> On 13 Jul 2023, at 22:09, Tomas Vanek <tom_...@users.sourceforge.net> wrote:
>
> It should *NOT* resume. The main purpose of a breakpoint is to stop the
> execution.
Actually "to stop the execution **during a debug session**".
> Connecting OpenOCD to a target implies enabling the debug mode of the device
> so IMO there is nothing like "standalone" mode.
I don't know about OpenOCD specifics, but, according to Arm:
"Semihosting is a mechanism that enables code running on an ARM target or
emulator to communicate with and use the Input/Output facilities on a host
computer. The host must be running the emulator, or a debugger that is attached
to the ARM target."
For the full definition, please see:
-
https://github.com/ARM-software/abi-aa/blob/main/semihosting/semihosting.rst#3introduction
In its essence, semihosting **is** "standalone", it works like a monitor (for
example RDIMON), and, even when implemented by a debugger, it does not need to
run in a manned debug session, it just needs to intercept the specific BKPT
instructions, the rest should run unconstrained.
If OpenOCD does not yet implement this mode, it can be improved.
A few years ago when I re-wrote the semihosting code in OpenCOD I needed it
only for the Eclipse OpenOCD debug plug-ins, i.e. for use during debug
sessions.
Now I moved my development environment outside Eclipse, and I would need it to
run independently of any debug sessions.
> I would also guess that the problem is a silicon errata in some versions of
> Cortex-M7, not yet documented.
> M7 is the first Cortex-M core with the double-issue pipeline so probably the
> execution of BKPT and interrupt at the same time
> could lead to debug halt at the interrupt context.
Possibly. And the STM32H7, being newer, does not have this bug.
> BTW combining semihosting with interrupts is not a good practice anyway.
> Except some basic tests the interrupts are unusable due
> to giant latencies caused by debug halts.
Hmmm... Can you be more specific on this? What halts do you mean? I definitely
do not set any breakpoints in my tests, and I expect the interrupts to run at
full speed.
Liviu
