Re: [PATCH 0/1] MR4002: oleaut32: Fix typelib cache race condition.

[Huw Davies (＠huw)]

On Wed Nov  8 12:38:27 2023 +0000, Connor McAdams wrote:
> I guess we could, but that feels a bit clunky IMO. We'd be
> entering/exiting the typelib cache CS every time we AddRef/Release the
> typelib interface, which seems to happen often.
> It's a simpler solution, but the current solution avoids that potential
> issue. E.g, I don't think we want to cause all AddRef calls on typelib
> interfaces to block while waiting on a new typelib to be inserted into
> the cache for example.

I like simple :-)  I doubt that the contention is going to be a big issue here - if it turns out to be the case we can revisit this.