add module_t::trap_taking_match and trigger_t::trap_taking_match for
checking itrigger after taking traps

triggers: hardware should ignore writes that set dmode to 1 if the previous trigger has both dmode of 0 and chain of 1

triggers: refactor: remove return value of mcontrol_t::tdata1_write(...) and trigger_with_tdata2_t::tdata2_write(...)

Write ignorance, i.e. dmode and chain, are checked in module_t

Mapping of previous and revised match_result_t data type:
-------------------------------------------
   Previous                 Revised
MATCH_NONE               fire=false (don't care timing)
MATCH_MATCH_FIRE_BEFORE  fire=true,timing=TIMING_BEFORE
MATCH_MATCH_FIRE_AFTER   fire=true,timing=TIMING_AFTER

chain() -> get_chain()
chain_bit -> chain

The tdata1.dmode is defined at the same level of tdata1.type. As a
result, all triggers has a tdata1.dmode field (even the disabled
trigger).

With the above consideration, we move the tdata1.dmode checking and
setting to the same level of tdata1.type, i.e., in module_t.

dmode() -> get_dmode()
trigger_t.dmode_bit -> mcontrol_t.dmode

store() -> get_store()
store_bit -> store

load() -> get_load()
load_bit -> load

execute() -> get_execute(), execute_bit -> execute

structuralize trigger's tdata CSRs for modularization

Add tinfo register.

Not very interesting while spike only supports one trigger type,
but #1128 is about to change that. Without tinfo, it can become quite
slow for a debugger to discover which types are supported.

If a variable other than `i` is passed as `inx`, `VI_LOOP_ELEMENT_SKIP`
will still check `i` in the `else` case, ignoring `inx`.

Since `i` is the only value ever passed, drop the argument to avoid any
future confusion.

Signed-off-by: Dylan Reid <dgreid@rivosinc.com>

Signed-off-by: Dylan Reid <dgreid@rivosinc.com>

When resuming from debug mode, clear mstatus.MPRV if the new privilege mode is less than M-mode (#1149)