On Tue, Sep 03, 2019 at 11:32:40PM -0700, Christoph Hellwig wrote: > After going back to this from the next patch I think I spotted a > behavior difference: xlog_state_do_iclog_callbacks only returns true, > and xlog_state_do_callback only increments loopdidcallbacks if > ic_callbacks was non-emty. But we already dropped the block just > to check that it is empty, so I think we need to keep the old > behavior. IIUC, you are saying that loopdidcallbacks is not a variable that tracks whether callbacks were run, but whether the icloglock was released or not. Ok, I'll go fix that up, and name the variable appropriately. Cheers, Dave. -- Dave Chinner david@xxxxxxxxxxxxx
