tpm-correct-tpm-timeouts-to-jiffies-conversion.patch -> 2.6.27

tpm-correct-tpm-timeouts-to-jiffies-conversion.patch -> 2.6.27

[Jiri Slaby]

Hi,

is it possible to push
tpm-correct-tpm-timeouts-to-jiffies-conversion.patch
(the patch header follows) into 2.6.27, please?

From: Marcin Obara <marcin_obara@users.sourceforge.net>

This patch fixes timeouts conversion to jiffies, by replacing
msecs_to_jiffies() calls with usecs_to_jiffies().  According to TCG TPM
Specification Version 1.2 Revision 103 (pages 166, 167) TPM timeouts and
durations are returned in microseconds (usec) not in miliseconds (msec).

Signed-off-by: Marcin Obara <marcin_obara@users.sourceforge.net>
Cc: Marcel Selhorst <tpm@selhorst.net>
Cc: Kylene Jo Hall <kjhall@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Re: tpm-correct-tpm-timeouts-to-jiffies-conversion.patch -> 2.6.27

[Andrew Morton]

On Thu, 11 Sep 2008 00:11:37 +0200
Jiri Slaby <jirislaby@gmail.com> wrote:

> Hi,
> 
> is it possible to push
> tpm-correct-tpm-timeouts-to-jiffies-conversion.patch
> (the patch header follows) into 2.6.27, please?

Sure.

> From: Marcin Obara <marcin_obara@users.sourceforge.net>
> 
> This patch fixes timeouts conversion to jiffies, by replacing
> msecs_to_jiffies() calls with usecs_to_jiffies().  According to TCG TPM
> Specification Version 1.2 Revision 103 (pages 166, 167) TPM timeouts and
> durations are returned in microseconds (usec) not in miliseconds (msec).
> 
> Signed-off-by: Marcin Obara <marcin_obara@users.sourceforge.net>
> Cc: Marcel Selhorst <tpm@selhorst.net>
> Cc: Kylene Jo Hall <kjhall@us.ibm.com>
> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Why?  I wasn't aware that this fixed anything which anyone had observed
(pokes tongue out at the changelog).

Re: tpm-correct-tpm-timeouts-to-jiffies-conversion.patch -> 2.6.27

[Marcin Obara]

2008/9/11 Andrew Morton <akpm@linux-foundation.org>:
> On Thu, 11 Sep 2008 00:11:37 +0200
> Jiri Slaby <jirislaby@gmail.com> wrote:
>
>> Hi,
>>
>> is it possible to push
>> tpm-correct-tpm-timeouts-to-jiffies-conversion.patch
>> (the patch header follows) into 2.6.27, please?
>
> Sure.
>
>> From: Marcin Obara <marcin_obara@users.sourceforge.net>
>>
>> This patch fixes timeouts conversion to jiffies, by replacing
>> msecs_to_jiffies() calls with usecs_to_jiffies().  According to TCG TPM
>> Specification Version 1.2 Revision 103 (pages 166, 167) TPM timeouts and
>> durations are returned in microseconds (usec) not in miliseconds (msec).
>>
>> Signed-off-by: Marcin Obara <marcin_obara@users.sourceforge.net>
>> Cc: Marcel Selhorst <tpm@selhorst.net>
>> Cc: Kylene Jo Hall <kjhall@us.ibm.com>
>> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
>
> Why?  I wasn't aware that this fixed anything which anyone had observed
> (pokes tongue out at the changelog).
>

This fixes i.e.: long hang while loading TPM driver, if TPM chip
starts in "Idle" state instead of "Ready" state.
Without this patch - 'modprobe' may hang for 30 seconds or more.
Please, push this patch into 2.6.27.

Marcin

Re: tpm-correct-tpm-timeouts-to-jiffies-conversion.patch -> 2.6.27

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 918 bytes --]

On Thu, 11 Sep 2008 12:03:52 +0200, Marcin Obara said:

> > Why?  I wasn't aware that this fixed anything which anyone had observed
> > (pokes tongue out at the changelog).
> >
> 
> This fixes i.e.: long hang while loading TPM driver, if TPM chip
> starts in "Idle" state instead of "Ready" state.
> Without this patch - 'modprobe' may hang for 30 seconds or more.
> Please, push this patch into 2.6.27.

I personally don't care whether this goes into .27 or waits till .28.  However,
I *would* appreciate it if we make sure that whenever it goes go upstream, we
also send tpm-work-around-bug-in-broadcom-bcm0102-chipset.patch at the same
time.

I have to admit that it's not exactly confidence inspiring - if the Broadcom
chip gets timeout units wrong, what *other* issues lurk in its silicon?  It's
bad enough when you find bugs in a RAID chipset or mouse hardware - it's even
worse when it's a security chip.. ;)



[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: tpm-correct-tpm-timeouts-to-jiffies-conversion.patch -> 2.6.27

[Andrew Morton]

On Thu, 11 Sep 2008 23:10:08 -0400 Valdis.Kletnieks@vt.edu wrote:

> On Thu, 11 Sep 2008 12:03:52 +0200, Marcin Obara said:
> 
> > > Why?  I wasn't aware that this fixed anything which anyone had observed
> > > (pokes tongue out at the changelog).
> > >
> > 
> > This fixes i.e.: long hang while loading TPM driver, if TPM chip
> > starts in "Idle" state instead of "Ready" state.
> > Without this patch - 'modprobe' may hang for 30 seconds or more.
> > Please, push this patch into 2.6.27.
> 
> I personally don't care whether this goes into .27 or waits till .28.  However,
> I *would* appreciate it if we make sure that whenever it goes go upstream, we
> also send tpm-work-around-bug-in-broadcom-bcm0102-chipset.patch at the same
> time.
> 
> I have to admit that it's not exactly confidence inspiring - if the Broadcom
> chip gets timeout units wrong, what *other* issues lurk in its silicon?  It's
> bad enough when you find bugs in a RAID chipset or mouse hardware - it's even
> worse when it's a security chip.. ;)
> 

hm, it's rather a lot of patching just to fix a once-off 30-second
pause.  On second thoughts, let's do it all in 2.6.28.