this makes the assumption that DPs will be v1 or higher, for SWD-DP scans this is
guaranteed, but on JTAG-SCANS it may not be true, DPv0 does not have DPIDR
implemented and reads are UNPREDICTABLE
Signed-off-by: Rafael Silva <perigoso@riseup.net>
add missing fetching of targetid in jlink probes
clarify how targetid is being read
handle idcode as debug_port_id
use targetid when available to identify device in probe routine
Expect signal integrity errors when using jumper cables. Often probes switch
the SWJ GPIO with highest speed, resulting in possible reflections. Additional
ground wires may help. If there is isolation between probe and targets,
additional ground wires are a must or ground shift will wrack the transfer!
E.g. AP1 on a STM32WLE5 points to a ROM table, but access to the ROM table
via AP1 hangs forever.
- Substantial reduce timeout when wait for a response. Valid access should
succeed fast.
- Abort AP access to free DP for other accesses
- Don't throw exception, only set dp->fault
- React on higher level
- RP2040 show both DPs
- Multidrop test with STM32L552 and STM32H745 allows selection
with "-m 0x4500041" (H7), "-m 1" (L552) or "-m 0x01002927" (RP2040)
Especially needed when leaving the debugger or during debug unit power-up.
ARM Debug Interface Architecture Specification ADIv5.0 to ADIv5.2
tells to clock the data through SW-DP to either :
- immediate start a new transaction
- continue to drive idle cycles
- or clock at least 8 idle cycles
Implement last option to favour correctness over slight speed decrease
Implement only for adapters where we assemble the seq_out_parity in our code,
as on firmware, ftdi and jlink. Hopefully the high level adapters do it right.
Reverts 2c33cde63f and
cde7726b87
Firmware BMP with ENABLE_DEBUG=1 will print WARN and INFO as before.
PC-Hosted BMPwill alway print to stderr. Warn is printed unconditional,
INFO, GDB, TARGET, DONGLE and WIRE will print if their appropriate bit in
cl_debuglevel is set via the -v verbose command line argument.
INFO will go to stdout with -t or -l.
This is will make debugging earier if this does happen, rather than
dereferencing the null pointer (or passing it to memcpy, or worse).
blackmagic PR #475
These extra cycles are needed by some CPU, e.g. STM32L0x1 to cross the SWCLK
/HCLK domains. Revert insufficient #373 also tackling that problem.
Thanks to Thorsten von Eicken for pointing out.
dragonmux
Rafael Silva
mean
Uwe Bonnes
Björn Mellström
Valmantas Paliksa
Richard Meadows