Since mono_time is now 64-bit, the utility functions interfacing with
mono_time should also be 64-bit so precision isn't lost.
Fixed build errors related to printing the now int64_t result of
stopwatch_duration_[m|u]secs in various places.
BUG=b:237082996
BRANCH=All
TEST=Boot dewatt
Change-Id: I169588f5e14285557f2d03270f58f4c07c0154d5
Signed-off-by: Rob Barnes <robbarnes@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/66170
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Add option to resource allocator v4 that restores the top-down
allocation approach at the domain level.
This makes it easier to handle 64-bit resources natively. With
the top-down approach, resources that can be placed either above
or below 4G would be placed above, to save precious space below
the 4G boundary.
Change-Id: Iaf463d3e6b37d52e46761d8e210034fded58a8a4
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41957
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Reviewed-by: Subrata Banik <subratabanik@google.com>
CL:3825558 changes all vb2_digest and vb2_hash functions to take a new
hwcrypto_allowed argument, to potentially let them try to call the
vb2ex_hwcrypto API for hash calculation. This change will open hardware
crypto acceleration up to all hash calculations in coreboot (most
notably CBFS verification). As part of this change, the
vb2_digest_buffer() function has been removed, so replace existing
instances in coreboot with the newer vb2_hash_calculate() API.
Due to the circular dependency of these changes with vboot, this patch
also needs to update the vboot submodule:
Updating from commit id 18cb85b5:
2load_kernel.c: Expose load kernel as vb2_api
to commit id b827ddb9:
tests: Ensure auxfw sync runs after EC sync
This brings in 15 new commits.
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: I287d8dac3c49ad7ea3e18a015874ce8d610ec67e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/66561
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jakub Czapiga <jacz@semihalf.com>
Aligning the "memory" ranges in devicetree is supposedly only needed on
very old arm32 kernels. So let's get rid of it.
Incidentally this fixes smaller than 1MB memory regions where the size
would end up being 0.
Change-Id: Ibbf5e331c79ed4ae3ed8dd37bf7a974d2412ce12
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/65607
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
There are too many "FIT" in firmware land. In order to reduce possible
confusion of CBFS_TYPE_FIT with the Intel Firmware Interface Table, this
patch renames it to CBFS_TYPE_FIT_PAYLOAD (including the cbfstool
argument, so calling scripts will now need to replace `-t fit` with `-t
fit_payload`).
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: I826cefce54ade06c6612c8a7bb53e02092e7b11a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64735
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Currently no smihandler uses heap.
coreboot's heap manager also is quite limited in what it will
free (only the latest alloc). This makes it a bad idea to use it inside
the smihandler, as depending on the alloc usage the heap might actually
be full at some point, breaking the smihandler.
This also reduces the ramstage by 448 bytes on google/vilboz.
Change-Id: I70cd822be17c1efe13c94a9dbd2e1038808b9c56
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64521
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Nico Huber <nico.h@gmx.de>
On some x86 targets it the bootblock is loaded via a different
mechanism, like via the AMD PSP or Intel IFWI. Some payloads need that
pointer so add it to cbfs.
Note that on Intel APL this file is not used, which is why the
bootblock still needs to contain the pointer in the ARCH_X86 part.
It is not worth it to add logic to specifically deal with APL as this is
a legacy feature anyway.
For AMD non-car platform this fixes cbfs access in SeaBIOS.
Change-Id: If46e80e3eed5cc3f59964ac58e507f927fc563c4
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64483
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
There are efforts to have bootflows that do not follow a traditional
bootblock-romstage-postcar-ramstage model. As part of that CBMEM
initialisation hooks will need to move from romstage to bootblock.
The interface towards platforms and drivers will change to use one of
CBMEM_CREATION_HOOK() or CBMEM_READY_HOOK(). Former will only be called
in the first stage with CBMEM available.
Change-Id: Ie24bf4e818ca69f539196c3a814f3c52d4103d7e
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63375
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Add 'lb_fill_pcie' function to pass PCIe information from coreboot to
libpayload, and add CB_ERR_NOT_IMPLEMENTED to the cb_err enum for the
__weak function.
ARM platform usually does not have common address for PCIe to access the
configuration space of devices. Therefore, new API is added to pass the
base address of PCIe controller for payloads to access PCIe devices.
TEST=Build pass and boot up to kernel successfully via SSD on Dojo
board, here is the SSD information in boot log:
== NVME IDENTIFY CONTROLLER DATA ==
PCI VID : 0x15b7
PCI SSVID : 0x15b7
SN : 21517J440114
MN : WDC PC SN530 SDBPTPZ-256G-1006
RAB : 0x4
AERL : 0x7
SQES : 0x66
CQES : 0x44
NN : 0x1
Identified NVMe model WDC PC SN530 SDBPTPZ-256G-1006
BUG=b:178565024
BRANCH=cherry
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: I6cdce21efc66aa441ec077e6fc1d5d1c6a9aafb0
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63251
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Shelley Chen <shchen@google.com>
Builds are failing sporadically with:
src/lib/master_header_pointer.c:5:10: fatal error: fmap_config.h: No such file or directory
5 | #include <fmap_config.h>
| ^~~~~~~~~~~~~~~
Correct the filename in the Makefile from header_pointer.c to
master_header_pointer.c so that there's a dependency from
master_header_pointer.c to fmap_config.h.
Change-Id: I41bcb2a21fdbc48f09d5b6be3e211ca56607d849
Signed-off-by: Reka Norman <rekanorman@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/64431
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Reviewed-by: Patrick Georgi <patrick@coreboot.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The makefiles don't like cbfs file names with spaces in them so update
the file name with '_' instead of spaces. To keep the master header at
the top of cbfs, add a placeholder.
This removes the need to handle the cbfs master header in cbfstool.
This functionality will be dropped in a later CL.
On x86 reserve some space in the linker script to add the pointer.
On non-x86 generate a pointer inside a C struct file.
As a bonus this would actually fix the master header pointer mechanism
on Intel/APL as only the bootblock inside IFWI gets memory mapped.
TESTED on thinkpad X201: SeaBIOS correctly finds the cbfs master
header.
Change-Id: I3ba01be7da1f09a8cac287751497c18cda97d293
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/59132
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
hardwaremain.c is the common ramstage entry to all platforms so move
out ACPI code generation (x86 specific) to boot state hooks.
Another reason to do this is the following:
On some platforms that start in dram it makes little sense to have
separate stages. To reduce the complexity we want to call the ramstage
main function instead of loading a full stage. To make this scheme
more maintainable it makes sense to move out as much functionality
from the 'main' function as possible.
Change-Id: I613b927b9a193fc076ffb1b2a40c617965ce2645
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63414
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Replace 'struct lb_uint64' with 'typedef __aligned(4) uint64_t
lb_uint64_t', and remove unpack_lb64/pack_lb64 functions since it's no
longer needed.
Also replace 'struct cbuint64' with 'cb_uint64_t' and remove
'cb_unpack64' in libpayload for compatible with lb_uint64_t.
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Change-Id: If6b037e4403a8000625f4a5fb8d20311fe76200a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/63494
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Prevents bad things from happening later when these new nodes are used.
This issue is hard to observe because:
1. Heap is zero-initialized, so you need to use allocated memory
filling it with non-zero values, free, allocate it again, use
uninitialized.
2. Most of allocated memory is not freed.
3. Implementation of free() does something only for one last malloc'ed
block, making most of freed memory unavailable for future
allocation.
Change-Id: I38a7ec1949d80f7a2564fac380ce94de6056a0c7
Signed-off-by: Sergii Dmytruk <sergii.dmytruk@3mdeb.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62928
Reviewed-by: Krystian Hebel <krystian.hebel@3mdeb.com>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This change will allow the SMI handler to write to the cbmem console
buffer. Normally SMIs can only be debugged using some kind of serial
port (UART). By storing the SMI logs into cbmem we can debug SMIs using
`cbmem -1`. Now that these logs are available to the OS we could also
verify there were no errors in the SMI handler.
Since SMM can write to all of DRAM, we can't trust any pointers
provided by cbmem after the OS has booted. For this reason we store the
cbmem console pointer as part of the SMM runtime parameters. The cbmem
console is implemented as a circular buffer so it will never write
outside of this area.
BUG=b:221231786
TEST=Boot non-serial FW with DEBUG_SMI and verified SMI messages are
visible when running `cbmem -1`. Perform a suspend/resume cycle and
verify new SMI events are written to the cbmem console log.
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Change-Id: Ia1e310a12ca2f54210ccfaee58807cb808cfff79
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62355
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Karthik Ramasubramanian <kramasub@google.com>
The ACPI RSDP can only be found in:
- legacy BIOS region
- via UEFI service
On some systems like ARM that legacy BIOS region is not an option, so
to avoid needing UEFI it makes sense to expose the RSDP via a coreboot
table entry.
This also adds the respective unit test.
Change-Id: I591312a2c48f0cbbb03b2787e4b365e9c932afff
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62573
Reviewed-by: Lance Zhao
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
cb_err_t was meant to be used in place of `enum cb_err` in all
situations, but the choice to use a typedef here seems to be
controversial. We should not be arbitrarily using two different
identifiers for the same thing across the codebase, so since there are
no use cases for serializing enum cb_err at the moment (which would be
the primary reason to typedef a fixed-width integer instead), remove
cb_err_t again for now.
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: Iaec36210d129db26d51f0a105d3de070c03b686b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62600
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
This patch aims to make timestamps more consistent in naming,
to follow one pattern. Until now there were many naming patterns:
- TS_START_*/TS_END_*
- TS_BEFORE_*/TS_AFTER_*
- TS_*_START/TS_*_END
This change also aims to indicate, that these timestamps can be used
to create time-ranges, e.g. from TS_BOOTBLOCK_START to TS_BOOTBLOCK_END.
Signed-off-by: Jakub Czapiga <jacz@semihalf.com>
Change-Id: I533e32392224d9b67c37e6a67987b09bf1cf51c6
Reviewed-on: https://review.coreboot.org/c/coreboot/+/62019
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Tim Wawrzynczak <twawrzynczak@chromium.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Now that the console system itself will clearly differentiate loglevels,
it is no longer necessary to explicitly add "ERROR: " in front of every
BIOS_ERR message to help it stand out more (and allow automated tooling
to grep for it). Removing all these extra .rodata characters should save
us a nice little amount of binary size.
This patch was created by running
find src/ -type f -exec perl -0777 -pi -e 's/printk\(\s*BIOS_ERR,\s*"ERROR: /printk\(BIOS_ERR, "/gi' '{}' ';'
and doing some cursory review/cleanup on the result. Then doing the same
thing for BIOS_WARN with
's/printk\(\s*BIOS_WARNING,\s*"WARN(ING)?: /printk\(BIOS_WARNING, "/gi'
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: I3d0573acb23d2df53db6813cb1a5fc31b5357db8
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61309
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Lance Zhao
Reviewed-by: Jason Glenesk <jason.glenesk@gmail.com>
In order to provide the same loglevel prefixes and highlighting that
were recently introduced for "interactive" consoles (e.g. UART) to
"stored" consoles (e.g. CBMEM) but minimize the amont of extra storage
space wasted on this info, this patch will write a 1-byte control
character marker indicating the loglevel to the start of every line
logged in those consoles. The `cbmem` utility will then interpret those
markers and translate them back into loglevel prefixes and escape
sequences as needed.
Since coreboot and userspace log readers aren't always in sync,
occasionally an older reader may come across these markers and not know
how to interpret them... but that should usually be fine, as the range
chosen contains non-printable ASCII characters that normally have no
effect on the terminal. At worst the outdated reader would display one
garbled character at the start of every line which isn't that bad.
(Older versions of the `cbmem` utility will translate non-printable
characters into `?` question marks.)
Signed-off-by: Julius Werner <jwerner@chromium.org>
Change-Id: I86073f48aaf1e0a58e97676fb80e2475ec418ffc
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61308
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Chromebooks normally run with non-serial enabled firmware because
writing to the UART console is very slow. This unfortunately makes
debugging boot errors more difficult. We tend to rely on port 80s and/or
the vboot recovery code.
When CONSOLE_CBMEM_DUMP_TO_UART is selected it will dump the entire
cbmem console to the UART whenever `vboot_reboot()` is called. We don't
incur any boot time penalty in the happy path, but still retain the
ability to access the logs when an error occurs.
The previous implementation was using a hard coded UART index and
`get_uart_baudrate` was always returning 0 since `CONFIG_TTYS0_BAUD`
wasn't defined. This change makes it so the UART console properties are
available when CONSOLE_CBMEM_DUMP_TO_UART is set. This results in the
following .config diff:
+CONFIG_UART_FOR_CONSOLE=0
+CONFIG_TTYS0_BASE=0x3f8
+CONFIG_TTYS0_LCS=3
+CONFIG_CONSOLE_SERIAL_115200=y
+CONFIG_TTYS0_BAUD=115200
This functionality is especially helpful on Guybrush. PSP Verstage is
run on S0i3 resume. Today, if there is an error, the cbmem console is
lost since it lives in the PSP SRAM.
BUG=b:213828947, b:215599230
TEST=Build non-serial guybrush FW and verify no serial output happens in
happy path. Inject a vboot error and perform an S0i3 suspend/resume.
Verify CBMEM console gets dumped to the correct UART.
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Change-Id: I997942204603362e51876a9ae25e493fe527437b
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61305
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Building libgfxinit with Debian’s toolchain – latest test with *gnat-11*
11.2.0-13 from Debian sid/unstable – the build fails with the error
below.
E: Invalid reloc type: 10
E: Unable to create rmodule from 'build/cbfs/fallback/ramstage.debug'.
Debian’s toolchain is built without enabling PIE by default.
So, explicitly pass `-fno-pie` to `ADAFLAGS_common` to be independent
from how the toolchain was built.
TEST=*gnat* 11.2.0-13 successfully. builds
purism/librem_cnl/variants/librem_mini with libgfxint.
With the coreboot toolchain `make BUILD_TIMELESS=1` produces the
same `build/coreboot.rom` for `BOARD_PURISM_LIBREM_MINI_V2=y` on
top of commit 50251400d2 (sb/intel/common/firmware: Reword
me_cleaner warning) with and without the change.
Change-Id: I6661937906d95c130c6099f598d61b21e958fd85
Signed-off-by: Paul Menzel <pmenzel@molgen.mpg.de>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/43759
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
When running in verstage before bootblock, the PSP (ARM co-processor) is
running with limited SRAM. It needs to stash the verstage console data
internally until DRAM is brought up. Once DRAM is brought up the data is
stashed in a "transfer buffer" region. In the current design, we are
using the same region for the transfer buffer and the
preram_cbmem_console region. This has the following downsides:
1) The pre-x86 buffer needs to be large enough to hold all the
verstage, bootblock and romstage console logs.
2) On AMD platforms, the PSP verstage is signed. Changing the size of
preram_cbmem_console after the fact will result in a mismatch of the
transfer buffer between verstage and bootblock.
This CL adds a new method that allows SoC specific code to copy the
CBMEM console in the transfer buffer to the active CBMEM console.
BUG=b:213828947
TEST=Boot guybrush and no longer see
*** Pre-CBMEM romstage console overflowed, log truncated!
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Change-Id: Idc0ab8090db740e0d1b3d21d8968f26471f2e930
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61099
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kangheui Won <khwon@chromium.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
This function is similar to cbmem_dump_console_to_uart except it uses
the normally configured consoles. A console_paused flag was added to
prevent the cbmem console from writing to itself.
BUG=b:213828947
TEST=Boot guybrush
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Change-Id: I3fe0f666e2e15c88b4568377923ad447c3ecf27e
Reviewed-on: https://review.coreboot.org/c/coreboot/+/61011
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Karthik Ramasubramanian <kramasub@google.com>
