This is a reland of
commit c4be70f6ff ("commonlib/list: Support circular list").
In some use cases, we want to add items to the linked list and then
iterate over them with the insertion order. With the current API, the
call site needs to either use the inefficient list_append() function to
append items to the end of the list, or manually maintain a "tail"
node pointer.
To support that use case, add an internal helper function _list_init()
to initialize the list as a circular one with a placeholder head node.
_list_init() is automatically called within list_insert_after() and
list_append(). In list_insert_before(), an assertion is added to avoid
an insertion before the head node (which should be invalid). The
implementation ensures that the list is initialized as a circular one
whenever the first element is added. That also allows all call sites to
be auto-upgraded to the "circular list" implementation without any
modification.
Modify list_for_each() to support circular lists, and improve
list_append() efficiency by inserting the new node before the
placeholder head node. Also add a few assertions in the implementation.
Add a new test case to test iterating over an empty list.
Note that '(uintptr_t)ptr + (uintptr_t)offsetof(typeof(*(ptr)), member)'
was used instead of the simpler '&((ptr)->member)' because GCC9+ assumes
that the address can never be NULL. See commit 88991caf00
("include/list.h: Add support for GCC9+") for details. Now, with the
new list_for_each() implementation, that pointer value can never be
NULL.
Change-Id: Idc22887cce71284c9028dce10eeef9cc16669028
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90962
Reviewed-by: Julius Werner <jwerner@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jakub "Kuba" Czapiga <czapiga@google.com>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
In a follow-up patch (CB:90962), the list will be changed to a circular
one, and list_node fields 'next' and 'prev' will become private to the
implementation.
To allow smooth transition to circular lists for all call sites, add the
following functions to the list API:
- list_is_empty()
- list_next()
- list_prev()
- list_first()
- list_last()
- list_length()
All list API call sites are expected to use the public API instead of
the raw 'next' and 'prev' pointers.
Change-Id: Ib1040f5caab8550ea52db9b55a074d7d79c591e5
Signed-off-by: Yu-Ping Wu <yupingso@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90961
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jakub "Kuba" Czapiga <czapiga@google.com>
Reviewed-by: Julius Werner <jwerner@chromium.org>
This reverts commit c4be70f6ff.
Reason for revert: The CL caused a hang in Depthcharge on
Google/Quartz.
BUG=b:479143030
TEST=Verify boot on Google/Quartz.
Change-Id: I38087d0b2dd218dfb32a02c343b199708bb47d49
Signed-off-by: Kapil Porwal <kapilporwal@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90956
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
In some use cases, we want to add items to the linked list and then
iterate over them with the insertion order. With the current API, the
call site needs to either use the inefficient list_append() function to
append items to the end of the list, or manually maintain a "tail"
node pointer.
To support that use case and make the change backward compatible, add a
helper list_init() function to initialize the list as a circular linked
list. list_init() is automatically called within list_insert_after() and
list_append(). In list_insert_before(), an assertion is added to avoid
an insertion before the head node (which should be invalid). The
implementation ensures that the list is initialized as a circular one
whenever the first element is added. That also allows all call sites to
be auto-upgraded to the "circular list" implementation without any
modification.
Modify list_for_each() to support circular lists, and improve
list_append() efficiency by inserting the new node before the
placeholder head node. Also add a few assertions in the implementation.
Add a new test case to test iterating over an empty list.
Note that '(uintptr_t)ptr + (uintptr_t)offsetof(typeof(*(ptr)), member)'
was used instead of the simpler '&((ptr)->member)' because GCC9+ assumes
that the address can never be NULL. See commit 88991caf00
("include/list.h: Add support for GCC9+) for details. Now, with the
new list_for_each() implementation, that pointer value can never be
NULL.
Change-Id: I8451f711d4e522e239c241b3943e00070896dec9
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90799
Reviewed-by: Julius Werner <jwerner@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
A later patch changes the .config file which we use for our tests.
However that causes the PLATFORM_USES_FSP2_0 option to be enabled, which
in turn causes build errors in our tests, because the function is
obviously not defined in our tests. Create a stub function of sorts like
we do for other coreboot table entries.
It also moves the declaration of the `lb_string_platform_blob_version`
function to coreboot_tables.h, since it doesn't belong in the FSP header
file. Because of that we can also remove the
`if (CONFIG_PLATFORM_USES_FSP2_0)`
check, which makes the code a bit cleaner.
Signed-off-by: Maximilian Brune <maximilian.brune@9elements.com>
Change-Id: I7721dfe4d287b2274a383bb7e5337b85a0f3f148
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90830
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jakub "Kuba" Czapiga <czapiga@google.com>
In our current tree this is always the case. The coreboot code (as far
as I know) doesn't have a hard requirement on that. But if these values
differ then it is usually always a mistake made by the programmer and it
is hard to catch since the value don't really depend on each other.
So until a time in which there comes a platform which needs a flashmap
that doesn't cover the whole flash, this check is introduced.
For that purpose we need to replace the default .config file for tests,
because otherwise the check won't match.
The config file that is used now is based on the fact that we use the
same mainboard for the `.config` as for the `fmap_config.h` in
`tests/include/tests/lib/fmap/fmap_config.h`
Signed-off-by: Maximilian Brune <maximilian.brune@9elements.com>
Change-Id: I8bc05a17a2630516ede949660b4fc428f199f3ab
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90264
Reviewed-by: Jakub "Kuba" Czapiga <czapiga@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
"-I$(cmockasrc)/include" should be added to TEST_CFLAGS if we are
building cmocka from source (i.e., USE_SYSTEM_CMOCKA is NOT 1). Fix the
condition in Makefile.common.
Change-Id: I957066fb24f03712a5b4b396aa9e04f3861940ee
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90798
Reviewed-by: Hsuan-ting Chen <roccochen@google.com>
Reviewed-by: Jakub "Kuba" Czapiga <czapiga@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
From the Jenkins result of CB:90798, it appears that the generated
junit-tests_lib_ux_locales-test(tests).xml is not a valid XML file
possibly due to incorrect quotes handling by cmocka.
Therefore, in the UX_LOCALES_GET_TEXT_TEST macro definition, replace
`#_expect` with `_expect`, so that the `name` field of the CMUnitTest
struct won't contain double quotes.
Change-Id: Idfec437ae627208031854694e66ca79e22132385
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90801
Reviewed-by: Jakub "Kuba" Czapiga <czapiga@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The issues still appears even after commit 65833355ca ("tests: Disable
gcov warnings"). So, disable the HTML generation completely until the
issue is figured out.
Change-Id: I683889ff8a0769697154079f99fe1d6ff9773281
Signed-off-by: Felix Singer <felixsinger@posteo.net>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90578
Reviewed-by: Matt DeVillier <matt.devillier@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Alicja Michalska <ahplka19@gmail.com>
This adds the option to compress ramstage and payloads with zstd.
zstd compressed ramstages are typically +5% bigger than lzma compressed
ramstages. The decompressor .text section grows by 20KiB and the
decompressor needs 16KiB more heap than the lzma decompressor.
To use less heap inside the zstd decompressor the build time define
ZSTD_DECODER_INTERNAL_BUFFER is used.
Quote:
The build macro `ZSTD_DECODER_INTERNAL_BUFFER` can be set to control
the amount of extra memory used during decompression to store
literals. This defaults to 64kB. Reducing this value reduces the
memory footprint of `ZSTD_DCtx` decompression contexts, but might
also result in a small decompression speed cost
TEST=Booted on Lenovo X220 with zstd ramstage showed no disadvantage
over a bigger internal buffer used.
TEST=Booted on Lenovo X220. The zstd decompressor is twice as fast
as the lzma decompressor.
cbmem -t shows:
- finished ZSTD decompress (ignore for x86) 79,444 (24,494)
- finished LZMA decompress (ignore for x86) 94,971 (45,545)
TEST=Booted on QEMU Q35, QEMU aarch64 virt, QEMU riscv RV64 with
zstd compressed ramstage.
Change-Id: Ic1b1f53327c598d07bd83d4391e8012d41696a16
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/69893
Reviewed-by: Julius Werner <jwerner@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Newer gcov/lcov versions shipped in CI container images throw a warning
and thus cause the CI to fail. It's unclear how to fix this warning at
the moment, but gcov isn't critical anyway. So disable this specific
warning for now, so that we can roll out new CI images.
Excluding file '/home/coreboot/node-root/workspace/test_coreboot/payloads/libpayload/tests/libcbfs/cbfs-lookup-test.c'
lcov: WARNING: (inconsistent) /home/coreboot/node-root/workspace/test_coreboot/payloads/libpayload/libcbfs/cbfs.c:79: unexecuted block on non-branch line with non-zero hit count. Use "geninfo --rc geninfo_unexecuted_blocks=1 to set count to zero.
(use "lcov --ignore-errors inconsistent,inconsistent ..." to suppress this warning)
Excluding file '/home/coreboot/node-root/workspace/test_coreboot/payloads/libpayload/tests/libcbfs/cbfs-lookup-test.c'
[snip]
Message summary:
1 warning message:
inconsistent: 1
genhtml: ERROR: (corrupt) unable to read trace file 'build/coverage/tests.info': genhtml: ERROR: (inconsistent) "/home/coreboot/node-root/workspace/test_coreboot/payloads/libpayload/libcbfs/cbfs.c":77: function 'cbfs_unmap' is not hit but line 79 is.
To skip consistency checks, see the 'check_data_consistency' section in man lcovrc(5).
(use "genhtml --ignore-errors inconsistent ..." to bypass this error)
(use "genhtml --ignore-errors corrupt ..." to bypass this error)
make[1]: *** [tests/Makefile.mk:277: coverage-report] Error 1
make: *** [util/testing/Makefile.mk:103: what-jenkins-does] Error 2
Change-Id: I2c52c53fbe856a8bca062f34c576fdfda3818f2b
Signed-off-by: Felix Singer <felixsinger@posteo.net>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/90554
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Matt DeVillier <matt.devillier@gmail.com>
Use 0 instead of NULL when testing or initializing uintptr_t
fields, since NULL is only defined for pointers.
"data.3.lzma.bin" file is regenerated using:
util/cbfs-compression-tool compress file.bin /tmp/file.lzma.bin lzma
dd if=/tmp/file.lzma.bin of=file.lzma.bin skip=8 ibs=1
Change-Id: Ia1e7cc5052e842c7ef97ec7cec34919bbe1e2228
Signed-off-by: Elyes Haouas <ehaouas@noos.fr>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/89356
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jakub "Kuba" Czapiga <czapiga@google.com>
Add a flag "USE_SYSTEM_CMOCKA" in the Makefile of tests. (default 1)
If USE_SYSTEM_CMOCKA=1, we will check if the system has Cmocka module,
and link it directly. If the system doesn't have Cmocka, we will set the
flag to 0 and print a warning message.
If USE_SYSTEM_CMOCKA=0, we will build Cmocka from 3rdparty source code.
BUG=none
TEST=make unit-tests -j
TEST=USE_SYSTEM_CMOCKA=0 make unit-tests -j
BRANCH=none
Signed-off-by: roccochen@chromium.com <roccochen@chromium.org>
Change-Id: I091784ca541e2590e3db0a18ceea83e7895ed0c2
Reviewed-on: https://review.coreboot.org/c/coreboot/+/79019
Reviewed-by: Jakub "Kuba" Czapiga <czapiga@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Now that we require the FMAP to start at offset 0 in the flash, we can
assume this across the entire codebase and therefore simplify it on
several ends.
Signed-off-by: Maximilian Brune <maximilian.brune@9elements.com>
Change-Id: Ieb1a23f9c0ae8c0e1c91287d7eb6f7f0abbf0c2c
Reviewed-on: https://review.coreboot.org/c/coreboot/+/86771
Reviewed-by: Erik van den Bogaert <ebogaert@eltan.com>
Reviewed-by: Frans Hendriks <fhendriks@eltan.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Shuo Liu <shuo.liu@intel.com>
For simplicity we are going to impose this restriction to coreboot.
Note however that this is only a restriction for coreboot itself. The
FMAP tool itself is still a generic tool that does not require the FMAP
to start at offset 0.
Add an defacto empty fmap_config.h to our test cases, since fmap.h now
includes fmap_config.h.
Signed-off-by: Maximilian Brune <maximilian.brune@9elements.com>
Change-Id: Iba04ebdcd5557664a865d2854028dd811f052249
Reviewed-on: https://review.coreboot.org/c/coreboot/+/86770
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jakub "Kuba" Czapiga <czapiga@google.com>
This commit adds a new UX locale message to display a warning when the
battery is critically low.
The message informs the user about the low battery and indicates that
the system is shutting down.
This change ensures that users are notified before the system
unexpectedly shuts down due to low battery.
BUG=b:339673254
TEST=Built and booted google/brox.
Change-Id: I75c7a0d4d439901098c7f17a1dc90355307116ac
Signed-off-by: Subrata Banik <subratabanik@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/86284
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Karthik Ramasubramanian <kramasub@google.com>
Reviewed-by: Julius Werner <jwerner@chromium.org>
This patch refactors the `ux_locales_get_text` API to handle fallback
text (English) internally, rather than relying on the caller. It
introduces message IDs for lookups, enabling the API to locate both
the UX locale name and fallback text based on the ID.
With this patch, `ux_locales_get_text` API locates UX locales message
based on message ID.
`ux_locales_get_text` retrieves fallback text message depending
upon the message ID if UX locales is not available.
This centralizes fallback handling and simplifies adding future
messages without per-SoC duplication.
BUG=b:339673254
TEST=Built and booted google/brox. Verified eSOL display.
Change-Id: I4952802396265b9ee8d164d6e43a7f2b3599d6c0
Signed-off-by: Subrata Banik <subratabanik@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/86283
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Karthik Ramasubramanian <kramasub@google.com>
Reviewed-by: Julius Werner <jwerner@chromium.org>
As suggested by the linter:
Prefer 'unsigned long' over 'unsigned long int' as the int is unnecessary
Link: https://qa.coreboot.org/job/coreboot-untested-files/lastSuccessfulBuild/artifact/lint.txt
Cc: Jakub Czapiga <czapiga@google.com>
Change-Id: Ia41f0674f4abab285d89ed5101b3805975f3f381
Signed-off-by: Ariel Otilibili <otilibil@eurecom.fr>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/85788
Reviewed-by: Maximilian Brune <maximilian.brune@9elements.com>
Reviewed-by: Elyes Haouas <ehaouas@noos.fr>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This to fix the error when using C23:
cannot cast an object of type 'nullptr_t' to 'uintptr_t' (aka 'unsigned long')
return (uintptr_t)NULL;
^
Change-Id: Ibdc8794513a508fc61a5046692f854183c36b781
Signed-off-by: Elyes Haouas <ehaouas@noos.fr>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/84893
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jakub Czapiga <czapiga@google.com>
The underlying IMD function already returns an integer which indicates
success or failure.
This removes the need to have initialized variables that need to be
checked for NULL later. In some cases this actually adds the appropriate
check for returned values.
Dying is appropriate if cbmem is not found as it is essential to the
bootflow.
Change-Id: Ib3e09a75380faf9f533601368993261f042422ef
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/84039
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Undefined behavior in unit-tests is no fun. assert_string_equal()
expects properly zero-terminated strings. None of the encoded test
strings contain a termination, hence add it manually.
Without this change, the test was often failing with a wrong error
message:
[==========] tests_lib_b64_decode-test(tests): Running 1 test(s).
[ RUN ] test_b64_decode
[ ERROR ] --- "AB" != "AB"
[ LINE ] --- tests/lib/b64_decode-test.c:38: error: Failure!
[ FAILED ] test_b64_decode
[==========] tests_lib_b64_decode-test(tests): 1 test(s) run.
Probably due to unprintable characters in the string. No idea why
my system is more susceptible to this issue.
Change-Id: Id1bd2c3ff06bc1d4e5aa21ddd0f1d5802540999d
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/84088
Reviewed-by: Jakub Czapiga <czapiga@google.com>
Reviewed-by: Nicholas Sudsgaard <devel+coreboot@nsudsgaard.com>
Reviewed-by: Paul Menzel <paulepanter@mailbox.org>
Reviewed-by: Matt DeVillier <matt.devillier@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The current region_end() implementation is susceptible to overflow
if the region is at the end of the addressable space. A common case
with the memory-mapped flash of x86 directly below the 32-bit limit.
Note: This patch also changes console output to inclusive limits.
IMO, to the better.
Change-Id: Ic4bd6eced638745b7e845504da74542e4220554a
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/79946
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
When `maxlen` is large (such as SIZE_MAX), the `end` pointer will
overflow, causing strnlen() to incorrectly return 0.
To not make the implementation over-complicated, fix the problem by
using a counter.
BUG=b:359951393
TEST=make unit-tests -j
BRANCH=none
Change-Id: Ic9d983b11391f5e05c2bceb262682aced5206f94
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83914
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Karthik Ramasubramanian <kramasub@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Mario Scheithauer <mario.scheithauer@siemens.com>
Add strlen() and strnlen() to commonlib/bsd by rewriting them from
scratch, and remove the same functions from coreboot and libpayload.
Note that in the existing libpayload implementation, these functions
return 0 for NULL strings. Given that POSIX doesn't require the NULL
check and that other major libc implementations (e.g. glibc [1]) don't
seem to do that, the new functions also don't perform the NULL check.
[1] https://github.com/bminor/glibc/blob/master/sysdeps/i386/strlen.c
Change-Id: I1203ec9affabe493bd14b46662d212b08240cced
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/83830
Reviewed-by: Maximilian Brune <maximilian.brune@9elements.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Already included <types.h> is supposed to provide <limits.h>. See
`Documentation/contributing/coding_style.md` section `Headers and includes`
Change-Id: I945eeeeccb16851f64d85cf5c67ea6e256082e11
Signed-off-by: Elyes Haouas <ehaouas@noos.fr>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/82724
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jakub Czapiga <czapiga@google.com>
<stdio.h> header is used for input/output operations (such as printf,
scanf, fopen, etc.). Although some input/output functions can manipulate
strings, they do not need to directly include <string.h> because they
are declared independently.
Change-Id: Ibe2a4ff6f68843a6d99cfdfe182cf2dd922802aa
Signed-off-by: Elyes Haouas <ehaouas@noos.fr>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/82665
Reviewed-by: Yidi Lin <yidilin@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This adds some helper functions for FDT, since more and more mainboards
seem to need FDT nowadays. For example our QEMU boards need it in order
to know how much RAM is available. Also all RISC-V boards in our tree
need FDT.
This also adds some tests in order to test said functions.
Signed-off-by: Maximilian Brune <maximilian.brune@9elements.com>
Change-Id: I2fb1d93c5b3e1cb2f7d9584db52bbce3767b63d8
Reviewed-on: https://review.coreboot.org/c/coreboot/+/81081
Reviewed-by: Julius Werner <jwerner@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
<string.h> is supposed to provide <stdarg.h> and <stdio.h>
Change-Id: I021ba535ba5ec683021c4dfc41ac18d9cebbcfd2
Signed-off-by: Elyes Haouas <ehaouas@noos.fr>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/81853
Reviewed-by: Eric Lai <ericllai@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
<device/device.h> is supposed to provide <device/{path,resource}.h>
Change-Id: I2ef82c8fe30b1c1399a9f85c1734ce8ba16a1f88
Signed-off-by: Elyes Haouas <ehaouas@noos.fr>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/81830
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <ericllai@google.com>
This patch moves commonlib/stdlib.h -> commonlib/bsd/stdlib.h, since
all code is BSD licensed anyway.
It also moves some code from libpayloads stdlib.h to
commonlib/bsd/stdlib.h so that it can be shared with coreboot. This is
useful for a subsequent commit that adds devicetree.c into commonlib.
Also we don't support DMA on arm platforms in coreboot (only libpayload)
therefore `dma_malloc()` has been removed and `dma_coherent()` has been
moved to architecture specific functions. Any architecture that tries to
use `dma_coherent()` now will get a compile time error. In order to not
break current platforms like mb/google/herobrine which make use of the
commonlib/storage/sdhci.c controller which in turn uses `dma_coherent` a
stub has been added to arch/arm64/dma.c.
Signed-off-by: Maximilian Brune <maximilian.brune@9elements.com>
Change-Id: I3a7ab0d1ddcc7ce9af121a61b4d4eafc9e563a8a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/77969
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Sometimes when a test doesn't work it's convenient to run it through
GDB. This patch adds a variable you can set on the make commandline to
conveniently enable all the compiler flags needed to make that work.
Change-Id: I3ac80ad095e0b72cc3176cbf915d1f390cd01558
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/81112
Reviewed-by: Jakub Czapiga <czapiga@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yu-Ping Wu <yupingso@google.com>
It is needed in order to move device_tree.c into commonlib in a
subsequent commit.
Signed-off-by: Maximilian Brune <maximilian.brune@9elements.com>
Change-Id: I16eb7b743fb1d36301f0eda563a62364e7a9cfec
Reviewed-on: https://review.coreboot.org/c/coreboot/+/77968
Reviewed-by: Julius Werner <jwerner@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The cmocka problem of sanitizing XML strings has been fixed in CB:80382.
Therefore the helper macros UX_LOCALES_GET_TEXT_FOUND_TEST() and
UX_LOCALES_GET_TEXT_NOT_FOUND_TEST() can be merged into one.
TEST=make unit-tests JUNIT_OUTPUT=y -j
Change-Id: Ic3199e2a061550282fb08122943994c835845543
Signed-off-by: Yu-Ping Wu <yupingso@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/80621
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Eric Lai <ericllai@google.com>
Reviewed-by: Hsuan-ting Chen <roccochen@google.com>
It seems that we have some applications where we need to calculate a GCD
in 64 bits. Now, we could instantiate the algorithm multiple times for
different bit width combinations to be able to use the most efficient
one for each problem... but considering that the function usually only
gets called once per callsite per stage, and that software emulation of
64-bit division on 32-bit systems doesn't take *that* long either, we
would probably usually be paying more time loading the second instance
of the function than we save with faster divisions. So let's just make
things easy and always do it in 64-bit and then nobody has to spend time
thinking on which version to call.
Change-Id: I028361444c4048a0d76ba4f80c7334a9d9983c87
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/80319
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Yidi Lin <yidilin@google.com>
This patch adds a few more test cases for the IP checksum algorithm to
catch more possible corner cases (large data with more than 64K carries,
unaligned data, checksum addition with offset, etc.).
Change-Id: I39b4d3f1bb833894985649872329eec88a02a22c
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/80252
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Jakub Czapiga <czapiga@google.com>
This patch moves the IP checksum algorithm into commonlib to prepare for
it being shared with libpayload. The current implementation is ancient
and pretty hard to read (and does some unnecessary questionable things
like the type-punning stuff which leads to suboptimal code generation),
so this reimplements it from scratch (that also helps with the
licensing).
This algorithm is prepared to take in a pre-calculated "wide" checksum
in a machine-register-sized data type which is then narrowed down to 16
bits (see RFC 1071 for why that's valid). This isn't used yet (and the
code will get optimized out), but will be used later in this patch
series for architecture-specific optimization.
Change-Id: Ic04c714c00439a17fc04a8a6e730cc2aa19b8e68
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/80251
Reviewed-by: Yidi Lin <yidilin@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Jakub Czapiga <czapiga@google.com>
It's what this function family is defined to do, we currently don't
usually run into the case (see: not too many die() instances going
around), it's more useful to try to recover, and the JPEG parser can run
into it if the work buffer size exceeds the remaining heap, whereas its
sole user (the bootsplash code) knows what to do when seeing a NULL.
Use xmalloc() if you want an allocation that either works or dies.
tl;dr: That code path isn't usually taken. Right now it crashes. With
this patch it _might_ survive. There is a use-case for doing it like
that now.
Change-Id: I262fbad7daae0ca3aab583fda00665a2592deaa8
Signed-off-by: Patrick Georgi <patrick@coreboot.org>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/80226
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin L Roth <gaumless@gmail.com>
Reviewed-by: Eric Lai <ericllai@google.com>
The .inc suffix is confusing to various tools as it's not specific to
Makefiles. This means that editors don't recognize the files, and don't
open them with highlighting and any other specific editor functionality.
This issue is also seen in the release notes generation script where
Makefiles get renamed before running cloc.
Signed-off-by: Martin Roth <gaumless@gmail.com>
Change-Id: Id09eafd293a54198aab87281f529749325df8b07
Reviewed-on: https://review.coreboot.org/c/coreboot/+/80122
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
The .inc suffix is confusing to various tools as it's not specific to
Makefiles. This means that editors don't recognize the files, and don't
open them with highlighting and any other specific editor functionality.
This issue is also seen in the release notes generation script where
Makefiles get renamed before running cloc.
The rest of the Makefiles will be renamed in following commits.
Signed-off-by: Martin Roth <gaumless@gmail.com>
Change-Id: Idaf69c6871d0bc1ee5e2e53157b8631c55eb3db9
Reviewed-on: https://review.coreboot.org/c/coreboot/+/80063
Reviewed-by: Felix Singer <service+coreboot-gerrit@felixsinger.de>
Reviewed-by: Maximilian Brune <maximilian.brune@9elements.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
This reverts commit 7713a2f295.
Reason for revert: breaks main branch
Change-Id: I2749bea9369c222e510b838e278c7797d5dce56e
Signed-off-by: Felix Singer <felixsinger@posteo.net>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/78852
Reviewed-by: Martin L Roth <gaumless@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Patrick Georgi <patrick@coreboot.org>
