Apologies if this is a wrong forum.
Keil's proprietary tracing can pack 4 characters in a 32-bit word.
The data registers of ITM are 32-bit, so it looks to me very attractive to print 4 characters at once.
Is there any implementation that supports this, maybe some fork of OpenOCD?
A "mainstream" OpenOCD seems to interpret 16 or 32 bit data as "wide" unicode characters. It's the time to embrace UTF-8 there. So Unicode will work too.
Any thoughts?
Are there solutions that utilize high bits of ITM_SendChar data, such as multiplexing output to N "streams"?
Regards,
Pavel A.