[S] Change in flashrom[master]: dummyflasher: use new API to register shutdown function

[Alexander Goncharov (Code Review)]

Attention is currently required from: Edward O'Callaghan, Anastasia Klimchuk.

Alexander Goncharov has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/72408 )

Change subject: dummyflasher: use new API to register shutdown function
......................................................................

Patch Set 2:

(1 comment)

File dummyflasher.c:

https://review.coreboot.org/c/flashrom/+/72408/comment/9e789263_e0994291 
PS2, Line 1416: data->refs_cnt += ((dummy_buses_supported & BUS_PROG) != BUS_NONE);
              : 	data->refs_cnt += ((dummy_buses_supported & BUS_NONSPI) != BUS_NONE);
              : 	data->refs_cnt += ((dummy_buses_supported & BUS_SPI) != BUS_NONE);
              : 
              : #if 0
              : 	data->refs_cnt += (dummy_buses_supported & BUS_PROG) ? 1 : 0;
              : 	data->refs_cnt += (dummy_buses_supported & BUS_NONSPI) ? 1 : 0;
              : 	data->refs_cnt += (dummy_buses_supported & BUS_SPI) ? 1 : 0;
              : #endif
> I have another idea! :) What if you add `refs_cnt++` into every if below. […]

I thought about it at first sight. But this approach may potentially cause flashrom to crash.

If the dummy programmer registers more than one bus, then the next accident can happen => If the first condition the programmer enters fails because `register_*_master()` could not register shutdown (`register_shutdown()`), then it will call shutdown immediately (`mst->shutdown(data)`, opaque.c:55). And since we only have one owner in refs_cnt at that moment, data will be released.

When the dummy tries to enter the second condition and increase `data->refs_cnt`, flashrom will crash because this memory isn't valid anymore.

This should be mentioned in the code, but I cannot come up with brief comment for that.