massive file corruption with reiser or NFS

Hallo all,

I got massive file corruptions with the kernels mentioned in the subject. I can reproduce it every time.

What did I do?

The kernel can't find files or directories which have been created seconds before. If I start configure of some program for example, the conftest-Directory can't be found again which is created during configure; datas can't be read in files (if it could open the file). I'm than getting an IO-error. No matter if it's local on reiser or remote via NFS. Always the same problem.

If you want to mount some partitions, like mount /boot, the /boot-Directory can't be found (IO-error). If you try it again and again, it will suddenly work :-(.

I tried to compile the 2.4.5ac6-Kernel under itself. It ended in massive errors while reading the sources.
I rebooted the machine (with a lot of errors while unmounting) with kernel 2.2.19 and tried to compile the above mentioned kernel again. I got a lot of other errors -> 2.4.5 destroyed the files! I had to do a rm -R to get rid of the whole tree. After newcreation of the tree, the compiling under kernel 2.2.19 worked fine.

I tried to compile the kernel with egcs 1.1.2 and gcc 2.95.3 - no matter. I tried to compile with or without APIC - no matter.

Do you have any suggestions how to compile the kernel to get it working and to locate the problem?

Some additional infos to my machine:
512 MB RAM
AMD Athlon 800
no overclocking

lspci -v
00:00.0 Host bridge: VIA Technologies, Inc. VT8371 [KX133] (rev 02)
Flags: bus master, medium devsel, latency 0 Memory at d6000000 (32-bit, prefetchable)
Capabilities: [a0] AGP version 2.0

00:01.0 PCI bridge: VIA Technologies, Inc. VT8371 [PCI-PCI Bridge] (prog-if 00 [Normal decode])
Flags: bus master, 66Mhz, medium devsel, latency 0
Bus: primary=00, secondary=01, subordinate=01, sec-latency=0 I/O behind bridge: 0000c000-0000cfff
Memory behind bridge: d4000000-d5ffffff
Prefetchable memory behind bridge: d0000000-d3ffffff
Capabilities: [80] Power Management version 2

00:07.0 ISA bridge: VIA Technologies, Inc. VT82C686 [Apollo Super] (rev 22)
Subsystem: VIA Technologies, Inc. VT82C686/A PCI to ISA Bridge
Flags: bus master, stepping, medium devsel, latency 0

00:07.1 IDE interface: VIA Technologies, Inc. VT82C586 IDE [Apollo] (rev 10) (prog-if 8a [Master SecP PriP])
Flags: bus master, medium devsel, latency 32
I/O ports at d000
Capabilities: [c0] Power Management version 2

00:07.2 USB Controller: VIA Technologies, Inc. VT82C586B USB (rev 10)
(prog-if 00 [UHCI]) Subsystem: Unknown device 0925:1234
Flags: bus master, medium devsel, latency 32, IRQ 9 I/O ports at d400
Capabilities: [80] Power Management version 2

00:07.4 Host bridge: VIA Technologies, Inc. VT82C686 [Apollo Super ACPI] (rev
30)
Flags: medium devsel
Capabilities: [68] Power Management version 2

00:07.5 Multimedia audio controller: VIA Technologies, Inc. VT82C686 [Apollo Super AC97/Audio] (rev 20)
Subsystem: VIA Technologies, Inc.: Unknown device 4511
Flags: medium devsel, IRQ 5
I/O ports at dc00
I/O ports at e000
I/O ports at e400
Capabilities: [c0] Power Management version 2

00:08.0 Ethernet controller: Silicon Integrated Systems [SiS] SiS900 10/100 Ethernet (rev
02)
Subsystem: Silicon Integrated Systems [SiS] SiS900 10/100 Ethernet
Adapter
Flags: bus master, medium devsel, latency 32, IRQ 10 I/O ports at e800
Memory at d9000000 (32-bit, non-prefetchable)
Capabilities: [40] Power Management version 1

00:09.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139 (rev 10)
Subsystem: Realtek Semiconductor Co., Ltd. RT8139
Flags: bus master, medium devsel, latency 32, IRQ 11
I/O ports at ec00
Memory at d9001000 (32-bit, non-prefetchable)

01:00.0 VGA compatible controller: ATI Technologies Inc Rage 128 PF (prog-if 00 [VGA])
Subsystem: ATI Technologies Inc: Unknown device 0008
Flags: bus master, stepping, 66Mhz, medium devsel, latency 32, IRQ 10 Memory at d0000000 (32-bit, prefetchable)
I/O ports at c000
Memory at d5000000 (32-bit, non-prefetchable)
Capabilities: [50] AGP version 2.0
Capabilities: [5c] Power Management version 2

My .config-file (without the options wich have not set):

#
# Automatically generated by make menuconfig: don't edit #
CONFIG_X86=y
CONFIG_ISA=y
CONFIG_UID16=y

#
# Code maturity level options
#
CONFIG_EXPERIMENTAL=y

#
# Loadable module support
#
CONFIG_MODULES=y
CONFIG_MODVERSIONS=y
CONFIG_KMOD=y

#
# Processor type and features
#
CONFIG_MK7=y
CONFIG_X86_WP_WORKS_OK=y
CONFIG_X86_INVLPG=y
CONFIG_X86_CMPXCHG=y
CONFIG_X86_XADD=y
CONFIG_X86_BSWAP=y
CONFIG_X86_POPAD_OK=y
CONFIG_RWSEM_XCHGADD_ALGORITHM=y
CONFIG_X86_L1_CACHE_SHIFT=6
CONFIG_X86_TSC=y
CONFIG_X86_GOOD_APIC=y
CONFIG_X86_USE_3DNOW=y
CONFIG_X86_PGE=y
CONFIG_X86_USE_PPRO_CHECKSUM=y
CONFIG_NOHIGHMEM=y
CONFIG_MTRR=y
CONFIG_X86_UP_APIC=y
CONFIG_X86_UP_IOAPIC=y
CONFIG_X86_LOCAL_APIC=y
CONFIG_X86_IO_APIC=y

#
# General setup
#
CONFIG_NET=y
CONFIG_PCI=y
CONFIG_PCI_GOANY=y
CONFIG_PCI_BIOS=y
CONFIG_PCI_DIRECT=y
CONFIG_PCI_NAMES=y
CONFIG_SYSVIPC=y
CONFIG_SYSCTL=y
CONFIG_KCORE_ELF=y
CONFIG_BINFMT_ELF=y
CONFIG_PM=y
CONFIG_APM=m
CONFIG_APM_DO_ENABLE=y
CONFIG_APM_DISPLAY_BLANK=y
CONFIG_APM_RTC_IS_GMT=y

#
# Memory Technology Devices (MTD)
#

#
# Parallel port support
#
CONFIG_PARPORT=m
CONFIG_PARPORT_PC=m
CONFIG_PARPORT_PC_FIFO=y
CONFIG_PARPORT_1284=y

#
# Plug and Play configuration
#

#
# Block devices
#
CONFIG_BLK_DEV_FD=m
CONFIG_BLK_DEV_LOOP=m

#
# Multi-device support (RAID and LVM)
#

#
# Networking options
#
CONFIG_PACKET=m
CONFIG_UNIX=m
CONFIG_INET=y
CONFIG_IP_MULTICAST=y

#
# QoS and/or fair queueing
#

#
# Telephony Support
#

#
# ATA/IDE/MFM/RLL support
#
CONFIG_IDE=y

#
# IDE, ATA and ATAPI Block devices
#
CONFIG_BLK_DEV_IDE=y
CONFIG_BLK_DEV_IDEDISK=y
CONFIG_IDEDISK_MULTI_MODE=y
CONFIG_BLK_DEV_IDESCSI=m
CONFIG_BLK_DEV_IDEPCI=y
CONFIG_IDEPCI_SHARE_IRQ=y
CONFIG_BLK_DEV_IDEDMA_PCI=y
CONFIG_IDEDMA_PCI_AUTO=y
CONFIG_BLK_DEV_IDEDMA=y
CONFIG_BLK_DEV_VIA82CXXX=y
CONFIG_IDEDMA_AUTO=y
CONFIG_BLK_DEV_IDE_MODES=y

#
# SCSI support
#
CONFIG_SCSI=m
CONFIG_BLK_DEV_SR=m
CONFIG_SR_EXTRA_DEVS=2
CONFIG_CHR_DEV_SG=m
CONFIG_SCSI_DEBUG_QUEUES=y
CONFIG_SCSI_MULTI_LUN=y
CONFIG_SCSI_CONSTANTS=y

#
# SCSI low-level drivers
#

#
# Fusion MPT device support
#

#
# IEEE 1394 (FireWire) support
#

#
# I2O device support
#

#
# Network device support
#
CONFIG_NETDEVICES=y

#
# ARCnet devices
#
CONFIG_DUMMY=m

#
# Ethernet (10 or 100Mbit)
#
CONFIG_NET_ETHERNET=y
CONFIG_NET_PCI=y
CONFIG_EEPRO100=y
CONFIG_8139TOO=m
CONFIG_SIS900=m

#
# Ethernet (1000 Mbit)
#

#
# Wireless LAN (non-hamradio)
#

#
# Token Ring devices
#

#
# Wan interfaces
#

#
# Amateur Radio support
#

#
# IrDA (infrared) support
#

#
# ISDN subsystem
#

#
# Old CD-ROM drivers (not SCSI, not IDE)
#

#
# Input core support
#

#
# Character devices
#
CONFIG_VT=y
CONFIG_VT_CONSOLE=y
CONFIG_SERIAL=m
CONFIG_UNIX98_PTYS=y
CONFIG_UNIX98_PTY_COUNT=256
CONFIG_PRINTER=m

#
# I2C support
#
CONFIG_I2C=m
CONFIG_I2C_ALGOBIT=m
CONFIG_I2C_PHILIPSPAR=m
CONFIG_I2C_ELV=m
CONFIG_I2C_VELLEMAN=m
CONFIG_I2C_ALGOPCF=m
CONFIG_I2C_CHARDEV=m

#
# Mice
#
CONFIG_MOUSE=m
CONFIG_PSMOUSE=y

#
# Joysticks
#

#
# Watchdog Cards
#

#
# Ftape, the floppy tape device driver
#
CONFIG_AGP=m
CONFIG_AGP_VIA=y
CONFIG_DRM=y
CONFIG_DRM_TDFX=y

#
# Multimedia devices
#

#
# File systems
#
CONFIG_REISERFS_FS=y
CONFIG_FAT_FS=m
CONFIG_MSDOS_FS=m
CONFIG_UMSDOS_FS=m
CONFIG_VFAT_FS=m
CONFIG_TMPFS=y
CONFIG_ISO9660_FS=m
CONFIG_JOLIET=y
CONFIG_PROC_FS=y
CONFIG_DEVPTS_FS=y
CONFIG_EXT2_FS=m
CONFIG_UDF_FS=m
CONFIG_UDF_RW=y

#
# Network File Systems
#
CONFIG_NFS_FS=m
CONFIG_SUNRPC=m
CONFIG_LOCKD=m

#
# Partition Types
#
CONFIG_MSDOS_PARTITION=y
CONFIG_NLS=y

#
# Native Language Support
#
CONFIG_NLS_DEFAULT="iso8859-1"
CONFIG_NLS_CODEPAGE_437=m
CONFIG_NLS_CODEPAGE_737=m
CONFIG_NLS_CODEPAGE_775=m
CONFIG_NLS_CODEPAGE_850=m
CONFIG_NLS_CODEPAGE_852=m
CONFIG_NLS_CODEPAGE_855=m
CONFIG_NLS_CODEPAGE_857=m
CONFIG_NLS_CODEPAGE_860=m
CONFIG_NLS_CODEPAGE_861=m
CONFIG_NLS_CODEPAGE_865=m
CONFIG_NLS_CODEPAGE_869=m
CONFIG_NLS_ISO8859_1=m
CONFIG_NLS_ISO8859_2=m
CONFIG_NLS_ISO8859_3=m
CONFIG_NLS_ISO8859_4=m
CONFIG_NLS_ISO8859_15=m

#
# Console drivers
#
CONFIG_VGA_CONSOLE=y
CONFIG_VIDEO_SELECT=y

#
# Frame-buffer support
#

#
# Sound
#
CONFIG_SOUND=y
CONFIG_SOUND_VIA82CXXX=m

#
# USB support
#

#
# Kernel hacking
#

You cannot use NFS on reiserfs unless you apply the knfsd patch. Look at www.namesys.com.

Rasmus

-- [ Rasmus 'Møffe' Bøg Hansen ] --------------------------------------
If you only have a hammer
everything looks like a nail
-------------------------------- [ moffe at amagerkollegiet dot dk ] --0

Am Samstag, 2. Juni 2001 12:52 schrieb Rasmus Bøg Hansen:

Thank you very much for your advice.

I tested your suggestion and run the machine without NFS-mounted devices - it seems to be working fine.

Anyway - I'm wondering why I didn't get any problem until 2.4.4ac10 with this configuration without the appropriate patch on the client or on the server?

I'm a little bit confused now about this patch.
Do I need this knfsd-patch for the NFS-server or just for the clients or for both?

Thank you for your advice,
Andreas Hartmann

The problem only happens when the clients do an operation on a file that has gone out of cache on the server. Under light load, this might happen very rarely.

You only need the patch on the server.

-chris

Am Samstag, 2. Juni 2001 18:42 schrieben Sie:

The load didn't change. YOu can forget the load, it's very small. It's my private server and I'm doing always the same thing via NFS - compiling e.g. This has been working fine until 2.4.4.ac10, afterwards it has been broken.

My experiences today are others: I need the patch on both, the server and the client (both 2.4.5) to get it working. See the other mailing to Alan in the list.

Regards,
Andreas Hartmann

Ok, there are two different problems here. The patch you posted to l-k is a generic NFS fix for 2.4.5. ext2 would need this too.

If you are serving NFS from your reiserfs disk, you need an additional patch on the server only (this is the one I was talking about). Checkout the FAQ on www.namesys.com for all the details.

-chris

Am Samstag, 2. Juni 2001 20:33 schrieb Chris Mason:

Just for my understanding:
While I used 2.2.19 without patch on the server (it serves only reiser-based data) and 2.4.[1234]ac[...] on the client (without patch), it has been working because of the light load of the server.

Since 2.4.4ac11 something has been broken in the NFS for 2.4.5, so I ran into problems.

When this conclusion is right, the following combination should work for light load, as I usually have it:
unpatched 2.2.19 on the server (as it has been working with 2.2.4ac10) and NFS-patched 2.4.5. The test showed, that it is working.

The actual combination 2.4.5 on the server and on the client with the mentioned NFS-patch is the same situation as with the unpatched 2.2.19 on the server and the NFS-patch only on the 2.4.5-client.

I hope, that there will be soon a 2.4.5 knfsd-patch for my server, because this is the secure way! And I hope, the broken NFS on the client in the 2.4.5-Kernels will be fixed soon - maybe in the next ac-patch?

Regards,
Andreas Hartmann

Which other 2.4 trees have you tried ?

Does booting with ide=nodma help ? [only in -ac]

Am Samstag, 2. Juni 2001 18:19 schrieben Sie:

I had the following situations:

NFS server:
linux 2.2.19

NFS Client:
linux 2.4.[32]ac[...],
linux 2.4.4ac[1-...]
[1-10] have been working fine. Beginning with ac11, I got the problems I wrote. During this time, I never used any knfsd-patch.

The following is the combination, which seems to be working fine:

NFS Server:
linux 2.2.19 with knfsd-patch or linux 2.4.5 with the following knfsd-Patch from Gergely Tamas <[email protected]> (I got it from the mailinglist of reiser) (there is no patch for ac6):

--------------------------------------------------------------------------------------

@@ -1044,6 +1044,8 @@
                                inode->i_state|=I_FREEING;
                                inodes_stat.nr_inodes--;
                                spin_unlock(&inode_lock);
+                               if (inode->i_data.nrpages)
+                                       truncate_inode_pages(&inode->i_data, 

0);
                                clear_inode(inode);
                        }
                }

@@ -753,6 +753,8 @@
 
        nfs_zap_caches(dir);
        error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name);
+       if (!error)
+               dentry->d_inode->i_nlink -= 2;
 
        return error;
 }

@@ -870,6 +872,8 @@
        error = NFS_PROTO(dir)->remove(dir, &dentry->d_name);
        if (error < 0)
                goto out;
+       if (inode)
+               inode->i_nlink--;
 
  out_delete:
        /*

------------------------------------------------------------------------

I patched the original 2.4.5-sources.

NFS Client:
linux 2.4.5 with knfsd-patch.

I need the patch on both the server and the client to get it working.

I tested the following combination:

Server
2.2.19 without knfsd-Patch

Client
2.4.5ac6 without knfsd-Patch but ide=nodma

Result:
IO-Errors as I wrote in my initial posting.

Regards
Adnreas Hartmann