Comparing against Hudson code, I do agree with your changes. I wander why no side effect was found before. Maybe some of the code (with exception of SPI prefetch) could simply be removed (modern OS would re-initialize them as needed)?

I'll await an answer, otherwise I would approve it (if you say leave it there, I'll accept and approve).

View Change

To view, visit change 32551. To unsubscribe, or for help writing mail filters, visit settings.