Re: [patch/RFA] multiarch INSTRUCTION_NULLIFIED

[Orjan Friberg <orjan dot friberg at axis dot com>]

Mark Kettenis wrote:
> That's why we had the STEP_SKIPS_DELAY macro that makes sure the
> single-step is done without any breakpoints inserted.  This was
> replaced by single_step_through_delay which presumably will also be
> used on a new CRISv32 port.

Yes, CRISv32 basically has the MIPS situation in reverse: if you single-step 
into a delay slot (belonging to a branch instruction) and single-step again, the 
branch instruction will be re-executed.  As a consequence, if there is a 
breakpoint at the branch instruction itself, we need to step again before 
re-inserting it (otherwise we'd just hit it again).  (I like to think of 
single_step_through_delay as "did we just step into a situation where we need to 
step again".)

--
Orjan Friberg
Axis Communications