TEST=Boot from scarlet, and mipi panel works
Change-Id: I52f8f8f966034f5273d7c2e673e5ebdd9dccf748
Signed-off-by: Nickey Yang <nickey.yang@rock-chips.com>
Reviewed-on: https://review.coreboot.org/19700
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
The coreboot had no supported the different frequency for gru yet.
e.g:
we can't support the bob to run ddr 800M for rev3 board and
run 928M for rev4 board.
So, in order to support the 800M and 928M ddr frequency for bob different
boards. We will use the ram_id and board_id to select the board on bob.
Change-Id: I613050292a09ff56f4636d7af285075e32259ef4
Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Reviewed-on: https://review.coreboot.org/19558
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Spread Spectrum Modulator (SSMOD) is a fully-digital circuit used to
modulate the frequency of the Silicon Creations’ Fractional PLL in order
to reduce EMI.
We need to turn the DPLL spread spectrum feature on to
reduce the EMI noise for DDR on bob.
Change-Id: I75461d4235bcf55324e6664a1220754e770b4786
Signed-off-by: Xing Zheng <zhengxing@rock-chips.com>
Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Reviewed-on: https://review.coreboot.org/19557
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
This reverts commit 39b633b26d.
Commit was accidentally pushed too early and broke the tree.
I'll repush the original.
Change-Id: Iaca6d43cc8fc0959565d5d151a330c0c7ba38309
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/19596
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
TEST=Boot from scarlet, and mipi panel work
Change-Id: Id5f81867ea50f72cc0bc13074627134e0dc198ba
Signed-off-by: Nickey Yang <nickey.yang@rock-chips.com>
Reviewed-on: https://review.coreboot.org/19476
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Board Scarlet doesn't use usbphy1.
BUG=b:37685249
TEST=boot Scarlet, check the firmware log, and confirm
no errors about USB1
Change-Id: I66e0d8a235cc9057964f7abca32bc692d41e88fd
Signed-off-by: Philip Chen <philipchen@google.com>
Reviewed-on: https://review.coreboot.org/19489
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
BUG=b:35647967
TEST=boot from bob
Change-Id: I756513f02ac13e159d5b8b1ac2346fa42cf3c219
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: cf18ed7b8fdf11594f812e5c48a2bd0fde5cb820
Original-Change-Id: I50c053ab7a6f6c14daee4fb2ab1cdcaeee2d67da
Original-Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/452286
Original-Commit-Ready: Caesar Wang <wxt@rock-chips.com>
Original-Tested-by: Caesar Wang <wxt@rock-chips.com>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/19434
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
In the safety considerations, we should make sure the slot of SD is
enabled first, since we want to the power switch of corresponding is
powered up.
The different boards have the different power switch for sdmmc.
Some power switch IC need turn on delay for long time.
let's move the slot power of SD to romstage and avoid explicit delays
or per-board.
BRANCH=none
BUG=b:35813418, b:35573103
TEST=check the signal for children of gru, and boot up from sd card.
Change-Id: Id164e4c4c900c6b1ca0251fc27db4cd36c56f6ff
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: ea1b01cc13628033b85251dbb44407f075efdc85
Original-Change-Id: I48ab543143d3de9be46608fc12d78e09decf8d79
Original-Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/447076
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/19430
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
The CR50 TPM can do both SPI and I2C communication. However,
there's situations where policy needs to be applied for CR50
generically regardless of the I/O transport. Therefore add
MAINBOARD_HAS_TPM_CR50 to encompass that. Additionally,
once the mainboard has selected CR50 TPM automatically select
MAINBOARD_HAS_TPM2 since CR50 TPM is TPM 2.0.
Change-Id: I878f9b9dc99cfb0252d6fef7fc020fa3d391fcec
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/19370
Reviewed-by: Furquan Shaikh <furquan@google.com>
Tested-by: build bot (Jenkins)
Kevin's center logic isn't super clean so it needs 925 mV for center
logic. All newer gru variants only need 900 mV.
BRANCH=gru
BUG=b:37429075
TEST=Reboot tests
Change-Id: I8c3bd6c245700b23c27cd5758c35c9993f801cb4
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/479463
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/19357
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
It seems that we should only ever run at 900mV on center logic.
Changing it to 950mV before might have just masked over problems that
are now fixed.
BRANCH=none
BUG=chrome-os-partner:56940
TEST=on kevin, run
stressapptest -M 1536 -s 1000
Change-Id: I5a09b1b403df800396bb2f2e8c76d14a4519d44a
Signed-off-by: Derek Basehore <dbasehore@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/391032
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Commit-Queue: Lin Huang <hl@rock-chips.com>
Tested-by: Lin Huang <hl@rock-chips.com>
Reviewed-on: https://review.coreboot.org/19356
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
Scarlet don't have eDP and MIPI driver is not ready, skipping
display for now or else Scarlet would be stuck in
reading eDP HPD because there even not power for it.
TEST=boot to kernel on Scarlet
Change-Id: I02ab4ef21bf77b98414f537aca57b46c11922348
Signed-off-by: Shunqian Zheng <zhengsq@rock-chips.com>
Reviewed-on: https://review.coreboot.org/19237
Tested-by: build bot (Jenkins)
Reviewed-by: Julius Werner <jwerner@chromium.org>
BUG=b:35583511
TEST=check i2c bus 0 initializes from ap console log
Change-Id: Ibb6709159f5ed28ad0b62397d2ddb504dec55167
Signed-off-by: Philip Chen <philipchen@google.com>
Reviewed-on: https://review.coreboot.org/19105
Tested-by: build bot (Jenkins)
Reviewed-by: Julius Werner <jwerner@chromium.org>
This patch attempts to finish the separation between CONFIG_VBOOT and
CONFIG_CHROMEOS by moving the remaining options and code (including
image generation code for things like FWID and GBB flags, which are
intrinsic to vboot itself) from src/vendorcode/google/chromeos to
src/vboot. Also taking this opportunity to namespace all VBOOT Kconfig
options, and clean up menuconfig visibility for them (i.e. some options
were visible even though they were tied to the hardware while others
were invisible even though it might make sense to change them).
CQ-DEPEND=CL:459088
Change-Id: I3e2e31150ebf5a96b6fe507ebeb53a41ecf88122
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/18984
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The virtualized developer switch was invented five years ago and has
been used on every vboot system ever since. We shouldn't need to specify
it again and again for every new board. This patch flips the Kconfig
logic around and replaces CONFIG_VIRTUAL_DEV_SWITCH with
CONFIG_PHYSICAL_DEV_SWITCH, so that only a few ancient boards need to
set it and it fits better with CONFIG_PHYSICAL_REC_SWITCH. (Also set the
latter for Lumpy which seems to have been omitted incorrectly, and hide
it from menuconfig since it's a hardware parameter that shouldn't be
configurable.)
Since almost all our developer switches are virtual, it doesn't make
sense for every board to pass a non-existent or non-functional developer
mode switch in the coreboot tables, so let's get rid of that. It's also
dangerously confusing for many boards to define a get_developer_mode()
function that reads an actual pin (often from a debug header) which will
not be honored by coreboot because CONFIG_PHYSICAL_DEV_SWITCH isn't set.
Therefore, this patch removes all those non-functional instances of that
function. In the future, either the board has a physical dev switch and
must define it, or it doesn't and must not.
In a similar sense (and since I'm touching so many board configs
anyway), it's annoying that we have to keep selecting EC_SOFTWARE_SYNC.
Instead, it should just be assumed by default whenever a Chrome EC is
present in the system. This way, it can also still be overridden by
menuconfig.
CQ-DEPEND=CL:459701
Change-Id: If9cbaa7df530580a97f00ef238e3d9a8a86a4a7f
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/18980
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Going forward it's important to note when a CR50 is expected
to be present in the system. Additionally, this Kconfig addition
provides symmetry with the equivalent i2c Kconfig option.
BUG=b:35775104
Change-Id: Ifbd42b8a22f407534b23459713558c77cde6935d
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/18680
Reviewed-by: Furquan Shaikh <furquan@google.com>
Tested-by: build bot (Jenkins)
Gru/Kevin use 933 MHz (actually 928 MHz for better jitter) as max sdram
frequency, while bob uses 800 MHz.
It's normal some variants can't meet 928 MHz SI requirement and hence
have to use a lower freq as spec.
BUG=chrome-os-partner:61001
BRANCH=gru
TEST=check dpll is 800 MHz on bob
Change-Id: I6d19a351f25d1f48547715ce57c3a87d9505f6f1
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 8176bfea52422c713f144ffec419752aeca66db2
Original-Change-Id: I46afba8d091f1489feeb20cafc44decaa81601fc
Original-Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Original-Signed-off-by: Shunqian Zheng <zhengsq@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/420208
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Original-Commit-Queue: Shasha Zhao <Sarah_Zhao@asus.com>
Original-Tested-by: Shasha Zhao <Sarah_Zhao@asus.com>
Original-(cherry picked from commit eba5dff79eeedae5ff608d2d8d297ccf9c13cb55)
Original-Reviewed-on: https://chromium-review.googlesource.com/448277
Original-Reviewed-by: Derek Basehore <dbasehore@chromium.org>
Reviewed-on: https://review.coreboot.org/18581
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Follow up to https://review.coreboot.org/#/c/18460/
Change-Id: Ic3aada2acf3051622698e10d2e764050e16480d5
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: https://review.coreboot.org/18475
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
According to USB 2.0 Spec Table 7-7, the High-speed squelch
detection threshold Min 100mV and Max 150mV, and we set USB
2.0 PHY0 and PHY1 squelch detection threshold to 150mV by
default, so if the amplitude of differential voltage envelope
is < 150 mV, the USB 2.0 PHYs envelope detector will indicate
it as squelch.
On Kevin board, if we connect usb device with Samsung U2 cable,
we can see that the impedance of U2 cable is too big according
to the eye-diagram test report, and this cause serious signal
attenuation at the end of receiver, the amplitude of differential
voltage falls below 150mV.
This patch aims to reduce the PHY0 and PHY1 otg-ports squelch
detection threshold to 125mV (host-ports still use 150mV by
default), this is helpful to increase USB 2.0 PHY compatibility.
BRANCH=gru
BUG=chrome-os-partner:62320
TEST=Plug Samsung U2 cable + SEC P3 HDD 500GB/Galaxy S3 into
Type-C port, check if the USB device can be detected.
Change-Id: Ia0a2d354781c2ac757938409490f7c4eecdffe61
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 7d74311c25762668386061234df0562f84b7203e
Original-Change-Id: Ib20772f8fc2484d34c69f5938818aaa81ded7ed8
Original-Signed-off-by: William wu <wulf@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/431015
Original-Commit-Ready: Caesar Wang <wxt@rock-chips.com>
Original-Tested-by: Inno Park <ih.yoo.park@samsung.com>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/18462
Reviewed-by: Martin Roth <martinroth@google.com>
Tested-by: build bot (Jenkins)
As David commented the "Bob and other follow-ons match Gru, Kevin should
be the special case here", and update the calculations value for gru/bob
board.
From the actual tests, some regulator voltage than the actual set of less
than 20mv on bob board. (e.g: little-cpus and Center-logic) Update the
{min, max} regulator voltage for Bob board. Make sure we get the accurate
voltage.
BUG=chrome-os-partner:61497
BRANCH=none
TEST=boot up Bob, measure the voltage for little cpu and C-logic.
Change-Id: Iad881b41d67708776bfb681487cf8cec8518064e
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 25e133815f49018e7496c75077b8559c207350a4
Original-Change-Id: I3098c742c7ec355c88f45bd1d93f878a7976a6b4
Original-Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Original-Signed-off-by: Shasha Zhao <Sarah_Zhao@asus.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/424523
Original-Reviewed-by: David Schneider <dnschneid@chromium.org>
Original-Reviewed-by: Brian Norris <briannorris@chromium.org>
Original-Signed-off-by: Shasha Zhao <Sarah_Zhao@asus.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/430403
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/18460
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
The children of Gru should share the benefits. In the real world, Bob can't
pass the eye diagram tests.
BUG=chrome-os-partner:62714
BRANCH=firmware-gru-8785.B
TEST=build coreboot
Change-Id: I2470bbc81acdaf2458d660dca5dc307cc3038f83
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: d0cb3e718a7571f602a00c08a42019851634e7fd
Original-Change-Id: I0ccb48bb52eb770ccc9c8c265b07df46b0308dd3
Original-Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/440745
Original-Reviewed-by: Douglas Anderson <dianders@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/441468
Reviewed-on: https://review.coreboot.org/18461
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Change-Id: I538c28fb1bc412947ef9df947fa3f6a3312aeb4b
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: https://review.coreboot.org/18322
Tested-by: build bot (Jenkins)
Reviewed-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Martin Roth <martinroth@google.com>
There will be more follow-up changes.
BUG=chrome-os-partner:62377
BRANCH=None
TEST=emerge-scarlet coreboot libpayload
Change-Id: I9ca45598ff0ab12bf8063d16a86be564cf509390
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: a020a9ba1228b15599e202972df0096f58b1b31c
Original-Change-Id: I4804239483f8b35bc3703aa62c2a8fd642e0234a
Original-Signed-off-by: philipchen <philipchen@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/433039
Original-Commit-Ready: Philip Chen <philipchen@chromium.org>
Original-Tested-by: Philip Chen <philipchen@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/18296
Tested-by: build bot (Jenkins)
Reviewed-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
The same GPIOs are used on both platforms, definitions are added an a
new .h to make it easier to re-use them across the code.
BRANCH=none
BUG=chrome-os-partner:51537
TEST=panel backlight still enabled on Gru as before. The rest of the
GPIOs are used in the upcoming patches.
Change-Id: I54ef3e8dd79670bdb037baeec91430113d11bcc1
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: c58788026f28af52c650da0159b93d97269ca4a9
Original-Change-Id: I1a6c5b5beb82ffcc5fea397e8e9ec2f183f4a7e0
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/346219
Original-Tested-by: Shunqian Zheng <zhengsq@rock-chips.com>
Reviewed-on: https://review.coreboot.org/18176
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
The commit 0ba3b2593b0c ("gru: Tuning USB 2.0 PHY to increase
compatibility") bypass ODT to set the max driver strength for
the Type-C otg-port, it works well on otg-port when connected
with USB2.0 devices.
Unfortunately, because the Type-C otg-port and host-port are
consisted in one USB2 PHY, so bypass ODT will have an effect
on both host-port and otg-port. I have tested the host-port
eye-diagram, the result shows that if we bypass ODT, the host-
port eye-diagram height will become to high, more than 500mv,
this may cause USB 2.0 high-speed enumeration failure.
This patch bypass ODT for host-port separately, and then we
can reduce the host-port driver strength without affecting
the otg-port driver strength.
BRANCH=gru
BUG=chrome-os-partner:60727
TEST=Boot system, run 'lsusb' command and check if the usb camera
and usb bluetooth are on usb 2.0 hub or usb 1.1 hub. If they are
on usb 1.1 hub, the issue happens. If not, try to run camera app
and then close camera app, repeat until find that the usb camera
is on the usb 1.1 hub.
Change-Id: Ib693e2a6f2113c06692a7bfee22d85b67ee3b165
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 5ea7660b7b05080b76fc5ca5af3fa18552a03491
Original-Change-Id: Ia1f12182929673c5726df9f77f0903469b5c957a
Original-Signed-off-by: William wu <wulf@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/425739
Original-Commit-Ready: Douglas Anderson <dianders@chromium.org>
Original-Tested-by: Douglas Anderson <dianders@chromium.org>
Original-Tested-by: Inno Park <ih.yoo.park@samsung.com>
Original-Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-on: https://review.coreboot.org/18126
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
We may support different sdram sizes on one board in future, so
we need to calculate sdram sizes from sdram drvier.
BRANCH=None
BUG=None
TEST=boot kevin
Change-Id: I43e8f164ecdb768c051464b4dbc7d890df8055d0
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 3c4d8b3cb647b2f9cebc416c298817c16d49330e
Original-Change-Id: I95d5ef34de9d79ebca3600dc7a4b9e14449606ff
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/411600
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/17629
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Gru only uses USB 2.0 in firmware to avoid all the madness associated
with Type-C port orientation and USB 3.0 tuning. We do this by isolating
the SuperSpeed lines in the Type-C PHY so it looks like they aren't
connected to the device.
Unfortunately, some devices seem to already get "locked" into SuperSpeed
mode as soon as they detect Rx terminations once, and can never snap out
again on their own. Since the terminations are already connected during
power-on reset we cannot disable them fast enough to prevent this, and
the only solution we found to date is to power-cycle the whole USB port.
Now, Gru's USB port power is controlled by the EC, and unfortunately we
have no direct host command to control it. We do however have a command
to force a certain USB PD "role", and forcing our host into "sink" mode
makes it stop sourcing power to the port. So for lack of a saner
solution we'll use this to work around our problem.
BRANCH=gru
BUG=chrome-os-partner:59346
TEST=Booted Kevin in recovery mode, confirmed that my "problem stick"
gets detected immediately (whereas previously I had to unplug/replug
it). Booted Kevin to OS in both developer and normal mode and confirmed
that USB still seems to work.
Change-Id: Ib3cceba9baa170b13f01bd5c01bd413be5b441ba
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: cd695eda33299e50362f1096c46f2f5260c49036
Original-Change-Id: I2db3d6d3710d18a8b8030e94eb1ac2e931f22638
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/413031
Reviewed-on: https://review.coreboot.org/17628
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Instead of putting all the functions inline just put the
current implementation into a C file. That way all the implementation
innards are not exposed.
Lastly, fix up the fallout of compilation units not including the
headers they actually use.
Change-Id: I01fd25d158c0d5016405b73a4d4df3721c281b04
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/17648
Tested-by: build bot (Jenkins)
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-by: Furquan Shaikh <furquan@google.com>
When testing USB 2.0 compatibility with different kinds
of USB 2.0 devices on Kevin board, we find that some
USB HDDs (e.g. seagate SRD00F1 1TB HDD) and some smart
phones (e.g. galaxy A5 smart phone) can't be detected.
And according to the error log, this issue is related
to USB 2.0 PHY signal problem.
For the USB HDD, error log is:
[ 592.557724] usb 5-1: new high-speed USB device number 2 using xhci-hcd
[ 592.847735] usb 5-1: new high-speed USB device number 3 using xhci-hcd
[ 593.473720] usb 5-1: new high-speed USB device number 6 using xhci-hcd
[ 594.187717] usb 5-1: new high-speed USB device number 9 using xhci-hcd
[ 595.020717] usb 5-1: new high-speed USB device number 13 using xhci-hcd
[ 595.284730] usb 5-1: new high-speed USB device number 14 using xhci-hcd
[ 595.574816] usb 5-1: new high-speed USB device number 15 using xhci-hcd
The log shows that HDD failed to high-speed handshake.
For the smart phone, error log is:
[ 1145.661625] usb 5-1: new high-speed USB device number 2 using xhci-hcd
[ 1145.771674] usb 5-1: device descriptor read/64, error -71
[ 1145.979752] usb 5-1: device descriptor read/64, error -71
[ 1146.187721] usb 5-1: new high-speed USB device number 3 using xhci-hcd
[ 1146.301754] usb 5-1: device descriptor read/64, error -71
[ 1146.509750] usb 5-1: device descriptor read/64, error -71
[ 1146.717722] usb 5-1: new high-speed USB device number 4 using xhci-hcd
[ 1146.724393] usb 5-1: Device not responding to setup address.
[ 1146.930795] usb 5-1: Device not responding to setup address.
[ 1147.137720] usb 5-1: device not accepting address 4, error -71
[ 1147.246644] usb 5-1: new high-speed USB device number 5 using xhci-hcd
[ 1147.253336] usb 5-1: Device not responding to setup address.
[ 1147.459786] usb 5-1: Device not responding to setup address.
[ 1147.665712] usb 5-1: device not accepting address 5, error -71
[ 1147.671789] usb usb5-port1: unable to enumerate USB device
The log shows that smart phone failed to read device
descriptor, error -71 may be caused by PHY signal problem.
This patch aims to tune USB 2.0 PHY with the following
parameters to support USB HDD, smart phone and some other
potential USB 2.0 devices.
1. Disable the pre-emphasize in chirp state to avoid
high-speed handshake failure.
2. Bypass ODT auto compensation to enable set max driver
strength manually. (Bit[42] of usbphy_ctrl register is
1'b1 for bypass, and Bit[41:37] of usbphy_ctrl register
is 5'b10000 for max driver strength).
3. Bypass ODT auto refresh, and set the max bias current
tuning reference. (Bit[57] of usbphy_ctrl register is
1'b1 for bypass, and Bit[52:50] of usbphy_ctrl register
is 3b'100 for max bias current tuning reference).
We have done the USB 2.0 compliance test and compatibility test
with this patch, it works well.
BRANCH=gru
BUG=chrome-os-partner:59623
TEST=plug/unplug USB HDD or smart phone in Type-C port,
check if they can be detected successfully.
Change-Id: I275c2236b8e469bfd04e9184d007eb095657225e
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 7735c514d4136978133c2299f2f58da8320bb89f
Original-Change-Id: I4e6c10faa1c03af9880a89afe4731a7065eb1e4e
Original-Signed-off-by: William wu <wulf@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/409856
Original-Commit-Ready: Eddie Cai <eddie.cai.rk@gmail.com>
Original-Tested-by: Cindy Han <cindy.han@samsung.com>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/17566
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Instead of defining the same functions for reading/clearing boot-mode
switches from EC in every mainboard, add a common infrastructure to
enable common functions for handling boot-mode switches if
GOOGLE_CHROMEEC is being used.
Only boards that were not moved to this new infrastructure are those
that do not use GOOGLE_CHROMEEC or which rely on some mainboard specific
mechanism for reading boot-mode switches.
BUG=None
BRANCH=None
TEST=abuild compiles all boards successfully with and without ChromeOS
option.
Change-Id: I267aadea9e616464563df04b51a668b877f0d578
Signed-off-by: Furquan Shaikh <furquan@chromium.org>
Reviewed-on: https://review.coreboot.org/17449
Tested-by: build bot (Jenkins)
Reviewed-by: Julius Werner <jwerner@chromium.org>
This changes memory to only do CA training with one pattern,
0xfffff/0x00000 and to also make sure CA training waits for all of the
captures during training.
BRANCH=none
BUG=chrome-os-partner:56940
TEST=boot kevin and run
stressapptest -M 1500 -s 1000
Change-Id: I0982674b4f4415f4d7865923ced93fa09bdd877e
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 75cdd911cea9c4e5744fd04505b260fa5755513c
Original-Change-Id: I3b86e6d4662c6fbbf9ddef274fce191a367904e5
Original-Signed-off-by: Derek Basehore <dbasehore@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/410320
Original-Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-on: https://review.coreboot.org/17383
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
This adds a new CA training pattern for all of the supported
frequencies. This pattern increases the hold time on CA.
BRANCH=none
BUG=chrome-os-partner:57845
TEST=boot kevin and run:
while true; do sleep 0.1; memtester 500K 1 > /dev/null; done
for several hours
Change-Id: Ie5958cf67c16247ef90ee261da9faef4ffa5b339
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 8babeafe75bffcb2dab17eb007b4f5bb0eb42606
Original-Change-Id: I7f7652f88e43dc9b2f6069e60514931bf7582ed1
Original-Signed-off-by: Derek Basehore <dbasehore@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/403547
Original-Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-on: https://review.coreboot.org/17382
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
framebuffer address is dynamically chosen by libpayload now, so there's
no need to configure it in coreboot.
CQ-DEPEND=CL:401402
BUG=chrome-os-partner:58675
BRANCH=none
TEST=Boot from kevin, dev screen is visible
Change-Id: I9f1e581d5c63b3579b26be22ce5c8d1e71679f6f
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: b3b6675420592c30e1e0abc8f8e9dd6ed5abd04c
Original-Change-Id: I7e3162f24a4dc426fe4e10d74865cf0042c80db5
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/401401
Original-Commit-Ready: Julius Werner <jwerner@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/17109
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
To enable DDR Dynamic Voltage and Frequency Scaling (DVFS) we need to
train alternative configurations first, so do the training and store the
values.
BUG=None
BRANCH=None
TEST=Boot from kevin
Change-Id: I944a4b297a4ed6966893aa09553da88171307a42
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 94533ff3ba21bcb0ace00bedcf0cebb89a341be2
Original-Change-Id: I4a98bc0db5553d154fedb657e35b926a92aa80c7
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/386596
Original-Commit-Ready: Douglas Anderson <dianders@chromium.org>
Original-Reviewed-by: Derek Basehore <dbasehore@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/17104
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
There are some sdram configurations that are no longer used. Drop them.
BUG=None
BRANCH=None
TEST=None
Change-Id: Ib6d2d58c3071147a3095bc1ed7fa7b02c748e1a5
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 111d375005ec6a3b91e47acdd676e8f1644c931c
Original-Change-Id: I5f9278093f02e785b2894faa8e8cf09ecec20325
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/399122
Original-Commit-Ready: Douglas Anderson <dianders@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/17103
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
We found sdram may fail in pctl_cfg(), so we check the status in this
function. If it exceeds 100ms still in this function, we will restart
the system. We also found there are rare chances DDR training fails,
so also restart system in that case.
BUG=chrome-os-partner:57988
BRANCH=None
TEST=coreboot resets on failure and eventually the system comes up
Change-Id: Icc0688da028a8f4f81eafe36bbaa79fdf2bcea74
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 89e45f8352f62e19a203316330aba14ccc5c8b11
Original-Change-Id: If4e78983abcfdfe1e0e26847448d86169e598700
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/397439
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/17045
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
We found that Kevin board PHY0 and PHY1 eye-diagram margin
is not enough to make compliance test pass, and the PHY0 USB
SI is worse than PHY1, because of the higher PCB impedance.
For PHY0, we can't improve the eye-diagram by SW PHY tuning,
so we need to reduce the RBIAS resistance from 133 ohm to 115
ohm, it can help to increase the eye-height.
For PHY1, we can improve the eye-diagram by setting the max
pre-emphasis level.
And after the above change, the USB2 signal amplitude will
become larger at the test point near to SOC USB2 PHY, in order
to avoid mis-trigger the disconnect detection (650mV), we need
to disable pre-emphasize in eop state.
BRANCH=None
BUG=chrome-os-partner:53863
TEST=do USB 2.0 compliance test for Kevin C0 and C1 port.
Change-Id: I95c0acd79623aeca9a0ae077b1dd3836d91fe561
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: de3cdef128966d76e7d8e2ebd641763b911c3ad5
Original-Change-Id: I00cb325b9938e4276cc77b5d6f5faa7023379608
Original-Signed-off-by: William wu <wulf@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/390615
Original-Commit-Ready: Julius Werner <jwerner@chromium.org>
Original-Tested-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/16911
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
GPIO1_B3 (WLAN_MODULE_RST#) defaults as a pull-up input, but it is also
"pulled up" by 1.8V_WLAN. However, 1.8V_WLAN remains low for some time
during early boot. This leaves the signal floating somewhere in the
middle.
This has two potential issues:
(1) we're leaking some power for some (hopefully) short period of time
(2) we are possibly screwing with the Wifi power sequence; we aren't
supposed to deassert PDn (i.e., MODULE_RST#) until all the rails
have fully ramped for some period of time
Neither of the above issues are likely to be significant, but it is nice
to fix, I expect.
BRANCH=gru
BUG=chrome-os-partner:54026
TEST=measure WLAN_MODULE_RST# on scope at boot time
Change-Id: Ia6af9ad6954ad8feeda33015e3f205842380939e
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 0e890a2787bf034d3358a33fc88c2dd8078593ab
Original-Change-Id: I120e26ad0ca486a326874986e142dcaee965b62d
Original-Signed-off-by: Brian Norris <briannorris@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/388009
Original-Reviewed-by: Douglas Anderson <dianders@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/16882
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Martin Roth <martinroth@google.com>
PHY_PER_CS_TRAINING is being enabled when DDR frequency >= 666.
For per cs training, the controller should consider the PHY
delay line switch time and there should be more cycles to
switch the delay line, so update the W2W_DIFFCS_DLY_ value
from 0x1 to 0x5.
BRANCH=none
BUG=chrome-os-partner:56940
TEST=do memtester on kevin board, and pass
Change-Id: I00df2d4724b0b77f3e7565809fb35bbd2ff01ea5
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: c135ea3e33d810ed322d947eb8d512d1ac119cfc
Original-Change-Id: I81b99cbc085769b7028e770509d79bd8d550820b
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/387506
Original-Reviewed-by: Douglas Anderson <dianders@chromium.org>
Original-Reviewed-by: Derek Basehore <dbasehore@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/16721
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
To save power when entering suspend, gpios 2 to 4 need to be set
to input and 'pull none' mode.
Pass the APIO configuration to ATF so it can do a proper job here.
BRANCH=None
BUG=chrome-os-partner:56423
TEST=run suspend_stress_test on kevin board
Change-Id: Id57fe8f622ae3f9c2bc7e58be89518b2b846cd37
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 9c42082d1ca9a6baa735821382d3e83c1f8dc9ad
Original-Change-Id: Iaf441e8e34c5591ffe7c65f6533fcf0b733ff5ac
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/378475
Original-Commit-Ready: Caesar Wang <wxt@rock-chips.com>
Original-Tested-by: Caesar Wang <wxt@rock-chips.com>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/16720
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
We need to disable some regulators when the device goes into suspend.
This means that we need to pass some gpios to bl31, and disable these
gpios when bl31 runs the suspend function.
BRANCH=None
BUG=chrome-os-partner:56423
TEST=enter suspend, measure suspend gpio go to low
[pg: also update arm-trusted-firmware to match]
Change-Id: Ia0835e16f7e65de6dd24a892241f0af542ec5b4b
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 0f3332ef2136fd93f7faad579386ba5af003cf70
Original-Change-Id: I03d0407e0ef035823519a997534dcfea078a7ccd
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/374046
Original-Commit-Ready: Caesar Wang <wxt@rock-chips.com>
Original-Tested-by: Caesar Wang <wxt@rock-chips.com>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/16719
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
To improve sdram 800MHz and 933MHz stability, we
need to modify write leveling flow to get the
proper write leveling value.
BUG=chrome-os-partner:56940
BRANCH=none
TEST=Boot from kevin on 933MHz, and do stressapptest
Change-Id: I5b24c93d4a57917fb9af7e5e2a95d8423ccbaa7e
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: d84bf25b3e5de373c7913e6d534a810cb984b3fd
Original-Change-Id: I87efddf628c3683fcb85d6875e029cf3cbc482be
Original-Signed-off-by: Jianqun Xu <jay.xu@rock-chips.com>
Original-Signed-off-by: Xing Zheng <zhengxing@rock-chips.com>
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/384292
Original-Commit-Ready: Julius Werner <jwerner@chromium.org>
Original-Tested-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/16716
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Since there's currently a limitation in coreboot's code that prevents
more than 4KB to be used by the eventlog anyway, this patch shrinks the
available RW_ELOG area in the FMAP for Gru down to 4KB. This may prove
prudent later if we ever resolve that limitation, so that tools can rely
on the area in the FMAP being the same as the area actually used by the
read-only firmware code on these boards.
BRANCH=gru
BUG=chrome-os-partner:55593
TEST=Booted Kevin, confirmed that eventlog got written normally. Ran a
reboot loop to exhaust eventlog space, confirmed that the shrink code
kicks in as expected before reaching 4KB.
Change-Id: I3c55d836c72486665a19783fe98ce9e0df174b6d
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 05efb82ca00703fd92d925ebf717738e37295c18
Original-Change-Id: Ia2617681f9394e953f5beb4abf419fe8d97e6d3e
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/384585
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-by: Simon Glass <sjg@google.com>
Reviewed-on: https://review.coreboot.org/16715
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
We found some boards are not stable when sdram is run at 933Mhz.
Before we can fix it, we need to lower the sdram frequency to 800MHz.
In this patch we modify the DQS delay from 0x280 to 0x260 and extend
the DQS window.
BRANCH=None
BUG=chrome-os-partner:56940
TEST=Booted Kevin.
Change-Id: I68561c4aa4d9ab66acfa3515a42d696157aff759
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 877a7f6ad22a5bde9f9e458bcb65f133f2f001bd
Original-Change-Id: I5eab6bbe96f0dae095c5353403292022e7a25421
Original-Signed-off-by: Lin Huang <hl@rock-chips.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/382724
Original-Commit-Ready: Douglas Anderson <dianders@chromium.org>
Original-Tested-by: Douglas Anderson <dianders@chromium.org>
Original-Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-on: https://review.coreboot.org/16709
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
This patch moves the big CPU cluster initialization on the RK3399 from
the clock init bootblock function into ramstage. We're only really doing
this to put the cluster into a sane state for the OS, we're never
actually taking it out of reset ourselves... so there's no reason to do
this so early.
Also cleaned up the interface for rkclk_configure_cpu() a bit to make it
more readable.
BRANCH=None
BUG=chrome-os-partner:54906
TEST=Booted Kevin.
Change-Id: I568b891da0abb404760d120cef847737c1f9e3ec
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: bd7aa7ec3e6d211b17ed61419f80a818cee78919
Original-Change-Id: Ic3d01a51531683b53e17addf1942441663a8ea40
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/377541
Original-Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-on: https://review.coreboot.org/16698
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Several of the special function pins we're using in firmware have a
pre-assigned pull-up or pull-down on power-on reset. We don't want those
to interfere with any of the signaling we're trying to do on those pins,
so this patch disables them.
Also do some house-cleaning to group the bootblock code better, and
change the setup code for all SPI and I2C buses to first initialize the
controller and then mux the pins... I assume this might be a little
safer (in case the controller peripheral has some pins in a weird state
before it gets fully initialized, we don't want to mux it through too
early).
BRANCH=None
BUG=chrome-os-partner:52526
TEST=Booted Kevin.
Change-Id: I4d5bd3f7657b8113d90b65d9571583142ba10a27
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: f8f7fd56e945987eb0b1124b699f676bc68d0560
Original-Change-Id: I6bcf2b9a5dc686f2b6f82bd80fc9a1a245661c47
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/382532
Reviewed-on: https://review.coreboot.org/16711
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Increase the SPI bus speed to speed up boot time. The maximum supported
speed at 1.8V is 37.5MHz, and 33MHz is the next lowest convenient speed,
given the clock parents.
BUG=chrome-os-partner:56556
BRANCH=none
TEST=boot on gru and see that things still work correctly. Total time
spent on reading from SPI reduces from 185ms to 141ms.
Change-Id: I71436c9e343b18360fa63d528dea5cfcfbc831e6
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: d7576f6e53e407af61160be142c3d589e864a8cf
Original-Change-Id: I55a19f523817862e081d23469e94fd795456dd67
Original-Signed-off-by: Simon Glass <sjg@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/381313
Original-Commit-Ready: Julius Werner <jwerner@chromium.org>
Original-Tested-by: Simon Glass <sjg@google.com>
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/16708
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
