I've just compared the performance of stock Linux kernel 3.1.1-1 which I just got via Arch updates versus the 3.1.1-7 PF kernel which I compiled trying my best to enable the best options for my machine. While the benchmark is old and is actually created to measure CPU and memory performance across different systems, I was at struggle to find a better tool to compare the kernels.
Ultimately, the difference was in the range of 0.07-0.7% for the test totals which, I guess, is within the statistical deviation. So, hardly conclusive at all. One thing, though - it appears that the PF kernel was losing in almost all of the test, though again, by a very narrow margin.
Either I have no experience in custom-compiling kernels whatsoever (well, at least it boots!) and the PF patches are totally useless, or the stock kernel is so good now that it doesn't really need any tweaking. We'll see how the PF performs in daily tasks.
Ultimately, the difference was in the range of 0.07-0.7% for the test totals which, I guess, is within the statistical deviation. So, hardly conclusive at all. One thing, though - it appears that the PF kernel was losing in almost all of the test, though again, by a very narrow margin.
Either I have no experience in custom-compiling kernels whatsoever (well, at least it boots!) and the PF patches are totally useless, or the stock kernel is so good now that it doesn't really need any tweaking. We'll see how the PF performs in daily tasks.
No comments:
Post a Comment