Add CFL defaults for VR TDC config and provide Iccmax for additional Xeon
CPUs tested on the Prodrive/Hermes board.
Based on the following Intel documents:
* Document Number 570805 (XEON E EDS Vol 1)
* Document Number 337344 (CFL Datasheet Vol 1)
* Document Number 571264 (CFL CNP PDG)
Change-Id: I681de076318fb647c44cc8b8c42eb297018cc540
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/40856
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Christian Walter <christian.walter@9elements.com>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
This makes future changes easier to review.
Change-Id: I5d67801a46a1613fbc7f813e94933fa30c1b92df
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41507
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Christian Walter <christian.walter@9elements.com>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
Read the PCH Strap Length field in FLMAP1 as described in the
"SPI Programming Guide" and print the number of fields specified there.
This code dumps the following straps:
* Intel GM45: 8 straps
* Intel C216: 72 straps
* Intel C240: 360 straps
Add a new function to easily set PCH straps, which is useful for debugging.
Change-Id: Ieb7891b214d82c984379794de9b3fe1a6d0d3466
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41693
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Philipp Hug <philipp@hug.cx>
Reviewed-by: Christian Walter <christian.walter@9elements.com>
Reviewed-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
Added device hid info to the MST and LSPCON devices on
kaisa, duffy and noibat.
BRANCH=None
BUG=b:156546414
TEST=None
Signed-off-by: Shiyu Sun <sshiyu@chromium.org>
Change-Id: I7b54512cd88e7280374c188315cabc2fba197f69
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42369
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Sam McNally <sammc@google.com>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
This reverts commit aac79e0b8f.
Reason for revert: This massively slows down the boot process because
the LAPIC delivery mode for the APs is not set anymore. Plus, not all
review comments were fully addressed, yet this got merged in anyway.
Change-Id: If9bae6aae0d4d1f21b067a7d970975193c2b16d5
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42166
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Add two fields for the ACPI logical device to the ACPI code. They are
used for taking asus/p3b-f board out of suspend by keyboard or mouse.
Change-Id: Icaadfea6a4dce7a2d665e8d89a024359975f8b2c
Signed-off-by: Keith Hui <buurin@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41096
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
When Linux is booted, the kernel reports
"do_IRQ: 1.55 No irq handler for vector"
So far it comes with payloads SeaBIOS and depthcharge, not with
Grub. We assume Grub does something to avoid this problem.
AMD bug tracker system (JIRA PLAT-21393) says the APs can not be set
EXTINT delivery mode.
In Intel 64 and IA-32 Architectures Software Developer’s Manual volume
3A, see chapter 10.5.1 Local Vector Table, it says:
"The APIC architecture supports only one ExtINT source in a system,
usually contained in the compatibility bridge. Only one processor in the
system should have an LVT entry configured to use the ExtINT delivery
mode."
Tested on mandolin (Picasso) board, the error in dmesg is gone.
The bug 153677727 has two parts.
1. Soft lockup
2. do_IRQ 1.55.
The soft lockup issued has been fixed by
https://review.coreboot.org/c/coreboot/+/41128
BUG=b:153677727
TEST=mandolin
Change-Id: I2956dcaad87cc1466deeca703748de33390b7603
Signed-off-by: Zheng Bao <zheng.bao@amd.com>
Signed-off-by: Zheng Bao <fishbaozi@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42219
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
mainboard_intr_data table mixed hexadecimal notation with
both small and capital letters. Now, it is unified to capitals only.
Signed-off-by: Piotr Kleinschmidt <piotr.kleinschmidt@3mdeb.com>
Change-Id: Icd8cf4324e72e87e7e98869872785523fb4e1809
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42388
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Michał Żygowski <michal.zygowski@3mdeb.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The only board that builds this does not have ACPI S3 support.
Also the code is wrong.
Change-Id: Ifb8e0ae5b6d862fa6a52b8e08197a84e7da4be36
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42357
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
The call made at mp_ops.post_mp_init() generally
uses four different names. Unify these with followups.
smm_southbridge_enable(SMI_EVENTS)
smm_southbridge_enable_smi()
hudson_enable_smi_generation()
enable_smi_generation()
Furthermore, some platforms do not enable power button
SMI early. It may be preferred to delay the enablement,
but fow now provide global_smi_enable_no_pwrbtn() too.
Change-Id: I6a28883ff9c563289b0e8199cd2ceb9acd6bacda
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42355
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
There are no side-effects in calling acpi_is_wakeup_s3()
and apm_control() is a no-op with HAVE_SMI_HANDLER=n.
Change-Id: Ia9195781955cc5fa96d0690aa7735fc590e527e4
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41986
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Attempts to write to APM_CNT IO port should always be guarded
with a test to verify SMI handler has been installed.
Immediate followup removes redundant HAVE_SMI_HANDLER tests.
Change-Id: If3fb0f1a8b32076f1d9f3fea9f817dd4b093ad98
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41971
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Patrick Rudolph <siro@das-labor.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
At least one mobile 945 series northbridge supports 4 threads, because
the dual-core Atom 330 CPU supports Hyper-threading. Therefore, we use
that as the default for this chipset.
Change-Id: I899ed1644d9b2da4fc72f09233a421200770110d
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41845
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The gm45 northbridge supports at most 4 threads. However, the only two
mobile Core 2 Quad models are not BGA956, so account for that as well.
Change-Id: Ie198ac4c366ec0bd53ddb337b6f9c03c331c73f5
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41844
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Pineview has at most 4 threads.
Change-Id: I0f45f002d0bab0345bc061ac3c7a29237a536cc5
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41843
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
ULT only has 4 threads, but we are not changing it here to preserve
binary reproducibility.
Change-Id: I041c5dff2de514244f9c919c4c475cca979c34ce
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41842
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
LGA775 CPUs can have at most 4 threads, and Eaglelake supports them.
As this socket is also used by other chipsets, temporarily place this
symbol into the northbridge scope until all chipsets are factored out.
Change-Id: I6e01363d995e135815cc70779e0cd5baf806cf60
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41841
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Arrandale CPUs have at most 4 threads.
Change-Id: Ifecbf5583011ff5e36c576d582a6276bc9b72803
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41840
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
cbmem is not online when vboot runs before the bootblock. Update the
macro to reflect that.
BUG=b:158124527
TEST=Build & boot psp_verstage on trembyle
Signed-off-by: Martin Roth <martin@coreboot.org>
Change-Id: I6fb4ad04f276f2358ab9d4d210fdc7a34a93a5bb
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42066
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
The assumption up to this point was that if the system had an x86
processor, verstage would be running on the x86 processor. With running
verstage on the PSP, that assumption no longer holds true, so exclude
pieces of code that cause problems for verstage on the PSP.
As a generalization, remove all-y for CONFIG_ARCH_xx guarded
makefiles.
BUG=b:158124527
TEST=Build and boot on Trembyle
Signed-off-by: Martin Roth <martin@coreboot.org>
Change-Id: Ia7dcfed699ee1c0cd5a5250431c5f05bf6d8b9c7
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42221
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Add the AMD supplied code (modified to work with GCC) to the vendorcode
directory. Verstage will be running on the PSP as a userspace
application under the bootloader, which is what bl_uapp signifies.
AMD is still working on documentation for the entire PSP userspace
application interface.
BUG=b:158124527
TEST=Build & boot psp_verstage on Trembyle
Signed-off-by: Martin Roth <martin@coreboot.org>
Change-Id: Ie740c89afe2277eff279fc5c94f88ffd43a78a37
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42060
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Because the PSP maps the MMIO addresses that are used to non-
deterministic addresses, the accesses need to be able to find
the address at runtime.
BUG=b:158124527
TEST=Build & boot with Trembyle
Signed-off-by: Martin Roth <martin@coreboot.org>
Change-Id: I68305e0f31956c57bfdee42025bdfe938703e82d
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42061
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Exclude pieces of console code from the vboot if running before
bootlock. The PSP verstage code will re-implement some of
these in its own code.
BUG=b:123887623
TEST=Build with following patches
Signed-off-by: Martin Roth <martin@coreboot.org>
Change-Id: Ifc9fb0810e0816fe0a68e52287eda6145043a619
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41815
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
For AMD's family 17h, verstage can run as a userspace app in the PSP
before the X86 is released. The flags for this have been made generic
to support any other future systems that might run verstage before
the main processor starts.
Although an attempt has been made to make things somewhat generic,
since this is the first and currently only chip to support verstage
before bootblock, there are a number of options which might ultimately
be needed which have currently been left out for simplicity. Examples
of this are:
- PCI is not currently supported - this is currently just a given
instead of making a separate Kconfig option for it.
- The PSP uses an ARM v7 processor, so that's the only processor that
is getting updated for the verstage-before-bootblock option.
BUG=b:158124527
TEST=Build with following patches
Signed-off-by: Martin Roth <martin@coreboot.org>
Change-Id: I4849777cb7ba9f90fe8428b82c21884d1e662b96
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41814
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
A looong time ago when cache_as_ram.S was built into romstage,
the stage was also linked twice. First at a fixed low address
and then again relocated at the final execute-in-place address.
Change-Id: Ic624feef6794f2c24e38459a45583d84fc07a484
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42347
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
When adding XIP stages on x86, the -P parameter was used to
pass a page size that covers the entire file to add. The same
can now be achieved with --pow2page and we no longer need to
define a static Konfig for the purpose.
TEST: Build asus/p2b and lenovo/x60 with "--pow2page -v -v" and
inspect the generated make.log files. The effective pagesize is
reduced from 64kB to 16kB for asus/p2b giving more freedom
for the stage placement inside CBFS. Pagesize remained at 64kB
for lenovo/x60.
Change-Id: I5891fa2c2bb2d44077f745619162b143d083a6d1
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41820
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Keith Hui <buurin@gmail.com>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
For add-stage command, --pow2page is equivalent of passing
-P log2ceil(sizeof stage). The sizeof stage can be hard to
determine in Makefile to be passed on the commandline.
Change-Id: If4b5329c1df5afe49d27ab10220095d747024ad6
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41832
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
This pulls in a newer version of the PSP-related blobs.
Change-Id: I6ff39260e9697512f78eb68435bd17ea83af35d5
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42346
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Picasso, Dali, and Pollock iGPU share the same PCI device ID, but need
different video BIOSes. This checks the vendor & device IDs along with
the revision and selects the correct video BIOS to use.
Also add the second VGA BIOS for Raven2-based SoCs and change all VGA
BIOS IDs to the format including the revision number.
Since SeaBIOS still expects the CBFS file name without the revision ID,
it won't find the VBIOS any more. As a temporary workaround add the
VBIOS for the silicon it will run on as VGA_BIOS_DGPU_*.
Change-Id: I8f48ecc3fbffddd21d1f830fbee26a09ac351e1c
Signed-off-by: Martin Roth <martinroth@chromium.org>
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-on: https://chromium-review.googlesource.com/2040455
Reviewed-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-by: Matt Papageorge <matt.papageorge@amd.corp-partner.google.com>
Reviewed-by: Justin Frodsham <justin.frodsham@amd.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41562
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
• Based on FSP EAS v2.1 – Backward compatibility is retained.
• Add multi-phase silicon initialization to increase the modularity of the
FspSiliconInit() API.
• Add FspMultiPhaseSiInit() API
• FSP_INFO_HEADER changes
o Added FspMultiPhaseSiInitEntryOffset
• Add FSPS_ARCH_UPD
o Added EnableMultiPhaseSiliconInit, bootloaders designed for
FSP 2.0/2.1 can disable the FspMultiPhaseSiInit() API and
continue to use FspSiliconInit() without change.
FSP 2.2 Specification:
https://www.intel.com/content/www/us/en/intelligent-systems/intel-firmware-support-package/intel-fsp-overview.html
Change-Id: If7177a267f3a9b4cbb60a639f1c737b9a3341913
Signed-off-by: Subrata Banik <subrata.banik@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41728
Reviewed-by: Srinidhi N Kaushik <srinidhi.n.kaushik@intel.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
If there is no installed audio daughter board on volteer then the
HDA driver in the kernel will crash on resume. In order to prevent
this disable the PCI device when AUDIO=NONE probe match is true.
BUG=b:147462631
TEST=boot on volteer and ensure that the PCI device at 0:1f.3 is gone
Change-Id: I4a436e1b76418030bf635427e490b54a713fdd33
Signed-off-by: Duncan Laurie <dlaurie@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41217
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Sathyanarayana Nujella <sathyanarayana.nujella@intel.com>
Reviewed-by: Srinidhi N Kaushik <srinidhi.n.kaushik@intel.com>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Zork family does not use OEM binary and so this change drops the
configs required for adding this binary.
Change-Id: Id38c67030e4055ab16934d1a900ee1cea5843b54
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42336
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
This way drivers can wait for their devices to be enabled.
I also rewrote enable_aoac_devices to take advantage of
wait_for_aoac_enabled.
BUG=b:153001807
TEST=Trembyle builds
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Change-Id: I8e653c857e164f90439e0028e08aa9608d4eca94
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42326
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
If the OS sets the target device state to D3, we need to clear it so we
can reestablish register access.
BUG=b:153001807
TEST=Boot trembyle with I2C powered off and see it power back on.
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Change-Id: If9bd1b7cfa7b8d074226c4dcdefc1a44cad8b940
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42325
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
This functionality is needed in the PSP and I can't include all of
southbridge.c.
BUG=b:153001807
TEST=Made sure trembyle still compiles
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Change-Id: I3a38c655588d7836e1bd033e958a505774de871e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42324
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The legacy UARTs are supposed to default to off according to the
documentation (PPR for AMD Family 17h Model 18h). But legacy UART Range_0
is enabled after reset. The PSP might be enabling it or the documentation
might be wrong.
Having it enabled causes problems though. We have ACPI nodes defining
MMIO UARTs, and the kernel also probes for legacy UARTs. This results in
two drivers accessing the same device, one via MMIO and one via IO. I
suspect this was the cause of the garbage serial output.
Before the change you would see the following in the console:
[ 0.741108] serial8250: ttyS3 at I/O 0x2e8 (irq = 3, base_baud = 115200) is a 16550A
After this change, we no longer see it.
BUG=b:152079780, b:157858890
TEST=Boot trembyle and make sure serial is still working.
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Change-Id: I9d837e449b961dbb55d1301d2107838e26b3f892
Reviewed-on: https://review.coreboot.org/c/coreboot/+/42323
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Patrick Rudolph