> After looking at the code, it looks like it was a false alarm on my side. > The I2C interrupts are disabled here and other threads should be locked out > by the I2C core locks. So it doesn't make sense to hold IRQs disabled during > a possibly long polling loop in rcar_i2c_bus_barrier(). So I'll just repost > the revert patch (if still needed). Nope. Thanks for looking into it.
Attachment:
signature.asc
Description: Digital signature
