Fix WP polarity on Mario

Fix WP polarity on Mario

Note that both the current and boot settings are incorrect as reported
by the BIOS.  This fixes both.

Change-Id: Iebd2b4ac91232444e538f6e2763a22cb227f2e4f

BUG=chrome-os-partner:2078
TEST=manual

Run crossystem on Mario and Alex with WP enabled, disabled and check 
Verify true hardware WP status by doing:
   flashrom --wp-enable
   flashrom --wp-status
   flashrom --wp-disable
   flashrom --wp-status
On a system with hardware WP enabled, the disable step should fail and write
protect will still report enabled.

Committed: http://chrome-svn/viewvc/chromeos?view=rev&revision=c80fe65

[Bill Richardson, 2011-02-17 19:05:48]

LGTM with a couple of nits.

http://codereview.chromium.org/6531035/diff/1/host/lib/crossystem.c
File host/lib/crossystem.c (right):

http://codereview.chromium.org/6531035/diff/1/host/lib/crossystem.c#newcode86
host/lib/crossystem.c:86: /* Returns non-zero if the FWID starts with the
specified string. */
s/non-zero/true/ ?

http://codereview.chromium.org/6531035/diff/1/host/lib/crossystem.c#newcode182
host/lib/crossystem.c:182: if (FwidStartsWith("Mario.") && -1 != value)
reverse the order of this expression. you don't have to compare if we can't read
the GPIO.

http://codereview.chromium.org/6531035/diff/1/host/lib/crossystem.c#newcode186
host/lib/crossystem.c:186: if (FwidStartsWith("Mario.") && -1 != value)
ditto

[Hung-Te, 2011-02-18 01:21:32]

Such design (value = 1 - value) means if someday we have a new Mario firmware
which fixed the ACPI table, then crossystem will report the wrong value.

Fixing by hard-coding polarity seems more reasonable to me, because the value
will be consistent no matter if Mario BIOS has fixed it or not.

[Randall Spangler, 2011-02-18 01:30:12]

On Thu, Feb 17, 2011 at 5:21 PM, <hungte@chromium.org> wrote:

> Such design (value = 1 - value) means if someday we have a new Mario
> firmware
> which fixed the ACPI table, then crossystem will report the wrong value.
>

We're safe, since we won't be fixing Mario - since Mario would then behave
differently for RO vs. RW firmware.

Prior to release, we should fix bugs in firmware so that it works properly.
 Post release, if the bug is in the RO firmware, we may need to add
workarounds in crossystem.

Thanks,
    Randall


> Fixing by hard-coding polarity seems more reasonable to me, because the
> value
> will be consistent no matter if Mario BIOS has fixed it or not.
>
>
> http://codereview.chromium.org/6531035/
>