Iterating a \Google\Protobuf\Internal\Message PHP object segfaults when using the protobuf extension

What version of protobuf and what language are you using?
Version: 4.31.1
Language: PHP

What operating system (Linux, Windows, ...) and version?
Linux 6.1.0-37-cloud-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.140-1 (2025-05-22) x86_64 GNU/Linux

What runtime / compiler are you using (e.g., python version or gcc version)
PHP 8.4.1 with protobuf extension 4.31.1

What did you do?
Steps to reproduce the behavior:

1. Define a protobuf message

syntax = "proto3";

package example;

message ProtoMessage {
  string msg = 1;
}

2. Compile with libprotoc

$ protoc --version
libprotoc 3.21.12
$ protoc --proto_path=. --php_out=build/gen message.proto
$ tree build/gen
build/gen/
├── Example
│   └── ProtoMessage.php
└── GPBMetadata
    └── Message.php

3. Run this script

<?php

require 'build/gen/GPBMetadata/Message.php';
require 'build/gen/Example/ProtoMessage.php';

echo "iterating generated php class ProtoMessage\n";

$proto_msg = new \Example\ProtoMessage();
foreach ($proto_msg as $key => $value) {
    assert(false);
}

What did you expect to see
The process does not emit a segv

What did you see instead?
Actually the process emits a segv

$ php segv.php
iterating generated php class ProtoMessage
Segmentation fault (core dumped)

Make sure you include information that can help us debug (full error message, exception listing, stack trace, logs).
gdb backtrace

(gdb) backtrace
#0  ZEND_FE_RESET_R_SPEC_CV_HANDLER () at ./Zend/zend_hash.h:317
#1  0x0000559b7e901690 in zend_vm_call_opcode_handler (ex=ex@entry=0x7fe88d214020)
    at ./Zend/zend_vm_execute.h:68448
#2  0x00007fe88d636958 in php_pcov_trace (execute_data=0x7fe88d214020)
    at /tmp/pear/temp/pcov/pcov.c:264
#3  php_pcov_execute_ex (execute_data=0x7fe88d214020) at /tmp/pear/temp/pcov/pcov.c:305
#4  0x0000559b7e900b63 in zend_execute (op_array=op_array@entry=0x7fe88d26e000, 
    return_value=return_value@entry=0x0) at ./Zend/zend_vm_execute.h:64236
#5  0x0000559b7e9681a0 in zend_execute_script (type=type@entry=8, 
    retval=retval@entry=0x0, file_handle=file_handle@entry=0x7ffd94dc1fc0)
    at ./Zend/zend.c:1934
#6  0x0000559b7e806cb5 in php_execute_script_ex (primary_file=<optimized out>, 
    retval=retval@entry=0x0) at ./main/main.c:2574
#7  0x0000559b7e806fd7 in php_execute_script (primary_file=<optimized out>)
    at ./main/main.c:2614
#8  0x0000559b7e969f2c in do_cli (argc=2, argv=0x559ba41a7480)
    at ./sapi/cli/php_cli.c:935
#9  0x0000559b7e41a879 in main (argc=2, argv=0x559ba41a7480) at ./sapi/cli/php_cli.c:1310
(gdb) 

Anything else we should know about your project / environment
Almost similar issue: https://github.com/protocolbuffers/protobuf/issues/7319

Reproduce the crash with segv.php using the generated build/gen/Example/ProtoMessage.php and build/gen/GPBMetadata/Message.php under PHP 8.4.1 and protobuf extension 4.31.1. Start from the foreach iteration entry point and the supplied gdb backtrace, then compare with issue 7319. Done means iterating the generated ProtoMessage no longer causes a segmentation fault.