* JTAG: Revert to Chisel._ for Issue 1160
* JTAG: Revert to Chisel._ for Issue 1160
* jtag: revert everything to Chisel._
* jtag: Revert all modules to Chisel._ vs chisel3, due to FIRRTL issues with chisel3 generated code
After detecting a corrupted BTB, don't speculatively update it until
the next non-speculative fetch. This prevents the frontend from replaying
forever.
The reason for the :=? operator was for when you have an adapter chain
whose direction of cardinality you could not know. We used explicit
directives to tell these compositions which way to go.
Unfortunately, that makes the API leaky. You think the chain of adapters
is just one adapter, but you have to use strange Cardinality scopes to
use it. That's just bad.
The new :*=* just automagically figures it out from the graph.
This removes the mostly obsolete 'numIn/Out' range restrictions on nodes.
It also makes it possible to connect optional crossbars that disappear.
val x = TLXbar()
x := master
slave := x
val y = TLXbar()
x :=* y // only connect y if it gets used
This will create crossbar x, but crossbar y will disappear.
RegField.bytes updates only those bytes which are written every cycle.
However, there was a bug that it would try to return the updated value on reads.
This led to another TL-spec violating combinational path, just like the Debug module.
When passed a Wire, WNotify outputs that wire on reads wire => d_bits.
Furthermore, it updates the Wire when a write occures d_ready => wire.
These registers should be returning undefined value on read, anyway.
This esoteric bug manifests if a tag-read error occurs when a FENCE.I is
executed, even if the error was correctable. Subsequently, an attempt to
flush a dirty line may flush the wrong line's data.
This esoteric bug manifests if a tag-read error occurs when a FENCE.I is
executed, even if the error was correctable. Subsequently, an attempt to
flush a dirty line may flush the wrong line's data.
Putting the common DTS nodes into a shared object makes them get
emitted only one time. Plus it's better style.
timebase-frequency should really have been in the cpu nodes in the
first place according to the spec anyway. I was foolishly trying to
save bytes. However, now we really want it there in case it differs.
Deallocation can change repl_way, which violates the assumption that it
remains constant throughout refill.
The workaround described in commit 3db066303b
still suffices, provided only the hart that owns the ITIM changes the ITIM
allocation.
This subsumes commit 3db066303b.
...instead of on the master side of the system bus.
People inheriting from HasTileMasterPort might need to add
`masterNode := tileBus.node` to their Tile child class.
* debug: Update macros from spec
* debug: some corrections in the auto-generated files
* debug: update renamed fields
* Debug: implement the implicit ebreak option for small program buffers
* debug: clean up some unused code and add more require() explanations
* debug: make implicit ebreak false
* debug: Add the havereset/haveresetack functionality
* debug: program buffer can still be 16 even if there is an implicit ebreak
This is a follow-up to PR #1108.
Rather than increasing the number of transactions we allow to be inflight,
instead just block TL when early source re-use happens. This is a better
fix since it means we don't pay mostly wasted downstream hardware to handle
an additional transaction inflight that almost never happens.
* TLToAXI4: fix WaR for single-source FIFO masters
* TLToAXI4: fix potential counter overflow => WaR hazard
If you have a FIFO master with 2^n-1 sources that performs early
source re-use, the old code could potentially break FIFO order.
- Put correctness responsibility on Frontend, not IBuf, for improved
separation of concerns. Frontend must detect case that the BTB
predicts a taken branch in the middle of an instruction.
- Pass BTB information down pipeline unconditionally, fixing case that
screws up the branch history when the BTB misses and the instruction
is misaligned.
- Remove jumpInFrontend option; it's now unconditional.
- Default to one-bit counters in the BHT. For tiny BHTs like these, it's
more resource efficient to have a larger index space than to have
hysteresis.
Fragmenter: add a third case for earlyAck (PutFulls only)
It seems quite common to have a device that is backed by ECC. When
performing a multibeat PutPartial, these devices can exhibit their
first error on the last beat (if it had an incomplete write mask
for that beat, which required read-write-modifying corrupted data).
Generally, these devices have ECC granularity <= the bus width. In
those cases, if you send a PutFull, the first beat carries the
error value for the whole burst. Consider:
If the PutFull was below the granularity, it was a single beat.
If the PutFull was multi-beat, it exceeds the granularity.
Therefore, an important variation on the earlyAck optimization is
the case where only PutFulls receive an earlyAck.
When switching ports, the bypass stalls new messages until all
outstanding messages have received their responses. However, this
stall must NOT stop the remaining beats of a partially sent request.
* tilelink: ToAXI4 should format it's error message
* WithStatelessBridge: mark the memory bus incoherent and cacheable
... and hope that the user doesn't put more than one master down.
* JTAG: Use sorted map for stability
Otherwise the generated FIRRTL/Verilog is non deterministic
* jtag : parens for clarity
* jtag: Use deterministic ListMap and sort for stability
* JTAG: use slightly clearer SortedMap (clearer to me anyway)
* jtag: whitespace cleanup
From the AHB spec:
"A slave only has to provide valid data when a transfer completes with an OKAY
response. ERROR responses do not require valid read data."
This makes it possible to treat chained composition associatively.
x := y :=? z :=* a ...
It also makes it easy to chain multiple optional adapters:
node :=? (Seq(a, b) ++ c ++ d)
* CSR provides a decode port to check for an illegal instruction.
* This commit now allows for multiple instructions in decode to get this
illegal instruction information.
* This commit leverages the existing decodeWidth parameter. This will
potentially over-provision the number of decode ports needed for
RVC-enabled cores.
Closes#1068