This patch adds a test for raiden_debug_spi and lots of libusb wraps.
libusb.h becomes required for tests to build and run, since new tests
are using libusb structs in depth and opaque symbols not sufficient
anymore.
BUG=b:181803212
TEST=builds and ninja test
Change-Id: I880a8637ab02de179df9169c1898230bce4dc1c7
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57918
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
__xstat and __fxstat variants of stat/fstat are invoked under
chromium chroot. For all existing tests it is sufficient for
stat/fstat to "do nothing, return 0", so new wraps do just that.
Test which needs __xstat: linux_mtd lifecycle.
Tests which need __fxstat:
read_chip_test_success
read_chip_with_dummyflasher_test_success
write_chip_test_success
write_chip_with_dummyflasher_test_success
Without this patch tests above fail under chromium chroot.
BUG=b:181803212
TEST=running tests on three different environments,
1) stat64/fstat64 (ninja tests in upstream tree)
2) stat64/fstat64 (ninja tests in chromium tree)
2) __xstat64/__fxstat64 (emerge with tests in chromium tree)
Change-Id: I4c5c243acde09dc5bb6b2a14042fcd23a49707db
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/58103
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Tiger Lake has very low ICCRIBA (TGL=0x11, CNL=0x34 and CML=0x34) and
detects as unknown chipset compatible with 300 series chipset. Add a
new enum CHIPSET_500_SERIES_TIGER_POINT and treat it identically to
CHIPSET_400_SERIES_COMET_POINT. There are some exceptions though,
ICCRIBA is no longer present n descriptor content so a new union has
been defined for new fields and used in descriptor guessing.
freq_read field is not present on Tiger Lake, moreover in CannonPoint
and Comet Point this field is used as eSPI/EC frequency, so a new
function to print read frequency has ben added. Finally Tiger lake
boot straps include eSPI, so a new bus has been added for the new
straps.
TEST=Flash BIOS region on Intel i5-1135G7
Signed-off-by: Michał Żygowski <michal.zygowski@3mdeb.com>
Change-Id: I28f3b6fe9f8ce9e976a6808683f46b6f4ec72bdd
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55578
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This patch adds two tests and initialises page_size in mock chip
chip_W25Q128_V. page_size was not needed for previous tests
(erase and read). page_size only needed to execute writing on chip
with dummyflasher, so it is added here.
BUG=b:181803212
TEST=ninja test
Change-Id: I6f0336613ab16a7e59857006496e3590ddb14d00
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/58357
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Clean up the feature target by outsourcing the test to an own variable.
Change the print output and don't write to the build-details file.
This is in preparation for further changes.
Change-Id: I18fc27252afb49fa7d1f2787faee2b5b669275aa
Signed-off-by: Thomas Heijligen <thomas.heijligen@secunet.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/58627
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Clean up the feature target by outsourcing the test to an own variable.
Change the print output and don't write to the build-details file.
HAS_CLOCK_GETTIME=no replaces DISABLE_CLOCK_GETTIME=yes
This is in preparation for further changes.
Change-Id: Ie1f43b3d5a8ad79bff3f9bbc21f359ec35abc42a
Signed-off-by: Thomas Heijligen <thomas.heijligen@secunet.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/58618
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
These warnings are printed at error level so they are displayed with
every invocation of flashrom. This clutters the flashrom output in the
usual case. Move warnings to debug level, add newline and clean up text.
TEST=Deploy to guybrush, observe messages are only seen when --verbose
is enabled
BUG=None
BRANCH=None
Change-Id: Idf5e735b9e504c943bf93a428da64976d723eb2c
Signed-off-by: Rob Barnes <robbarnes@google.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/58529
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Clean up the compiler target by outsourcing the test to an own variable.
Change the print output and don't write to the build-details file.
This is in preparation for further changes.
Change-Id: I3d6f08ef030744c772b4ec0dc2c9e614fb90461d
Signed-off-by: Thomas Heijligen <thomas.heijligen@secunet.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/58616
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
ich_descriptors_tool/Makefile: have the same behavior as the main
flashrom Makefile
- only set gcc explicit on MinGW HOST_OS
- don't fallback to gcc if CC was not set
- set CFLAGS and EXEC_SUFFIX for TARGET_OS, not for HOST_OS
Change-Id: I353c3de250167994a4aea1edfef57d839e900d78
Signed-off-by: Thomas Heijligen <thomas.heijligen@secunet.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/58468
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Successful return value for fread is the number of items read, and
default behaviour for all wraps is to return success.
This worked previously because all existing tests have custom mocks
for fread, so default behaviour hasn't been used. However next patch
in this chain adds new test which needs default fread wrap.
BUG=b:181803212
TEST=ninja test
Change-Id: I17d82d281a87129843f547b87c18f52aca23314d
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/58356
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Missing wrap for libusb_init has been discovered while working
on the test for raiden_debug_spi. Both dediprog (existing test) and
raiden_debug_spi (new test, see later in this chain) are using
libusb_init, so it definitely needs to be wrapped and added to io_mock.
Why tests worked before, without the wrap: my understanding is that
real libusb_init was called for dediprog test? Given that tests
definitely should not call any real libusb functions, wrap is needed.
BUG=b:181803212
TEST=builds and ninja test
Change-Id: I51c9cb96db1afb3298f4d098df96509d3cb3c046
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57917
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
In preparation for adding more tests with lots of libusb wraps,
existing wraps are extracted into separate file.
BUG=b:181803212
TEST=builds and ninja test
Change-Id: I17c1cc0189fcb742a99ad1b0fd615ab95504a74a
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57916
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
tests.c is growing and needs to be split, specifically all
libusb wraps need to be extracted into their own file. See later
in the chain a lot more wraps are added for libusb functions. To
be able to split it, current_io needs to be moved one level up,
to be visible across tests. This allows having multiple files with
wraps, and all the wraps can use current_io.
BUG=b:181803212
TEST=builds and ninja test
Change-Id: I5327b5de430afe13a8cc931c8b4b188dcb8c8cf6
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57915
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
LOG_ME macro is very generic and can be useful anywhere in tests.
Previously was only used in scope of tests.c. With time more tests
are added, and more files, LOG_ME needs to be visible everywhere.
BUG=b:181803212
TEST=builds and ninja test
Change-Id: If7f3d256161bc8b81e996328e445cccab9a82174
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57914
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
This patch adds NON_ZERO macro and not_null function into io_mock.h,
so that they can be used anywhere in tests. Common usage for not_null
is to indicate a valid pointer, where it doesn't matter what the
pointer is, only matters it is not null. Common usage of NON_ZERO is
to indicate a valid file descriptor, where it only matters the
descriptor is non-zero integer.
New features replace all usages of previous MOCK_HANDLE.
This patch corrects return value from __wrap_ioctl to be successful
by default. It used to be MOCK_HANDLE, but should be 0. Included in
this patch because this is also a replacement of MOCK_HANDLE.
BUG=b:181803212
TEST=builds and ninja test
Change-Id: I5ad6ee4aa9091447c6c9108c92bf7f6e755fca48
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57269
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Replace the remaining uses by the plain macros.
Windows, Linux and MacOS are the only systems with explicit macros.
Only the Windows macro is used in several places. The others can easily
be handled like all other systems.
Change-Id: I23ec8fdcff8202d10ddf5a7520432e5b26b01cd4
Signed-off-by: Thomas Heijligen <thomas.heijligen@secunet.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/58276
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
This allows programmer to register shutdown function in spi_master
struct, which means there is no need to call register_shutdown in init
function, since this call is now a part of register_spi_master.
As a consequence of using new API, two things are happening here:
1) No resource leakage anymore in case register_shutdown() would fail,
2) Fixed propagation of register_spi_master() return values.
BUG=b:185191942
TEST=test in CB:57918
(Nikolai) tested probe/read/write with a servo micro and puff board
Change-Id: I8927224779f24d1fda088991337e54d7272775a6
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Tested-by: Nikolai Artemiev <nartemiev@google.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57975
Reviewed-by: Nikolai Artemiev <nartemiev@google.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This patch prepares the programmer to use new API which allows to
register shutdown function in spi_master struct. See also later
patch in this chain, where the programmer is converted to new API.
BUG=b:185191942
TEST=test in CB:57918
Change-Id: I567e24c9d39e24a4a1634ddcbdd05930760afcdc
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57974
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Spi data for this programmer now contains pointer to spi_master
struct, and can be used in configure_protocol.
This patch is making init code of raiden_debug_spi a bit more
consistent with other programmers, for example it allows to remove
data assignment `spi_config->data = data` from init function. This
assignment was only needed for configure_protocol function, and
now it is not needed anymore.
BUG=b:185191942
TEST=test in CB:57918
Change-Id: I5df0609efcd1fdb274b8cc84536253d5dbae5270
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57973
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
For this programmer, spi_master struct is dynamically allocated
and needs to be freed on shutdown. Adding a pointer to spi data
struct allows to link spi_master from spi data, which in turn allows
to pass spi data into shutdown function. As a result, both
register_spi_master and register_shutdown use the same data, and
this unblocks moving raiden_debug_spi to new API (like all the
other spi masters in commit a69c5196d2).
See further patches in this chain.
BUG=b:185191942
TEST=test in CB:57918
Change-Id: I5400802aa6961538df12fcd5339f7e2a5db145a1
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57972
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
This reintroduces a reworked version of the GPIOL pin control first
introduced in commit 3207844 (CB:49637), which was reverted in commit
6518cf3 (CB:55692) due to breakage.
This change introduces a new argument `gpiolX` to allow use of the four
GPIOL pins either as generic gpios or as additional CS# signal(s). `X`
specifies the GPIOL pin (0-3) to be set to one of [HLC] with the
following meaning:
* H - set the pin as output high
* L - set the pin as output low
* C - use the pin as additional CS# signal
The third value, `C`, aims to replace the parameter `csgpiol`, that is
now marked as deprecated and can be removed at some point in the future.
`gpiol` and `csgpiol` are mutually exclusive and use of both results in
an error.
Multiple pins may be set by specifying the parameter multiple times.
Documentation was updated/added accordingly.
Test: All pin levels/modes have been verified to behave correctly with a
logic analyzer.
Change-Id: I3989f0f9596c090de52dca67183b1363dae59d3a
Signed-off-by: Alan Green <avg@google.com>
Signed-off-by: Michael Niewöhner <foss@mniewoehner.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57810
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
This partially undoes changes made in commit cd9b7b427
(ich_descriptors: Normalize chipset detection) to re-add explicit
matching of each chipset with one or more strap length values.
Since ranges are checked explicitly, the `warn_if` parameter to
warn_peculiar_desc() is no longer necessary and is removed.
Change-Id: Ica49477492876810a6fa212768b1ab9e8c12001f
Signed-off-by: David Hendricks <ddaveh@amazon.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/57793
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
