[J-core] Equiv of a datasheet?

D. Jeff Dionne Jeff at SE-Instruments.com
Fri Jan 6 21:14:25 EST 2017

GDB itself should be compiled to run on the host system, not the target.   Basically, --target=sh2-elf has worked for quite some time, and the debug stub complied into the standard boot rom will talk to it.   We have a JTAG server that can talk to that GDB also.

The problem with both is (at least) GDB isn't really reliable around certain instructions, notably branches.  JTAG has another problem with tracking pipeline state, but that is j-core bugs not general SH GDB breakage.

gdbserver has a few other problems that Rich ran into.

GDB is my favoured debugging method, having it not work reliably is a pain...


> On Jan 7, 2017, at 09:38, Rob Landley <rob at landley.net> wrote:
>> On 01/06/2017 12:31 PM, Don A. Bailey wrote:
>> This is great! Thanks for the information. 
>> Does gdb work on sh2 (j2)? I can't for the life of me seem to build gdb
>> that supports either sh2 or sh4. The closest I got was building
>> gdbserver from the gdb-5.3 package, but then I couldn't actually cross
>> compile gdb, itself. 
> That's a question for Rich Felker, but unfortunately he's out this week
> due to a house fire:
>  https://twitter.com/RichFelker/status/817103182283280386
> My understanding is that gdb works but has the usual nommu limitations
> (breakpoints can be put on code but not data access, etc), but I don't
> use gdb unless I have no other choice. (Over the years I've stuck print
> statements in u-boot's dram init code and uClibc's dynamic linker
> self-relocation code. I have a black belt in debugging via printf, or
> printk, or write(), or
> https://balau82.wordpress.com/2010/02/28/hello-world-for-bare-metal-arm-using-qemu/
> or...)
> Rob
> _______________________________________________
> J-core mailing list
> J-core at lists.j-core.org
> http://lists.j-core.org/mailman/listinfo/j-core

More information about the J-core mailing list