[staging] phison Module broken: ExpressCard SSD not usable

Bug #669343 reported by Mathieu Simon
26
This bug affects 5 people
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
Won't Fix
Low
Unassigned

Bug Description

**** Fixed with 2.6.37-rc3 kernel *****

Description: Ubuntu 10.10
Release: 10.10

I have been using a Verbatim 32GB SSD ExpressCard that is attached via PCIe as a IDE drive.The card works out of the box in lucid but in maverick dmesg just throws a stack trace.
Shutdown and reboot are blocked until the card is removed.

The card needs the "phison" module.

Ubuntu lucid amd64 (good)
[ 23.003156] pciehp 0000:00:1c.3:pcie04: Card present on Slot(3)
[ 23.130232] pci 0000:05:00.0: reg 10 io port: [0x00-0x07]
[ 23.130265] pci 0000:05:00.0: reg 14 io port: [0x00-0x03]
[ 23.130291] pci 0000:05:00.0: reg 18 io port: [0x00-0x07]
[ 23.130317] pci 0000:05:00.0: reg 1c io port: [0x00-0x03]
[ 23.130343] pci 0000:05:00.0: reg 20 io port: [0x00-0x0f]
[ 23.130603] pci 0000:05:00.0: no hotplug settings from platform
[ 23.130778] pata_acpi 0000:05:00.0: enabling device (0000 -> 0001)
[ 23.130790] pata_acpi 0000:05:00.0: PCI INT A -> GSI 19 (level, low) -> IRQ 19
[ 23.130847] pata_acpi 0000:05:00.0: setting latency timer to 64
[ 23.130875] pata_acpi 0000:05:00.0: PCI INT A disabled
[ 23.137678] phison: module is from the staging directory, the quality is unknown, you have been warned.
[ 23.138392] phison_e-box 0000:05:00.0: PCI INT A -> GSI 19 (level, low) -> IRQ 19
[ 23.138445] phison_e-box 0000:05:00.0: setting latency timer to 64
[ 23.138920] scsi4 : phison_e-box
[ 23.139000] scsi5 : phison_e-box
[ 23.139033] ata5: PATA max UDMA/100 cmd 0x2010 ctl 0x2020 bmdma 0x2000 irq 19
[ 23.139035] ata6: PATA max UDMA/100 cmd 0x2018 ctl 0x2024 bmdma 0x2008 irq 19
[ 23.970550] ata5.00: ATA-7: ExpressCard SSD, PER1.36, max UDMA/133
[ 23.970553] ata5.00: 62881792 sectors, multi 0: LBA
[ 23.990561] ata5.00: configured for UDMA/100
[ 23.990664] scsi 4:0:0:0: Direct-Access ATA ExpressCard SSD PER1 PQ: 0 ANSI: 5
[ 23.990816] sd 4:0:0:0: Attached scsi generic sg2 type 0
[ 23.990918] sd 4:0:0:0: [sdb] 62881792 512-byte logical blocks: (32.1 GB/29.9 GiB)
[ 23.990955] sd 4:0:0:0: [sdb] Write Protect is off
[ 23.990957] sd 4:0:0:0: [sdb] Mode Sense: 00 3a 00 00
[ 23.990976] sd 4:0:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 23.991087] sdb: sdb1
[ 23.992355] sd 4:0:0:0: [sdb] Attached SCSI disk

Then the partitions can be mounted - all fine,

With maverick the output is rather long, the dmesg paste can be found here:
http://pastebin.com/gZXyF6FN

Seems to be a regression? :-/

tags: added: maverick regression-release
summary: - phison Module broken: ExpressCard SSD not usable
+ phison Module broken: ExpressCard SSD not usable [staging]
Revision history for this message
Fabio Marconi (fabiomarconi) wrote :

dmesg pasted, i assign to linux

summary: - phison Module broken: ExpressCard SSD not usable [staging]
+ [staging] phison Module broken: ExpressCard SSD not usable
affects: ubuntu → linux (Ubuntu)
Changed in linux (Ubuntu):
status: New → Confirmed
Revision history for this message
Mathieu Simon (mathieu-simon) wrote :

Just for Information: Latest mainline 2.6.36 also shows same problem, if you want I can try latest 2.6.35 mainline to see whether this is because of a Debian or Ubuntu Patch / Config?

uname -a
Linux 2.6.36-020636-generic #201010210905 SMP Thu Oct 21 09:08:58 UTC 2010 GNU/Linux

Revision history for this message
Mathieu Simon (mathieu-simon) wrote :

Another update I tried the mainline Kernel Image from 2.6.32.25. Since 2.6.32 series are the base for lucids' kernel.
This time worked without flaws, plugged in the card - et voilà

Details: Linux 2.6.32-02063225-generic #201010290905 SMP Fri Oct 29 09:08:55 UTC 2010 x86_64 GNU/Linux

I will try to find out the more exact version where the driver broke.

These ExpressCards are quire fast and handy but a nightmare when it comes to drivers ;-)

Revision history for this message
Mathieu Simon (mathieu-simon) wrote :

I have tracked down the problem with mainline kernels up until 2.6.34.y where it remained working.
According to kernel.org, no changes between 2.6.35 and .36 but they differ by two patches from .34

2.6.37-rc1 adds one patch by gregk which addresses this exact problem, see
https://bugzilla.kernel.org/show_bug.cgi?id=19872

The patch can be found here:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=cf10700bf8047f0668dd874b607f89516612e6c7

I'm looking forward to see this little patch integrated on the Maverick series kernel to fix this problem.

Markus Korn (thekorn)
Changed in linux (Ubuntu):
status: Confirmed → Triaged
importance: Undecided → Low
Revision history for this message
Florian Hofmann (fhaust) wrote :

Is there any chance that we can get the patch integrated into Ubuntu 10.10?

There have been 2 new kernel versions since i encountered this problem and after all it is one line in the kernel that changes a function call to the correct api.

Another problem is that installing Ubuntu from CD onto a PCIExpress SSD card is not possible because of this.

Just wanted to give a shout,
Florian

Revision history for this message
Fabio Marconi (fabiomarconi) wrote :

hello Florian
Have you already tryied the new upstream mainline kernel from:
http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.37-rc2-maverick/
Installing _all.deb ,header and image for your architecture.
Thanks
Fabio

Revision history for this message
Mathieu Simon (mathieu-simon) wrote :

Hi Fabio and Florian

I can confirm that the phison module works again using a 2.6.37-rc3 kernel. A dmesg output looks sane compared to what I got using 2.6.35 and .36-based kernels. Output looks same/similar to what I got under 2.6.34 where the driver worked last time.

uname -a
Linux n-sim-01 2.6.37-020637rc3-generic #201011220905 SMP Mon Nov 22 09:08:55 UTC 2010 x86_64 GNU/Linux

I'd really like to see this bug fixed in maverick's kernel images, essentially because it resulted out of a merge error that happened upstream and it's fix is trivial.

Revision history for this message
Florian Hofmann (fhaust) wrote : Re: [Bug 669343] Re: [staging] phison Module broken: ExpressCard SSD not usable
Download full text (3.6 KiB)

I second that. The .37-rc3 mainline kernel works as expected.
But i also vote for fixing it in maverick ... not everyone likes to
run bleeding edge packages.

2010/12/6 Mathieu Simon <email address hidden>:
> Hi Fabio and Florian
>
> I can confirm that the phison module works again using a 2.6.37-rc3
> kernel. A dmesg output looks sane compared to what I got using 2.6.35
> and .36-based kernels. Output looks same/similar to what I got under
> 2.6.34 where the driver worked last time.
>
> uname -a
> Linux n-sim-01 2.6.37-020637rc3-generic #201011220905 SMP Mon Nov 22 09:08:55 UTC 2010 x86_64 GNU/Linux
>
> I'd really like to see this bug fixed in maverick's kernel images,
> essentially because it resulted out of a merge error that happened
> upstream and it's fix is trivial.
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/669343
>
> Title:
>  [staging] phison Module broken: ExpressCard SSD not usable
>
> Status in “linux” package in Ubuntu:
>  Triaged
>
> Bug description:
>  Description:  Ubuntu 10.10
> Release:        10.10
>
> I have been using a Verbatim 32GB SSD ExpressCard that is attached via PCIe as a IDE drive.
> The card works out of the box in lucid but in maverick dmesg just throws a stack trace.
> Shutdown and reboot are blocked until the card is removed.
>
> The card needs the "phison" module.
>
> Ubuntu lucid amd64 (good)
> [   23.003156] pciehp 0000:00:1c.3:pcie04: Card present on Slot(3)
> [   23.130232] pci 0000:05:00.0: reg 10 io port: [0x00-0x07]
> [   23.130265] pci 0000:05:00.0: reg 14 io port: [0x00-0x03]
> [   23.130291] pci 0000:05:00.0: reg 18 io port: [0x00-0x07]
> [   23.130317] pci 0000:05:00.0: reg 1c io port: [0x00-0x03]
> [   23.130343] pci 0000:05:00.0: reg 20 io port: [0x00-0x0f]
> [   23.130603] pci 0000:05:00.0: no hotplug settings from platform
> [   23.130778] pata_acpi 0000:05:00.0: enabling device (0000 -> 0001)
> [   23.130790] pata_acpi 0000:05:00.0: PCI INT A -> GSI 19 (level, low) -> IRQ 19
> [   23.130847] pata_acpi 0000:05:00.0: setting latency timer to 64
> [   23.130875] pata_acpi 0000:05:00.0: PCI INT A disabled
> [   23.137678] phison: module is from the staging directory, the quality is unknown, you have been warned.
> [   23.138392] phison_e-box 0000:05:00.0: PCI INT A -> GSI 19 (level, low) -> IRQ 19
> [   23.138445] phison_e-box 0000:05:00.0: setting latency timer to 64
> [   23.138920] scsi4 : phison_e-box
> [   23.139000] scsi5 : phison_e-box
> [   23.139033] ata5: PATA max UDMA/100 cmd 0x2010 ctl 0x2020 bmdma 0x2000 irq 19
> [   23.139035] ata6: PATA max UDMA/100 cmd 0x2018 ctl 0x2024 bmdma 0x2008 irq 19
> [   23.970550] ata5.00: ATA-7: ExpressCard SSD, PER1.36, max UDMA/133
> [   23.970553] ata5.00: 62881792 sectors, multi 0: LBA
> [   23.990561] ata5.00: configured for UDMA/100
> [   23.990664] scsi 4:0:0:0: Direct-Access     ATA      ExpressCard SSD  PER1 PQ: 0 ANSI: 5
> [   23.990816] sd 4:0:0:0: Attached scsi generic sg2 type 0
> [   23.990918] sd 4:0:0:0: [sdb] 62881792 512-byte logical blocks: (32.1 GB/29.9 GiB)
> [   23.990955] sd 4:0:0:0: [sdb] Write Protect is off
> [   23.990957] sd 4:0:0:0: [sd...

Read more...

description: updated
Revision history for this message
Joseph Salisbury (jsalisbury) wrote : Unsupported series, setting status to "Won't Fix".

This bug was filed against a series that is no longer supported and so is being marked as Won't Fix. If this issue still exists in a supported series, please file a new bug.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

Changed in linux (Ubuntu):
status: Triaged → Won't Fix
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.