You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Regression · php-src-strict · Reflection / PHP 8.4 math builtins
Problem
Under PHP_COMPILER_PROFILE=8.4, the IEEE float builtins fadd / fsub / fmul / fmax / fmin / nextafter are callable (positional works) but Reflection reports 0 parameters and no return type. Zend named args (num1 / num2, fmaxvalue / ...values) therefore fail with Unknown named parameter.
Distinct from #25459 (max/min/pow/fdiv/fmod) and #27600 (fpow/bcdivmod).
Maintainer note (2026-08-07): live php:8.4-cli 8.4.24 and php:8.5-cli 8.5.8 report function_exists('fadd'|'fmax'|'nextafter'|…) === false. These are not Zend stub surfaces to Reflection-fix — they are phantoms. Tracking removal/gating in #28565. This issue’s “match Zend float stubs” premise should be closed or rewritten against #28565 done-when.
Activity
PurHur commentedon Aug 7, 2026
Maintainer note (2026-08-07): live
php:8.4-cli8.4.24 andphp:8.5-cli8.5.8 reportfunction_exists('fadd'|'fmax'|'nextafter'|…) === false. These are not Zend stub surfaces to Reflection-fix — they are phantoms. Tracking removal/gating in #28565. This issue’s “match Zend float stubs” premise should be closed or rewritten against #28565 done-when.PurHur commentedon Aug 7, 2026
Closing as not planned: Zend 8.4/8.5 do not ship these builtins. Phantom removal tracked in #28565.