Bug 29452

(switched to email.  Please respond via emailed reply-to-all, not via the
bugzilla web interface).

2.6.37->2.6.37.1 tpm_tis regression.  Presumably also present in
2.6.38-rcX.  Presumably caused by one of

commit 44489516c52b3b76d9b2a0e670a26b6e64938ddf
Author: Stefan Berger <stefanb@linux.vnet.ibm.com>
Date:   Tue Jan 11 14:37:29 2011 -0500

    tpm_tis: Use timeouts returned from TPM
    
    commit 9b29050f8f75916f974a2d231ae5d3cd59792296 upstream.

commit 926636ad472462044f06af264031f4a776504971
Author: Rajiv Andrade <srajiv@linux.vnet.ibm.com>
Date:   Fri Nov 12 22:30:02 2010 +0100

    TPM: Long default timeout fix
    
    commit c4ff4b829ef9e6353c0b133b7adb564a68054979 upstream.



On Sat, 19 Feb 2011 18:58:06 GMT
bugzilla-daemon@bugzilla.kernel.org wrote:

> https://bugzilla.kernel.org/show_bug.cgi?id=29452
> 
>            Summary: 2.6.37.1 breaks tpm_tis
>            Product: Drivers
>            Version: 2.5
>     Kernel Version: 2.6.37.1
>           Platform: All
>         OS/Version: Linux
>               Tree: Mainline
>             Status: NEW
>           Severity: normal
>           Priority: P1
>          Component: Other
>         AssignedTo: drivers_other@kernel-bugs.osdl.org
>         ReportedBy: mnl@mnl.de
>         Regression: No
> 
> 
> After upgrading from 2.6.37 to 2.6.37.1 there is some problem with tpm_tis.
> During boot I get:
> 
> Feb 19 19:22:42 lipp-nb kernel: [    9.664344] tpm_tis 00:04: 1.2 TPM
> (device-id 0xB, rev-id 16)
> Feb 19 19:22:42 lipp-nb kernel: [   10.000017] tpm_tis 00:04: Operation Timed
> out
> 
> The real problem occurs when I later try to go to standby:
> 
> 
> Feb 19 13:00:25 lipp-nb kernel: [  174.292169] tpm_tis 00:04: Operation Timed
> out
> Feb 19 13:00:25 lipp-nb kernel: [  174.292177] legacy_suspend():
> pnp_bus_suspend+0x0/0x70 returns -62
> Feb 19 13:00:25 lipp-nb kernel: [  174.292181] PM: Device 00:04 failed to
> suspend: error -62
> 
> I'm sorry, but that's all there is. The only additional information I have is
> that it is an infineon device (tpm_infineon gets loaded as well). After
> unloading all tpm* modules, stand by works again.
>

Please verify if reverting commit 44489516c52b3b76d9b2a0e670a26b6e64938ddf
fixes the problem.

On Wed, Feb 23, 2011 at 02:40:30PM -0800, Andrew Morton wrote:
> 
> (switched to email.  Please respond via emailed reply-to-all, not via the
> bugzilla web interface).
> 
> 2.6.37->2.6.37.1 tpm_tis regression.  Presumably also present in
> 2.6.38-rcX.  Presumably caused by one of
> 
> commit 44489516c52b3b76d9b2a0e670a26b6e64938ddf
> Author: Stefan Berger <stefanb@linux.vnet.ibm.com>
> Date:   Tue Jan 11 14:37:29 2011 -0500
> 
>     tpm_tis: Use timeouts returned from TPM
>     
>     commit 9b29050f8f75916f974a2d231ae5d3cd59792296 upstream.
> 
> commit 926636ad472462044f06af264031f4a776504971
> Author: Rajiv Andrade <srajiv@linux.vnet.ibm.com>
> Date:   Fri Nov 12 22:30:02 2010 +0100
> 
>     TPM: Long default timeout fix
>     
>     commit c4ff4b829ef9e6353c0b133b7adb564a68054979 upstream.

Already known, and reverted in 2.6.37.2-rc1.

On Wednesday, February 23, 2011, Andrew Morton wrote:
> 
> (switched to email.  Please respond via emailed reply-to-all, not via the
> bugzilla web interface).
> 
> 2.6.37->2.6.37.1 tpm_tis regression.  Presumably also present in
> 2.6.38-rcX.  Presumably caused by one of
> 
> commit 44489516c52b3b76d9b2a0e670a26b6e64938ddf
> Author: Stefan Berger <stefanb@linux.vnet.ibm.com>
> Date:   Tue Jan 11 14:37:29 2011 -0500
> 
>     tpm_tis: Use timeouts returned from TPM
>     
>     commit 9b29050f8f75916f974a2d231ae5d3cd59792296 upstream.

This one has been reverted from .38-rc and will be reverted in .37.2.

So, it would be good to verify if the issue is still present in Linus' current.

> commit 926636ad472462044f06af264031f4a776504971
> Author: Rajiv Andrade <srajiv@linux.vnet.ibm.com>
> Date:   Fri Nov 12 22:30:02 2010 +0100
> 
>     TPM: Long default timeout fix
>     
>     commit c4ff4b829ef9e6353c0b133b7adb564a68054979 upstream.
> 
> 
> 
> On Sat, 19 Feb 2011 18:58:06 GMT
> bugzilla-daemon@bugzilla.kernel.org wrote:
> 
> > https://bugzilla.kernel.org/show_bug.cgi?id=29452
> > 
> >            Summary: 2.6.37.1 breaks tpm_tis
> >            Product: Drivers
> >            Version: 2.5
> >     Kernel Version: 2.6.37.1
> >           Platform: All
> >         OS/Version: Linux
> >               Tree: Mainline
> >             Status: NEW
> >           Severity: normal
> >           Priority: P1
> >          Component: Other
> >         AssignedTo: drivers_other@kernel-bugs.osdl.org
> >         ReportedBy: mnl@mnl.de
> >         Regression: No
> > 
> > 
> > After upgrading from 2.6.37 to 2.6.37.1 there is some problem with tpm_tis.
> > During boot I get:
> > 
> > Feb 19 19:22:42 lipp-nb kernel: [    9.664344] tpm_tis 00:04: 1.2 TPM
> > (device-id 0xB, rev-id 16)
> > Feb 19 19:22:42 lipp-nb kernel: [   10.000017] tpm_tis 00:04: Operation
> Timed
> > out
> > 
> > The real problem occurs when I later try to go to standby:
> > 
> > 
> > Feb 19 13:00:25 lipp-nb kernel: [  174.292169] tpm_tis 00:04: Operation
> Timed
> > out
> > Feb 19 13:00:25 lipp-nb kernel: [  174.292177] legacy_suspend():
> > pnp_bus_suspend+0x0/0x70 returns -62
> > Feb 19 13:00:25 lipp-nb kernel: [  174.292181] PM: Device 00:04 failed to
> > suspend: error -62
> > 
> > I'm sorry, but that's all there is. The only additional information I have
> is
> > that it is an infineon device (tpm_infineon gets loaded as well). After
> > unloading all tpm* modules, stand by works again.
> > 
> 
>

I'd like to take another attempt at having the patch applied - likely once the next window opens.

The crux here is with non-compliant devices that return timeouts in milliseconds rather than microseconds, so timeouts are a factor of 1000 shorter than they are supposed to be. I didn't know that the Infineon TPM had the same problem.

Would it be possible that you then show the output of the following sysfs entries

cat /sys/devince/pnp0/00:06/pcrs
cat /sys/devince/pnp0/00:06/caps

to verify that the device works and

cat /sys/devices/pnp0/00:06/timeouts

to see what its timeouts are? That would be helpful. Thanks.

OK, I loaded the modules manually (I have them blacklisted currently). syslog says

Feb 24 08:53:21 lipp-nb kernel: [69439.820201] tpm_inf_pnp 00:04: Found GTPM with ID IFX0102
Feb 24 08:53:21 lipp-nb kernel: [69439.820255] tpm_inf_pnp 00:04: TPM found: config base 0xfe00, data base 0xfe80, chip version 0x000b, vendor id 0x15d1 (Infineon), product id 0x000b (SLB 9635 TT 1.2)


$ cat /sys/devices/pnp0/00:04/pcrs
does not show anything

$ cat /sys/devices/pnp0/00:04/caps
Manufacturer: 0x49465800
TCG version: 1.1
Firmware version: 0.0

Fixed by: 8d1dc20 Revert "TPM: Long default timeout fix"
e587137 Revert "tpm_tis: Use timeouts returned from TPM"