The GPIO configuration matches the one from the original BIOS.
Some configs don't make much sense, but I kept it as is so it
would match (such as a NC pin with TX set to 1, or RXINV enabled).
Remove unnecessary early GPIO config.
Change-Id: Iaec8630cef9a523fb2e2503143aa4aa72fbedc1f
Signed-off-by: Youness Alaoui <youness.alaoui@puri.sm>
Reviewed-on: https://review.coreboot.org/19934
Reviewed-by: Matt DeVillier <matt.devillier@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Like other devices using ENE embedded controllers, the librem13v2
requires this config option for the PS2 keyboard and mouse
(trackpad) to function properly.
Change-Id: Ifba13b93a1fe2e76b2790d1c273fd9e2b5368ab0
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/19933
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Add the initial directory for the port of the Librem 13 v2.
The base implementation was copied from the google/chell directory
and the chell references were replaced. spd directory was removed
since the RAM is not soldered on the MB. The Kconfig, board_info.txt
and devicetree.cb files were modified to match the Librem 13 v2
hardware information. The romstage.c, mainboard.c, Makefile.in and
dsdt.asl were modified to remove chromeos specific code. The boardid.c,
chromeos.c, chromeos.fmd, cmos.layout, ec.c, ec.h and smihandler.c
files were removed from the tree, and the acpi directory was replaced
with the acpi directory from the purism/librem13 board.
These changes allow us to remove the references to chromeos specific
code and allow coreboot to compile when the librem13v2 board is selected.
Change-Id: I24263fde18fcea70163dbdc59df6ea1d98c97af8
Signed-off-by: Youness Alaoui <youness.alaoui@puri.sm>
Reviewed-on: https://review.coreboot.org/19932
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
The Role-Based Error Reporting is not a configurable field,
it's a read only field in the Device Capability register.
This code was old and comes from commit f6eb88ad but evidently
is not useful in any way. The PCIe Specification [1] states
that it's read-only and must always be set to 1.
I have also done tests on purism/librem13 hardware, trying to
change that value, without any success.
[1]: PCI Express Base Specification Revision 3.0
Page 612
Change-Id: I729617a5c6f4f52dfc4c422df78379b309066399
Signed-off-by: Youness Alaoui <youness.alaoui@puri.sm>
Reviewed-on: https://review.coreboot.org/19901
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Philippe Mathieu-Daudé <philippe.mathieu.daude@gmail.com>
Reviewed-by: Martin Roth <martinroth@google.com>
The ASPM options (L1 substates, CLKREQ support, Common Clock and ASPM)
are hardcoded for broadwell chips, but some boards may not support
these ASPM options even if the SoC does support it (non-wired CLKREQ
pin for example).
This is required to disable L1 substates on the Purism/Librem 13 which
seems to have issues with NVMe drives falling into L1.2 state and not
being able to exit that state.
Change-Id: I2c7173af1d482cccdc784e3fa44ecbb5d38ddc34
Signed-off-by: Youness Alaoui <youness.alaoui@puri.sm>
Reviewed-on: https://review.coreboot.org/19899
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
The PCIe specification[1] describes a race condition that
can occur when using the Retrain Link bit in the Link
Control Register.
The race condition is avoided by checking the retrain link
bit in the link status register and waiting until it is
set to 0, before initiating a new link retraining.
[1] PCI Express Base Specification Revision 3.0
Page 633
Change-Id: I9d5840fb9a6e63838b5a4084d3bbe483f1d870ed
Signed-off-by: Youness Alaoui <youness.alaoui@puri.sm>
Reviewed-on: https://review.coreboot.org/19556
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Martin Roth <martinroth@google.com>
A simple rename of the directory and the config values
and string in Kconfig/Kconfig.name/board_info.txt
It will be less confusing for users since the first models
are referred to as 'v1' everywhere now.
Change-Id: I23fa977717230c2001868741bb684e9633a2c0bb
Signed-off-by: Youness Alaoui <youness.alaoui@puri.sm>
Reviewed-on: https://review.coreboot.org/19931
Reviewed-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Adapted from Chromium commit 8fbe1e7 for soc/braswell
(also review.coreboot.org/#/c/20060/); same issue affects
baytrail as well.
This patch recollects TOLM accessing; as Aaron recalled some
core_msr_script turns off access to TOLM register, he suggests
to store tolm to avoid getting back a zero while setting acpi
nvs space.
Change-Id: Ib26d4fe229b3f7d8ee664f5d89774d1f4a997f51
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/20081
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Each I2C controller should have a unique pair of DMA request lines,
and DMA channels should be assigned incrementally, rolling over as
necessary.
Source: Intel Baytrail/ValleyView UEFI reference code
Change-Id: Icc9b27aaa14583d11d325e43d9165ddda72ca865
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/20080
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Each I2C controller should have a unique pair of DMA request lines,
and DMA channels should be assigned incrementally, rolling over as
necessary.
Source: Intel Braswell UEFI reference code
Change-Id: I1d97b5a07bf732c27caf57904c138b120b93ca81
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/20079
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Commit 7c634ae8 (msrtool: added support for Intel CPUs) adds `no-pic` to
the compiler flags.
GCC 7.0.1 20170316 fails to built with the error below.
```
/usr/bin/ld: msrtool.o: relocation R_X86_64_32 against `.rodata.str1.1' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: msrutils.o: relocation R_X86_64_32 against `.rodata.str1.1' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: sys.o: relocation R_X86_64_32 against `.rodata.str1.1' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: linux.o: relocation R_X86_64_32 against `.rodata.str1.1' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: freebsd.o: relocation R_X86_64_32S against `.data' can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: final link failed: Nonrepresentable section on output
```
Removing the flag causes the build to succeed with GCC 7, 6.3, and clang
4.0.
Change-Id: I3d7aed27ce7f84aa27305c68e2d5f14607c58ec8
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: https://review.coreboot.org/18907
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Stefan Tauner <stefan.tauner@gmx.at>
Reviewed-by: David Hendricks <david.hendricks@gmail.com>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
For an unknown reason, the I2C ACPI devices were placed
under \SB intead of \SB.PCI0, as with all other non-Atom
based Intel platforms. While Linux is tolerant of this,
Windows is not. Correct by moving I2C ACPI devices where
they belong.
Also, adjust I2C devices at board level for intel/strago
and google/cyan as to not break compilation.
Change-Id: Iaf8211bd86d6261ee8c4d9c4262338f7fe19ef43
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: https://review.coreboot.org/20055
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Switch from lapic to tsc.
Allows timestamps to be used in coreboot, as there's a reference
clock available to calculate correct time units.
Clean Kconfig, remove duplicated lapic code and include tsc dir for
LGA1155 boards.
Tested on Lenovo T430.
Change-Id: I849ca2b3908116d9d22907039cd6e4464444b1d1
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
Reviewed-on: https://review.coreboot.org/20044
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The RTC RX6110SA has several configuration options which might be
interesting to set. To make this setup independent of the driver itself
but let it still be configurable on mainboard level, add more
configuration options to the chip driver.
Change-Id: I7f8b2aa7cd001a887f271be36f655e10e60e778b
Signed-off-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-on: https://review.coreboot.org/20084
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Mario Scheithauer <mario.scheithauer@siemens.com>
GCC version 7 is being a bit picky about pointer and integer comparison
by default, which triggers a crossgcc build error.
This backports a patch from upstream GCC to fix the issue.
Change-Id: I8b1e806c10604c0df080ac5edc667bf1141e2c17
Signed-off-by: Paul Kocialkowki <contact@paulk.fr>
Reviewed-on: https://review.coreboot.org/20103
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The new config choice is called RUN_FSP_GOP. Some things had to happen
on the road:
* Drop confusing config GOP_SUPPORT,
* Add HAVE_FSP_GOP to chipsets that support it,
* Make running the GOP an option for FSP2.0 by returning 0
in random VBT getters.
Change-Id: I92f88424004a4c0abf1f39cc02e2a146bddbcedf
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: https://review.coreboot.org/19815
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Provide all gfx init methods as a Kconfig `choice`. This elimates the
option to select native gfx init along with running a Video BIOS. It's
been only theoretically useful in one corner case: Hybrid graphics
where only one controller is supported by native gfx init. Though I
suppose in that case it's fair to assume that one would use SeaBIOS to
run the VBIOS.
For the case that we want the payload to initialize graphics or no
pre-boot graphics at all, the new symbol NO_GFX_INIT was added to the
choice. If multiple options are available, the default is chosen as
follows:
* NO_GFX_INIT, if we add a Video BIOS and the payload is SeaBIOS,
* VGA_ROM_RUN, if we add a Video BIOS and the payload is not SeaBIOS,
* NATIVE_VGA_INIT, if we don't add a Video BIOS.
As a side effect, libgfxinit is now an independent choice.
Change-Id: I06bc65ecf3724f299f59888a97219fdbd3d2d08b
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: https://review.coreboot.org/19814
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
MAINBOARD_FORCE_NATIVE_VGA_INIT is to be selected instead of the user
option MAINBOARD_DO_NATIVE_VGA_INIT. The distinction is necessary to
use the latter in a choice.
Change-Id: I689aa5cadea9e1091180fd38b1dc093c6938d69c
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: https://review.coreboot.org/19813
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Update libgfxinit to the latest master. Changes:
* Remove trailing whitespace in debug output.
* Change some types to make it verify with SPARK Pro.
* Add Broxton (Apollo Lake) support for eDP/DP/HDMI.
* Add Linux user-space test tool `gfx_test`.
* Add a README describing libgfxinit and the build process.
TEST=Booted lenovo/t420 and verified that internal and
external displays are working.
Change-Id: I4d0e23b8a254234173461b831585eae58d3af58e
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: https://review.coreboot.org/20091
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Update libhwbase to the current master. Some noteworthy changes:
* Add prerequisites for upcoming Apollo Lake support in libgfxinit.
* Add some support for Linux user-space for libgfxinit's `gfx_test`.
* Fix compilation with GCC 7.
Change-Id: If3c65065ef9a2ff6fce221939fda43c9e30c1eb8
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: https://review.coreboot.org/20090
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
As of Change-Id: I780d34ded2c1e3737ae1af685c8c2da832842e7c the
reference clock can be 100Mhz.
Decode the register and use the reference clock to calculate
the selected DDR frequency.
Tested on Lenovo T430.
Change-Id: I8481564fe96af29ac31482a7f03bb88f343326f4
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
Reviewed-on: https://review.coreboot.org/19995
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Fix regression introduced by commit 5c026445
(drivers/intel/wifi: Add support for generating SSDT table)
In case the regular PCI path is taken, there're no chip_ops and the code
will segfault. The bug was covered by other bugs that caused this code
to never execute.
Add NULL pointer checks and only fill in device name if one is provided.
Tested on Lenovo T430 and wifi card 8086:0085.
Change-Id: I84e804f033bcd3af1a7f76670275fdf5159d381f
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
Reviewed-on: https://review.coreboot.org/20082
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Because the help block uses significant whitespace to determine whether
or not text is inside the help block, a mixture of spaces and tabs
confuses the parser.
If there's an unrecognized line, and the previous line was inside a help
block, it's likely that this line is too.
Additionally, this was found with a line that started ' configuration',
and threw a perl warning about an uninitialized value because the parser
thought this was the start of a new config line, but couldn't find the
symbol. Now we make sure that config statements have whitespace after
the 'config' statement.
Change-Id: I46375738a18903b266ea9fff3102a1a91235e609
Signed-off-by: Martin Roth <gaumless@gmail.com>
Reviewed-on: https://review.coreboot.org/19155
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
- Turn the check for help text with no indentation from a warning to
an error.
- Show an error if the help text is at the same indentation level as
the 'help' keyword.
Change-Id: Ibf868c83e2a128ceb6c4d3da7f2cf7dc237054e6
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/19851
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
This patch fixes ACPI debug print issue reported internally
while using APRT asl method. Potentially some junk characters
gets added into final print buffer due to LPSS MMIO register
space is 32 bit width and ADBG is one byte at a time.
TEST=Built and boo eve to ensure to be able to get ASL console
log without any corruption.
Change-Id: I0b6af789c0ffc79f7fee4652b4aa6a125b182296
Signed-off-by: Subrata Banik <subrata.banik@intel.com>
Reviewed-on: https://review.coreboot.org/20009
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The coreboot sites support HTTPS, and requests over HTTP with SSL are
also redirected. So use the more secure URLs, which also saves a
request most of the times, as nothing needs to be redirected.
Run the command below to replace all occurences.
```
$ git grep -l -E 'http://(www.|review.|)coreboot.org'
| xargs sed -i 's,http://\(.*\)coreboot.org,https://\1coreboot.org,g'
```
Change-Id: If53f8b66f1ac72fb1a38fa392b26eade9963c369
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: https://review.coreboot.org/20034
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The comma-separated PCI vendor and device ID is used to associate the
VGA BIOS to the video device by using it as the file name of the VGA
Option ROM.
Change-Id: I755554eeb9a560d034d6e8fe49de619d800ea045
Signed-off-by: Paul Menzel <pmenzel@molgen.mpg.de>
Reviewed-on: https://review.coreboot.org/18741
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Device ID is read from HP Elitebook 2760p.
Based on:
- superio/smsc/kbc1100 (LDNs, keyboard, EC)
- DSDT from OEM firmware (COM1 and mailbox)
- Datasheet "KBC1122 Priliminary DS Rev. 0.8"
Change-Id: Id172ae42411a6d42a4ae7c7f30f96aeda3e6c384
Signed-off-by: Iru Cai <mytbk920423@gmail.com>
Reviewed-on: https://review.coreboot.org/18480
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
GCC7 has a new feature called -Wimplicit-fallthrough enabled by
default which checks for fallthrough in switch statements which is a
common error. When a fallthrough is actually intended a comment saying
so will satisfy GCC.
Fixes cbfstool not building with GCC7.
Change-Id: I83252fc96be7ce0971d4251b0fc88fbbd7440e71
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/20036
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
