On Tue, 2019-06-18 at 12:20 -0400, Steven Rostedt wrote: > Heads up. Johannes found a regression that this code above causes > reading a 1.7G file where the above code makes it go from reading it in > 5 seconds, to over 2 minutes! Actually, I didn't even let it finish ... It was something like 4.5 seconds to not finishing in 4.5 minutes, then I gave up. My prior version was commit 1ad32c24746. The file is 1.7G, but that's perhaps not the interesting part? A little under 620k events - by my standards, that's a small file :-) > I bisected it down to this commit (saw a small, but not so drastic > performance issue it my tests). I asked Johannes to revert the commit, > and he said that it removed the regression. With that reverted I actually see ~2 seconds now, which is 2x faster than before :-) > He said he'll report this in the Bugzilla, but we should be looking at > changing the realloc() logic. https://bugzilla.kernel.org/show_bug.cgi?id=203927 johannes
|