Hopefully this stops .depends.mk from being generated more than
necessary, as presumably the fact that compile and simulator both add
new files to build_dir (and Verilog_RTL for Verilator) causes the
timestamp to change and thus make to think .depends.mk is stale.
We have a lot of ?s floating around that needlessly constrain logic. As
with Flute, enable optimisations on these in the hope that timing and/or
area improve further.
When DEBUG_WEDGE is defined, expose the last committed and next in the
reorder buffer PC and corresponding instruction via DMI registers, since
even when the core is wedged and we can't read GPRs etc we can still
interact with the debug module itself. Hopefully this proves useful for
debugging wedges.
Also use RVFI_DII not RVFIDII in the directory names.
This makes everything match Piccolo/Flute rather than having Toooba be a
weird, inconsistent and plain wrong.
This one has the advantage of being able to be called with the same
flags as bsc, rather than needing to pass things through special
environment variables. As a result, revert all our changes to dealing
with BSC_COMPILATION_FLAGS (some of this diff therefore looks strangely
formatted, but it's to match upstream verbatim, and should be left that
way to minimise diffs and avoid conflicts).
New method 'debug_external_interrupt_req' to support emulation of a
debug module starts at P3_Core interface and is plumbed all the way in
to the CSR register MIP as interrupt [14]. The corresponding MIE[14]
is always 1, so it is never masked. Still todo: should not be masked
by MSTATUS interrupt-enables either. Also expanded
interrupt-detection logic, mcause etc. to extend up to interrupt 14.
Builds in standalone mode, runs ISA tests.
Builds in src_SSITH_P3, generating RTL.
