1
0
Commit Graph

3508 Commits

Author SHA1 Message Date
Wesley W. Terpstra
d2421654c4 tilelink2: refactor address into addr_hi on ABC and addr_lo on CD
We need addr_lo in order to properly convert widths.
As part of the refactoring, move all methods out of the Bundles
2016-09-06 23:46:44 -07:00
Yunsup Lee
b76612f357 relax contraint on adding AddrMapEntry to AddrMap (#248)
now you can add them in any order.  there's an explicit check at the end to figure out whether there are overlapping regions.
2016-09-06 21:53:55 -07:00
Howard Mao
7504498dff Merge pull request #247 from ucb-bar/replseqmem_pr
Bump FIRRTL for Sequential Memories
2016-09-06 17:22:18 -07:00
Megan Wachs
e95fe646a3 mem_gen failure doesn't create the target 2016-09-06 16:29:29 -07:00
Howard Mao
bbef3a8d3e Merge pull request #246 from ucb-bar/fix-configstring-printout-problem
fix configstring printout with no memory
2016-09-06 15:31:39 -07:00
Megan Wachs
48098f5e2d Bump FIRRTL to instantiate Sequential Memory Macros 2016-09-06 14:48:28 -07:00
Megan Wachs
1fec9807f6 allow override of vlsi_mem_gen script 2016-09-06 14:44:12 -07:00
Wesley W. Terpstra
aae4230627 tilelink2: fix bugs found by Megan in Legacy converter 2016-09-06 13:12:33 -07:00
Yunsup Lee
56d81b0034 fix configstring printout with no memory 2016-09-06 10:40:11 -07:00
Wesley W. Terpstra
54ab14cd9d tilelink2: statically optimize numBeats for simple managers 2016-09-05 22:11:03 -07:00
Wesley W. Terpstra
314d6ebd6f tilelink2: stricter TransferSizes requirements 2016-09-05 22:10:28 -07:00
Wesley W. Terpstra
56170c605c tilelink2: be more forgiving in what Legacy TL requires 2016-09-05 21:12:51 -07:00
Wesley W. Terpstra
3167539331 tilelink2: Narrower must be little-endian 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
ded246fb95 tilelink2: relax max transfer size; the real requirement is not exceeding alignment 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
cf0291061d tilelink2: fix a bug in UIntToOH1 triggered if the size was too big 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
9f45212c95 tilelink2: Fragmenter needs to update subaddress 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
757d46279e tilelink2: expand data correctly in D channel narrower 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
0faa8c4051 tilelink2: fix Xbar bug where Mux1H broke FSM if only one manager 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
a0c25880c7 tilelink2: Monitor should check mask of reconstructed request 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
df32cc3887 tilelink2: be careful; apply Andrew's masking trick everywhere 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
fb262558ee tilelink2: helper objects should pass source line from where they were invoked 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
1a081b4dd5 tilelink2: Monitor should report which TL connection was the problem 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
cb54df0a8a tilelink2: tie off unused channels 2016-09-05 20:58:41 -07:00
Wesley W. Terpstra
68e64a9859 tilelink2: clarify ready-valid use of RegisterRouter 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
e3b3543841 tilelink2: ensure RegFields don't exceed their bounds 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
8343070639 tilelink2: detect 1-bit overflow in register definitions 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
a1fc01fd6d tilelink2: prevent mapping the same register twice 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
81162a2dc9 tilelink2: support attaching a DecoupledIO directly to a register 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
6a378e79e3 tilelink2: allow 0-stage backpressure in combinational regmap 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
4746cf00ce tilelink2: move files to new uncore directory 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
e034775bfa tilelink2: use the fancy new hasData functions 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
11b0272d91 tilelink2: create optimized hasData method on edges (statically evaluates if known) 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
5db7ae262b tilelink2: first version of Narrower (only supports uncached IO) 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
b004d54d71 tilelink2: add a Fragmenter adapter
God that was a pain in the ass!
2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
ecc3c2a4b2 tilelink2: more efficient one-hot circuits 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
3d84795641 tilelink2: use LazyModule(new ...) just like Chisel Module(new ...) 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
2069ca5d8d tilelink2: pass sourceInfo using implicits in Monitor 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
935b53f3bf tilelink2: explicitly check that fixed fields never change in multibeat 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
e652cd155b tilelink2: edge parameters on the same link had better match 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
c411a3e77f tilelink2: simpler sizes requirement for users to understand 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
ab998c08f1 tilelink2: save some hardware in HintHandler if no BCE 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
18e7d4cd65 tilelink2: make it possible to write Node-only adapters 2016-09-05 20:58:40 -07:00
Wesley W. Terpstra
4a401fc480 tilelink2: add a Buffer adapter to insert pipeline stages 2016-09-05 20:58:39 -07:00
Wesley W. Terpstra
50f0dee69e tilelink2: add an IdentityNode for adapters that change nothing 2016-09-05 20:58:39 -07:00
Wesley W. Terpstra
9cd2991fb3 tilelink2: AddressSet always has an assigned base address
The consensus seems to be that TileLink should not be assigning
addresses dynamically. The reasons:

1. We can come up with another scheme for assigning addresses that is
   independent of TileLink.  This decoupling is good, because it would
   allow us to use the same mechanism for different buses in the SoC.

2. The informational flow of addresses is more likely to naturally follow
   the module hierarchy than the TileLike bus topology. Thus, it seems
   better to pass address parameterization using Module constructors.

3. Addresses are still checked by TileLink, so using a Module-centric
   flow for addresses will not pose a correctness concern.

4. An address need only be provided to a slave on its construction and
   TileLink parameterization spreads this globally. Thus, the burden to
   manually assign an address is low.
2016-09-05 20:58:39 -07:00
Wesley W. Terpstra
ae2bc4da21 tilelink2: refactor RegField into interface and implementation 2016-09-05 20:58:39 -07:00
Wesley W. Terpstra
d6727abbbc tilelink2: rename Operations to Edges (as it only includes Edges) 2016-09-05 20:58:39 -07:00
Wesley W. Terpstra
ee3e31cb23 tilelink2: refactor TLNodes into a seperate file 2016-09-05 20:58:39 -07:00
Wesley W. Terpstra
69b3de92a8 tilelink2: decouple BaseNode from TileLink bus (so it can be reused) 2016-09-05 20:58:39 -07:00
Wesley W. Terpstra
c785375276 tilelink2: use 'connect' instead of TL-specific 'tl' to connect nodes 2016-09-05 20:58:39 -07:00