A Wolverine installer dies in many interesting ways if one tries to turn on bad blocks checking; at least on a machine with PDC20265 controller. Effects vary - I got an instant spontaneous reboot, a backtrace from anaconda which, unfortunately, I could not save to a floppy as everything just died and attempt to repeat failed, just a silent lockup... You name it. On one occasion I was left with enough of shell to be able to save /tmp/syslog file to a floppy. It is attached to this report. BTW - using a boot floppy made from a boot.img from Wolverine and RH7 CD for starting an installation has the same effects. Note: the machine on which this behaviour was observed is not mine and pretty soon I may not have an access to it. :-( Michal michal
Created attachment 11040 [details] syslog file saved from a failed installation attempt
What is a PDC20265 controller?
> What is a PDC20265 controller? IDE multichannel controller from Promise. See also #29472 and #29508. Here is 'lspci -vv' output for this device: 00:11.0 Unknown mass storage controller: Promise Technology, Inc. 20265 (rev 02) Subsystem: Promise Technology, Inc.: Unknown device 4d33 Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- Latency: 32 Interrupt: pin A routed to IRQ 10 Region 0: I/O ports at 8000 [size=8] Region 1: I/O ports at 7800 [size=4] Region 2: I/O ports at 7400 [size=8] Region 3: I/O ports at 7000 [size=4] Region 4: I/O ports at 6800 [size=64] Region 5: Memory at d9000000 (32-bit, non-prefetchable) [size=128K] Expansion ROM at <unassigned> [disabled] [size=64K] Capabilities: [58] Power Management version 1 Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-) Status: D0 PME-Enable- DSel=0 DScale=0 PME-
*** Bug 29472 has been marked as a duplicate of this bug. ***
*** Bug 29499 has been marked as a duplicate of this bug. ***
This defect is considered MUST-FIX for Florence Gold release
I have an Asus A7V motherboard with the same Promise controller and I've had no problems installing and using it as my test machine for anaconda development. What brand motherboard do you have, and what version of the BIOS are you running?
It is Asus A7V (small world) with Award ACPI BIOS 1005C. Which BIOS version you are using?
Here are reports which seems to be related to what I am seeing: http://www.uwsg.indiana.edu/hypermail/linux/kernel/0102.2/0888.html http://www.uwsg.indiana.edu/hypermail/linux/kernel/0102.3/0022.html
Will try with the latest BIOS, I have 1004D.
BUILD: Wolverine AND qa0228.0 HARDWARE: ASUS A7V --Promise PDC20265 controller--Maxtor 53073UG (30Gig) and CDROM BIOS: Award ACPI Rev. 1005C lspci -v: 00:11.0 Unknown mass storage controller: Promise Technology, Inc. 20265 (rev 02) Subsystem: Promise Technology, Inc.: Unknown device 4d33 Flags: bus master, medium devsel, latency 32, IRQ 10 I/O ports at 9000 [size=8] I/O ports at 8800 [size=4] I/O ports at 8400 [size=8] I/O ports at 8000 [size=4] I/O ports at 7800 [size=64] Memory at e1000000 (32-bit, non-prefetchable) [size=128K] Expansion ROM at <unassigned> [disabled] [size=64K] Capabilities: [58] Power Management version 1 Performed: NFS-Custom w/ bad block check -- completed successfully post-install -- OK GUI-CDROM w/ bad block check -- completed successfully post-install -- OK Configurations: Installations completed successfully with devices attached to ATA100-primary (hd) and ATA-100-secondary (cdrom); and with devices on ATA100-primary (hd) and IDE-primary (cdrom).
Please check my latest comments to #29508. I still have to check the whole installation but this may be it.
Adding Bob to CC list since he's investigating Promise problems.
Arjan and I have been working with the settings on these motherboards trying to determine the cause of the problem. Here's what we've come up with: V = (PCI Dram Master Pipeline Request, Enhanced CPU to PCI Write, PCI Burst, Post Write Retry Fail) V = 1101, default settings in BIOS v. 1003, 1004, 1004D, fails with FS errors. 1100, FS errors 1011, FS errors 0101, FS errors 0010, no FS errors So, it is some combination of settings which is causing the problems, but we know a setting which appears to work.
Few days ago a new BIOS version 1007 showed up. Unfortunately I do not see much difference after an update (with an exception that with a new BIOS if "System Performance" is set to "Normal" then BIOS also displays "Normal" and not "Optimal" like with 1005C). Still with an "Optimal" setting a system disintegrates after a short while in many interesting ways. OTOH with "Normal", and with "Legacy USB support" turned off, I went through roughly 40 GB worth of disk with 'badblocks' checks and it passed.
The via-corruption patch in kernel 0.1.21 has fixed this problem. The machine ran a minimal Cerberus test (including all the FS tests) with no corruption.