Should finish be uninterruptible?

[codygman]

This change seems to fix the symptom I get a couple levels up the abstraction chain in Persistent:

finish :: Connection
       -> IO ()
finish (Conn fp _) = do
    uninterruptibleMask_ $ finalizeForeignPtr fp

yesodweb/persistent#1199

I discovered this fixed that symptom here: yesodweb/persistent#1199 (comment)

[codygman]

I also see that odbc uses withAsyncBound here and says:

-- | Run an action in a bound thread. This is neccessary due to the
-- interaction with signals in ODBC and GHC's runtime.

I'm not sure if this applies to postgresql-libpq.

Edit: Here is what the async docs have to say about forkOS, which is what makes withAsyncBound different:

However, forkOS creates a bound thread, which is necessary if you need to call foreign (non-Haskell) libraries that make use of thread-local state, such as OpenGL (see Control.Concurrent).

So I guess the question is:

Does Connection, I guess the fp here, make use of thread-local state?

[codygman]

Still not sure about this, but this also fixes the symptom I linked to in my initial description:

finish :: Connection
       -> IO ()
finish (Conn fp _) = do
    -- uninterruptibleMask_ $ finalizeForeignPtr fp
    runInBoundThread $ finalizeForeignPtr fp

I'm assuming this is safer since the exceptions won't be both masked and uninterruptible in the Bound thread but I'm honestly kinda new to low-level Haskell and the FFI.

[phadej]

This change seems to fix the symptom I

What symptom?

I don't see a problem with finalizeForeignPtr of connection being interrupted.

At worst it should only delay the release of resources and not cause any other issues.