Andrew Waterman
97e74aec3a
Merge RTC and PRCI
2016-06-27 23:06:07 -07:00
Howard Mao
ec5b9dfc86
make sure trackers can handle case where there are no caching clients
2016-06-27 16:29:51 -07:00
Howard Mao
a93a70c8ec
make sure merged voluntary releases are handled properly
2016-06-27 11:40:32 -07:00
Andrew Waterman
354b81c8fe
Remove legacy HTIF things
...
The SCR file is gone, too, because it is tightly coupled. The
general concept could be revived as a module that somehow connects
to (or is contained by) the debug module.
2016-06-23 13:17:11 -07:00
Andrew Waterman
f57524e0c1
Remove FENCE.I from debug ROM; specialize for RV64
2016-06-23 00:01:26 -07:00
Howard Mao
e3391b36b2
get rid of MuxBundle now that MuxCase and MuxLookup are fixed
2016-06-21 10:43:44 -07:00
Howard Mao
719fffff40
make sure updates from irel and iacq gated by tracker allocation
2016-06-17 17:15:02 -07:00
Howard Mao
b75b6fdcda
make sure no-data voluntary releases get tracked
2016-06-17 17:15:02 -07:00
Howard Mao
ebe95fa827
fix wmask buffer clearing in L2 agents
2016-06-16 15:34:31 -07:00
Howard Mao
aba13cee7f
fix BRAM slave so that it can correctly take all TileLink requests
2016-06-16 15:34:31 -07:00
Howard Mao
e716661637
make sure merged no-alloc put still allocs if original put allocs
2016-06-16 15:34:31 -07:00
Howard Mao
7e43b1d889
fix mistaken dequeueing from roq in TileLink unwrapper
2016-06-16 15:34:31 -07:00
Howard Mao
2789e60b6b
fix ignt_q logic
2016-06-16 15:18:58 -07:00
Henry Cook
16bfbda3c9
Refactor the TransactionTracker logic in all the L2 TileLink Managers.
...
They now share common sub-transactions within traits, and use a common
set of state transitions and scoreboarding logic. Tracker allocation
logic has also been updated. No changes to external IOs or the TileLink protocol.
A new bufferless Broadcast hub is also included, but does not yet pass fuzzing checks.
2016-06-16 15:18:48 -07:00
mwachs5
2d2096e509
Add smaller ROM/RAM for 32-bit debug ( #60 )
2016-06-15 15:07:43 -07:00
Palmer Dabbelt
e5cfc2dac1
Add a Smi to TileLink converter ( #59 )
...
I'm trying to get someone to attach their stuff to Rocket Chip for the
upcoming tapout. TileLink sounded too complicated, but Smi went over
well. Since the mmioNetwork in Rocket Chip is based on TileLink, it
seemed like the easiest thing to do was to write a TileLink to Smi
converter so people could use it.
It turns out there was already one inside the groundtest unit tests, so
I just moved that into uncore (it was inlined into a test case so you
couldn't actually use it before). Internally the converter uses Nasti,
but I figured that's good enough for now.
2016-06-10 14:04:28 -07:00
Megan Wachs
cee0cf345e
[debug] Update Debug ROM contents to write F..F to RAM in case of exception
2016-06-09 14:05:30 -07:00
Wesley W. Terpstra
a1ebc73477
tilelink: don't accidentally make a malformed union
...
Closes #55
2016-06-09 10:44:00 -07:00
Wesley W. Terpstra
31b72625aa
ahb: allow no-ops to progress also when a slave is !hready
2016-06-09 10:41:12 -07:00
Wesley W. Terpstra
7014eef339
ahb: fix bugs found using comparatortest
2016-06-09 10:41:11 -07:00
mwachs5
93c1b17b52
[debug] Remove erroneous buffer on SB read data ( #56 )
2016-06-08 23:31:13 -04:00
Albert Ou
5151570894
Fix valid signal for multibeat grants
2016-06-08 15:13:39 -07:00
Howard Mao
f421e2ab11
fix TileLinkWidthAdapter
2016-06-08 09:58:23 -07:00
Wesley W. Terpstra
324cabc494
tilelink: wmask was double the width it should be
...
When amo_offset = UInt(0), UIntToOH(amo_offset) = "b01", not b"1".
This meant that the amo wmask was double wide, making wmask() fat.
2016-06-07 14:04:01 -07:00
Howard Mao
2d66ac93d3
make sure HastiRAM cuts off the correct number of bits for word address
2016-06-06 09:26:51 -07:00
Andrew Waterman
dd85f2410f
Avoid need for cloneType
2016-06-05 23:47:56 -07:00
Andrew Waterman
631e3e2dd9
Make PRCI a singleton, not per-tile
...
Some stuff is densely packed in the address space (e.g. IPI regs),
so needs to be on the same TileLink slave port
2016-06-05 23:06:21 -07:00
Andrew Waterman
be7500e4a9
Update PLIC addr map
2016-06-05 23:04:51 -07:00
Megan Wachs
b832689642
Correct Debug ROM contents
2016-06-05 19:35:25 -07:00
Megan Wachs
605fb5b92f
[debug]: fix issue with subword select logic
2016-06-05 19:31:07 -07:00
Megan Wachs
3e8322816b
Correct DMINFO Fields
2016-06-05 19:29:50 -07:00
Megan Wachs
7e550ab07c
[debug] rocket: fix for issue 121, correct debug ROM and stall logic
2016-06-05 19:29:44 -07:00
Andrew Waterman
2e88ffc364
Cope with changes to AddrMap
2016-06-03 13:48:09 -07:00
Andrew Waterman
f1745bf142
Allow PLIC nPriorities=0 (priority fixed at 1)
2016-06-02 13:48:29 -07:00
Andrew Waterman
b7ca2145b3
Fix PLIC control bug when !grant.ready
2016-06-02 13:47:59 -07:00
Andrew Waterman
0866b4c045
Can't assign to Vec literals
2016-06-01 23:36:34 -07:00
Andrew Waterman
20e1de08da
Avoid chisel2 pitfall
...
This code is erroneously flagged as incompatible with chisel3.
In fact, it is correct in both chisel2 and chisel3. D'oh.
2016-06-01 23:35:49 -07:00
Andrew Waterman
5629fb62bf
Avoid bitwise sub-assignment
2016-06-01 21:59:02 -07:00
Andrew Waterman
9518b3d589
Fix arithmetic in ROM row count
2016-06-01 21:59:02 -07:00
Andrew Waterman
8e80d1ec80
Avoid floating-point arithmetic where integers suffice
2016-06-01 21:59:02 -07:00
Wesley W. Terpstra
11b3cee07a
Ahb tweaks ( #50 )
...
* ahb: handle tlDataBytes==1 and tlDataBeats==1 gracefully
I only now learned that chisel does not handle 0-width wires properly
and that log2Up and log2Ceil differ on 1. Fix-up code to handle this.
* ahb: optionally disable atomics => optimize to nothing
Trust the compiler the compiler to optimize away unused logic.
2016-06-01 16:42:39 -07:00
mwachs5
740a6073f6
Add Debug Module ( #49 )
...
* Add Debug Module
* [debug] Remove unit tests, update System Bus register addresses, parameterize nComponents
* [debug] Update Debug ROM contents to match updated addresses
2016-06-01 16:33:33 -07:00
Howard Mao
e8408f0a8a
fix HastiRAM
2016-06-01 10:33:59 -07:00
Andrew Waterman
6d82c0d156
Add M_FLUSH_ALL command
2016-05-31 19:25:31 -07:00
Andrew Waterman
8afdd7e3da
Work around PutBlocks draining into data array prematurely
2016-05-26 23:08:05 -07:00
Andrew Waterman
391a9b9110
Use buses, rather than crossbars, by default in TLInterconnect
...
We should eventually parameterize this, of course.
2016-05-26 16:10:42 -07:00
Andrew Waterman
b6d26e90f8
Add generic TileLink width adapter
2016-05-26 15:59:42 -07:00
Andrew Waterman
8139f71dfb
Work around Chisel2 bug
...
This code is correct, but Chisel2 erroneously flags it as a Chisel3
compatibility error because it looks like Vec(Reg) when factor=1.
2016-05-26 12:37:31 -07:00
Andrew Waterman
22568de5f3
Work around another zero-width wire limitation
2016-05-25 21:42:02 -07:00
Andrew Waterman
e2755a0f0a
Work around zero-width wire limitation in HTIF
2016-05-25 20:39:53 -07:00