The Missing Table Rows

Medium

· Backend

Incident

SEV-1

BUG-FDIV

Priority: Recall-grade · reported by verification

Time to fix: 20 min

Failing checks on the shipped build

  • a table hole falls back to real division

    unsupported operand type(s) for *: 'int' and 'NoneType'

How to work this incident

  1. Read the report: write down the expected behaviour and the actual one in a sentence each.
  2. Reproduce it: run the visible tests and find the failing check that matches the report.
  3. Read the logs and the failing checks for the first value that is wrong — the earliest wrong value is nearer the cause than the loudest error.
  4. Follow the request through the handler, the validation and the service it calls; find the first step whose output disagrees with the report.
  5. Fix the cause rather than the symptom, run the visible tests again, and submit against the hidden ones — then write the root cause down.

Inspired by

Intel Pentium · 1994

Inspired by the 1994 Pentium FDIV bug: five entries missing from a division lookup table inside the chip, wrong answers on rare inputs, a $475M recall. The table here is locked — some entries really are missing. The lookup path must fall back to real division when it hits a hole.

Fix fastdiv.py.

Code map

table = bug_require("./table")

def estimate(a, d):
    return round(a * table.RECIPROCALS[d], 6)

Sign in to edit the file, run the tests and submit a fix

A free account — no card, no trial clock. You come straight back here.

The Missing Table Rows — Python Bug Hunt

Inspired by the 1994 Pentium FDIV bug: five entries missing from a division lookup table inside the chip, wrong answers on rare inputs, a $475M recall.

  • Language: Python
  • Layer: Backend
  • Difficulty: Medium
  • Concepts: Lookup Tables, Null Safety
  • Modelled on: Intel Pentium · 1994
  • Visible tests: table hits divide fast and exact; a table hole falls back to real division
  • Reward: 50 XP for a complete fix

Incident

  • Severity: SEV-1 (priority: Recall-grade)
  • Ticket: BUG-FDIV
  • Reported by: verification
  • Time to fix: 20 minutes, the target for a medium incident

Failing checks on the shipped build

  • a table hole falls back to real division — unsupported operand type(s) for *: 'int' and 'NoneType'

How to work this incident

  1. Read the report: write down the expected behaviour and the actual one in a sentence each.
  2. Reproduce it: run the visible tests and find the failing check that matches the report.
  3. Read the logs and the failing checks for the first value that is wrong — the earliest wrong value is nearer the cause than the loudest error.
  4. Follow the request through the handler, the validation and the service it calls; find the first step whose output disagrees with the report.
  5. Fix the cause rather than the symptom, run the visible tests again, and submit against the hidden ones — then write the root cause down.

Briefing

Inspired by the 1994 Pentium FDIV bug: five entries missing from a division lookup table inside the chip, wrong answers on rare inputs, a $475M recall. The table here is locked — some entries really are missing. The lookup path must fall back to real division when it hits a hole.

Fix fastdiv.py.

Bug report

BUG-FDIV · Priority: Recall-grade · Reported by: verification

estimate(a, d) — divide a by digit d (1..9) using the reciprocal table:

  • table hit -> round(a * reciprocal, 6)
  • table hole (None) -> round(a / d, 6) fallback

Observed: dividing by 7 crashes with a TypeError — exactly the class of rare input the table forgot.

Logs

[calc] 4195835 / 3145727 path -> TypeError: unsupported operand

The code as shipped

src/chip/fastdiv.py (editable)

table = bug_require("./table")

def estimate(a, d):
    return round(a * table.RECIPROCALS[d], 6)

Read-only context: src/chip/table.py.

Open the hunt to start the incident clock, edit the files, run the visible tests and submit against the hidden ones, then write the root cause down. More Python bug hunts.