","underflow","overflow"][:N]),None,sorted(["invalid","divzero","inexact","underflow","overflow"][:N])])
check('empty operation', solve(["inexact"] if N%2 else ["invalid"],[],[],[]), [["inexact"] if N%2 else ["invalid"],["inexact"] if N%2 else ["invalid"],None,["inexact"] if N%2 else ["invalid"]])
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
sticky survives[['invalid'], ['inexact', 'invalid'], None, ['inexact', 'invalid']][['invalid'], ['inexact', 'invalid'], None, ['inexact', 'invalid']]Passed
overflow implies inexact[[], ['inexact', 'overflow'], None, ['inexact', 'overflow']][[], ['inexact', 'overflow'], None, ['inexact', 'overflow']]Passed
underflow implies inexact[[], ['inexact', 'underflow'], None, ['inexact', 'underflow']][[], ['inexact', 'underflow'], None, ['inexact', 'underflow']]Passed
priority[[], ['inexact', 'invalid', 'overflow'], 'invalid', ['inexact', 'invalid', 'overflow']][[], ['inexact', 'invalid', 'overflow'], 'invalid', ['inexact', 'invalid', 'overflow']]Passed
old flag no retrap[['invalid'], ['inexact', 'invalid'], None, ['inexact', 'invalid']][['invalid'], ['inexact', 'invalid'], None, ['inexact', 'invalid']]Passed
clear after trap[[], ['divzero'], 'divzero', []][[], ['divzero'], 'divzero', []]Passed
selective clear[['inexact', 'invalid'], ['inexact', 'invalid'], None, []][['inexact', 'invalid'], ['inexact', 'invalid'], None, ['invalid']]Failed
masked flag still sticky[[], ['divzero'], None, ['divzero']][[], ['divzero'], None, ['divzero']]Passed
five distinct flag sets[[], ['invalid'], None, ['invalid']][[], ['invalid'], None, ['invalid']]Passed
empty operation[['inexact'], ['inexact'], None, ['inexact']][['inexact'], ['inexact'], None, ['inexact']]Passed

SHA-256 / 6bb41fe575e29b4adc973e956820f43b3b001550051f228f6a287c48a49936c8

2 / The unsuccessful fix

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json
import math
import struct
N = 1
observations = []
def solve(sticky, enabled, raised, clear):
    priority=['invalid','divzero','overflow','underflow','inexact']
    prior=set(sticky)
    new=set(raised)
    if 'overflow' in new: new.add('inexact')
    if 'underflow' in new: new.add('inexact')
    accrued=prior|new
    pending=new & set(enabled)
    trap=next((flag for flag in priority if flag in pending),None)
    remaining=accrued & set(clear)
    return [sorted(prior),sorted(accrued),trap,sorted(remaining)]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('sticky survives', solve(["invalid"],[],["inexact"],[]), [["invalid"],["inexact","invalid"],None,["inexact","invalid"]])
check('overflow implies inexact', solve([],[],["overflow"],[]), [[],["inexact","overflow"],None,["inexact","overflow"]])
check('underflow implies inexact', solve([],[],["underflow"],[]), [[],["inexact","underflow"],None,["inexact","underflow"]])
check('priority', solve([],["overflow","invalid"],["overflow","invalid"],[]), [[],["inexact","invalid","overflow"],"invalid",["inexact","invalid","overflow"]])
check('old flag no retrap', solve(["invalid"],["invalid"],["inexact"],[]), [["invalid"],["inexact","invalid"],None,["inexact","invalid"]])
check('clear after trap', solve([],["divzero"],["divzero"],["divzero"]), [[],["divzero"],"divzero",[]])
check('selective clear', solve(["invalid","inexact"],[],[],["inexact"]), [["inexact","invalid"],["inexact","invalid"],None,["invalid"]])
check('masked flag still sticky', solve([],[],["divzero"],[]), [[],["divzero"],None,["divzero"]])
check('five distinct flag sets', solve([],[],["invalid","divzero","inexact","underflow","overflow"][:N],[]), [[],sorted(["invalid","divzero","inexact","underflow","overflow"][:N]),None,sorted(["invalid","divzero","inexact","underflow","overflow"][:N])])
check('empty operation', solve(["inexact"] if N%2 else ["invalid"],[],[],[]), [["inexact"] if N%2 else ["invalid"],["inexact"] if N%2 else ["invalid"],None,["inexact"] if N%2 else ["invalid"]])
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
sticky survives[['invalid'], ['inexact', 'invalid'], None, []][['invalid'], ['inexact', 'invalid'], None, ['inexact', 'invalid']]Failed
overflow implies inexact[[], ['inexact', 'overflow'], None, []][[], ['inexact', 'overflow'], None, ['inexact', 'overflow']]Failed
underflow implies inexact[[], ['inexact', 'underflow'], None, []][[], ['inexact', 'underflow'], None, ['inexact', 'underflow']]Failed
priority[[], ['inexact', 'invalid', 'overflow'], 'invalid', []][[], ['inexact', 'invalid', 'overflow'], 'invalid', ['inexact', 'invalid', 'overflow']]Failed
old flag no retrap[['invalid'], ['inexact', 'invalid'], None, []][['invalid'], ['inexact', 'invalid'], None, ['inexact', 'invalid']]Failed
clear after trap[[], ['divzero'], 'divzero', ['divzero']][[], ['divzero'], 'divzero', []]Failed
selective clear[['inexact', 'invalid'], ['inexact', 'invalid'], None, ['inexact']][['inexact', 'invalid'], ['inexact', 'invalid'], None, ['invalid']]Failed
masked flag still sticky[[], ['divzero'], None, []][[], ['divzero'], None, ['divzero']]Failed
five distinct flag sets[[], ['invalid'], None, []][[], ['invalid'], None, ['invalid']]Failed
empty operation[['inexact'], ['inexact'], None, []][['inexact'], ['inexact'], None, ['inexact']]Failed

SHA-256 / 7df35ea396cbec0bc67676b0d49d5b34e344886a9ef06dc918f26db24d69fcd6

HELD IN THE MEMBER ARCHIVE

The verified repair and its recorded checks are member-only.

This mechanism has 10 recorded checks per implementation. The open-access tier publishes the failure and the unsuccessful fix; the repaired source that passes every check, and the observations that prove it, are available to members.

Every case sharing this mechanism uses the same contract and the same repair, so this one record is held back for all of them.

Member access is invitation-based. Sign in with your invited account to inspect the repair.

Sign in to the archive ↗

Verification & scope

Controlled binary64 or explicitly stipulated miniature format; no hardware exception flags or platform floating environment are modeled. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-29T14:39:37.358853+00:00.

Case digest / 27b791cdd1f0c48e10e6d5c32def03863ee0570872074fb0c16b83625a067a81