i386/71395: Data corrupted on Serverworks CG-SL chipset

[Horacy Kapala]

>Number:         71395
>Category:       i386
>Synopsis:       Data corrupted on Serverworks CG-SL chipset
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    freebsd-i386
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Sun Sep 05 10:30:17 GMT 2004
>Closed-Date:
>Last-Modified:
>Originator:     Horacy Kapala
>Release:        4.10
>Organization:
>Environment:
# uname -a
FreeBSD serve.foo 4.10-STABLE FreeBSD 4.10-STABLE #0: Wed Aug 25 10:37:47 EDT 2004     root at serve.foo:/usr/src/sys/compile/BOGUS  i386

>Description:
My FreeBSD has problems with supporting motherboard on ServerWorks CG-SL chipset. 

I've data corrupted when reading(writing?) big files. I discovered this when I was unable to unpack big gzip archive. Some tests:

I tried to generate md5 checksum a few times for one and the same file. Checksums aren't match!!!
MD5 (testfile.tar.gz) = c588d7e9d005b117060be9cf2e788ce9
MD5 (testfile.tar.gz) = 7961a831f93b2ef58168becce154f864
MD5 (testfile.tar.gz) = 7a656381e538efe28eaccd194ea863cc
MD5 (testfile.tar.gz) = c588d7e9d005b117060be9cf2e788ce9 

The testfile is 1GB of size. I can't reproduce this error with 100MB file (tried 50 times).

The motherboard is TYAN Tiger GC-SL (S2727). It was replaced to brand new with newest BIOS. Problem still exists. HTT was disabled in BIOS - also it wasn't helpful. I tried test on two HDDs - both a new ones.

Last I connected external PROMISE ATA/133 IDE controller. I am still afftected by this problem!!! Seems to be a FreeBSD bug in supportng all chipset CG-SL.

My dmesg:

Sep  4 05:18:51 serve /kernel: Timecounter "i8254"  frequency 1193182 Hz
Sep  4 05:18:51 serve /kernel: CPU: Intel(R) Xeon(TM) CPU 2.40GHz (2392.04-MHz 686-class CPU)
Sep  4 05:18:51 serve /kernel: Origin = "GenuineIntel"  Id = 0xf29  Stepping = 9
Sep  4 05:18:51 serve /kernel: Features=0xbfebfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,
MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE>
Sep  4 05:18:51 serve /kernel: real memory  = 1610612736 (1572864K bytes)
Sep  4 05:18:51 serve /kernel: avail memory = 1563869184 (1527216K bytes)
Sep  4 05:18:51 serve /kernel: Programming 16 pins in IOAPIC #0
Sep  4 05:18:51 serve /kernel: IOAPIC #0 intpin 2 -> irq 0
Sep  4 05:18:51 serve /kernel: Programming 16 pins in IOAPIC #1
Sep  4 05:18:51 serve /kernel: Programming 16 pins in IOAPIC #2
Sep  4 05:18:51 serve /kernel: FreeBSD/SMP: Multiprocessor motherboard: 2 CPUs
Sep  4 05:18:51 serve /kernel: cpu0 (BSP): apic id:  0, version: 0x00050014, at 0xfee00000
Sep  4 05:18:51 serve /kernel: cpu1 (AP):  apic id:  6, version: 0x00050014, at 0xfee00000
Sep  4 05:18:51 serve /kernel: io0 (APIC): apic id:  7, version: 0x000f0011, at 0xfec00000
Sep  4 05:18:51 serve /kernel: io1 (APIC): apic id:  8, version: 0x000f0011, at 0xfec01000
Sep  4 05:18:51 serve /kernel: io2 (APIC): apic id:  9, version: 0x000f0011, at 0xfec02000
Sep  4 05:18:51 serve /kernel: Preloaded elf kernel "kernel" at 0xc034a000.
Sep  4 05:18:51 serve /kernel: Warning: Pentium 4 CPU: PSE disabled
Sep  4 05:18:51 serve /kernel: Pentium Pro MTRR support enabled
Sep  4 05:18:51 serve /kernel: md0: Malloc disk
Sep  4 05:18:51 serve /kernel: Using $PIR table, 12 entries at 0xc00f3ce0
Sep  4 05:18:51 serve /kernel: npx0: <math processor> on motherboard
Sep  4 05:18:51 serve /kernel: npx0: INT 16 interface
Sep  4 05:18:51 serve /kernel: pcib0: <ServerWorks host to PCI bridge(unknown chipset)> on motherboard
Sep  4 05:18:51 serve /kernel: IOAPIC #1 intpin 6 -> irq 2
Sep  4 05:18:51 serve /kernel: IOAPIC #1 intpin 1 -> irq 5
Sep  4 05:18:51 serve /kernel: IOAPIC #1 intpin 14 -> irq 9
Sep  4 05:18:51 serve /kernel: IOAPIC #0 intpin 11 -> irq 10
Sep  4 05:18:51 serve /kernel: IOAPIC #0 intpin 10 -> irq 11
Sep  4 05:18:51 serve /kernel: pci0: <PCI bus> on pcib0
Sep  4 05:18:51 serve /kernel: atapci0: <Promise TX2 ATA133 controller> port 0xb000-0xb00f,0xb400-0xb403,0xb800-0xb807,0xc000-0xc003,0xc400-
0xc407 mem 0xfe9f8000-0xfe9fbfff irq 2 at device 6.0 on pci0
Sep  4 05:18:51 serve /kernel: ata2: at 0xc400 on atapci0
Sep  4 05:18:51 serve /kernel: ata3: at 0xb800 on atapci0
Sep  4 05:18:51 serve /kernel: fxp0: <Intel 82551 Pro/100 Ethernet> port 0xc800-0xc83f mem 0xfe9a0000-0xfe9bffff,0xfe9fd000-0xfe9fdfff irq 5
 at device 9.0 on pci0
Sep  4 05:18:51 serve /kernel: fxp0: Ethernet address 00:e0:81:25:cf:05
Sep  4 05:18:51 serve /kernel: inphy0: <i82555 10/100 media interface> on miibus0
Sep  4 05:18:51 serve /kernel: inphy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
Sep  4 05:18:51 serve /kernel: pci0: <ATI Mach64-GR graphics accelerator> at 10.0 irq 9
Sep  4 05:18:51 serve /kernel: atapci1: <ServerWorks CSB6 ATA100 controller (channel 0+1)> port 0xd000-0xd00f,0xd400-0xd403,0xd800-0xd807,0x
e000-0xe003,0xe400-0xe407 irq 10 at device 14.0 on pci0
Sep  4 05:18:51 serve /kernel: ata0: at 0x1f0 irq 14 on atapci1
Sep  4 05:18:51 serve /kernel: ata1: at 0x170 irq 15 on atapci1
Sep  4 05:18:51 serve /kernel: isab0: <PCI to ISA bridge (vendor=1166 device=0203)> at device 15.0 on pci0
Sep  4 05:18:51 serve /kernel: isa0: <ISA bus> on isab0
Sep  4 05:18:51 serve /kernel: atapci2: <ServerWorks CSB6 ATA100 controller (channel 0+1)> port 0xffa0-0xffaf,0x374-0x377,0x170-0x177,0x3f4-
0x3f7,0x1f0-0x1f7 at device 15.1 on pci0
Sep  4 05:18:51 serve /kernel: ohci0: <OHCI (generic) USB controller> mem 0xfe9fe000-0xfe9fefff irq 11 at device 15.2 on pci0
Sep  4 05:18:51 serve /kernel: usb0: OHCI version 1.0, legacy support
Sep  4 05:18:51 serve /kernel: usb0: <OHCI (generic) USB controller> on ohci0
Sep  4 05:18:51 serve /kernel: usb0: USB revision 1.0
Sep  4 05:18:51 serve /kernel: uhub0: (0x1166) OHCI root hub, class 9/0, rev 1.00/1.00, addr 1
Sep  4 05:18:51 serve /kernel: uhub0: 4 ports with 4 removable, self powered
Sep  4 05:18:51 serve /kernel: pcib1: <ServerWorks host to PCI bridge(unknown chipset)> on motherboard
Sep  4 05:18:51 serve /kernel: pci1: <PCI bus> on pcib1
Sep  4 05:18:51 serve /kernel: pcib2: <Host to PCI bridge> on motherboard
Sep  4 05:18:51 serve /kernel: pci2: <PCI bus> on pcib2
Sep  4 05:18:51 serve /kernel: orm0: <Option ROMs> at iomem 0xc0000-0xc7fff,0xc8000-0xc97ff,0xc9800-0xcbfff on isa0
Sep  4 05:18:51 serve /kernel: pmtimer0 on isa0
Sep  4 05:18:51 serve /kernel: fdc0: ready for input in output
Sep  4 05:18:51 serve /kernel: fdc0: cmd 3 failed at out byte 1 of 3
Sep  4 05:18:51 serve /kernel: atkbdc0: <Keyboard controller (i8042)> at port 0x60,0x64 on isa0
Sep  4 05:18:51 serve /kernel: atkbd0: <AT Keyboard> flags 0x1 irq 1 on atkbdc0
Sep  4 05:18:51 serve /kernel: kbd0 at atkbd0
Sep  4 05:18:51 serve /kernel: vga0: <Generic ISA VGA> at port 0x3c0-0x3df iomem 0xa0000-0xbffff on isa0
Sep  4 05:18:51 serve /kernel: sc0: <System console> at flags 0x100 on isa0
Sep  4 05:18:51 serve /kernel: sc0: VGA <16 virtual consoles, flags=0x300>
Sep  4 05:18:51 serve /kernel: sio0 at port 0x3f8-0x3ff irq 4 flags 0x10 on isa0
Sep  4 05:18:51 serve /kernel: sio0: type 16550A
Sep  4 05:18:51 serve /kernel: sio1: configured irq 3 not in bitmap of probed irqs 0
Sep  4 05:18:51 serve /kernel: ppc0: parallel port not found.
Sep  4 05:18:51 serve /kernel: APIC_IO: Testing 8254 interrupt delivery
Sep  4 05:18:51 serve /kernel: APIC_IO: Broken MP table detected: 8254 is not connected to IOAPIC #0 intpin 2
Sep  4 05:18:51 serve /kernel: APIC_IO: routing 8254 via 8259 and IOAPIC #0 intpin 0
Sep  4 05:18:51 serve /kernel: IP packet filtering initialized, divert disabled, rule-based forwarding enabled, default to accept, unlimited
 logging
Sep  4 05:18:51 serve /kernel: DUMMYNET initialized (011031)
Sep  4 05:18:51 serve /kernel: SMP: AP CPU #1 Launched!
Sep  4 05:18:51 serve /kernel: ad4: 78167MB <Maxtor 6Y080L0> [158816/16/63] at ata2-master UDMA133
Sep  4 05:18:51 serve /kernel: ad6: 78167MB <Maxtor 6Y080L0> [158816/16/63] at ata3-master UDMA133
Sep  4 05:18:51 serve /kernel: Mounting root from ufs:/dev/ad4s1a

>How-To-Repeat:
Probably try any CG-SL motherboard...
>Fix:
Enabling PIO4 mode is semi-helpful. There are less errors, but there are still exist.
>Release-Note:
>Audit-Trail:
>Unformatted: