With this, register_opaque_master can take care of register_shutdown
as well, and every opaque master only needs to call
register_opaque_master instead of calling both register_opaque_master
and register_shutdown.
Next patches in the chain convert opaque masters to use new API.
BUG=b:185191942
TEST=builds and ninja test from CB:56413
Change-Id: I34183e6bafc787eec54ee4a26b73a40803f3ce99
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56823
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Param is only used in the first half of init function, and it is
local, so there is no need to keep it until the end. This makes
handling error paths in the second half of init function shorter,
because those paths can just return 1 instead of going to a label.
BUG=b:185191942
TEST=builds and ninja test from CB:56413
Change-Id: I126a8d06297ef4d42bc93a73f7067ccc1352d1e9
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56822
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Previously shutdown function was registered conditionally for 82580,
only if the device was not UNPROG_DEVICE. This patch moves the check
for UNPROG_DEVICE into shutdown function itself, so that shutdown
function can be always registered for 82580.
This also fixes a memory leak in nicintel_ee_shutdown_82580.
No changes for i210 device init/shutdown, only for 82580.
And very importantly this unlocks API change which plans to move
register_shutdown inside register_opaque_master, similar to what's
done in CB:56103
BUG=b:185191942
TEST=builds
Change-Id: I5c729a3a63d0106e65525a6a77b2f9104c96847f
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56821
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
This patch prepares spi masters to use new API which allows to
register shutdown function in spi_master struct. See also later
patch in this chain, where spi masters are converted to new API.
BUG=b:185191942
TEST=builds and ninja test
Comparing flashrom binary before and after the patch,
make clean && make CONFIG_EVERYTHING=yes VERSION=none
binary is the same
Change-Id: I50716686552b4ddcc6089d5afadb19ef59d9f9b4
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56101
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
With this, register_spi_master can take care of register_shutdown
as well, and every spi master only needs to call register_spi_master
instead of calling both register_spi_master and register_shutdown.
Testing:
In dummyflasher, comment out free(data) in shutdown. Test fails with error:
../dummyflasher.c:949: note: block 0x55e0727a6e40 allocated here
ERROR: dummy_init_and_shutdown_test_success leaked 1 block(s)
Means, shutdown function is invoked for drivers with "old" API
(so, transitioning from old to new API is not breaking anything).
Next patches in the chain converts spi masters to use new API.
BUG=b:185191942
TEST=builds and ninja test
Change-Id: I2dc80dceca2f8204bcd0dad1f51753d7e79f1af5
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55932
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
In some earlier refactoring the init function for this programmer was
broken such that it never succeeds, since the return value was never set
to a value other than a generic failure. Fix it to return success unless
there is a problem.
TEST=flashrom -p realtek_mst_i2c_spi:bus=n --flash-size now works
when n is a bus that has a connected MST.
Change-Id: I0596014fc9b76b4d11de230f9f5c414c1321dff1
Signed-off-by: Peter Marheine <pmarheine@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56799
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This patch adds a test and mocks for two pci functions that
nicreltek is using.
Main reason for this is to have at least one test for par master
(currently there are none), in preparation for a change like
CB:56103 but for par masters.
BUG=b:181803212
TEST=ninja test
Change-Id: Iaed14fe1d83c8eb45ec185ebd3f1c97cb81941f4
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56470
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
This adds to the search-by-serial functionality with
search-by-description (product string). This is useful when e.g. one has
multiple FTDIs in a system and wants the serial numbers to reflect the
system-level serial number, and the description to reflect the
subcomponent names.
Tested manually by running with both serial and description searches, on
a machine with multiple FTDIs plugged in. Ensured that when two devices
with the same vid/pid/serial number are plugged in, description can
be used to differentiate.
Verifed no-description, no-serial, one FTDI plugged in base case works.
Original version of this code used the original single "arg" char*, but
on further thought, this wasn't worth the readability and functionality
losses. The new version with arg2 gets rid of several lines of code, the
gotos, and adds the ability to filter by both description and serial
simultaneously.
Change-Id: Ib4be23247995710900175f5f16e38db577ef08fa
Signed-off-by: Harry Johnson <johnsonh@waymo.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56164
Reviewed-by: Anastasia Klimchuk <aklm@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
write_aai is required to register a spi master, if it is not set in
spi_master struct then register_spi_master returns ERROR_FLASHROM_BUG.
Masters in this patch did not have it set in the struct, and
register_spi_master always returned an error for them. However return
value of register_spi_master was ignored, so this was hard to notice.
Next patch in the chain checks return value of register_spi_master.
BUG=b:185191942
TEST=builds and ninja test
Change-Id: I712e74e11244e1f0ab8d8e245fcd5207ce211219
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56102
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Three more tests in this patch where layout does / does not pass
sanity checks.
BUG=b:193584590
TEST=ninja test
Change-Id: Ia21585d60443b2741f3868d7887476090e35f79b
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56440
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
This patch is making layout test output more compact, spanning
less vertical lines while logging the same information.
No changes made to the test functionality.
BUG=b:193584590
TEST=ninja test
Change-Id: I36a39505f8db508672cde5a848684a3a22f5ccc4
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56439
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
The flashrom project no longer uses freenode. To avoid having outdated
man pages in the future, the contact methods are now listed in the wiki.
Change-Id: I75e8f43c50dc4c3feede0250334a877cdaac8103
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56031
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
There are no tests for layout, it would be great to add some.
Also partially inspired by
commit 06a89d713951a2e08ef8fb698a7688357baa83d1
and commit c9039fc27916c03e21ba91365d01e6bc49503053
BUG=b:193584590
TEST=ninja test
Change-Id: I7aa8dc0c9bc5a22fe5deea757eea1a151b969cea
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56324
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This is a known issue in cmocka (see
https://github.com/clibs/cmocka/issues/17) where cmocka does not
recognise memory allocation happening inside strdup, and then later
throws an error when the memory is freed. If the issue is fixed at
some point, this workaround can be removed.
Given that cmocka already overrides malloc, calloc, realloc, free,
adding strdup there seems fine.
Existing tests now can (and have to) free the memory they allocated
by strdup, and this is in the same patch.
BUG=b:193584590
TEST=ninja test
Change-Id: I56aef6b342752d80995c36ab075b12198fc101d9
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56323
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
This patch adds mocks for libusb functions. To avoid dependency on
libusb.h, libusb symbols for context and device handle are redefined.
Real libusb functions are never called in tests anyway, cmocka wraps
work with this without complaints.
BUG=b:181803212
TEST=builds and ninja test
Change-Id: I38508dfb6d7c24d42522f22fcae0c5e410c5f7ea
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55934
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
The function is only ever used within flashrom.c.
BUG=none
BRANCH=none
TEST=builds
Change-Id: I81f1cdb9df98c151201390edeb69c74defe7881f
Signed-off-by: Edward O'Callaghan <quasisec@google.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56295
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Anastasia Klimchuk <aklm@chromium.org>
Just add the name to the existing entry, as usual it is supposed to be
compatible.
Change-Id: I59c8067f15b5ceac5a2e2f8fe93431a465f17e23
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/56054
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Also one case that doesn't work on Android.
Change-Id: I17381172f8df213bbddccedec2ef288326e49ecf
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55877
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
`mark_unsupported` disables all programmers passed as arguments and
adds them to the UNSUPPORTED_FEATURES list. It's conveniently used
with our dependency lists, e.g.
$(call mark_unsupported,$(NEEDS_LIBUSB1) $(NEEDS_LIBFTDI))
Change-Id: I6c9b66dfa89ec3d71675576d03fcebfdea30ce7b
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55873
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
For each feature or library that we treat as dependency, list which
programmers depend on it.
Change-Id: I6929181f18213e1434cbbc7ba2d2b0b04bd46a6a
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55872
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
This reverts commit cf1e8f7b904ddf639b29b2e1bb0ad681c1aad76d.
It's an indirect, spurious dependency. Why clutter the build system
with it? Instead, the functions these programmers actually depend
on should be factored into a compilation unit that doesn't need PCI.
No matter this change, compilation with `CONFIG_INTERNAL=no` is
broken because of these programmers right now.
Change-Id: I134eb7d74b44f92dfdea3c764f39bc3673739fa0
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55869
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Let's assume this function is only called when overlapping, included
regions are fatal.
Change-Id: Ie969e0538d302ccd58d3fec3921265ed3621eaa5
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55999
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Anastasia Klimchuk <aklm@chromium.org>
Fixes a regression introduced by commit 06a89d713 (layout: Introduce
layout_next()). A negation was dropped by accident which resulted in
comparisons with later, not-included regions instead of included ones.
Change-Id: I18168b9bd9a7d0fe2ba04fb091d1600f65d6dd5b
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55997
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Normalize the fallback paths:
* Always end with the newest, assumed compatible chipset.
* Perform tighter checks when it's about warnings only.
* If two chipsets seem compatible, always return the same
(this is currently the case for 8/9 series and 300/400
series which we can't distinguish).
Change-Id: Ic5a5fee870202173b3a9813b03ec261e8ee45155
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55651
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Michał Żygowski <michal.zygowski@3mdeb.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Only warn if the `freq_read` setting looks odd but don't override
our previous guess. The `freq_read` check was taken from `ifdtool`
but seems less reliable than our own detection scheme.
Change-Id: I658d76ec2567d1d660a18d0b0ae71c744e603e8f
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55650
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Michał Żygowski <michal.zygowski@3mdeb.com>
Detection based on ICCRIBA and FMSBA became a little messy lately.
However, there's a new static difference: Since 300 series (Cannon
Point), there is an MDTBA field in FLUMAP1 that has always been 0
(reserved) before. Taking this into account, we can relax the checks
on ICCRIBA.
Change-Id: I587ad1abe390843d4a9e74431b6fc4b63f8ba512
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55647
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Michał Żygowski <michal.zygowski@3mdeb.com>
Shutdown function was covering two different jobs here: 1) the actual
shutdown which is run at the end of the driver's lifecycle and
2) cleanup in cases when initialisation failed. Now, shutdown is only
doing its main job (#1), and the driver itself is doing cleanup
when init fails (#2).
The good thing is that now resources are released/closed immediately
in cases when init fails (vs shutdown function which was run at some
point later), and the driver leaves clean space after itself if init
fails.
And very importantly this unlocks API change which plans to move
register_shutdown inside register master API, see
https://review.coreboot.org/c/flashrom/+/51761
BUG=b:185191942
TEST=builds
Change-Id: I3273da907614a042d50090338c337dfd64695354
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55887
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Return values for buspirate_sendrecv come from serialport_write
and serialport_read, and those return 1s for any error. No need
to explicity assign ret = 1, because it is already 1 for error.
Follow up on commit 751afa88a7f22487069665c5d206ae490f158231
where this idea was suggested.
BUG=b:185191942
TEST=builds
Change-Id: I31fd70f607dc965d5cac1cd0116faa447dbc177a
Signed-off-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/55818
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
* Parameter names like `flashctx` for a `struct flashrom_flashctx`
don't add any value.
* `const` qualification of parameters is meaningless in forward
declarations. Arguments are always passed by copy and an API
user does not need to know what callees do with their copy.
Change-Id: Iadcc1670ff86578a400dec9e804d6dda93e0fcf0
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/54288
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Rename it to layout_sanity_checks() as that is what it does and
let it work on the currently active layout instead of the global
layout.
Change-Id: Ifae3480d4bd68c939c291f05734544e93f00306c
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/54285
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Anastasia Klimchuk <aklm@chromium.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
