[Girgias]

No description provided.

[SakiTakamachi]

I was curious and took a look at this, but it's definitely not due to a conflict between tests. Even if I run the php file generated from phpt alone, a transaction error occurs.

It seems that the switching of dbh->in_txn is probably insufficient, and there are places where it remains true and does not return.

[Girgias]

I was curious and took a look at this, but it's definitely not due to a conflict between tests. Even if I run the php file generated from phpt alone, a transaction error occurs.

It seems that the switching of dbh->in_txn is probably insufficient, and there are places where it remains true and does not return.

Thank you for investigating! So do you think this is a problem with the Firebird driver or with the test?

[SakiTakamachi]

Thank you for investigating! So do you think this is a problem with the Firebird driver or with the test?

I believe it is an implementation issue of pdo_firebird. The test is perfectly fine and has told us about bugs in firebird.

I'm currently working on fixing this issue. I'll be able to submit a pull request in about an hour.

[SakiTakamachi]

I got it working for now. I'll think of a better way to write it, and I'll add comments where they are needed.

#12657
#12741