make CFG survive post-reload splitting: SH breakage

[Jan Hubicka]

> This didn't used to matter because split_all_insns() didn't care about
> basic block info at this point in the compilation.  Your patch changed
> all that, and now we do care about basic block information:
I see, I didn't get the idea that machine description may call split_all_insns
directly.
SH, IA64 and m68hc11 does this.

Cleanest soltuion to me seems to be to add function split_all_insns_noflow
that would do the job for these targets.

I will try to come with patch shortly.

Honza