In message <2DDDD55D-4445-4565-9384-4E4BE7B41D7D@xxxxxxx>, Kyle Moffett writes: > On Sep 26, 2007, at 09:40:20, Erez Zadok wrote: [...] > > Recently we've done a full audit of the entire code, and added un/ > > likely where we felt that the chance of succeeding is 95% or better > > (e.g., error conditions that should rarely happen, and such). > > Actually due to the performance penalty on some systems I think you > only want to use it if the chance of succeeding is 99% or better, as > the benefit if predicted is a cycle or two and the harm if > mispredicted can be more than 50 cycles, depending on the CPU. *That's* the information I was looking for, Kyle: what's the estimated probability I should be using as my guideline. I used 95% (20/1 ratio), and you're telling me I should use 99% (100/1 ratio). The difference between the number of cycles saved/added is very compelling. Given that I certainly agree with you that I'm using un/likely too much. I'll re-evaluate and update my patch series then. Thanks, Erez. - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
