GitLab

r8152: Support partial reads for jumbo frame ethernet packets

See merge request !3

Julius Werner authored Jun 23, 2020 and Ricardo Cañuelo Navarro committed Dec 11, 2020

The RealTek R8153 dongle will only send individual USB packets up to a
certain maximum packet size (not quite clear what this is, possibly
2KB). Usually this seems to be enough for our TFTP needs, but depending
on TFTP server configuration and network path MTU they may occasionally
get bigger. This patch makes sure the driver will not choke if it
receives a 9KB jumbo frame in small pieces.

BRANCH=None
BUG=b:159263757
TEST=ODM said this fixes their problem.
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: I9c56de5c15ab6b2017a9a3ee8fa907f28737c0af
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/2261078

Reviewed-by: Aaron Durbin <adurbin@google.com>

Guillaume Tucker authored Jan 30, 2018 and Guillaume Tucker committed Oct 30, 2020

Print the git branch name as well as the build date and time just
before starting the Depthcharge interactive prompt.  This helps
diagnosing issues with a quick way to check the firmware build being
used.
Signed-off-by: Guillaume Tucker <guillaume.tucker@collabora.com>

Guillaume Tucker authored Jan 30, 2018 and Guillaume Tucker committed Oct 30, 2020

Show the length of data being downloaded over TFTP and line returns to
make it easier to follow.  This helps with diagnosing issues.
Signed-off-by: Guillaume Tucker <guillaume.tucker@collabora.com>

Enable CONFIG_CLI and set the interactive prompt to "grunt:" for test
automation in LAVA.
Signed-off-by: Guillaume Tucker <guillaume.tucker@collabora.com>

commit 2b719239

 upstream.

Add an optional argument to provide the name of a ramdisk file to
download separately from the kernel image.  When provided, it on to
the netboot() function.

BRANCH=none
BUG=b:154536870
TEST=tftpboot with extra ramdisk argument
Signed-off-by: Guillaume Tucker <guillaume.tucker@collabora.com>
Change-Id: Ib3382a019cdef3c75fcd826cafb734eb82c85933
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/2288613


Commit-Queue: Guenter Roeck <groeck@chromium.org>
Tested-by: Guenter Roeck <groeck@chromium.org>
Reviewed-by: Guenter Roeck <groeck@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>

commit 72b9c3ba

 upstream.

While arm/arm64 have support for the FIT format which can include a
ramdisk separate from the main kernel image, other architectures such
as x86 only have support for plain kernel images (bzImage) in
Depthcharge.  Loading a ramdisk over TFTP therefore needs to be done
separately.  For this purpose, add a ramdisk argument to netboot()
with the name of an optional separate file to be downloaded as a
ramdisk and pass it on via the boot_info structure.

BRANCH=none
BUG=b:154536870
TEST=boot on x86
Signed-off-by: Guillaume Tucker <guillaume.tucker@collabora.com>
Change-Id: I251df7a56589c0240761dd5caab174f535bf9ce9
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/2288612

Reviewed-by: Julius Werner <jwerner@chromium.org>
Commit-Queue: Guenter Roeck <groeck@chromium.org>
Tested-by: Guenter Roeck <groeck@chromium.org>

commit 28593919

 upstream.

When a ramdisk image is provided in the boot_info structure, copy it
at an appropriate address according to the x86 Linux kernel boot
protocol and pass it on via the kernel header.

BRANCH=none
BUG=b:154536870
TEST=boot on x86
Signed-off-by: Guillaume Tucker <guillaume.tucker@collabora.com>
Change-Id: I2bc1dc2cc8f5788af18b5af8f556c2c9a2100c95
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/2288611

Reviewed-by: Duncan Laurie <dlaurie@google.com>
Reviewed-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Guenter Roeck <groeck@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Commit-Queue: Guenter Roeck <groeck@chromium.org>
Tested-by: Guenter Roeck <groeck@chromium.org>

make sure nuwani and treeya are not support FW update

BUG=b:150799568
BRANCH=none
TEST=build nuwani bios and ec and DUT can boot up successfully.
Signed-off-by: peichao.wang <peichao.wang@bitland.corp-partner.google.com>
Change-Id: Id8279a93c65f581c100213ff9e156197825ed329
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/2106829

Reviewed-by: Martin Roth <martinroth@google.com>
Commit-Queue: Martin Roth <martinroth@google.com>
Tested-by: Martin Roth <martinroth@google.com>
Auto-Submit: Martin Roth <martinroth@google.com>

nuwani variant use anx3447 and leverage to Treeya

BUG=b:150799568
BRANCH=none
TEST=build nuwani bios which includes the latest ec, it can boot
     successfully to Chrome system and observe that anx3429 error
     message do not appear in depthcharge output from debug
     console.

Change-Id: Iaf1cd196ed1701bf512ba9e095a9901e43f642c4
Signed-off-by: peichao.wang <peichao.wang@bitland.corp-partner.google.com>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/2102894

Reviewed-by: Martin Roth <martinroth@google.com>
Commit-Queue: Martin Roth <martinroth@google.com>
Tested-by: Martin Roth <martinroth@google.com>
Auto-Submit: Martin Roth <martinroth@google.com>

Treeya variant use anx3447 as MB PD tcpc controller and it does
not support OCM and need not firmware update. Differentiate it
with anx3429 by sku_id.

BUG=b:140010592, b:139836700, b:137849739
BRANCH=none
TEST=build treeya bios which includes the latest ec, it can boot
     successfully to Chrome system and observe that anx3429 error
     message do not appear in depthcharge output from debug
     console.

Change-Id: I289b4ed9045ab7e589edfe6b43d2e94a75198606
Signed-off-by: Paul Ma <magf@bitland.corp-partner.google.com>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1771166

Reviewed-by: Edward Hill <ecgh@chromium.org>
(cherry picked from commit a542dd01)
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1782299


Commit-Queue: Edward Hill <ecgh@chromium.org>
Tested-by: Edward Hill <ecgh@chromium.org>

MMC_MODE_ should really be MMC_CAPS_ since it's only used by the caps
field. I also cleaned up other ones I found strange.

Generated with:
s/MMC_AUTO_CMD12/MMC_CAPS_AUTO_CMD12/g
s/MMC_MODE_HS_200MHz/MMC_CAPS_HS200/g
s/MMC_MODE_/MMC_CAPS_/g

s/MMC_HS_200MHZ/EXT_CSD_CARD_TYPE_HS200_1_8V/g
s/MMC_HS400/EXT_CSD_CARD_TYPE_HS400_1_8V/g

I'm honestly not quite sure what MMC_CAPS_HS is supposed to mean, or
what it encodes that the other CAPS don't.

BRANCH=grunt
BUG=b:122244718
TEST=Built grunt

Change-Id: Id5c83f4f2f197cd22549f521a2422919e51b8f9b
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1662621

Reviewed-by: Julius Werner <jwerner@chromium.org>
(cherry picked from commit 20efcd99)
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1670777

This allows us to enable the 200Mhz clock.

BRANCH=grunt
BUG=b:122244718
TEST=Verified tuning was working on grunt with both the 52Mhz clock and
the 200Mhz clock. A coreboot patch is required to change the clock
speed.
vb2_load_partition: read 8419 KB in 184ms at 45651 KB/s.

vb2_load_partition: read 8419 KB in 47ms at 175567 KB/s.

You can see the training log here: https://0paste.com/29923-475c04e0



Change-Id: Id21d7df08d498308b498189c5bc23e3875b3761e
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1655553

Reviewed-by: Julius Werner <jwerner@chromium.org>
(cherry picked from commit ba20d0b6)
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1661022

The Bayhub part has support for hardware tuning, but it only supports it
at 4 bits.

I added a new sdhci_send_tuning_cmd because sdhci_send_command_bounced
requires a data buffer, sets up DMA and waits for a different interrupt
then the tuning command returns. Instead of adding a bunch of
conditionals to sdhci_send_tuning_cmd, I added a new function.

BRANCH=grunt
BUG=b:122244718
TEST=Verified it compiles on grunt

Change-Id: I19ef97b474d1597a55136aa55d4f3e4f992938b9
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1655552

Reviewed-by: Julius Werner <jwerner@chromium.org>
(cherry picked from commit 6953df6f)
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1661021

For grunt we need to execute tuning in order to use HS200.

BRANCH=grunt
BUG=b:122244718
TEST=Made sure it compiles

Change-Id: Ifa512cb5e1621b143cadd6852c1661530f27d5c7
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1661019

Reviewed-by: Martin Roth <martinroth@chromium.org>
Commit-Queue: Martin Roth <martinroth@chromium.org>
Tested-by: Martin Roth <martinroth@chromium.org>

CQ-DEPEND=CL:1448957
BUG=chromium:903833,b:120086021
TEST=Tested altfw menu in bootloader
BRANCH=none

Change-Id: Ic79a74e65307517cc6eba816073d5ef990a02f3a
Signed-off-by: Mathew King <mathewk@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1448881

Reviewed-by: Julius Werner <jwerner@chromium.org>
(cherry picked from commit a3edaf91)
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1615320

Reviewed-by: Martin Roth <martinroth@chromium.org>
Commit-Queue: Martin Roth <martinroth@chromium.org>
Tested-by: Martin Roth <martinroth@chromium.org>

   Before power down the internal MCU of ANX3429, EC shall
   reset the internal MCU of ANX3429, otherwise it maybe
   cause I2C salve hang up

BRANCH=None
BUG=b:127532977
TEST=None

Change-Id: I1c711613353bc9556c02b5ab771cc87b7b92b231
Signed-off-by: Xin Ji <xji@analogixsemi.com>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1537754

Reviewed-by: Edward Hill <ecgh@chromium.org>
Commit-Queue: Edward Hill <ecgh@chromium.org>
Tested-by: Edward Hill <ecgh@chromium.org>

Enable multiboot support for booting Zircon kernels.

BUG=b:126618289
BRANCH=grunt
TEST=none

Change-Id: I7162e1bb048b80f48b854ac63e7c08dcaed7e041
Signed-off-by: Duncan Laurie <dlaurie@google.com>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/depthcharge/+/1496176


Commit-Queue: Martin Roth <martinroth@chromium.org>
Tested-by: Martin Roth <martinroth@chromium.org>
Reviewed-by: Martin Roth <martinroth@chromium.org>

1#. program OTP data with the whole firmware raw data, repeat 6 times. 2#. verify OTP data without ECC. 3#. if any problems happened in 1# or 2#, re-start update(step 1# and 2#) (maximum 3 times).

BUG=b:118690201
BRANCH=None
TEST=None

Change-Id: I8d950ae6e4c826863c34b2e76bdac7bf67082a70
Signed-off-by: Xin Ji <xji@analogixsemi.com>
Reviewed-on: https://chromium-review.googlesource.com/1355970


Commit-Ready: Kevin Chiu <kevin.chiu@quanta.corp-partner.google.com>
Tested-by: Einstein Cheng <echeng@analogixsemi.com>
Reviewed-by: Jett Rink <jettrink@chromium.org>
(cherry picked from commit f22a8c2f)
Reviewed-on: https://chromium-review.googlesource.com/c/1460280

Reviewed-by: S Wang <swang@analogix.corp-partner.google.com>
Reviewed-by: Edward Hill <ecgh@chromium.org>
Commit-Queue: Edward Hill <ecgh@chromium.org>
Tested-by: Edward Hill <ecgh@chromium.org>

BUG=b:113880780
BRANCH=none
TEST=Verified the boot log no longer contains the message

Change-Id: Ib093732ca0f277f263775a4b5c64e73f16d14a51
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1396173

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401416

The cr50 i2c tpm protocol uses different register ids then the TPM
specification. This will pretty print the registers to make it easier to
debug TPM transaction failures.

BUG=b:113880780
BRANCH=none
TEST=Enabled on grunt
cr50_i2c_read: 0x50 <= 06 [0xf00: TPM DID VID][4 bytes]
cr50_i2c_read: 0x50 => e0 1a 28 00

Change-Id: Id8f1e355490cb2bc2a334b3d4347dc06b20c899e
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1396172

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401415

This will help when debugging TPM errors.

BUG=b:113880780
BRANCH=none
TEST=Made sure it compiled for grunt

Change-Id: Ia6bfd06c2cd4fdbeb8bd44a4a34b4787b8eecf64
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1396171

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401414

Restructure it to look like the other methods and have it print error
info.

BUG=b:113880780
BRANCH=none
TEST=Grounded SCL and made sure the error was returned.
i2c_wait_for_bus_idle: timeout waiting for bus to become idle.
Status: 0x27, Interrupt Status: 0x110
RX FIFO Level: 0, TX FIFO Level: 0
Enable: 0x1, Enable Status: 0x1
Abort Source: 0x0
i2c_transfer: failed to start transfer.
i2c_disable: Timeout
Status: 0x27, Interrupt Status: 0x110
RX FIFO Level: 0, TX FIFO Level: 0
Enable: 0x0, Enable Status: 0x1
Abort Source: 0x0
cr50_i2c_read: Address write failed
Stress test failed, i2c error!

Change-Id: Id588b841a3ddebe51e34173e632c64f04c833625
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1395344

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401413

This will now correctly catch abort errors when finishing a transaction.

BUG=b:113880780
BRANCH=none
TEST=Verified errors are reported correctly

i2c transaction aborted: [0x1:Abrt7BAddrNoAck]
i2c_wait_for_stop_condition: Fatal Interrupt
i2c_xfer_finish: Error occurred while waiting for stop condition.
cr50_i2c_read: Address write failed
Stress test failed!

Change-Id: Id1fcfa35905226dbb2bb8faf684e291bd8fbd155
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1395343

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401412

BUG=b:113880780
BRANCH=none
TEST=Modified H1 to reset the I2CS every 150 ms and then ran an I2C
stress test. Verified that the abort source is printed.

i2c transaction aborted: [0x1000:ArbLost]
i2c_wait_for_rx_fifo_not_empty: Fatal Interrupt
i2c_transfer_segment: Transferring byte 3 of 4 failed.
i2c_transfer: transferring segment 0 of 1 failed.

Change-Id: I2d380d729b91d90a35e8819433f60181a4d7dd3c
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1395342

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401411

This structural refactor makes the logic easier to follow and sets the
code up to add additional error checking.

BUG=b:113880780
BRANCH=none
TEST=Ran i2c stress test on grunt.

Change-Id: I1609a8ef7ac6959df84674656d665e49d597a7be
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1395341

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401410

BUG=b:113880780
BRANCH=none
TEST=Ran stress test

Change-Id: I03a24b3b4347bb58f66cf5a75e40a1d0640e30ef
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1395340

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401409

BUG=b:113880780
BRANCH=none
TEST=Never saw this condition happen while testing.

Change-Id: I2cf0f3218bc28a7217582fee81e3d3144f176eb0
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1395339

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401408

The enum names are hard to guess, so comment what they mean.
Also added register offsets to make looking up the registers in
documentation easier.

BUG=b:113880780
BRANCH=none
TEST=None

Change-Id: I96e390ecc5ea4a62cb3222195e6c042ff9b50042
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1395338

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401407

I was getting the following error:
'main_fw' may be used uninitialized in this function

BUG=none
BRANCH=none
TEST=verified my compiler error went away.

Change-Id: I49c582a77f2c0943de4d5fc362b65f50767e3e27
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1395337

Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1401406

At present we call this function through an elaborate struct. Call the
function directly to avoid confusion. Also, this function returns 0 on
error, so add some documentation about the cases that occur here.

BUG=b:113880780
BRANCH=grunt
TEST=FW_NAME=careena USE="em100-mode" emerge-grunt --nodeps depthcharge

Change-Id: Iad638448254253018df0c057e8a94d820296cb95
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1384575

Reviewed-by: Raul E Rangel <rrangel@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1396166

This function may fail. Add logging to detect this condition even though
we cannot report it to the caller.

The cr50_i2c_write() call inside can also fail. For that case we have
already logged the error, and adding a short delay after the failure
should not make anything worse.

BUG=b:113880780
BRANCH=grunt
TEST=FW_NAME=careena USE="em100-mode" emerge-grunt --nodeps depthcharge

Change-Id: I810243dd92bd2eabf1fb60c13d7bf56aeba1cb14
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1384574

Reviewed-by: Raul E Rangel <rrangel@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1396165

We can probably not worry about errors in tpm_init() since the caller will
complain (and presumably reboot) if that fails. But there are a few other
cases which don't currently log. Fix these.

BUG=b:113880780
BRANCH=grunt
TEST=FW_NAME=careena USE="em100-mode" emerge-grunt --nodeps depthcharge

Change-Id: Ieb0559da4f2a9ecb668dc36b85390d73a1a8bfc8
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1384573

Reviewed-by: Raul E Rangel <rrangel@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1396164

This driver checks for an error condition of non-zero or < 0 in different
places. The functions are clearly documented as returning either 0 or -1
so non-zero is sufficient and easier to read.

Update the code to use non-zero consistently through. Add some missing
function comments also.

BUG=b:113880780
BRANCH=grunt
TEST=FW_NAME=careena USE="em100-mode" emerge-grunt --nodeps depthcharge

Change-Id: I1c951c1b442a69fdd4cc053269d2b694dc25a599
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1384572

Reviewed-by: Raul E Rangel <rrangel@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1396163

At present the payload functions take a cbfs_media * as an argument. But
at present we use the same area for all payload operations, so this is
unnecessary.

Add a new function to return the cbfs_media struct, and keep a single one
of these in payload.c to simplify the API.

BUG=b:120555612
BRANCH=grunt
TEST=manually test on grunt, pressing '1' to boot into U-Boot

Change-Id: I1f7eaaac185c5233760cee5e8b1d8cc6c9080524
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1370604

Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Raul E Rangel <rrangel@chromium.org>
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1396162

This function should set up the media information always, in case the
caller relies on it. This behaviour matches the documented function
header. Fix this.

BUG=b:120555612
BRANCH=grunt
TEST=manually test on grunt with Ctrl-L, 1 and see that it boots into
U-Boot

Change-Id: I43b1cf42a6d122d074348f5a21ffe4492f411e5e
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1370208

Reviewed-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
(cherry picked from commit 846882c3)
Reviewed-on: https://chromium-review.googlesource.com/c/1384564

Do not check the status of device during registration. Instead perform a
lazy evaluation of device status when the registered device is accessed.

BUG=chromium:896451
BUG=b:112306866
BRANCH=None
TEST=bootup to ChromeOS by connecting and disconnecting the USB
daughterboard

Change-Id: Iaae538ad0f59020a41be24349b8caf8b9f7d92f0
Signed-off-by: Karthikeyan Ramasubramanian <kramasub@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1289778

Reviewed-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Jett Rink <jettrink@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1351847

Reviewed-by: Martin Roth <martinroth@chromium.org>
Tested-by: Martin Roth <martinroth@chromium.org>

Currently capture_device_id returns success or failure after querying
EC for PD chip info. Update that function to return different states
of the device. The different states include: missing device, device is
not a parade chip, device is in corrupted state and device is in good
state.

BUG=chromium:896451
BRANCH=None
TEST=bootup to ChromeOS by connecting and disconnecting the USB
daughterboard

Change-Id: Id8d1c3abfcd8911ec2fcae569f1ac1766adb0c8a
Signed-off-by: Karthikeyan Ramasubramanian <kramasub@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1320152

Reviewed-by: Jett Rink <jettrink@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1351846

Reviewed-by: Martin Roth <martinroth@chromium.org>
Tested-by: Martin Roth <martinroth@chromium.org>

Move the ps8751_should_try_upgrade function around so as to avoid a
forward declaration in the follow-up patch.

BUG=chromium:896451
BRANCH=None
TEST=bootup to ChromeOS by connecting and disconnecting the USB
daughterboard

Change-Id: I7b02d610826e4826945ff2f9e10417493f6f3d58
Signed-off-by: Karthikeyan Ramasubramanian <kramasub@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1313230

Reviewed-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Jett Rink <jettrink@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1351845

Reviewed-by: Martin Roth <martinroth@chromium.org>
Tested-by: Martin Roth <martinroth@chromium.org>

There is no need to protect a registered device if the device is not
present at run-time. Update vboot_aux_fw driver to handle the
device not present scenario.

CQ-DEPEND=CL:1299993
BUG=chromium:896451
BRANCH=None
TEST=bootup to ChromeOS by connecting and disconnecting the USB
daughterboard

Change-Id: Ida8dad2f177a9c5e332924d7ef8ecdabed07b7fd
Signed-off-by: Karthikeyan Ramasubramanian <kramasub@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1299900

Reviewed-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Jett Rink <jettrink@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/1351844

Reviewed-by: Martin Roth <martinroth@chromium.org>
Tested-by: Martin Roth <martinroth@chromium.org>