Enable TPM in developer mode again.

Enable TPM in developer mode again.

Also fix a few comments, and make extra debugging work when compiled
in firmware.

BUG=none
TEST=make && make runtests; all pass

[semenzato, 2010-08-05 19:50:53]

This would look fine except I missed some earlier changes and I no longer
understand what the goal is, so I'll come over and talk to you.

http://codereview.chromium.org/3007036/diff/3001/4001
File firmware/lib/rollback_index.c (right):

http://codereview.chromium.org/3007036/diff/3001/4001#newcode340
firmware/lib/rollback_index.c:340: TlclSelfTestFull();
This has changed a lot and I don't understand why it's not calling SetupTPM. 
Maybe we can talk.

http://codereview.chromium.org/3007036/diff/3001/4002
File firmware/lib/tpm_lite/tlcl.c (right):

http://codereview.chromium.org/3007036/diff/3001/4002#newcode64
firmware/lib/tpm_lite/tlcl.c:64: #ifdef EXTRA_LOGGING
We should do something about VBDEBUG when compiling for user-level execution. 
Right now it does nothing, but I'd like to see those messages.

Also, you don't like #if instead of #ifdef?  The advantage of #if is that the
variable must always be defined, which makes it easier for the next programmer
to figure out WHERE to define it.

http://codereview.chromium.org/3007036/diff/3001/4003
File firmware/lib/vboot_firmware.c (right):

http://codereview.chromium.org/3007036/diff/3001/4003#newcode66
firmware/lib/vboot_firmware.c:66: VBDEBUG(("Unable to setup TPM.\n"));
This message may be too scary when the firmware is just reenabling the TPM.

[semenzato, 2010-08-05 19:58:09]

Ah OK, sorry, I missed the #ifdef DISABLE_ROLLBACK_TPM.  

LGTM

On 2010/08/05 19:50:53, semenzato wrote:
> This would look fine except I missed some earlier changes and I no longer
> understand what the goal is, so I'll come over and talk to you.
> 
> http://codereview.chromium.org/3007036/diff/3001/4001
> File firmware/lib/rollback_index.c (right):
> 
> http://codereview.chromium.org/3007036/diff/3001/4001#newcode340
> firmware/lib/rollback_index.c:340: TlclSelfTestFull();
> This has changed a lot and I don't understand why it's not calling SetupTPM. 
> Maybe we can talk.
> 
> http://codereview.chromium.org/3007036/diff/3001/4002
> File firmware/lib/tpm_lite/tlcl.c (right):
> 
> http://codereview.chromium.org/3007036/diff/3001/4002#newcode64
> firmware/lib/tpm_lite/tlcl.c:64: #ifdef EXTRA_LOGGING
> We should do something about VBDEBUG when compiling for user-level execution. 
> Right now it does nothing, but I'd like to see those messages.
> 
> Also, you don't like #if instead of #ifdef?  The advantage of #if is that the
> variable must always be defined, which makes it easier for the next programmer
> to figure out WHERE to define it.
> 
> http://codereview.chromium.org/3007036/diff/3001/4003
> File firmware/lib/vboot_firmware.c (right):
> 
> http://codereview.chromium.org/3007036/diff/3001/4003#newcode66
> firmware/lib/vboot_firmware.c:66: VBDEBUG(("Unable to setup TPM.\n"));
> This message may be too scary when the firmware is just reenabling the TPM.

[Luigi Semenzato, 2010-08-05 20:03:27]

And my email is getting rejected even though I have permission to do the code
review :-(

On 2010/08/05 19:58:09, semenzato wrote:
> Ah OK, sorry, I missed the #ifdef DISABLE_ROLLBACK_TPM.  
> 
> LGTM
> 
> On 2010/08/05 19:50:53, semenzato wrote:
> > This would look fine except I missed some earlier changes and I no longer
> > understand what the goal is, so I'll come over and talk to you.
> > 
> > http://codereview.chromium.org/3007036/diff/3001/4001
> > File firmware/lib/rollback_index.c (right):
> > 
> > http://codereview.chromium.org/3007036/diff/3001/4001#newcode340
> > firmware/lib/rollback_index.c:340: TlclSelfTestFull();
> > This has changed a lot and I don't understand why it's not calling SetupTPM.

> > Maybe we can talk.
> > 
> > http://codereview.chromium.org/3007036/diff/3001/4002
> > File firmware/lib/tpm_lite/tlcl.c (right):
> > 
> > http://codereview.chromium.org/3007036/diff/3001/4002#newcode64
> > firmware/lib/tpm_lite/tlcl.c:64: #ifdef EXTRA_LOGGING
> > We should do something about VBDEBUG when compiling for user-level
execution. 
> > Right now it does nothing, but I'd like to see those messages.
> > 
> > Also, you don't like #if instead of #ifdef?  The advantage of #if is that
the
> > variable must always be defined, which makes it easier for the next
programmer
> > to figure out WHERE to define it.
> > 
> > http://codereview.chromium.org/3007036/diff/3001/4003
> > File firmware/lib/vboot_firmware.c (right):
> > 
> > http://codereview.chromium.org/3007036/diff/3001/4003#newcode66
> > firmware/lib/vboot_firmware.c:66: VBDEBUG(("Unable to setup TPM.\n"));
> > This message may be too scary when the firmware is just reenabling the TPM.