It turns out that these patches almost apply in -current but while they are enough to allow 4.x to work, they are not sufficient to allow -current to work. there is a bunch of different code in -current that is just not there in 4.x and Geom is sticking its finger in the picture as well. I'd rather just get 4.x capable of handling a 2 TB drive and handle -current as a separate issue. 4.x will never need a partition > 1TB and will never handle a drive > 2TB without a LOT of work. but -current might. S it's not really worth making 4.x sysinit capable of doing it. These patches are sufficient for what we can do in 4.x,
On Thu, 14 Nov 2002, M. Warner Losh wrote: > In message: <[EMAIL PROTECTED]> > Julian Elischer <[EMAIL PROTECTED]> writes: > : If I get no complaints I'll commit these to 4.x. > : It's all different in 5.x so an MFC doesn't really work.. > > Is there any reason that you didn't just jump to int64_t for blocks > and such? You have a limit of 2T still with these patchs. I don't > know if the drivers would support more than this, but it wouldn't hurt > to have the upper layers know how to do it once there's driver > support... In current daddr_t is __int64_t, but only int32_t in > -stable. Of course, -stable can't support more than 2T, so maybe this > is moot. > > Ideally, you'd make sure that you can do this on -current with the > massively different code, and suggest patches if not. > > Warner > > To Unsubscribe: send mail to [EMAIL PROTECTED] > with "unsubscribe freebsd-hackers" in the body of the message > To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
