Disable stack protector for all boards/apps by default, not only
qmod-dfu. Use 'make STACK_PROTECTOR=1' to enable.
This was recommened by Eric:
"I'd argue that we do not want this in general, since it adds canaries
to all functions that deal with buffers, and therefore impacts the
overall timing in a non determinstic way depending on inlining and
optimizations, while contributing nothing in non debug builds."
Related: OS#5081
Change-Id: I30ad97f231ea5b401def650bc9adc7e9f2770df0
Prevent build failure on debian 9, ubuntu 20.04, 20.10, where
bin/qmod-dfu-flash.elf does not fit the ROM.
Fixes: OS#5081
Change-Id: I9fffe4c323094679062428f41a4246b1c1b30ca2
This reverts commit 4a29f64cbe.
The code replicates to a large extent what is already present in iso7816_fidi.c
and I have serious doubts about the correctness of the computation in
its iso7816_3_calculate_wt() function.
Change-Id: I80dab4401d13306d573a6a35ce8729d2acc141e4
This reverts commit 4a58c08d67.
The code replicates to a large extent what is already present in iso7816_fidi.c
and I have serious doubts about the correctness of the computation in
its iso7816_3_calculate_wt() function.
Change-Id: I3f26da4e9aa8d7b0f4b4b7992269cf365a643ec7
While this differs from tha naming in the schematics ({CLK,IO}_PHONE),
this matches the naming scheme used for USIM2 and the naming on other
boards.
Change-Id: I486b14260faec897e8c8698c4b7987bf36492497
this will become part of libosmocore since it it common to smart
card related projects (such as osmo-ccid-firmware)
Change-Id: I3d4c65d137fc4555fcb256443feadd1c695de73d
make V=1 can be used to echo all compilation commands, which is useful
because it allows IDEs to parse the gcc output in oder to properly index
the source files using the actual defines passed to the compiler.
Change-Id: I25c41dff89302a73ddd2a4aaba7cb14912fac3b8
For some reason undefined symbols were downgraded to warnings, which
means building a firmware that calls missing functions (= address zero)
was perfectly fine, which of course made development more exciting....
This applies to builtins, too, printf of one char gets downgraded to
putchar, which we don't have, so disable builtins.
Change-Id: I492f41ad4162b9d07b1881ae4aed019db2dff8b5
Don't create copies of firmware files with version strings appended in
the normal build. Only do this before uploading the firmware files.
I have verified that "make" before this change and
"make; contrib/prepare_upload.sh" after produce the same files.
Close: OS#4413
Change-Id: I118a4ff397a178281c26a6b98112fa66b6f049ab
Fix having the version set to UNKNOWN in all packages built by OBS. The
osmo-ci.git scripts generating the source packages to be built by OBS
generate a ".tarball-version" file with a version string like
"0.7.0.70-657c", but it did not get used because the path was wrong.
Related: OS#4413
Related: https://osmocom.org/projects/cellular-infrastructure/wiki/Git-version-gen
Change-Id: Ic0b06011a604beec7c1c907c2c6e4ae927456e2e
dfu flashing the ST12 was easy, but i was never able to
get ST34 into dfu mode. Changing the firmware so it resets
itself just like the octsim instead of starting a timer and
waiting for a reset from the host made it work every time for me.
Change-Id: Ida636ec925f40d6d56551f170150181350d03bbd
We are not using any PIO interrupts from DFU mode. It's only used in
the main application firmware (verified by "git grep PIO_ConfigureIt")
Change-Id: Id1447af519df3183061f3d3f156a8dd84789af16
On Ubuntu 20.04 when builiding dpkg packages, even when cross-compiling
firmware, gcc stack smashing protection is enabled. Let's provide what
is minimally required in order to sucessfully complete builds on such
platforms.
Change-Id: Ic2f68f16b0730e7b5db17c30effc29a2909d1997
Closes: OS#4687
The original code assumes that calls to PIO_ConfigureIt() are only
made once e.g. during board start-up. Hoewever, we call those
at USB SetConfiguration time, when we know which particular hardware
function we are supposed to perform. This means that after the host
has issued SetConfiguration more than a given number of times, the
code will assert() due to overflow of the static array.
Let's check if we already have allocated an array slot for a given pin
and reuse that allocated array bucket rather than allocating new ones
for the same pin.
Change-Id: I0c46d4b51eeebd58a8786d65e31e7a84e65b6a8e
Related: OS#4454
If we do this, the resulting USB code will fail on any of the
USB-IF Chapter 9 tests. EP0 should not be reset.
Change-Id: I070faf4cb7029d3ccfa6c63f8f04aa0f02657536
In dispatch_received_usb_msg(), we ran into an infinite loop if a
too long messages was received on the OUT EP. Let's break the loop.
Change-Id: I5325ed15d3dd79a42f8dac34d618e86b9334c301
Closes: OS#4429
Reading the Unique ID from flash is a rather tricky procedure: After the
STUI command has been issued, we cannot read normal flash anymore.
Rather, the unique ID is mapped at 0x00000000. This is unfortuantely
also where the exception vector table is stored.
EEFC_ReadUniqueID() is already linked to RAM, which is good. Hoewver,
if an Interrupt happens between STUI and SPUI, then we try to access
the vector table and code from flash, which is illegal. We run into
a hardfault and stay there until the watchdog resets the processor.
Change-Id: I3c4fad55b47e9013f6615a331983b3989ca805a7
Closes: OS#4428
As reported in https://osmocom.org/issues/4335, there appear to
be some cards / use cases in which the 512 byte sized ringbuffer is
insufficient. As we do have free RAM available, we can easily
increase the buffer size, despite not entirely knowing yet why
it needs to be *that* large.
Change-Id: Ie713d614ec5b334e9058d5d430e4bb660f5b8b69
Closes: OS#4335
this adds the DFU as application, allowing to flash the bootloader.
a USB DFU alternative is added to flash the bootloader partition.
when the DFU is started as bootloader, the partition/alternative
to flash the bootloader is marked as "not available", and
ineffective.
the same happens for the application partition when DFU is started
as application.
this distinction is make at compile time, not at runtime, because
of size restrictions (the bootloader was already close to the
16 kB limit).
*_dfu_flash.bin should not be mixed with *_dfu_dfu.bin.
*_dfu_dfu.bin should be flashed as application using the already
existing DFU bootloader.
once this images is started (as application), the *_dfu_flash.bin
should be flashed as bootloader using the DFU application.
once the DFU bootloader has been flashed, soft resetting
(not re-powering) will cause the bootloader to start, allowing to
flash the application with a normal image (e.g. not DFU),
replacing the DFU application.
this switch to DFU only happens after downloading (e.g. flashing).
it is planned to have the DFU application erase itself after
flashing, but this is currently not implemented.
Change-Id: Ic273bb593a7669111b0219fe301d7897419167c8
In the end we want to share most of this, irrespective of the "APP".
Let's avoid unnecessary differences.
Change-Id: Icf063d4ca79edf66ffbe8e87a915deb77dec478a
This unifies the printing of the welcome banner, and it also ensures
that all modes print all information (serial number, reset cause).
Furthermore the APP and BOARD #defines from the make environment are
also printed.
Change-Id: I7e6bc05cee4b9ec0fd9a05dc90ce0b26a5763e5a
This is a purely cosmetic change in terms of coding style, but it
fixes an [invalid] detection for unterminated value_string arrays
by our verify_value_string_arrays_are_terminated.py script:
Change-Id: I2f2370a673074f6bf5380106b6254b4aa1e8a792
ERROR: file contains unterminated value_string 'struct value_string iso7816_3_card_state_names[]': './deps/simtrace2/firmware/libcommon/source/card_emu.c'
ERROR: file contains unterminated value_string 'struct value_string tpdu_state_names[]': './deps/simtrace2/firmware/libcommon/source/card_emu.c'
If this feature flag is set by the host application, we will start
generating SIMTRACE_MSGT_BD_CEMU_STATUS towards the interrupt endpoint
on any changes to the SIM card status, such as change of VCC/CLK/RST
Change-Id: I832b4ac4aeb6152195d985067ffdd64fa5a7d4cc
Related: OS#4330
The new SIMTRACE_MSGT_BD_CEMU_CONFIG command can be used to set/change
a bitmask of optional features that can be enabled in the simtrace
by the host software. This is useful for enabling features in the
firmware which are not supported by older host utilities.
The firmware will return the actually set/configured features in
response.
The feature bitmask is not yet used anywhere; it will be by subequent
patches.
Change-Id: I7cdd3f9171dbed45de0089defe29d2b59044bd84
ch->pts.state is initialized every time we enter ISO_S_WAIT_TPDU
state, so there's no need to reset it at other locations. KISS.
Change-Id: I5ff8b2671c3c893cc5f6542d48b65487ae8c5be1
card_set_state(ch, ISO_S_WAIT_ATR) internally calls tc_etu_enable(),
so there's no need for the caller to do the same, too.
Change-Id: I2a2b4d64c40264623313a867957ffcefe57748f9
A function that is called to set the state should not in itself
contain logic to issue further state changes.
Let's shift the related block out to the end of card_emu_io_statechg(),
which is the only source of card_set_state() calls for the
WAIT_{POWER,CLK,RST} states anyway.
As an added benefit, the block of statements is now also executed if
there's no state change - something that was prevented by the
"if old == new" state guard at the top of card_set_state(). I believe
this may help us to cover more (non-standard) card activation sequences.
Change-Id: Ieefa1807099eb234cfd994bca83caaa0dcc919b6
We need to send zero-length packets on bulk endpoints whenever the
transfer size is exactly a multiple of the wMaxPacketSize.
See USB 2.0 Specification Section 5.8.3 titled "Bulk Transfer Packet
Size constraints."
Change-Id: Ice3842399d5a5c4a18383860f81074497c6e7c9b
Closes: OS#4331
When the Modem is asserting RST, the emulated card should set all
of its state back to default/initial values and release any buffers
it may still hold in its state.
Change-Id: I66eca6afe4ff7d900c5b75df1e3ec6d52f6ef214
Related: OS#4251
So far, we called card_emu_io_statechg() from interrupt context, which
was safe with the current code. However, as we want to trigger a more
comprehensive reset of data structures, including the flushing of
message queues, it is best to shift this processing from interrupt
context to main loop context.
Change-Id: I3983078396538f32bdff55611c41d2163be4d796
Related: OS#4251
Let's move all initialization of state that may change at runtime
to a separate function: card_handle_reset().
This has the advantage that the related function may also be called
at later/other instances, not just during initialization of card_emu.
Change-Id: I0954d192d001a3af5592010cc84a3a22aaf4b030
Related: OS#4251
We need at least one per IN/IRQ endpoint, as well as at least 3 for
every OUT endpoint. Plus some more depending on the application,
in case of cardem there could be one in uart_tx and one in uart_rx.
So for cardem, it should be 7 per slot, resulting 14 for dual-slot
devices like OWHW and QMOD.
Given that we have plenty of RAM available (utilization up from 49%
to 56% with this change), we can be on the comfortable size and
easily have 20 buffers for now.
Future MITM / triple-play applications likely need more.
Change-Id: I022ba3bb3e1f5fd0c364aafe1a23f8fd9b2d8499
Related: OS#4251
In Change-ID Ie15183f16b22193ffdaf01845db2eae4c7f43c17 we wanted
to limit the number of queue elements to 3, but actually implemented 4.
Change-Id: Ic5fb164f2dfd1c293c51bee8184fc34552267a54
Some boards like QMOD cannot simulate a virtual card insert signal
towards the modem (as the mPCIe connector doesn't have that signal).
In these situations, we were aborting processing of the related OUT
endpoint command and forgot to release the related memory.
Change-Id: I489097b7e332004e0f2beca0b22fa36d5d1d31b5
Related: OS#4251
This helps when debugging the firmware, as it shows the current
utliization of the 10-msgb-talloc pool.
Change-Id: Ib10c4396cd4c9c4a6257cf45886e367214787927
Related: OS#4251
