This adds an unconditional endpoint reset procedure to every SET_FEATURE(UnHalt).
It doesn't really make sense that this is required, *particularly* as
we *MUST NOT* set bEndpoint->bank to 0 here.
Without this patch, I'm observing the following problem:
Every first OUT transfer after a SET_INTERFACE + UNHALT on a bulk endpoint
is lost. "lost" means that it completes successfully on the host, can
be seen completing successfully with an ACK on a USB bus analyzer,
but still doesn't show up in the firmware. No Endpoint Interrupt
is generated.
This can be reproduced by calling libusb_set_interface_alt_setting()
from the host and then submitting a single OUT transfer.
Change-Id: I18ed530e617baddf76e8f9829512443ce2a76e0d
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
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
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
the USB trace now respects the global setting.
the verbosity is also decreased, not showing USB activity unless
debugging.
this also saves some space.
the reset cause is now printed.
the strings increase the bootloader size, but it already exceeded
the 16 kB limit when trace level is set to info.
Change-Id: I9ba08d4bb4f188f6e7a202ea86acb7a42a2054f3
now both partitions (bootloader and application) use a commonly
defined memory location to shared the DFU state (which includes
the magic value to know which part to start), instead of using
a hard coded value.
the bootloader size has now also been restricted to 16 kB.
this limitation is enforced so to not be able to create larger
images, which could be corrupted when flashing the application.
bootloader and application flashing have been successfully tested
on qmod st12 and st34.
Change-Id: I204bed7e9391602672ed894decec1fc12e879275
else it's too nosy while debugging other components, not often
used, and break the flow since it does not and a line.
Change-Id: I8920ff7c33b4c9fb174bb31a29334a63fcbede43
as for the main application firmware, the DFU bootloader firmware
now also has the unique chip ID as iSerial in the USB description,
and an additional empty USB configuration indicates the firmware
version (e.g. DFU bootloader version).
these are only visible when the device is in DFU mode.
Change-Id: I11a2cd8079fda374d816da180f39f1c33d10af60
moving the define after the header fixes the following warning:
from ./atmel_softpack_libraries/libchip_sam3s/source/exceptions.c:46:
libcommon/include/stdio.h:63:12: warning: redundant redeclaration of 'printf_sync' [-Wredundant-decls]
signed int printf_sync(const char *pFormat, ...);
^~~~~~~~~~~
./atmel_softpack_libraries/libchip_sam3s/source/exceptions.c:45:16: note: previous declaration of 'printf_sync' was here
#define printf printf_sync
^~~~~~~~~~~
libcommon/include/stdio.h:51:12: note: in expansion of macro 'printf'
signed int printf(const char *pFormat, ...);
Change-Id: I21a8de325e8f8b91297dd157f2d6a0f64434bb28
The default ISR (particularly the HardFault handler) print information,
but this information was not displayed on the console because the UART
IRQ is lower than some default blocking IRQ.
Allowing to set synchronous transfer corrects this.
The underlying Atmel exception library had to be modified to use the
synchronous output.
Making UART_PutChar always synchronous when called from an ISR is not
desired because we use TRACE_ macros is some ISR. The synchronous
output must be set explicitly.
Change-Id: I1b4ace5185cf2dc32684934ed12bf6a8682e9bad
Sometimes descriptor string 0xee is requested.
This is a mechanism used by Microsoft Windows to further identify the USB device.
Instead of stalling, as is the original code, leading to an USB reset, we send an empty packet.
I am not sure if sending an empty string would be better, but an empty packet seems sufficient.
e.g. in CCID mode we need to treat class-specific control requests,
and we want to do this in a way how the CCID code doesn't need to
understand about DFU.
We cannoy simply use the DFU runtime descriptor of the DFU mode, but we
have to use the descriptor of the specific currently-selected runtime
configuration. Let's iterate over the descriptors of a configuration
and find the DFU runtime descriptor in it.
Using the USBDFU_OverrideEnterDFU() function, a board/application can
define extra conditions when the system should boot in DFU mode, even if
it was not explicitly switched to DFU mode from the application.
The app/dfu/main.c uses this mechanism to boot into DFU mode if the
stack + reset vector addresses are not plausible (i.e. some random junk
appears to be flashed in the application partition) or if the user
places a jumper accross the RxD+TxD lines of the debug UART. The idea
is that the system can be recovered by placing this jumper and then
re-installing the application from DFU.
It is utterly unacceptable that Atmel is shipping example code in 2011,
which doesn't support parts that are shipped still in 2016. They would
have had five years to fix their code to implement the chip errata :(
We need to refer to g_dfu as a pointer from all code. In DFU mode, it
gets assigned to the address of _g_dfu, which is placed by the linker
script at the start of RAM. In runtime mode, the pointer is statically
defined to point at the start of RAM. The linker script for the runtime
(dfu environment) is adjusted to save the first 32 bytes for the _g_dfu
structure.
* we now actually route the EP0 control requests in DFU mode to the
correct handler (weak linker symbols are tricky)
* we now actually call code to read/write data from/to RAM/FLASH
This way we can easily check with 'diff' for differences in our code and
Atmel softpack. Also, this layout is more suitable for building various
different firmware images (e.g. factory-test, dfu-loader, main
application) for a variety of different boards (simtrace, owhw, qmod).
