Use monotonic timer to accumulate the time spent in
console code.
For bootblock and romstage, only stage total is reported.
For ramstage each boot_state is reported individually.
Change-Id: Id3998bab553ff803a93257a3f2c7bfea44c31729
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/36574
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>
This has nothing to do with console options.
This also improves the help text to reflect what it actually does.
Change-Id: I039f4f6bbe144769d6a362192b225838ed3d9d43
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/32042
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>
Link files to provide snprintf used by VBOOT code.
Change-Id: I040c3952c22893da5aae11b20a618aa4006c6c58
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/36053
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 discussed on the mailing list and voted upon, the coreboot project
is going to move the majority of copyrights out of the headers and into
an AUTHORS file. This will happen a bit at a time, as we'll be unifying
license headers at the same time.
Signed-off-by: Martin Roth <martin@coreboot.org>
Change-Id: I2f350cc3008b17516b5a42cdf07e28d2da5995e9
Reviewed-on: https://review.coreboot.org/c/coreboot/+/35179
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
ENV_ROMSTAGE_OR_BEFORE is a direct replacement for testing
defined(__PRE_RAM__) as a true statement instead of with the
help of the preprocessor.
Note that for x86, due to existence of ENV_POSTCAR and ENV_SMM,
ENV_ROMSTAGE_OR_BEFORE and ENV_RAMSTAGE are not the inverse of
each other.
Change-Id: Ibd2292f922ccb9e79d10ca9bc35797048d174287
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34939
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Almost all platforms force it on. Make it enabled by
default but under user control to optionally disable it.
Change-Id: I6b0f19c8bfd6ffed93023d57a1d28ca6acc06835
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34803
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Martin Roth <martinroth@google.com>
Implicit fall throughs are a perpetual source of bugs and Coverity Scan
issues, so let's squash them once and for all. GCC can flag implicit fall
throughs using the -Wimplicit-fallthrough warning, and this should
ensure no more enter the code base. However, many fall throughs are
intentional, and we can use the following comment style to have GCC
suppress the warning.
switch (x) {
case 1:
y += 1;
/* fall through */
case 2:
y += 2;
/* fall through - but this time with an explanation */
default:
y += 3;
}
This patch adds comments for all remaining intentional fall throughs,
and tweaks some existing fall through comments to fit the syntax that
GCC expects.
Change-Id: I1d75637a434a955a58d166ad203e49620d7395ed
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34297
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
It doesn't make sense to print these values for memory-mapped UARTs.
Change-Id: Ie2d9cf95f0b0fdcf601e74de799b1390c08f2335
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34247
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Commit b19946cc62 (console: Remove support for printing extra bases)
truncated the digits string to only print integers of up to base 16.
However, that string was also used to print the leading 'x' or 'X' for
hexadecimal integers and is now too short. Fix this to prevent an out
of bounds read.
Change-Id: Iab6470cc88f445f074cf7c0b675346b37f3f2375
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1402999
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34211
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Julius Werner <jwerner@chromium.org>
No reason why the files could not be used with ASEG.
Attempts to use malloc() from ASEG would still fail,
though, due the lack of heap.
Change-Id: Idf470ae84eb34c442e833925510b08d5314e7638
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34126
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
It is occasionally useful to print a uintmax_t or intmax_t, so add
support for the j specifier. This also makes defining the PRI* macros
in <inttypes.h> simpler.
Change-Id: I656e3992029199b48e62a9df2d56f54c34e4e10f
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34027
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>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
vtxprintf() can only print numbers in base 8, 10, and 16, so the
extra letters in the alphabet aren't needed.
Change-Id: I6a51c13f3298a597e801440f86bf698bdd8c736a
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34028
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Sometimes coreboot needs to compile external code (e.g.
vboot_reference) using its own set of system header files. When
these headers don't line up with C Standard Library, it causes
problems.
Create ctype.h header file. Relocate ctype.h functions from
string.h into ctype.h. Update source files which call ctype.h
functions accordingly.
Note that ctype.h still lacks five functions which are not used
in coreboot source:
isalnum, isalpha, iscntrl, isgraph, ispunct
BUG=b:124141368
TEST=make clean && make test-abuild
BRANCH=none
Change-Id: I31b5e8af49956ec024a392a73c3c9024b9a9c194
Signed-off-by: Joel Kitching <kitching@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33525
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
The prototype of vprintk() is currently declared unconditionally, which
prevents it from being used in situations where the console is disabled.
The code will compile correctly, but not link, since the definition in
console.c isn't being provided. This adds a shim around the declaration
so that, like printk(), a call to vprintk() in this situation will expand
to a no-op function instead.
Change-Id: Ib4a9aa96a5b9dbb9b937ff45854bf6a407938b37
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33181
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
die() currently only accepts a fixed message string, which is rather
inconvenient when there is extra information that would be helpful to
print in the error message. This currently requires an extra call to
printk(), which is somewhat awkward:
printk(BIOS_EMERG, "Bad table, opcode %d at %d", id, i);
die(""); // what do I say here?
die() already has a printk() inside it to print the error message, so
let's just make it variadic to combine the two.
die("Bad table, opcode %d at %d", id, i); // much better
Forwarding variadic arguments from one function to another is rather
tricky, so die_with_post_code() is redefined as a variadic macro
instead.
Change-Id: I28b9eac32899a1aa89e086e0d3889b75459581aa
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33153
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: Keith Short <keithshort@chromium.org>
vtxprintf.c seems to have been written before string.h was as fleshed
out as it is today -- this patch removes some custom implementation of
stuff we now have globally. It also makes the skip_atoi() function
globally available, because I need it somewhere else, and while we maybe
don't want a huge fully-featured string parsing library in coreboot,
being able to parse an integer is occasionally useful.
Change-Id: Iecb2b970aecfc768540d2bf8b3023445f54853a4
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/32858
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Alex Thiessen <alex.thiessen.de+coreboot@gmail.com>
Add a new post code POST_HW_INIT_FAILURE, used when coreboot fails to
detect or initialize a required hardware component.
BUG=b:124401932
BRANCH=sarien
TEST=build coreboot for sarien and arcada platforms
Change-Id: I73820d24b3e1c269d9d446a78ef4f97e167e3552
Signed-off-by: Keith Short <keithshort@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/32774
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Rudolph <siro@das-labor.org>
Now that we call post_code in other stages other than RAMSTAGE, we need
to guard the elog calls with the appropriate condition in order to
compile correctly.
Change-Id: I766c276f28d46492fb05e0e3be71853e21f4e8e0
Signed-off-by: Jett Rink <jettrink@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/32914
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Add a new helper function die_with_post_code() that generates a post
code and an error string prior to halting the CPU.
BUG=b:124401932
BRANCH=sarien
TEST=build coreboot for sarien and arcada platforms
Change-Id: I87551d60b253dc13ff76f7898c1f112f573a00a2
Signed-off-by: Keith Short <keithshort@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/32838
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This patch fixes up all code that would throw a -Wtype-limits warning.
This sometimes involves eliminating unnecessary checks, adding a few odd
but harmless casts or just pragma'ing out the warning for a whole file
-- I tried to find the path of least resistance. I think the overall
benefit of the warning outweighs the occasional weirdness.
Change-Id: Iacd37eb1fad388d9db7267ceccb03e6dcf1ad0d2
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/32537
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This patch is a raw application of
find src/ -type f | xargs sed -i -e 's/IS_ENABLED\s*(CONFIG_/CONFIG(/g'
Change-Id: I6262d6d5c23cabe23c242b4f38d446b74fe16b88
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/31774
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
For fast CBMEM console use minimum BIOS_DEBUG level.
For other consoles, Kconfig and/or nvram settings
apply.
Change-Id: Iff56a0a3182f258200cac80e013957d598cc2130
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/31370
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
With PCI_DEV() always defined it is no longer
necessary to exclude this code from building.
Change-Id: I58a6348750d240aa6024599f7b1af1449f31e8ac
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/31303
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Sometimes, it’s not clear, what log level is configured (in Kconfig and
CMOS), so print the log level in the banner.
coreboot-4.9-354-gff6e99cebe Tue Jan 15 15:23:20 UTC 2019 bootblock starting (log level: 7)...
Change-Id: I82c87ae90cd53fd47458fc6df3ef2c7f238f0f3d
Signed-off-by: Paul Menzel <pmenzel@molgen.mpg.de>
Reviewed-on: https://review.coreboot.org/c/30935
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Add a new Kconfig NO_BOOTBLOCK_CONSOLE to disable the BOOTBLOCK_CONSOLE
option completely. The commit message of fbb11cf (ARM: Separate the
early console (romstage) from the bootblock console.) states that it
doesn't work before romstage on Exynos 5420.
Change-Id: I9b56a52f2555b5233300f27031a9ef50e7ab7cea
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/30926
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Invert the default instead of selecting it everywhere. Restores the
ability to use its Kconfig prompt.
Beside Qemu targets, the only platforms that didn't select it seem
to be samsung/exynos5420, intel/cannonlake, and intel/icelake. The
latter two were about to be patched anyway.
Change-Id: I7c5b671b7dddb5c6535c97c2cbb5f5053909dc64
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/30891
Reviewed-by: Lijian Zhao <lijian.zhao@intel.com>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Under normal circumstances no printk() goes through until
console_hw_init() has completed. This is wanted behaviour,
except when you need to debug the setup of one of consoles.
Change-Id: Ifc2bb22bf930009ee229d4461f512ada3018307b
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/30558
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
The Kconfig declaration for FIXED_UART_FOR_CONSOLE was accidentally
placed inside an `if CONSOLE_SERIAL` in a96e66a (soc/intel: Clean mess
around UART_DEBUG).
TEST=Start a clean config, select intel/leafhill and disable serial
console. Confirm that config can be saved without error.
Change-Id: Ie41687e91af11a13697cbe25938dada2c74b40fb
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/30829
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Mario Scheithauer <mario.scheithauer@siemens.com>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Everything is wrong here, the Kconfig symbols are only the tip of the
iceberg. Based on Kconfig prompts the SoC code performed pad configu-
rations! I don't see why the person who configures coreboot should have
the board schematics at hand.
As a mitigation, we remove the prompts for UART_DEBUG, which is renamed
to INTEL_LPSS_UART_FOR_CONSOLE (because the former didn't really say
what it's about), and for UART_FOR_CONSOLE in case the former is selec-
ted.
Change-Id: Ibe2ed3cab0bb04bb23989c22da45299f088c758b
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/29573
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
The feature is used to enable PCI MMIO accesses behind
PCIe links (or bridges) before PCI enumeration has been
completed.
Add the feature for bootblock, verstage and postcar, it
is required with add-on PCIe serial cards for early
console output. It's up to the board specific code to
configure PCIe root port prior to calling console_init()
for this to work.
Remove feature from ramstage, it bypasses any resource
allocations and bus number assignments.
For the moment PCI configuration support before ramstage
is available only on ARCH_X86.
Also switch from device_t to pci_devfn_t.
Change-Id: I08acec68b6f17f4d73d30039cc41274492ea4f45
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/30496
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Remove all cases in code where we tested for
EARLY_CBMEM_INIT or LATE_CBMEM_INIT being set.
This also removes all references to LATE_CBMEM_INIT
in comments.
Change-Id: I4e47fb5c8a947d268f4840cfb9c0d3596fb9ab39
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/26827
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The dependencies of CONSOLE_SERIAL and DRIVERS_UART were somehow
backwards. Fix that. Now, CONSOLE_SERIAL depends on DRIVERS_UART,
because it's using its interface. The individual UART drivers
select DRIVERS_UART, because they implement the interface and
depend on the common UART code.
Some guards had to be fixed (using CONSOLE_SERIAL now instead of
DRIVERS_UART). Some other guards that were only about compilation
of units were removed. We want to build test as much as possible,
right?
Change-Id: I0ea73a8909f07202b23c88db93df74cf9dc8abf9
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/29572
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
CB:26053 changed coreboot's default loglevel from SPEW to DEBUG. This
may be the most reasonable choice for most users that are mostly
interested in the UART console. However, on Chrome OS devices the UART
is disabled for production configurations anyway, and instead they rely
heavily on the CBMEM console for remote debugging and bug reports. For
these kinds of cases more info is almost always better, and you can't
easily reproduce a remotely filed bug if you notice that you need some
info that is only provided by BIOS_SPEW. On the other hand, the cost of
logging extra info to the CBMEM console is pretty negligible.
Therefore, let's bump the loglevel for CONFIG_CHROMEOS in particular
back up to the maximum. (Unfortunately, it seems that you can't 'select'
a choice option from another option, so this has to go in the
console/Kconfig file.)
Change-Id: I50724e3f7f8f57fdbc5846f21babc71798b21b65
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/29144
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Its spreading copies got out of sync. And as it is not a standard header
but used in commonlib code, it belongs into commonlib. While we are at
it, always include it via GCC's `-include` switch.
Some Windows and BSD quirk handling went into the util copies. We always
guard from redefinitions now to prevent further issues.
Change-Id: I850414e6db1d799dce71ff2dc044e6a000ad2552
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/28927
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
It seems to be the only user of the USB debug driver. So having to
enable it separately seems wrong.
It still depends on the selection of the EHCI debug driver.
Change-Id: I5f5f38a912423d9b8f1e71ae875b6a14fdee651c
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/28892
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Add a __noreturn macro that wraps __attribute__((noreturn)) and replace
current users with the macro.
Change-Id: Iddd0728cf79678c3d1c1f7e7946c27375a644a7d
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/28505
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
As we suppress output now before console_init() is done, the log level
read at start of ramstage is always -1.
Change-Id: Ia078d647c47aaa41ca9f2df9cf8506148ef86538
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: https://review.coreboot.org/26832
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Fix regression after commit
6032018 console: only allow console messages after initialization
Fix it so that the two remaining platforms that are being
moved to EARLY_CBMEM_INIT have chance to send board-status
with non-dirty tags before and after the conversion is made.
This also leaves us with a record in the repository where
LATE_CBMEM_INIT was known to work on some platform.
Change-Id: Ie874f986a2c474bba117d7d6ae959decec8060a8
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/26743
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Masked ROMs are the silent killers of boot speed on devices without
memory-mapped SPI flash. They often contain awfully slow SPI drivers
(presumably bit-banged) that take hundreds of milliseconds to load our
bootblock, and every extra kilobyte of bootblock size has a hugely
disproportionate impact on boot speed. The coreboot timestamps can never
show that component, but it impacts our users all the same.
This patch tries to alleviate that issue a bit by allowing us to
compress the bootblock with LZ4, which can cut its size down to nearly
half. Of course, masked ROMs usually don't come with decompression
algorithms built in, so we need to introduce a little decompression stub
that can decompress the rest of the bootblock. This is done by creating
a new "decompressor" stage which runs before the bootblock, but includes
the compressed bootblock code in its data section. It needs to be as
small as possible to get a real benefit from this approach, which means
no device drivers, no console output, no exception handling, etc.
Besides the decompression algorithm itself we only include the timer
driver so that we can measure the boot speed impact of decompression. On
ARM and ARM64 systems, we also need to give SoC code a chance to
initialize the MMU, since running decompression without MMU is
prohibitively slow on these architectures.
This feature is implemented for ARM and ARM64 architectures for now,
although most of it is architecture-independent and it should be
relatively simple to port to other platforms where a masked ROM loads
the bootblock into SRAM. It is also supposed to be a clean starting
point from which later optimizations can hopefully cut down the
decompression stub size (currently ~4K on RK3399) a bit more.
NOTE: Bootblock compression is not for everyone. Possible side effects
include trying to run LZ4 on CPUs that come out of reset extremely
underclocked or enabling this too early in SoC bring-up and getting
frustrated trying to find issues in an undebuggable environment. Ask
your SoC vendor if bootblock compression is right for you.
Change-Id: I0dc1cad9ae7508892e477739e743cd1afb5945e8
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/26340
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Fix regression after commit:
6032018 console: only allow console messages after initialization
Bail out early on AP CPUs, this avoids use of CAR_GLOBAL when
we have SQUELCH_EARLY_SMP=y.
Change-Id: I506c5fbec43a6eb6f6d9362d62a040def9e1e7bb
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/26120
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The former default SPEW is very noisy, intentionally. It is usually
only useful to debug specific issues and doesn't carry much infor-
mational value.
Reducing the loglevel should also mitigate overflows in pre-CBMEM
console buffers.
Change-Id: Iebcd4681572c58f1d17085c5ef01a2dd49e981ca
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: https://review.coreboot.org/26053
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-by: Angel Pons
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Kyösti Mälkki