Subject: LFS
To: None <netbsd-users@netbsd.org>
From: Miles Nordin <carton@Ivy.NET>
List: netbsd-users
Date: 02/16/2005 14:13:03
--pgp-sign-Multipart_Wed_Feb_16_14:13:02_2005-1
Content-Type: text/plain; charset=US-ASCII

I think, no, it's not working in 2.0.

I tried it last night.  When I woke up, my Mac was crashed for some
other zstty reason:

db> bt
0x0064cef0: at zs_abort+0x70
0x0064cf00: at zstty_stint+0xd4
0x0064cf20: at zsc_intr_hard+0x124
0x0064cf40: at zshard+0x54
0x0064cf60: at ext_intr+0x1d4
0x0064cfa0: at trapstart+0x880
0xffffe980: at ADBDevTable+0x4b990e4c

so, maybe not LFS-related, but it caused the LFS to be uncleanly
dismounted.  When I fsck_lfs and remounted, it crashed again within a
few minutes, this time in LFS.  so I think maybe fsck_lfs is
incomplete, and the kernel code is not robust to corrupt filesystems.

panic: lfs_valloc: inuse inode 5317 on the free list
Stopped in pid 115.1 (pax) at   netbsd:cpu_Debugger+0x10:       lwz     r0, r1, 0
x14
db> bt
0xd52efb30: at panic+0x19c
0xd52efbc0: at lfs_valloc+0x278
0xd52efbf0: at VOP_VALLOC+0x44
0xd52efc20: at ufs_mkdir+0x98
0xd52efd90: at lfs_mkdir+0x60
0xd52efdb0: at VOP_MKDIR+0x44
0xd52efde0: at sys_mkdir+0x148
0xd52efed0: at syscall_plain+0xe0
0xd52eff40: user SC trap #136 by 0x418f42b8: srr1=0xf032
            r1=0xffffe690 cr=0x44004082 xer=0 ctr=0x418f42b0

BTW, sbp2 firewire layer also crashed after a couple hours, so i gave
up on that before these tests.

--pgp-sign-Multipart_Wed_Feb_16_14:13:02_2005-1
Content-Type: application/pgp-signature
Content-Transfer-Encoding: 7bit

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.6 (NetBSD)

iQCVAwUAQhObP4nCBbTaW/4dAQKL+AP+P4NIi+yD6ORrjt1CBBLnqkEUHUKUxNwo
AiIjSbaaa9GoXkAtgSFfoEH4cx6lstLATjR4KMcJO878D6ynd/9lDK37VXOaAiTZ
dGQ/hXMfvOpjh7cPKTVvaCtLBKBKsZSDeIHedDVvXlvS8PYmqd5rwzfrcwy0eZRv
v8HC5YciePg=
=aN1x
-----END PGP SIGNATURE-----

--pgp-sign-Multipart_Wed_Feb_16_14:13:02_2005-1--