Re: fsdb pledge

2016-08-20 Thread Theo Buehler
> this is kinda gross, but ok for now. I agree that it isn't very nice. How about this version that passes a flag to setup()? Unless I'm missing something, the only reason for the pledges being conditional on !hotroot() is the mount(2) call in fsck_ffs/main.c:336, where a ro root filesystem is r

Re: fsdb pledge

2016-08-20 Thread Ted Unangst
Theo Buehler wrote: > On Sat, Aug 20, 2016 at 01:28:23PM +0200, Jan Stary wrote: > > On current/amd64, fsdb(8) fails like this: > > > > # fsdb -f /dev/rsd0l > > ** /dev/rsd0l > > Editing file system `/dev/rsd0l' > > Last Mounted on /usr/xenocara > > current inode: directory > >

Re: fsdb pledge

2016-08-20 Thread Theo Buehler
On Sat, Aug 20, 2016 at 01:28:23PM +0200, Jan Stary wrote: > On current/amd64, fsdb(8) fails like this: > > # fsdb -f /dev/rsd0l > ** /dev/rsd0l > Editing file system `/dev/rsd0l' > Last Mounted on /usr/xenocara > current inode: directory > I=2 MODE=40755 SIZE=5

fsdb pledge

2016-08-20 Thread Jan Stary
On current/amd64, fsdb(8) fails like this: # fsdb -f /dev/rsd0l ** /dev/rsd0l Editing file system `/dev/rsd0l' Last Mounted on /usr/xenocara current inode: directory I=2 MODE=40755 SIZE=512 MTIME=Aug 8 19:27:56 2016 [0 nsec]