On Tue, Dec 16, 2014 at 06:53:19AM -0600, Chris Rorvick wrote: > On Tue, Dec 16, 2014 at 5:35 AM, Dan Carpenter <dan.carpenter@xxxxxxxxxx> wrote: > > On Tue, Dec 16, 2014 at 11:14:35AM +0000, Dilger, Andreas wrote: > > > > > > Sorry, that isn't right. Chris' patch is actually doing the right thing > > > to check for units > 1. > > > > It's not right because it discards the negative. > > I don't think this patch introduces a bug. If anything, it was already > there. The original code may be totally buggy. Who knows? Why are we passing negative numbers here anyway instead of just returning -EINVAL? But the new code is also buggy and not consistent with itself. In the original code if the user data is "-1k" or "-1024" that was treated the same. In the new code, "-1k" means negative 1024 because the user supplies units but "-1024" means positive 1024 because there are no units given. > > > The proposed change above discards "mult" > > > entirely, which breaks the users of this function that are not in this > > > file (e.g. osc_cached_mb_seq_write() or ll_max_cached_mb_seq_write()) > > > that have tunables in units of MB by default, but can also use parameters > > > with units like "4.5G" for convenience. > > > > I think you are confusing lprocfs_write_frac_helper() and > > lprocfs_write_frac_u64_helper(). There is only one caller for this > > function. > > By this logic, lprocfs_write_frac_u64_helper() should just be removed > and it's code should be folded into lprocfs_write_u64_helper(), no? > There are vast swathes of lustre code which need to be deleted but I haven't looked at this one. Probably. regards, dan carpenter _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
