On Thu, Dec 05, 2024 at 12:52:25AM -0600, Bill O'Donnell wrote: > > 1) Our vaunted^Wshitty review process didn't catch various coding bugs, > > and testing didn't trip over them until I started (ab)using precommit > > hooks for spot checking of inode/dquot/buffer log items. > > You give little time for the review process. I don't really think that is true. But if you feel you need more time please clearly ask for it. I've done that in the past and most of the time the relevant people acted on it (not always). > > 2) Most of the metadir/rtgroups fixes are for things that hch reworked > > towards the end of the six years the patchset has been under > > development, and that introduced bugs. Did it make things easier for a > > second person to understand? Yes. > > No. So you speak for other people here? > I call bullshit. You guys are fast and loose with your patches. Giving > little time for review and soaking. I'm not sure who "you" is, but please say what is going wrong and what you'd like to do better. > > > becoming rather dodgy these days. Do things need to be this > > > complicated? > > > > Yeah, they do. We left behind the kindly old world where people didn't > > feed computers fuzzed datafiles and nobody got fired for a computer > > crashing periodically. Nowadays it seems that everything has to be > > bulletproofed AND fast. :( > > Cop-out answer. What Darrick wrote feels a little snarky, but he has a very valid point. A lot of recent bug fixes come from better test coverage, where better test coverage is mostly two new fuzzers hitting things, or people using existing code for different things that weren't tested much before. And Darrick is single handedly responsible for a large part of the better test coverage, both due to fuzzing and specific xfstests. As someone who's done a fair amount of new development recently I'm extremely glad about all this extra coverage.
