The current method was only taking the cbfs path. Because
of this fsp.bin was never being utilized from the RW slots.
Using prog_locate() now provides both the cbfs and vboot
locate methods for free.
BUG=chrome-os-partner:44827
BRANCH=None
TEST=Built and booted glados.
Change-Id: I2b3e088326d5a965ad90806a7950b9f401ed57de
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11831
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Leave the SPI controller enabled upon boot block exit.
BRANCH=none
BUG=chrome-os-partner:44827
TEST=Build and run on kunimitsu
Change-Id: I5b10d7cc8d5d350282206abe6a945bab66f97ada
Signed-off-by: Lee Leahy <Leroy.P.Leahy@intel.com>
Reviewed-on: http://review.coreboot.org/11825
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Move base address into iomap.h. Use PCI symbols instead of SPI specific
symbols. Fix comments.
BRANCH=none
BUG=chrome-os-partner:44827
TEST=Build and run on kunimitsu
Change-Id: Id5d21603150b52fd1b71dd448105938bd6aff1a9
Signed-off-by: Lee Leahy <Leroy.P.Leahy@intel.com>
Reviewed-on: http://review.coreboot.org/11826
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
In order to support x86 verstage proper the work buffer
needs to live in cache-as-ram. However, after cache-as-ram
is torn down one still needs the verification results to
know which slot was selected. Though the platforms with
a dedicated SRAM can just use the work buffer in SRAM, the
x86 cache-as-ram platforms need a place to stash the
results. For that situation cbmem is employed. This works
because when cbmem is initialized cache-as-ram is still
enabled. The VBOOT_DYNAMIC_WORK_BUFFER case assumes
verified boot doesn't start until after cbmem is up. That
doesn't change, but it's a goal to get rid of that option
entirely once all other x86 platforms are moved over to
pre-romstage vboot.
BUG=chrome-os-partner:44827
BRANCH=None
TEST=Built and booted glados with pre-romstage verification
as well as VBOOT_DYNAMIC_WORK_BUFFER case.
Change-Id: I7eacd0edb2b6ca52b59b74075d17c00b50676d4c
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11821
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
On x86 the early stages are currently execute-in-place which
means they live in the memory-mapped spi flash. However, when
loading romstage from verstage the romstage is
execute-in-place so it's unnecessary to write over a read-only
media -- not to mention writing to read-only memory is wrong
to begin with.
BUG=chrome-os-partner:44827
BRANCH=None
TEST=Built and booted glados. Noted reduction of 20ms when
loading romstage.
Change-Id: I7cd399302a3925a05fbce82600b4c50ea66a0fcb
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11823
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Bump up the romstage size to allow more breathing room.
Change-Id: I4df7031d286c13797dccdf2f49d023bbf462fbb8
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11830
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The conditions in cbmem console for supporting verstage
were implicitly utilizing CONFIG_BOOTBLOCK_CONSOLE to handle
the cbmem console enablement. Fix it so verstage is a first
class citizen for deciding actions pertaining to cbmem console.
BUG=chrome-os-partner:44827
BRANCH=None
TEST=Built and booted glados using verstage. cbmem console
shows verstage output.
Change-Id: Iba79efd1c1d4056f1a105a5e10ffc95f3e69b597
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11820
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
For the purpose of isolating the work buffer logic
the surface area of the API was slimmed down. The
vb2_working_data structure is no longer exposed,
and the function signatures are updated accordingly.
BUG=chrome-os-partner:44827
BRANCH=None
TEST=Built and booted glados.
Change-Id: If64184a79e9571ee8ef9822cfce1eda20fceee00
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11818
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
For vboot1 there was an rmodule that was loaded and ran to
do the firmware verification. That's no longer used so remove
the last vestiges of VBOOT_STUB.
BUG=chrome-os-partner:44827
BRANCH=None
TEST=Built glados.
Change-Id: I6b41544874bef4d84d0f548640114285cad3474e
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11817
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
In order to introduce a verstage which performs vboot
verification the cache-as-ram environment needs to be
generalized and split into pieces that can be utilized
in romstage and/or verstage. Therefore, the romstage
pieces were removed from the cache-as-ram specific pieces
that are generic:
- Add fsp/car.h to house the declarations for functions in
the cache-as-ram environment
- Only have cache_as_ram_params which are isolated form the
cache-as-ram environment aside from FSP_INFO_HEADER.
- Hardware requirements for console initialization is done
in the cache-as-ram specific files.
- Provide after_raminit.S which can be included from a
romstage separated from cache-as-ram as well as one that
is tightly coupled to the cache-as-ram environment.
- Update the fallout from the API changes in
soc/intel/{braswell,common,skylake}.
BUG=chrome-os-partner:44827
BRANCH=None
TEST=Built and booted glados.
Original-Change-Id: I2fb93dfebd7d9213365a8b0e811854fde80c973a
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/302481
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Change-Id: Id93089b7c699dd6d83fed8831a7e275410f05afe
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11816
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The report_platform_info() and set_max_freq() are not being
used similarly on skylake and braswell. With the addition
of other SoCs I suspect a similar pattern will emerge. Instead
of having weak functions to ensure things link with the hardcoded
policy push these calls into their respective SoC homes.
For parity, both skylake and braswell were updated to be consistent
with the same calls prior to this patch.
BUG=chrome-os-partner:44827
BRANCH=None
TEST=Built and booted glados. Built braswell.
Original-Change-Id: I3371d09aff0629503254296955fef28d35754a38
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/303334
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Change-Id: I2de33632ed127cac52d7075cbad95cd6387a1b46
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11815
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Latest FSPv83 made some change related to UPD/VPD
need this patch to align those
BUG=None
TEST=Build and Boot Cyan System
BRANCH=strago-7287.B
CQ-DEPEND=CL:*226897
Original-Change-Id: I6395f3a1f4eecaef14fc4720b00252f9e6143fa3
Original-Signed-off-by: Subrata Banik <subrata.banik@intel.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/291394
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Tested-by: Hannah Williams <hannah.williams@intel.com>
Original-Commit-Queue: Hannah Williams <hannah.williams@intel.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/303137
Original-Commit-Ready: John Zhao <john.zhao@intel.com>
Original-Tested-by: John Zhao <john.zhao@intel.com>
Change-Id: I9920eea84b802699454850bfde489668201ffeb6
Signed-off-by: Subrata Banik <subrata.banik@intel.com>
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11813
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
On Skylake, mailbox interface is used to configure VRs, dropping direct msr
writing. With current fsp, svid/vr programming seems to be functional - no
errors are given in the svid transactions in boot, and hw engineer verified
the VRs on Kunimitsu. Additional tunnings might be needed later with power
testing.
24mhz calibration is no longer needed on Skylake due to bclk archtecture
change.
BRANCH=none
BUG=chrome-os-partner:45387
TEST=Built and boot on kunimitsu/glados, reboot, S3/resume verified.
Signed-off-by: robbie zhang <robbie.zhang@intel.com>
Original-Change-Id: If99b5758fcdba8604139c761a07403d4a5d2eb4c
Original-Reviewed-on: https://chromium-review.googlesource.com/301470
Original-Commit-Ready: Robbie Zhang <robbie.zhang@intel.com>
Original-Tested-by: Robbie Zhang <robbie.zhang@intel.com>
Original-Tested-by: Wenkai Du <wenkai.du@intel.com>
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Change-Id: I98acf78aac9c705614fb200f8c3313a89296fbf2
Signed-off-by: robbie zhang <robbie.zhang@intel.com>
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11811
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Instead of just passing bits, tsc_low, tsc_high, and an
opaque pointer to chipset context those fields are bundled
into a cache_as_ram_params struct. Additionally, a new
struct fsp_car_context is created to hold the FSP
information. These could be combined as the existing
romstage code assumes what the chipset_context values are, but
I'm leaving the concept of "common" alone for the time being.
While working in that area the ABI between assembly and C code
has changed to just pass a single pointer to cache_as_ram_params
struct. Lastly, validate the bootloader cache-as-ram region
with the Kconfig options.
BUG=chrome-os-partner:44676
BRANCH=None
TEST=Built and booted glados.
Original-Change-Id: Ib2a0e38477ef7c15cff1836836cfb55e5dc8a58e
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/300190
Original-Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Original-Reviewed-by: Leroy P Leahy <leroy.p.leahy@intel.com>
Change-Id: Ic5a0daa4e2fe5eda0c4d2a45d86baf14ff7b2c6c
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11809
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
It builds only on veyron_* which already select it, no need to ask user.
Change-Id: Ie508b9eade16e0f39073b23dc0da6b6d1e0a4c73
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/10380
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc@marcjonesconsulting.com>
board_id() returns an integer which is platform-specific. 0 for one port
is different from 0 for another port. So there is no default board_id()
and hence enabling it on boards other than urara would cause build failure.
Not enabling it on urara or just setting id to "(none)" as is default results
in board_id() = 0 which means urara and an error message on console.
Change-Id: I94618f36a75e7505984bbec345a31fe0fa9cc867
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/10379
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc@marcjonesconsulting.com>
Fixes linking error. Specifies that we're in text mode.
Change-Id: I7ad258961039c19e1491e2b3832b003671d8a5c7
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/11848
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc@marcjonesconsulting.com>
E.g. on my MacbookAir to generate spd.bin to be used
with coreboot I do:
./inteltool -S spd.bin
Change-Id: If165475ed3e1f3262a8926ef619128d25b1e2896
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/11847
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc@marcjonesconsulting.com>
MBa doesn't have a usable usbdebug port.
Change-Id: Ia8459daa5c9b9405c289954b28ecf1423b1f076c
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/11849
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc@marcjonesconsulting.com>
Update to commit 832bc6f1 (Remove microcode stored in C-array format),
which is the latest commit in the BLOBs repository.
Building the Lenovo X60 currently fails as the microcode file cannot be
found.
CREATE build/mainboard/lenovo/x60/cbfs-file.wkWhPK.out (from src/mainboard/lenovo/x60/cmos.default)
make: *** No rule to make target '3rdparty/blobs/cpu/intel/model_69x/microcode.bin', needed by 'build/cpu_microcode_blob.bin'. Schluss.
Change-Id: I40ebceec299f46c19fd60861d872adcd91df3610
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/11865
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Taken from cbfstool.
Change-Id: I4387900517dbfb1aa51ae6f679e26d0cf5b2acf8
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: http://review.coreboot.org/11808
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested on T60 with intel graphics.
Change-Id: Id74d0a1315749052e7313135242e6b64862aa5e1
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/5345
Tested-by: build bot (Jenkins)
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Only one value would work with corresponding gma code currently (which one
depends on board). Going forward, it's possible to compute which number can
be used, so there is no need to keep this info around.
Change-Id: Iadc77ef94b02f892860e3ae8d70a0a792758565d
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/11862
Tested-by: build bot (Jenkins)
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Based on the info by Felix Held.
Change-Id: Iab84dd8a0e3c942da20a6e21db5510e4ad16cadd
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/11857
Tested-by: build bot (Jenkins)
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
MBA has eDP and not LVDS, so it's not supported by our native init.
Change-Id: I489b7a98163b648f0e8000202117593c6b1aaf31
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/11842
Tested-by: build bot (Jenkins)
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
MBA has a soldered RAM without SPD, so you need to use stored SPD.
Change-Id: I0205e6c65ccbfe7764c12c815e60801a3c3623a5
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/11841
Tested-by: build bot (Jenkins)
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Just ran autoport on the data from MacBookAir4,2
Change-Id: Iba2a56a6846d81d29e6b090a9a31253ce240914d
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/11840
Tested-by: build bot (Jenkins)
Reviewed-by: Philipp Deppenwiese <zaolin@das-labor.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Make sure edge write test results are sane.
Check rn.all to make sure rn.start and rn.end are valid.
Most likely the following test is going to fail on the same
rank anyway.
Change-Id: Ifa601406e6c74ceb8d70063be5ce1bf6bc512c18
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/11247
Tested-by: build bot (Jenkins)
Reviewed-by: Vladimir Serbinenko <phcoder@gmail.com>
Don't disable PEG bits while turning on IGD.
Fixes PCI device enumeration of PEG devices.
Test system:
* Intel Pentium CPU G2130
* Gigabyte GA-B75M-D3H
Sidenote: This should be taken from a CMOS option instead.
Change-Id: I2d6522504e4404f2d57f9c319351d08317aefdcb
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/11058
Tested-by: build bot (Jenkins)
Reviewed-by: Vladimir Serbinenko <phcoder@gmail.com>
Activate PEG clock-gating only if all PEG devices are disabled.
Fixes system hang when trying to access PEG registers.
Test system:
* Intel Pentium CPU G2130
* Gigabyte GA-B75M-D3H
Change-Id: I7d62fbb83c16741965639cea1a0e4978d4e3d6da
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/11059
Tested-by: build bot (Jenkins)
Currently cbfstool would reject non-alpanumeric characters in
image names. Underscore is not alphanumeric and is used in some
default fmaps. This change allows image names to contain all
"printable" characters except spaces.
Change-Id: I6ba2b581d5623f5b028149ece0169892ea63fd04
Signed-off-by: Andrey Petrov <andrey.petrov@intel.com>
Reviewed-on: http://review.coreboot.org/11807
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Since we now have more freedom in the bootblock linking step it no
longer makes sense to use a monolithic bootblock.S. Code segments must
still be included as the order in bootblock.S determines code flow.
However, non-code flow related assembly stubs don't need to be directly
included in bootblock.S
Change-Id: I08e86e92d82bd2138194ed42652f268b0764aa54
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/11792
Tested-by: build bot (Jenkins)
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The code flow doesn't fall through to walkcbfs, as it does in the rest
of bootblock.S. Instead, walkcbfs is called (albeit via a jmp). The
linker cannot know this when walkcbfs.S is included directly.
When we use a CAR bootblock, we lose several hundred bytes because
walkcbfs is not garbage-collected, yet it isn't used. This problem
is solved by assembling walkcbfs.S separately, and linking it.
Change-Id: Ib3a976db09b9ff270b7677cb4f9db80b0b025e22
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/11785
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
As part of preparing for systems with non-memory-mapped media, we want
to be able to call into C code. This change allows us to link C code
directly into the bootblock. The steps of going from bootblock main()
to CAR setup to C code will be implemented in subsequent patches.
Note that a few files selected with bootblock-y will now be compiled
for the bootblock as well, but since we enabled garbage collection,
they will not be included in the final binary.
Change-Id: I5ca6dcaf176f5469c6a3bb925859399123493bc6
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/11783
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
The only difference between the ifeq/else/endif guarded rules is the
linker flags specific to x86. Add those flags to LDFLAGS_bootblock,
and only use one rule for bootblock.debug.
Change-Id: I986a93e0418f05fb273512d7efe0573052493332
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/11782
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The macro is defined in `util/cbmem/cbmem.c` too, so do the same here,
so that searching for that macro name shows all the usages.
Change-Id: I52e9fa414fbbe2012bc6d00312db528efba3e564
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/11803
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Tested-by: build bot (Jenkins)
Commit 47818b4d60
(fsp/cache_as_ram.inc and boards: Fix incorrect usage of POST_IO)
breaks the logic which decides whether FSP
could be found or not in cache_as_ram.inc.
Fix the error by inverting the logic of the test.
TEST=Bootet mc_tcu3 board
Change-Id: I993d3422ac406d204a53e4dc890210fb9a52469d
Signed-off-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-on: http://review.coreboot.org/11806
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Now that cbfs is adding more metadata in the cbfs file
header one needs to access that metadata. Therefore,
add struct cbfsf which tracks the metadata and data
of the file separately. Note that stage and payload
metadata specific to itself is still contained within
the 'data' portion of a cbfs file. Update the cbfs
API to use struct cbfsf. Additionally, remove struct
cbfsd as there's nothing else associated with a cbfs
region aside from offset and size which tracked
by a region_device (thanks, CBFS_ALIGNMENT!).
BUG=None
BRANCH=None
TEST=Built and booted through end of ramstage on qemu armv7.
Built and booted glados using Chrome OS.
Change-Id: I05486c6cf6cfcafa5c64b36324833b2374f763c2
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/11679
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The x86 bootblock linking is a mess. The bootblock is treated in
a very special manner, and never received the update to link-time
garbage collection.
On newer x86 platforms, the boot media is no longer memory-mapped.
That means we need to do a lot more setup in the bootblock. ROMCC is
unsuitable for this task, and walkcbfs only works on memory-mapped
CBFS. We need to revise the x86 bootflow for this new case.
The approach this patch series takes is to perform CAR setup in the
bootblock, and load the following stage (either romstage or verstage)
from the boot media. This approach is not new, but has been done on
our ARM ports for years.
Since we will be adding .c files to the bootblock, it is prudent to
use link-time garbage collection. This is also consistent to how we
do things on other architectures. Unification FTW!
Change-Id: I16b78456df56e0053984a9aca9367e2542adfdc9
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/11781
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Up to now, the multi-CBFS code path merely looked up files in the "boot
ro" image (ie. the default), disregarding the specified fmap region to
use for CBFS.
The code still relies on the master header being around, which on the
upside allows it to skip an offset at the beginning of the region (eg.
for ARM bootblocks).
This will change later (both the reliance on the master header and the
presence of the bootblock like this).
Change-Id: Ib2fc03eac8add59fc90b4e601f6dfa488257b326
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: http://review.coreboot.org/11805
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
POST_IO is a user-visible config bool. fsp_1_0/cache_as_ram.inc made a
mess of it, by forcing a build-time error when CONFIG_POST_IO was not
being set. fsp 1.0 boards ended 'select'ing this in their Kconfig.
Refactor fsp/cache_as_ram.inc handling of POST codes, and remove the
"select POST_IO" from boards that have it. Instead of implementing an
ad-hoc changing post code display and a delay based on port 0xed, just
encode the FSP failure code in the POST code. Since FSP failure codes
are > 16, we can encode the failure code in the lower nibble, and theirfailing function in the upper nibble.
Change-Id: Iaa3e6533e8406b16ec0689abd704984d79293952
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/8485
Tested-by: build bot (Jenkins)
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
There is no other guard to prevent this from being picked up when
building for other architectures.
Change-Id: I2039a289a4dd9970d5dd0f90d43d5d5c2a6d0a0b
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/11795
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The EM100Pro allows the debug console to be sent over the SPI bus.
This is not yet working in romstage due to the use of static variables
in the SPI driver code. It is also not working on chipsets that have
SPI write buffers of less than 10 characters due to the 9 byte
command/header length specified by the EM100 protocol.
While this currently works only with the EM100, it seems like it would
be useful on any logic analyzer with SPI debug - just filter on command
bytes of 0x11.
Change-Id: Icd42ccd96cab0a10a4e70f4b02ecf9de8169564b
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: http://review.coreboot.org/11743
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The binutils-2.25 has added some new line, making the hunk
move downward a little. The utility patch can fix the offset
with "fuzz" message. So, recreate the patch to avoid that
message.
Change-Id: Ie659a8faf923465f6d47f7c0c0bf903c5eb903ab
Signed-off-by: Zheng Bao <zheng.bao@amd.com>
Signed-off-by: Zheng Bao <fishbaozi@gmail.com>
Reviewed-on: http://review.coreboot.org/11639
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Aaron Durbin