On 08/26/2014 10:46 PM, Trond Myklebust wrote: <> > > So to return to the original question: could we please change the > layoutcommit in your patch so that it is asynchronous? > There is nothing change. The original message at the head of this thread, is Christoph's patch with the asynchronous layoutcommit. As a reply I posted what I had in my tree, with a sync operation. Because I wanted the recall to take 500 micro-seconds instead of like 300-1000 mili-seconds. When the STD supports my model specifically I do not see why not Don't you guys have recalls on the hot path? Sigh, another nail in the coffin of what was suppose to be a dam good STD. Thinking about it from day one you had a personal agenda against RECALLS. For some reason, which you never told me, you decided to crash and burn any proper use of LO_RECALLS. With that dreaded *synchronous* polling driven, forgetful--model recall. (One day we'll sit on a Bear and you will tell me why? I really want to know) Now that there is the last ever chance to make an half dissent citizen out of what was left of recalls, you are all on your back feet, against it. With religious mantras, when you have never sat and actually stared at the prints and experienced recalls for yourself. Your words Christoph I have a deja-vu its principles against good science again. Lets revert to your patch. Reviewed-by: Boaz Harrosh <bharrosh@xxxxxxxxxxx> -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
