<Sorry about the length.  I'm just trying to provide as much
info as I can.>

  Is anyone else experiencing bad behavior from NFS?  I built a
new system (sup -o, rm -rf kernel dir, config and build
kernel, reboot, cleandir userland, build userland, reboot),
in hopes that it was due to stale .o files or something, but to
no avail.

  The symptoms I've noticed:
  1.  If a directory is larger than 1024 bytes, a readdir(?) on
    it might not return anything:

% wc -c lSLMP*	<-- these are directories:
5120 lSLMP.1
1024 lSLMP.1.static
6144 lSLMP.16
...
6144 lSLMP.8
1024 lSLMP.static.1
5120 lSLMPtest.4
% printf "%s\n" */*
<only prints files in lSLMP.1.static/ and lSLMP.static.1/, both
 of which are 1024 byte directories>

    Doing an ls -la in lSLMP.8, for example, doesn't print
_anything_, not even . and ..!

    Under 1.3C, 1.3E, 1.3F, 1.3G (IIRC) it works as expected
(all i386 boxes).

    This filesystem happens to be mounted from a Linux box.  From
tcpdump -vv -s 1000 during `ls' in lSLMP.8, the server sends
back one readdir reply, followed by an ERR reply, followed by two
more ERR replies:

23:47:49.591810 marvin.3234575964 > sim6.nfs: 128 getattr fh Unknown/1 (ttl 64, id 8816)
23:47:49.598264 sim6.nfs > marvin.1019: . ack 494307729 win 32120 (DF) (ttl 64, id 38167)
23:47:49.619802 sim6.nfs > marvin.3234575964: reply ok 100 getattr DIR 40755 ids 1084/1001 sz 6144  (DF) (ttl 64, id 38168)
23:47:49.620001 marvin.3234575965 > sim6.nfs: 96 fsstat fh Unknown/1 (ttl 64, id 8818)
23:47:49.629489 sim6.nfs > marvin.3234575965: reply ok 52 fsstat tsize 8192 bsize 512 blocks 7916998 bfree 983556 bavail 573 956 (DF) (ttl 64, id 38171)
23:47:49.629752 marvin.3234575966 > sim6.nfs: 136 readdir fh Unknown/1 8192 bytes @ 0 (ttl 64, id 8819)
23:47:49.637243 sim6.nfs > marvin.3234575966: reply ok 1460 readdir offset 1 size 688326671 eof (DF) (ttl 64, id 38174)
23:47:49.638503 sim6.nfs > marvin.1869640494: reply ERR 1460 nop (DF) (ttl 64, id 38175)
23:47:49.638579 marvin.1019 > sim6.nfs: . ack 3072 win 14284 (ttl 64, id 8820)
23:47:49.639244 sim6.nfs > marvin.13: reply ERR 1080 nop (DF) (ttl 64, id 38176)
23:47:49.639340 marvin.1019 > sim6.nfs: . ack 4152 win 17200 (ttl 64, id 8821)
23:47:49.640411 sim6.nfs > marvin.7: reply ERR 1340 nop (DF) (ttl 64, id 38177)
23:47:49.691164 marvin.1019 > sim6.nfs: . ack 5492 win 15860 (ttl 64, id 8829)
23:47:54.121269 marvin.573243392 > sim6.nfs: 40 null (ttl 64, id 8831)
23:47:54.131618 sim6.nfs > marvin.573243392: reply ok 32 null (ttl 64, id 38805)
23:48:24.131246 marvin.1378549760 > sim6.nfs: 40 null (ttl 64, id 8883)


  2.  chmod for the sticky bit doesn't propogate from the flaky
    client to the server or other clients very fast:

% ls -l foo
-rwxrwxrwx  1 bgrayson  wheel  7 Sep  7  1997 foo*
% chmod 04777 foo
% ls -l foo
-rwsrwxrwx  1 bgrayson  wheel  7 Sep  7  1997 foo*
% ssh r2d2 ls -l foo
-rwxrwxrwx  1 bgrayson  wheel  7 Sep  7  1997 foo
% ssh r2d2 ls -l foo	<-- after waiting a few more seconds:
-rwsrwxrwx  1 bgrayson  wheel  7 Sep  7  1997 foo

  This filesystem is served by a 1.3F NetBSD box.  On the other
client machines, a chmod appears to happen instantly to all
other machines.

  This is the first time in a long time that a buildable
-current system has been broken for me, so it seems highly
likely that I've goofed in some way.  The last -current system we
had working was probably from July.  Thanks in advance!

  P.S.  This is probably totally unrelated, but 12 files in
/usr/src/gnu/usr.bin/gawk had 512 bytes of 0xf6 in the middle of
them, on a 512-byte boundary.  So far, no other files appear
affected.  Does this look familiar to anyone?  Is our disk going bad?

  Here are my kernel configs (std.UT holds stuff for all
machines, and MARVIN holds the marvin-specific stuff):
------------------------  std.UT  ----------------------------
#	
#	std.UT 
#

include "arch/i386/conf/std.i386"

##  bgrayson:  doubling maxusers to 64, to allow more processes.
maxusers	64		# estimated number of users
options 	I586_CPU
options 	I686_CPU
options 	XSERVER,UCONSOLE
options 	INSECURE	# insecure; allow /dev/mem writing for X

options 	RTC_OFFSET=0	# hardware clock is this many mins. west of GMT
options		KTRACE

options 	SYSVMSG		# System V-like message queues
options 	SYSVSEM		# System V-like semaphores
options 	SYSVSHM		# System V-like memory sharing
#  bgrayson bumped SHMMAXPGS -- BSPlib seems to need more than 1024 when
#  simulating more than 8 processors.
options 	SHMMAXPGS=4096	# 1024 pages is the default

options 	LKM		# loadable kernel modules

options 	COMPAT_NOMID	# compatibility with 386BSD, BSDI, NetBSD 0.8,
options 	COMPAT_09	# NetBSD 0.9,
options 	COMPAT_10	# NetBSD 1.0,
options 	COMPAT_11	# NetBSD 1.1,
options 	COMPAT_12	# NetBSD 1.2,
options 	COMPAT_13	# NetBSD 1.3,
options 	COMPAT_43	# and 4.3BSD
options 	COMPAT_386BSD_MBRPART # recognize old partition ID

options 	COMPAT_LINUX	# binary compatibility with Linux
options		LINUX_GCC_SIGNATURE	##  (temporary?) modification
options 	COMPAT_FREEBSD	# binary compatibility with FreeBSD

options 	EXEC_ELF32	# 32-bit ELF executables (SVR4, Linux)

file-system 	FFS		# UFS
file-system 	EXT2FS		# second extended file system (linux)
file-system 	MFS		# memory file system
file-system 	NFS		# Network File System client
file-system 	CD9660		# ISO 9660 + Rock Ridge file system
file-system 	MSDOSFS		# MS-DOS file system
file-system 	FDESC		# /dev/fd
file-system 	KERNFS		# /kern
file-system 	NULLFS		# loopback file system
file-system 	PORTAL		# portal filesystem (still experimental)
file-system 	PROCFS		# /proc
file-system 	UMAPFS		# NULLFS + uid and gid remapping
file-system 	UNION		# union file system

options 	NFSSERVER	# Network File System server
options 	FIFO		# FIFOs; RECOMMENDED

options 	INET		# IP + ICMP + TCP + UDP
options 	NS		# XNS
options 	ISO,TPIP	# OSI
options 	EON		# OSI tunneling over IP
options 	CCITT,LLC,HDLC	# X.25

config		netbsd	root on ? type ?

options 	PCIVERBOSE	# verbose PCI device messages

mainbus0 at root

pci*	at mainbus? bus ?
pci*	at pchb? bus ?
pci*	at ppb? bus ?
pchb*	at pci? dev ? function ?	# PCI-Host bridges
pcib*	at pci? dev ? function ?	# PCI-ISA bridges
ppb*	at pci? dev ? function ?	# PCI-PCI bridges

isa*	at mainbus?			# all other ISA
isa*	at pcib?			# ISA on PCI-ISA bridge
pcppi0	at isa?
sysbeep0 	at pcppi?

spkr0 at pcppi?		#  PC speaker

npx0	at isa? port 0xf0 irq 13	# math coprocessor

com0	at isa? port 0x3f8 irq 4	# standard PC serial ports
com1	at isa? port 0x2f8 irq 3
com2	at isa? port 0x3e8 irq 5
lpt0	at isa? port 0x378 irq 7	# standard PC parallel ports

fdc0	at isa? port 0x3f0 irq 6 drq 2	# standard PC floppy controllers
fd*	at fdc? drive ?

##  The ed driver no longer exists -- split into we and something else.
##  We use the we version.
we0	at isa? port 0x280 iomem 0xd0000 irq 9	# WD/SMC Ethernet

pss0	at isa? port 0x220 irq 7 drq 6  # Personal Sound System
sp0	at pss0 port 0x530 irq 10 drq 0	# 	sound port driver

sb0	at isa? port 0x220 irq 7 drq 1	# SoundBlaster
#spkr0	at pckbcport? port 0x61  #  Now at pcppi instead.

# PnP bus and devices should be declared last 
isapnp0	at isa? 
ep*	at isapnp?
sb*	at isapnp?
joy*	at isapnp?

audio* at sb?
audio* at sp?

include	"arch/i386/conf/GENERIC.local"

pseudo-device	ccd	4		# concatenated disk devices
pseudo-device	vnd	4		# paging to files

pseudo-device	bpfilter 8		# packet filter
pseudo-device	ipfilter		# ip filter
pseudo-device	loop	1		# network loopback
pseudo-device	ppp	2		# PPP
pseudo-device	sl	2		# CSLIP
pseudo-device	tun	2		# network tunneling over tty

pseudo-device	pty	64		# pseudo-terminals
pseudo-device	tb	1		# tablet line discipline
-----------------------------------------------------------------
---------------------------  MARVIN  ----------------------------
#	
#	MARVIN -- the guinea pig for the Chase group
#

include "arch/i386/conf/std.UT"

options 	DDB		# in-kernel debugger
#makeoptions	DEBUG="-g"	# compile full symbol table
options 	DIAGNOSTIC	# internal consistency checks
#options 	KTRACE		# system call tracing, a la ktrace(1)
##  The BIOSEXTMEM option is no longer needed, if the new boot blocks 
##  (from around Sep 97 or later) are on the boot disk.  bgrayson

vt0	at isa? port 0x60 irq 1
wdc0	at isa? port 0x1f0 irq 14	# ST506, ESDI, and IDE controllers
wdc1	at isa? port 0x170 irq 15
wd*	at wdc? drive ?

##  ATAPI support
atapibus* at wdc?
cd* at atapibus? drive ?

####  For trying out the NCR SCSI card in marvin:
ncr*	at pci? dev ? function ?	# NCR 53c8xx SCSI
####  For trying out orac's old SCSI card in marvin:
ahc*	at pci? dev ? function ?	# Adaptec 294x, aic78x0 SCSI controllers
options		SCSIVERBOSE
##  Bus
scsibus* at ncr?
scsibus* at ahc?
##  Devices...
sd*	at scsibus? target ? lun ?	# SCSI disk drives
st*	at scsibus? target ? lun ?	# SCSI tape drives
cd*	at scsibus? target ? lun ?	# SCSI CD-ROM drives
ch*	at scsibus? target ? lun ?	# SCSI autochangers
ss*	at scsibus? target ? lun ?	# SCSI scanners
uk*	at scsibus? target ? lun ?	# SCSI unknown
-------------------------------------------------------------