The SH512 code gets quite large when unrolled, about 20KB larger on x86.
This is a net loss on machines with slow SPI. Split this out into a
separate knob, and don't enable it on any architecture for now.
Also swap the code around so that we do #ifdef...#else...#endif instead
of #ifndef...#else...#endif.
BUG=chrome-os-partner:13961
BRANCH=none
TEST=manual
build and boot to kernel on link
U-Boot image size before this change:
text data bss dec hex filename
319403 8260 83988 411651 64803 u-boot
after:
293227 8260 85492 386979 5e7a3 u-boot
This is a saving of about 25KB.
Signed-off-by: Simon Glass <sjg@chromium.org>
Change-Id: I9fa7ea8eba6691d8a06df9374950303e6f2ce2fd
Reviewed-on: https://gerrit.chromium.org/gerrit/40155
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Queue: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
When V=1, the full command lines are printed. When V is not 1, then only a
small summary line is printed which shows what commands are being executed.
The command lines themselves are usually quite long and are overwhelming to
see fly by on the console. Abbreviated command lines are easier to read and
don't fill up your console so quickly.
This change is primarily targeted at vboot_fw.a and probably excludes some
things which could also be converted. The indentation between the action
string (OBJCOPY, CC, etc.) and the target is three spaces longer than "normal",
aka what's used in depthcharge, so that when this make is run from the other,
you can tell the difference between the commands run by each.
BUG=chrome-os-partner:8339
TEST=Built with and without V=1 and saw and did not see the full command
lines, respectively.
BRANCH=None
Change-Id: Ibee244c24dc44b8da109b8c23ac7273174836bb9
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/40011
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Che-Liang Chiou <clchiou@chromium.org>
Commit-Queue: Gabe Black <gabeblack@chromium.org>
Tested-by: Gabe Black <gabeblack@chromium.org>
This just adds a one-byte field in the nvstorage region for use in debugging
hard-to-catch errors. There's no official meaning or expectation for this
field. It's just a handy place to emit some information.
BUG=chrome-os-partner:11534
BRANCH=parrot
TEST=manual
Just change the value and ensure that it persists across a (working) reboot.
It's only updated at specific points under very exacting error conditions,
so all we really want to test is that it works as a place to store some
extra info.
crossystem recovery_subcode
crossystem recovery_subcode=14
reboot
crossystem recovery_subcode
The recovery_subcode byte is at index [6] of the VbNv.raw bytes that appear
when you press TAB, so you can find it there too:
VbNv.raw: 60 20 00 00 00 00 0e 00 00 00 00 00 00 00 00 65
Decimal 14 == 0x0e
Change-Id: I1930b8f81a03ab838dbee99a8d72c35a444efdfd
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/39803
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Make sure all Tlcl users benefit from the new retry logic.
BUG=None
TEST=daisy build, manual testing of racing tpmc loops
BRANCH=None
Change-Id: I8e9656a65b5d6b45694c1c8bceb95f54f7c751bb
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/39525
Reviewed-by: Luigi Semenzato <semenzato@chromium.org>
If there were any errors communicating with the TPM at the OS layer
(open, read, write failures), the library would immediately exit, not
allowing the caller to make any decisions about how to handle it. This
introduces a way to initialize the library so that errors will get passed
back up to the caller instead of unceremoniously exiting.
Setting the environment variable "TPM_NO_EXIT=1" enables the feature. To
avoid needing to implement supporting functions in all backends, the
feature is currently limited to just the Tlcl stub implementation.
In the case of mount-encrypted, it can now survive the kernel returning
read/write failures. In the past it had only worked around having open
fail, but that has now been replaced with more sensible logic instead of
the environment variable trickiness.
BUG=chrome-os-partner:15960
TEST=daisy built with an always-failing kernel driver, u-boot builds too
BRANCH=None
Change-Id: Ic7b217017537980f9c239d678067398613045676
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/38791
Reviewed-by: Luigi Semenzato <semenzato@chromium.org>
There are several places where the same recovery_reason was used to report
slightly different points of failure. Let's create some new codes instead.
Remember that recovery mode is handled by RO firmware, so if an updated RW
firmware uses one of the new error codes, pressing TAB at the recovery
screen will say "We have no idea what this means". That's not a bug. This CL
deprecates the original codes, so the fact that the RO firmware doesn't
recognize it just means it's a new code reported by a new RW BIOS.
BUG=chromium-os:36562
TEST=manual
BRANCH=parrot
Run
make && make runtests
It should pass. You can test some of the error cases on actual hardware by
using
crossystem recovery_reason=86
reboot
and pressing TAB at the recovery screen. For that example you should see the
message
recovery_reason: 0x56 TPM lock error in rewritable firmare
Change-Id: I123c781e6c6f6fe0284c4fd49f5f5a855eece7df
Reviewed-on: https://gerrit.chromium.org/gerrit/38652
Commit-Ready: Bill Richardson <wfrichar@chromium.org>
Tested-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Changing languages is terribly slow at the confirm screen, when
switching from dev to normal. Reduce sleep time to improve user
experience.
BUG=chrome-os-partner:15726
TEST=boot in dev, hit space, hit arrows rapidly to change language,
observe no lag.
BRANCH=butterfly, stout
Change-Id: I0943debc31d78dcfce87e7f7d4537ae47f5f8cfd
Reviewed-on: https://gerrit.chromium.org/gerrit/36956
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Commit-Ready: Shawn Nematbakhsh <shawnn@google.com>
Tested-by: Shawn Nematbakhsh <shawnn@google.com>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
The TPM on snow devices may boot in an unusable state. The workaround
is to detect this early and reboot. The workaround code prevents
an infinite reboot loop by counting the number of reboots and entering
recovery mode with this reason after a small threshold has been reached.
BUG=chromium:156655
TEST=no test! Not even compiled!
BRANCH=none
Change-Id: Ica2f14f8f7df8c46b7cbe5dbd578ba93c8f3a78c
Reviewed-on: https://gerrit.chromium.org/gerrit/36790
Tested-by: Luigi Semenzato <semenzato@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Ready: Luigi Semenzato <semenzato@chromium.org>
While not having a TPM was supported for non-Chrome devices, it was not
expected for Chrome devices. This adds logic to fail the TPM calls
before making them when the TPM is missing. The tpm_lite library doesn't
handle the TPM being missing, so we have to do this ourselves.
BUG=chrome-os-partner:15192
TEST=parrot build, verified operation after "mv /dev/tpm0 /dev/tpm0.bak"
BRANCH=none
Change-Id: I2f625305dce7fa698fcad33e412ee37c60da9bc2
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/35440
Reviewed-by: Luigi Semenzato <semenzato@chromium.org>
Reviewed-by: Will Drewry <wad@chromium.org>
Ctrl-U used to simply beep without messages for why it does not work (due to NV
data dev_boot_usb). Since the system is already in Developer mode, it should be
fine to provide some debug information otherwise we can spent time trying to
figure out why the firmware doesn't work.
BRANCH=all
BUG=chrome-os-partner:14474
TEST=flash image to Link, enter DEV and press Ctrl-U; gets beep and warning messages.
Change-Id: Iab20ecdb2e1c4e267b7257a7bd241006241ddf70
Reviewed-on: https://gerrit.chromium.org/gerrit/34406
Tested-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Ready: Hung-Te Lin <hungte@chromium.org>
We use gbb-flag-force-dev-switch-on in default firmware images to make
things easier for factory and some devs.
But when we request normal mode there should be some sort of warning/error
telling the user that this is not available, otherwise we can spent time trying
to figure out why the firmware doesn't work.
BRANCH=all
BUG=chrome-os-partner:14474
TEST=flash image to Link, set GBB flags to 0x39, boot to DEV screen
and press SPACE (TONORM); gets beep and warning messages.
Change-Id: Id48c12693c7575001fae7fad92a868cb5465e83d
Reviewed-on: https://gerrit.chromium.org/gerrit/34172
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Ready: Hung-Te Lin <hungte@chromium.org>
Tested-by: Hung-Te Lin <hungte@chromium.org>
When the cgpt utility complaints about parameter errors, it is
impossible to tell what exactly went wrong. This change consolidates
error definitions and adds a function to convert integer error values
into text messages.
BRANCH=none
BUG=none
TEST=manual
. emerge-link vbooot_reference
. copy generated `cgpt' to a Link device
. run command with wrong arguments with respect to the existing GPT:
localhost var # ./cgpt add -i 3 -b 3985408 -s 1757184 -t rootfs -l ROOT-A /dev/sda
ERROR: cgpt add: Starting LBA overlaps
ERROR: cgpt add: -i 3 -l ROOT-A -b 3985408 -s 1757184 -t 3CB8E202-3B7E-47DD-8A3C-7FF2A13CFCEC
. on the host, in the chroot in src/platform/vboot_reference run
$ make && make runtests
observe all tests succeed
Change-Id: Ibd23ca0430a875f70524adc99e0509b26ae699b2
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/34003
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This option is disabled per default and can be enabled with
crossystem dev_boot_legacy=1
or by setting the GBB flag
GBB_FLAG_FORCE_DEV_BOOT_LEGACY 0x00000080
BUG=chrome-os-partner:6108
TEST=crossystem dev_boot_legacy=1
boot to dev mode screen, press CTRL-L, see SeaBIOS start
(other CLs needed)
BRANCH=link
Signed-off-by: Stefan Reinauer <reinauer@chromium.org>
Change-Id: I593d2be7cff5ca07b8d08012c4514a172bd75a38
Reviewed-on: https://gerrit.chromium.org/gerrit/31265
Reviewed-by: Stefan Reinauer <reinauer@chromium.org>
Tested-by: Stefan Reinauer <reinauer@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Ready: Stefan Reinauer <reinauer@chromium.org>
In order to dual boot Windows and ChromeOS, Windows must
not find a GPT partition table on the disk. So change
ChromeOS to cope with an alternative signature "CHROMEOS"
instead of the standard "EFI PART"
BUG=chrome-os-partner:6108
TEST=rebuild chromeos, install it,
run cgpt legacy /dev/sda
dd if=/dev/sda of=/tmp/x bs=1k
hexdump -C /tmp/X
see the string CHROMEOS
BRANCH=link
Signed-off-by: Stefan Reinauer <reinauer@chromium.org>
Change-Id: Ia88eff33b9880bd73a78c1b8e026c1f8298c4557
Reviewed-on: https://gerrit.chromium.org/gerrit/31264
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Ready: Stefan Reinauer <reinauer@chromium.org>
Tested-by: Stefan Reinauer <reinauer@chromium.org>
This just chews up memory and wastes time on ARM, since the data is already
in memory.
BUG=chrome-os-partner:13492
BRANCH=snow
TEST=manual
Build and boot on snow with manually modified code, to see that the bmpfv
pointer is in the same region as the bmp region.
Build and boot on link and see that displaying screens is still fast.
Change-Id: I98349b73671e38fa6cace966b6953a2abf129fab
Reviewed-on: https://gerrit.chromium.org/gerrit/32629
Reviewed-by: Mike Truty <truty@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Che-Liang Chiou <clchiou@chromium.org>
Commit-Ready: Simon Glass <sjg@chromium.org>
This enum seems partially complete, and not used in vboot_reference.
Complete it and use it.
BUG=chrome-os-partner:13492
BRANCH=snow
TEST=manual
Build and boot through to recovery on snow. Run through the various
screens and check that they still appear correctly.
Change-Id: Ifca54d072457d9a0396a38026f44f8334efb9cf5
Reviewed-on: https://gerrit.chromium.org/gerrit/32628
Reviewed-by: Mike Truty <truty@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Commit-Ready: Simon Glass <sjg@chromium.org>
%L is, in some standard libraries like U-Boot's, a synonym for %ll which is
for long long integers, required by the C99 standard to be at least 64 bits.
For practical purposes that basically means %ll should be used with 64 bit
values. Since %L seems to be non-standard and, at least in U-Boot's case, %ll
is recognized in the same way, %ll seems preferable.
BUG=chrome-os-partner:8339
TEST=Booted ChromeOS using depthcharge and U-Boot. Booted with
depthcharge/libpayload which does not support %L and saw a number where %L had
been printed.
BRANCH=None
Change-Id: Id51fb5c9295e0dd65b42a5c0738eb34c8210a2b2
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/32660
Reviewed-by: Randall Spangler <rspangler@chromium.org>
On x86 U-Boot cannot see the power button, which means that the EC must
deal with it, and may power off the unit at any time. To get around this,
we write the vbcontext every time we change it.
Since this isn't a problem on ARM, and we want to avoid spurious writes
(due to delay and disk wear), make this code execute only on x86 machines.
BUG=chrome-os-partner:13717
BUG=chrome-os-partner:7689
BRANCH=snow,link
TEST=manual
On snow, see that the EC no longer gets MKBP messages to write the nv
context.
On link, manually add a print to U-Boot's nvstorage_write_disk() function
and see that changing language in recovery still causes a write.
Change-Id: I62508739c9fc3aca46fba58b196a8af45269af2a
Reviewed-on: https://gerrit.chromium.org/gerrit/32464
Commit-Ready: Tom Wai-Hong Tam <waihong@chromium.org>
Reviewed-by: Tom Wai-Hong Tam <waihong@chromium.org>
Tested-by: Tom Wai-Hong Tam <waihong@chromium.org>
Currently we check the keyboard each 250ms. This makes for a pretty choppy
experience when changing languages. Change to check every 20ms, without
changing the disk check interval (which remains 1s).
BUG=chrome-os-partner:13717
BRANCH=snow
TEST=manual
Boot into recovery
Try changing language on snow with left/right arrow and see that it updates
instantly.
Change-Id: I2ae411bc36fdb2badac11595b099bca43f116669
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/32463
Commit-Ready: Tom Wai-Hong Tam <waihong@chromium.org>
Reviewed-by: Tom Wai-Hong Tam <waihong@chromium.org>
Tested-by: Tom Wai-Hong Tam <waihong@chromium.org>
Rather than read the images from slow flash every time we need them, cache
them the first time and use that cache thereafter.
BUG=none
BRANCH=snow,link
TEST=manual
Go into recovery mode on link
See that we can display a new screen in roughly 20ms instead of the 250ms
it previously took on link.
Also tested on snow and shown to have no ill effects.
Change-Id: Ieb39c44bddeb6315da8983669f19f550888659bd
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/32462
Commit-Ready: Tom Wai-Hong Tam <waihong@chromium.org>
Reviewed-by: Tom Wai-Hong Tam <waihong@chromium.org>
Tested-by: Tom Wai-Hong Tam <waihong@chromium.org>
We have to define the function we need here, so that we can implement it in
U-Boot, then we can come back here and try to use it. Grr.
BUG=chrome-os-partner:11215
BRANCH=link
TEST=none
This just defines the function prototype. No change to test.
Change-Id: I38a19baa54c59c9744d20f743eb53260f2d19852
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31658
Reviewed-by: Randall Spangler <rspangler@chromium.org>
We need to know not only whether the HW WP pin is asserted, but whether the
flash chip has configured its software protection registers to actually
protect anything. This flag can be used to indicate that.
BUG=chrome-os-partner:13265
BRANCH=link
TEST=none
This just adds the flag. Nothing actually sets the flag yet, so there's
nothing to test.
Change-Id: Icba9945fb56eb3a4681486c630cbbdc9232485ef
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31642
Reviewed-by: Randall Spangler <rspangler@chromium.org>
At DEV screen:
- Space triggers TONORM
- Enter is ignored unless new GBB flag is set
At TONORM screen:
- Enter always means YES
- Space is ignored
So, if you hold the space key at the dev screen, you'll go to tonorm
and stay there until you press Enter or Esc. If you hold the Enter
key at the dev screen, nothing will happen.
Add a GBB flag to allow Enter to trigger the TONORM screen; this will
be used by FAFT testing.
BRANCH=all
BUG=chrome-os-partner:12699
TEST=manual
1. press enter at dev screen. nothing happens.
2. press space at dev screen. tonorm.
3. press space at tonorm. nothing happens.
4. press enter at tonorm. turns off dev mode.
Change-Id: I9f3128d5114e1486911cc4d76d0ccd5649de1680
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/30456
Reviewed-by: Tom Wai-Hong Tam <waihong@chromium.org>
This adds two new flags to crossystem:
clear_tpm_owner_request
clear_tpm_owner_done
The first one requests that the firmware clear the TPM owner on the
next boot. When the firmware does this, it will set
clear_tpm_owner_request=0, and set clear_tpm_owner_done=1. The OS can
use the done-flag as a hint that trusted things guarded by the TPM are
no longer trustable.
BUG=chromium-os:31974
TEST=manual
crossystem
// both flags initially 0
crossystem clear_tpm_owner_request=1
crossystem clear_tpm_owner_done=1
// request=1, done=0; done can be cleared but not set by crossystem
reboot
tpmc getownership
// owned=no
crossystem
// request=0, done=1
crossystem clear_tpm_owner_done=0
crossystem
// both flags 0 again
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Change-Id: I49f83f3c39c3efc3945116c51a241d255c2e42cd
Reviewed-on: https://gerrit.chromium.org/gerrit/25646
It's not yet possible to verify the kernel in an Chromium OS image
with the Sandbox Version of U-Boot due to the lack of keys. For now,
stub out the verification process and behave as if everything is ok:
Sandbox U-Boot is only interested in the selected kernel and boot mode
at this point.
BUG=chromium-os:32603
TEST=With this change, it's possible to get valid answers from
vboot_twostop command with Sanbox U-Boot.
Change-Id: I3b1142889657315675eacd3a1d1448aeee7ccb62
Signed-off-by: Taylor Hutt <thutt@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/30256
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Presently kernel load address and buffer size are programmed in the
u-boot device tree. There is no reason for this: the address and size
are part of the vboot encapsulation headers. Duplicating this
information hardcoded in the device tree does not bring any benefit
and is in fact harmful, as it is easy to get out of sync.
A better way of doing things is to derive kernel load address and size
from the appropriate vboot header. ARM people object to this, as they
want the very same kernel blob operate on devices with DRAM mapped to
different address ranges.
The suggested solution is to exclude the kernel memory section from
the device tree on the platforms where the load address could be
safely taken from the vboot header. In this case u-boot will pass
address of zero to vboot, which will know to derive the address/size
from the appropriate header. vboot then rewrites fields of the u-boot
supplied structure with actual address and size of the kernel blob.
There is no sanity check yet, as it is presumed that there is enough
memory to load any kernel and u-boot does not use the space above
0x100000 for at least 16 megabytes (the kernel partition size). On x86
platform the check could be verify that the top of the kernel space is
well below the stack.
BUG=chrome-os-partner:11994
TEST=manual
. with the appropriate u-boot change run a Link target through a
FAFT cycle, observe it succeed.
Change-Id: I3c2c2cefb1e31d16ac497a01894bf32638479ed7
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/29038
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
Commit-Ready: Bill Richardson <wfrichar@chromium.org>
Add a definition of kBootStateSHA1Digests[]. Without this, it is not
possible to build the Sanbox version of U-Boot.
BUG=chromium-os:32603
TEST=Allows vboot to link when using mocked TPM with U-Boot Sandbox
Change-Id: Ie84f4ba3f1c266ed8063fbf6aea0093dd21f638b
Signed-off-by: Taylor Hutt <thutt@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/30200
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
We need to alert user his change has been accepted for 5 seconds.
BUG=chrome-os-partner:11078
TEST=emerge-daisy vboot_reference-firmware vboot_reference chromeos-u-boot chromeos-boogimage
Change-Id: I6a3b13e7543479fbd6543d759093ecf501fb1d04
Reviewed-on: https://gerrit.chromium.org/gerrit/29982
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Ready: Hung-Te Lin <hungte@chromium.org>
Tested-by: Hung-Te Lin <hungte@chromium.org>
This is to avoid confusion with the canonical common.mk file that is
a CrOS build system.
BUG=chromium-os:33327
TEST=`cros_run_unit_tests --board x86-alex -p vboot_reference` still works
Change-Id: I4b6719d58a4a8ab44b62c23c0e2c45b154374958
Reviewed-on: https://gerrit.chromium.org/gerrit/29578
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Ready: Mike Frysinger <vapier@chromium.org>
Tested-by: Mike Frysinger <vapier@chromium.org>
Some EC's are slow to program their firmware. This could happen at any boot
as part of software sync. This adds the VB_INIT_FLAG_EC_SLOW_UPDATE input
flag to VbInit(), so that vboot knows to display the EC-is-being-programmed
screen when that flag is set.
BUG=chrome-os-partner:12255
TEST=none
This isn't activated yet by anything, so there's nothing to test.
Change-Id: Id4d12e463fcdee017b7c52b53a2facefc33b745f
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/29370
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This is more reliable than reading them through FDT/ACPI, since it reflects
the positions as shown to verified boot code.
Notes:
1. This affects ALL platforms with virtual dev switches (x86 AND arm)
2. The fix should have no effect on older platforms, but I haven't tested those.
BUG=chrome-os-partner:11805
TEST=manual
1. boot in normal mode.
devsw_boot = 0 # Developer switch position at boot
recovery_reason = 0 # Recovery mode reason for current boot
recoverysw_boot = 0 # Recovery switch position at boot
wpsw_boot = 1 # Firmware write protect hardware switch position at boot
2. boot in developer mode.
localhost ~ # crossystem
devsw_boot = 1 # Developer switch position at boot
recovery_reason = 0 # Recovery mode reason for current boot
recoverysw_boot = 0 # Recovery switch position at boot
wpsw_boot = 1 # Firmware write protect hardware switch position at boot
3. boot in developer-recovery mode using keyboard combo.
devsw_boot = 1 # Developer switch position at boot
recovery_reason = 2 # Recovery mode reason for current boot
recoverysw_boot = 1 # Recovery switch position at boot
wpsw_boot = 1 # Firmware write protect hardware switch position at boot
4. disable WP and reboot. wpsw_boot should be 0.
Change-Id: If4156b5e14c6923c5b331c7e5feaabbffe1dad37
Reviewed-on: https://gerrit.chromium.org/gerrit/29199
Commit-Ready: Randall Spangler <rspangler@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Tested-by: Randall Spangler <rspangler@chromium.org>
The SHA256 constants and some algo definitions are currently in the
.data segment, this wastes precious RAM on the EC where read-only data
are stored in flash only.
BUG=chrome-os-partner:12271
TEST=sudo emerge vboot_reference &&
emerge-daisy vboot_reference vboot_reference-firmware chromeos-ec
and manually check the content of the .data segment in the EC binary.
Change-Id: I67b8c0f2843312651e8ac287cea2f8cb5f075453
Reviewed-on: https://gerrit.chromium.org/gerrit/29273
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Ready: Vincent Palatin <vpalatin@chromium.org>
Some systems take a long time to program the EC firmware (because it's
behind a really slow I2C bus or something). This could happen at any boot as
part of software sync. If it's necessary to reprogram the EC on a slow
system, we need to display a BIOS screen to the user so he/she doesn't think
it's bricked.
This CL is just to add a new default bitmap so we'll have something to
display when we need it. Nothing actually uses it yet.
BUG=chrome-os-partner:12254
TEST=none
There should be no user-visible changes. We don't invoke this screen yet.
Change-Id: Icb922f933e2e921472dbdff7a26a3ca4d58fbba3
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/29241
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
We're still working out the correct behavior for the keyboard-based
dev-mode. Before this, we rebooted into recovery mode before asking if you
wanted to return to normal mode, so if you said "no, stay in dev-mode" you'd
still be at the recovery screen.
But now the confirmation for returning to normal mode happens at the
dev-mode screen, so you never get to the recovery screen (unless normal mode
won't boot, in which case you get there automatically).
BUG=chrome-os-partner:11707
TEST=manual
First, clear the GBB flags so that you can actually test the virtual dev
switch:
/usr/share/vboot/bin/set_gbb_flags.sh 0
reboot
It should come up in normal mode.
Activate keyboard-based dev-mode as before:
- three-finger salute
- Ctrl-D at the recovery screen
- Press ENTER when asked
Now, at the DEV screen, try to leave:
- Press SPACE. It should ask if you want to go to normal mode.
- Press ESC (no). You should be back at the DEV screen again.
- Press ENTER. It should ask if you want to go to normal mode.
- Press ESC (no). You should be back at the DEV screen again.
- Press SPACE. It should ask if you want to go to normal mode.
- Press ENTER (yes). It should reboot into normal mode.
Change-Id: I99af6e7b97fb61f943bd14c8c7166571b5ccf106
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/28872
BUG=chrome-os-partner:11887
TEST=manual
Boot with dev mode on. Boots normally.
Boot with dev mode on, hammering on ENTER key as soon as you release power.
Dev mode screen is shown.
Keep hammering. Dev screen will stay up as long as you do this.
Wait >2 sec.
Press ENTER. System reboots to TONORM screen.
Change-Id: I18e35b23c18a65637a84d3d1964b291e0cb5e8c5
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/28995
Reviewed-by: Tom Wai-Hong Tam <waihong@chromium.org>
And space at TONORM screen no longer confirms disabling dev mode
Added Ctrl+Refresh as an alias for tonorm screen request. U-boot will
need to be enhanced to support that. Until then, many FAFT tests will
break.
BUG=chrome-os-partner:11887
TEST=manual
1. Boot to DEV screen
2. Press Enter. Nothing happens.
3. Press Space. Goes to TONORM screen.
4. Press Space. Nothing happens.
5. Press Enter. Reboots with dev mode disabled.
Change-Id: I7f61c4001c668ac916f50f931a79a107752c83b5
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/28851
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
This could wrap before the assignment:
uint64_t = uint32_t * int;
Instead:
uint64_t = uint32_t;
uint64_t *= int;
BUG=chrome-os-partner:11643
TEST=none
Nothing to test or verify. If the security guys approve, it's fixed.
Change-Id: Ib7c9774998332ac1a29c4551bc039eaa999ee681
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/28841
Reviewed-by: Gaurav Shah <gauravsh@chromium.org>
So if the EC reboots when the lid is closed (for example, it wakes
from hibernate because the AC adapter was inserted), it won't leave
the AP powered on.
BUG=chrome-os-partner:11087
TEST=manual
Reboot EC with lid closed ('reboot hard').
Should finish software sync and shut down.
Reboot EC with lid open. Boots.
Hibernate EC and wake with lid open. Boots.
Hibernate EC and wake with power button press. Boots.
Reboot into recovery mode. Boots. Press power button; shuts down.
Reboot into recovery mode. Boots. Close lid; shuts down.
CQ-DEPEND=*22578
Change-Id: I09d5331222aa10b73518f9f574ec5a32d8e6ac23
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/28525
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Software sync will, of course, work perfectly, so these will never be needed.
BUG=none
TEST=none (hey, it all works perfectly)
Change-Id: I0014dfb99507c5eb00de73b77edb7538b598658f
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/28495
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
Necessary for updating snow EC
BUG=chrome-os-partner:11087
TEST=force an EC update by loading a slightly old EC and then a new BIOS
Change-Id: Id00257f8a67c08077a5b396cf120a056a7601671
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/28436
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
BUG=chrome-os-partner:11087
TEST=manual
1. Update EC to a new version
2. Rebuild EC code and chromeos-bootimage
3. Update BIOS (AP) RW firmware
4. Reboot. EC should get updated to the version in the BIOS iamge.
CQ-DEPEND=28414
Change-Id: I227d9bf7cc1f4984d6eb13b5372bcf4181ad0df5
Signed-off-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/28415
Reviewed-by: Bill Richardson <wfrichar@chromium.org>
vboot is passing in an index but the function expects an enum.
BUG=chrome-os-partner:11765
TEST=manual
Build with serial enabled, watch what happens. Before, it does this:
ec.c: VbExEcGetExpectedRW: Unrecognized EC firmware requested.
VbEcSoftwareSync() - VbExEcGetExpectedRW() returned 65536
VbSetRecoveryRequest(34)
After this change, it should get a little further:
ec.c: VbExEcGetExpectedRW: EC-RW image offset 0 size 0.
ec.c: VbExEcGetExpectedRW: EC image has bogus size.
VbEcSoftwareSync() - VbExEcGetExpectedRW() returned 65536
VbSetRecoveryRequest(34)
Sigh.
Change-Id: Ideb5f9d70bc5f3f202a8e04a6ec718e1d674d526
Signed-off-by: Bill Richardson <wfrichar@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/28341
Reviewed-by: Randall Spangler <rspangler@chromium.org>
