On Fri, Aug 30, 2019 at 02:25:13PM +0800, Chao Yu wrote: > On 2019/8/30 11:36, Gao Xiang wrote: > > As Dan Carpenter suggested [1], I have to remove > > all erofs likely/unlikely annotations. > > > > [1] https://lore.kernel.org/linux-fsdevel/20190829154346.GK23584@kadam/ > > Reported-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx> > > Signed-off-by: Gao Xiang <gaoxiang25@xxxxxxxxxx> > > I suggest we can modify this by following good example rather than removing them > all, at least, the fuzzed random fields of disk layout handling should be very > rare case, I guess it's fine to use unlikely. No, no... It's the opposite. Only use those annotations on fast paths where it's going to show up in benchmarks. On fast paths then remove all the debug code and really optimize the heck out of the code. We sacrifice readability for speed in places where it matters. regards, dan carpenter _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
