Bug 441652 - Problem with Fusion MPT
Problem with Fusion MPT
Product: Fedora
Classification: Fedora
Component: kernel (Show other bugs)
x86_64 Linux
low Severity urgent
: ---
: ---
Assigned To: Kernel Maintainer List
Fedora Extras Quality Assurance
Depends On:
  Show dependency treegraph
Reported: 2008-04-09 07:14 EDT by Patrick De Maziere
Modified: 2008-10-22 17:25 EDT (History)
2 users (show)

See Also:
Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of:
Last Closed: 2008-10-22 17:25:55 EDT
Type: ---
Regression: ---
Mount Type: ---
Documentation: ---
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---

Attachments (Terms of Use)
dmidecode result of problematic computer (27.68 KB, text/plain)
2008-04-09 07:14 EDT, Patrick De Maziere
no flags Details
Tar Gzipped file containing screenshots of the problem (7.56 MB, application/octet-stream)
2008-04-23 04:10 EDT, Patrick De Maziere
no flags Details

  None (edit)
Description Patrick De Maziere 2008-04-09 07:14:35 EDT
Description of problem:

I moved from the 2.6.23.xx series kernel to the 2.6.24.xx series kernel, notably
the, in order to be able to attach a USB disk array bigger than
2 TB (A LaCie Biggest Quadra 4TB), since the 2.26.23.xx and lower kernels report
that they cannot read devices > 2 TB, and return the so-called READ CAPACITY(16)
failed error :-( . Using my laptop with FC8 and kernel-2.6.24-3-50.fc8.i386, I
spotted that this kernel did the trick. I then started updating our workstations
 (Dell T7400, see dmidecode specs in attach) with kernel-2.6.24-3-50.fc8.x86_64,
but then the Fusion MPT base driver 3.04.06 starts giving errors, reporting that
"IOC is in fault state". After waiting a very long time, and some iterations
between "IOC is in fault state" and "IOC recovered" messages, he continues
booting, but then reports, as expected, that he could not find his disks. So, no
working computer after waiting a long time ....

The LSI device/Fusion info is (from /var/log/messages, under
kernel-, the working kernel thus)

Fusion MPT base driver 3.04.05
Copyright (c) 1999-2007 LSI Logic Corporation
Fusion MPT SAS Host driver 3.04.05
ACPI: PCI Interrupt 0000:05:00.0[A] -> GSI 56 (level, low) -> IRQ 56
mptbase: Initiating ioc0 bringup
ioc0: LSISAS1068E B3: Capabilities={Initiator}
scsi0 : ioc0: LSISAS1068E B3, FwRev=00142d00h, Ports=1, MaxQ=511, IRQ=56
scsi 0:0:0:0: Direct-Access     ATA      ST3750640AS      G    PQ: 0 ANSI: 5
sd 0:0:0:0: [sda] 1465149168 512-byte hardware sectors (750156 MB)

I see that the kernel- uses the Fusion MPT base driver 3.04.05,
and this gives thus no problem for booting, nor for working with the workstation

So, could you correct this problem by issuing a newer version of the Fusion MPT
driver/kernel that works, or tell me how
a) I can make a 2.6.23.xx FC8 series kernel read USB filesystems > 2TB
b) I can update the mptbase and related modules so that the 2.6.24.xx FC8
kernels boot my system correctly

Version-Release number of selected component (if applicable):
see above

How reproducible:

Steps to Reproduce:
1. install a 2.6.24.xx kernel
2. boot with this kernel
Actual results:
no working system

Expected results:
a working system :-)

Additional info:

In attach you find the dmidecode info. You can also go directly to the dell
website and request the specs of our machine, using the service tag 35L7H3J
Comment 1 Patrick De Maziere 2008-04-09 07:14:35 EDT
Created attachment 301779 [details]
dmidecode result of problematic computer
Comment 2 Patrick De Maziere 2008-04-09 07:50:57 EDT
Additional info:

Knowing that the LSISAS1068E is used in computer, I went to http://www.lsi.com/
and downloaded their latest mpt-driver: untarred the RHEL5-driver, did a
rpmbuild --rebuild of the src.rpm and then a rpm -ivh of the x86_64.rpms.
Installed without a flaw, but still only the 2.6.23.xx kernel led to a correct
boot, while the 2.6.24.xx still did not lead to a working system ....

Booting with the 2.6.24.xx gives also an additional message I forgot to mention
in the main text, namely:
Driver 'sd' needs updating - please use bus_type methods

Any idea ???
Comment 3 Chuck Ebbert 2008-04-10 19:26:48 EDT
Is the latest firmware installed on the controller?

You could try disabling PCI MSI with the kernel option "pci=nomsi" but that
probably won't help.

Can you take a picture of the screen and attach that so we can see the exact
error messages?
Comment 4 Patrick De Maziere 2008-04-16 11:02:14 EDT
I will deliver a description (digital photo) of the screen next week (April, 
22nd), since the computers are now performing analyses which take some days 
and I cannot stop them for now without risking my life due to the angry users 
that are awaiting their results :-)
Comment 5 Patrick De Maziere 2008-04-23 04:10:46 EDT
Created attachment 303450 [details]
Tar Gzipped file containing screenshots of the problem

As promised, the screen shots of the problematic SATA Controller.
I put them in  a zip. Apologies for the sometimes blurry images, but taking
quickly pictures from a scrolling non-flat screen is not that easy....

Hope you can solve the problem using this info.
Let me know if you need more info...

Comment 6 Chuck Ebbert 2008-04-23 22:44:45 EDT
Is the adapter firmware up-to-date? Many people report mysterious problems just
go away when they update the firmware...
Comment 7 Chuck Ebbert 2008-04-23 22:45:48 EDT
(In reply to comment #5)
> Created an attachment (id=303450) [edit]
> Tar Gzipped file containing screenshots of the problem

Please don't upload .tar.gz files -- we need to be able to just click on them.
Comment 8 Christopher D. Stover 2008-10-22 17:25:55 EDT
The information we've requested above is required in order
to review this problem report further and diagnose or fix the
issue if it is still present.  Since it has been thirty days or
more since we first requested additional information, we're assuming
the problem is either no longer present in the current Fedora release, or
that there is no longer any interest in tracking the problem.

Setting status to "CLOSED: INSUFFICIENT_DATA".  If you still
experience this problem after updating to our latest Fedora
release and can provide the information previously requested,
please feel free to reopen the bug report.

Thank you in advance.

Note You need to log in before you can comment on or make changes to this bug.