riscv/rocket-chip
1
0
Fork 0
Code Releases Activity
536 Commits 1 Branch 0 Tags 13 MiB
53ad8387cc
Commit Graph

510 Commits

Author SHA1 Message Date
Howard Mao
a551a12d70 add missing Wire wrap in BasicCrossbar 2015-08-05 17:05:31 -07:00
Andrew Waterman
eb6583d607 use cloneType in PhysicalNetworkIO 2015-08-05 16:47:49 -07:00
Andrew Waterman
798ddeb5f5 Chisel3 compatibility: use >>Int instead of >>UInt 
The latter doesn't contract widths anymore.
 2015-08-04 13:15:17 -07:00
Andrew Waterman
fb718f03c1 bump scala to 2.11.6 2015-08-03 19:50:58 -07:00
Andrew Waterman
77cf26aeba Chisel3: Flip order of := and <> 2015-08-03 18:53:39 -07:00
Andrew Waterman
121e4fb511 Flip direction of some bulk connects 2015-08-03 18:01:14 -07:00
Andrew Waterman
a21979a2fa Bits -> UInt 2015-08-03 18:01:06 -07:00
Andrew Waterman
9c7a41e8d3 Chisel3: bulk connect is not commutative 
We haven't decided if this is a FIRRTL limitation that we should relax,
or a backwards incompatibility we're forced to live with.  Should make
for lively debate.
 2015-08-01 21:09:00 -07:00
Andrew Waterman
6fc807f069 Chisel3: Avoid subword assignment 2015-08-01 21:08:35 -07:00
Andrew Waterman
6d574f8c1b Fix incompatible assignment 2015-07-31 00:59:34 -07:00
Andrew Waterman
377e17e811 Add Wire() wrap 2015-07-31 00:32:02 -07:00
Andrew Waterman
0686bdbe28 Avoid cross-module references 
You can't instantiate a Vec in one module and use it in another.
An idiosyncrasy of the Chisel2 implementation let this one slip by.
In this case, it's just a matter of using def instead of val.
 2015-07-30 23:49:06 -07:00
Andrew Waterman
8f7b390353 UInt-> Bits; avoid mixed UInt/SInt code 2015-07-30 23:49:06 -07:00
Andrew Waterman
6c391e3b37 Use UInt(0), not UInt(width=0), for constant 0 2015-07-30 23:49:06 -07:00
Jim Lawson
4c0f996808 Fix typo (juntion -> junctions). 2015-07-30 14:50:28 -07:00
Henry Cook
c70b495f6d moved buses to junctions repo 2015-07-29 18:04:30 -07:00
Henry Cook
8b1ab23347 update README.md 2015-07-29 11:49:21 -07:00
Henry Cook
4daa20b5fe simplify .sbt files 2015-07-29 11:49:20 -07:00
Andrew Waterman
a69c749249 Fix compilation with scala 2.11.6 
We forgot to specify return types on overloaded methods, and a previous
version of the scala compiler failed to flag this as an error.
 2015-07-28 16:24:45 -07:00
Andrew Waterman
f8ec6d6393 Chisel3 compatibility: use BitPat for don't-cares 
Also, call the UInt factory instead of the Bits one, for good measure.
 2015-07-28 02:46:23 -07:00
Andrew Waterman
0e06c941df Chisel3 compatibility fixes 2015-07-23 14:58:46 -07:00
Andrew Waterman
3c0475e08b Add Wire() wrap 2015-07-15 20:24:03 -07:00
Andrew Waterman
2d6b3b2331 Don't use clone 2015-07-15 18:06:27 -07:00
Andrew Waterman
276f53b652 Delete BigMem; it's not used anymore 2015-07-15 17:41:47 -07:00
Andrew Waterman
15cec0eab7 Vec(Reg) -> Reg(Vec) 2015-07-15 12:44:54 -07:00
Andrew Waterman
e76a9d3493 Chisel3: Don't mix Mux types 2015-07-11 14:05:39 -07:00
Andrew Waterman
5dc3da008e Use Chisel3 SeqMem construct 2015-07-11 13:36:26 -07:00
Henry Cook
fb91e3e1ab minor metadata API update (0.3.3) 2015-07-09 14:36:09 -07:00
Henry Cook
80ad1eac70 Update README.md 2015-07-08 19:05:18 -07:00
Andrew Waterman
55059632c4 Temporarily use HTIF to push RTC value to cores 2015-07-05 16:19:39 -07:00
Henry Cook
d7cb60e8fa L2 WritebackUnit bug fix 2015-07-02 13:52:40 -07:00
Andrew Waterman
b4e38192a1 Fix (?) L2$ miss bug 
The victim's metadata was incorrectly used for the new line.
 2015-06-24 18:01:56 -07:00
Andrew Waterman
ea76800d1a Fix data array reset bug 
io.resp.valid could have been valid the cycle after reset, causing the
write mask in the acquire tracker to have an erroneous value after reset.
This caused the L1 I$ to be refilled with the wrong data.

This probably only affects programs loaded with +loadmem and so shouldn't
matter for the EOS24 silicon.  It should only affect the first L2 xact,
which, in practice, would be an HTIF write to load the program.
 2015-06-11 15:28:23 -07:00
Henry Cook
f3a838cedf nasti converters, hub bugfix 2015-05-21 19:49:17 -07:00
Henry Cook
c202449e34 first version NASTI IOs 2015-05-14 15:29:49 -07:00
Henry Cook
90c9ee7b04 fix unalloc putblocks 2015-05-14 12:37:35 -07:00
Henry Cook
a7fa77c7fc track operand size for Gets 2015-05-13 23:28:18 -07:00
Henry Cook
172c372d3e L2 alloc cleanup 2015-05-12 17:14:06 -07:00
Henry Cook
5fdae2cb61 Merge branch 'master' of github.com:ucb-bar/uncore 2015-05-07 16:18:23 -07:00
Henry Cook
fc883b5049 rm index.html 2015-05-07 16:17:40 -07:00
Henry Cook
8362eba00f Merge branch 'gh-pages' 2015-05-07 16:16:13 -07:00
Henry Cook
aec24cf1a7 readme 2015-05-07 16:16:07 -07:00
Henry Cook
62b6f24798 Delete TileLink0.3.1Specification.pdf 2015-05-07 15:43:06 -07:00
Henry Cook
90ced93eeb Merge branch 'master' into gh-pages 2015-05-07 12:35:14 -07:00
Henry Cook
4cef8c9cd4 Added MemIOArbiter 2015-05-07 10:55:38 -07:00
Henry Cook
8832b454ce add plugins to make scala doc site and publish to ghpages 2015-04-29 15:34:56 -07:00
Henry Cook
1e05fc0525 First pages commit 2015-04-29 13:18:26 -07:00
Henry Cook
3673295d03 network shim cleanup 2015-04-27 16:59:30 -07:00
Henry Cook
09e30041ed Voluntary Writeback tracker rewrite 2015-04-27 12:56:33 -07:00
Henry Cook
11b5222d01 Refactored WritebackUnit 2015-04-21 22:23:04 -07:00
Henry Cook
4c7969b2b3 Metadata docs and api cleanup 2015-04-20 16:32:09 -07:00
Henry Cook
f66a9fd7a6 simplify ClientMetadata.makeRelease 2015-04-20 10:46:02 -07:00
Henry Cook
6d40a61060 TileLink scala doc and parameter renaming 2015-04-19 22:06:44 -07:00
Henry Cook
ba7a8b1752 TileLink refactor; TileLinkPorts now available. L2Banks no longer have unique ids (suitable for hierarhical P&R). 2015-04-17 16:55:20 -07:00
Henry Cook
ce3271aef2 refactor LNClients and LNManagers 2015-04-15 15:48:36 -07:00
Henry Cook
90f800d87d Grant bugfixes and more comments 2015-04-13 15:57:06 -07:00
Henry Cook
3cf1778c92 moved ecc lib to uncore, l2 checks for partial write masks when ecc is enabled 2015-04-06 12:22:23 -07:00
Henry Cook
9708d25dff Restructure L2 state machine and utilize HeaderlessTileLinkIO 2015-04-06 12:19:51 -07:00
Henry Cook
ced627f00a slight mod to pending_puts 
cleaner state transition logic
 2015-04-01 15:24:53 -07:00
Andrew Waterman
c941f0a68e New virtual memory implementation (Sv39) 2015-03-27 16:21:29 -07:00
Henry Cook
8959b2e81a TileLinkEnqueuer 2015-03-26 13:29:52 -07:00
Henry Cook
b7af610569 broadcast hub bugfix 2015-03-26 11:29:04 -07:00
Henry Cook
4176edaa34 clean up tracker allocation 2015-03-26 10:17:51 -07:00
Henry Cook
db5511300d Merge branch 'l2-subblock-merging' 2015-03-18 23:52:06 -07:00
Henry Cook
3cf033180f pending read fix 2015-03-18 22:41:09 -07:00
Henry Cook
004ad11af6 cleanup pending signals 2015-03-18 22:14:41 -07:00
Henry Cook
002851f836 disentangle is_hit logic 2015-03-18 21:11:40 -07:00
Henry Cook
b92ea60891 you can 'hit' with putblocks even when the tag doesn't match but you still better writeback 2015-03-18 19:32:46 -07:00
Henry Cook
fb8071c12d generous hit detection on PutBlocks 2015-03-18 18:49:32 -07:00
Henry Cook
19059bf0eb put data can be used for ignts 2015-03-18 18:28:03 -07:00
Henry Cook
1ff184bf62 first cut at optimized state transitions 2015-03-18 17:55:05 -07:00
Henry Cook
e325399c87 Re-split mem resp tag and data queues 2015-03-18 12:49:53 -07:00
Henry Cook
42aa4aa8ca Secondary miss param 2015-03-17 22:53:50 -07:00
Henry Cook
b364d387de Merge branch 'l2-subblock-merging' of github.com:ucb-bar/uncore into l2-subblock-merging 2015-03-17 22:46:54 -07:00
Henry Cook
825c4b2850 make ignts more eager 2015-03-17 22:44:53 -07:00
Yunsup Lee
aa5435800d fix get merging, and always turn it on 2015-03-17 22:43:00 -07:00
Yunsup Lee
f4f59464df fix pending_puts initialization 2015-03-17 21:44:26 -07:00
Henry Cook
0e4cf74d8a always merge Puts 2015-03-17 20:53:27 -07:00
Henry Cook
d48775eecb cleanup outdated comments 2015-03-17 20:31:23 -07:00
Henry Cook
638bace858 avoid reading data when write mask is full 2015-03-17 20:28:21 -07:00
Henry Cook
b9591b297c added s_wait_puts to L2AcquireTracker 2015-03-17 20:28:21 -07:00
Henry Cook
2d3f947a9c cleaned up finish counter 2015-03-17 20:28:21 -07:00
Yunsup Lee
9de5161d7a guard all writes to data ram with masks 2015-03-17 20:24:04 -07:00
Yunsup Lee
d14efce0b4 fix wmask_buffer initialization 2015-03-17 19:54:11 -07:00
Yunsup Lee
2d7375760d set pending_writes for puts 2015-03-17 18:35:04 -07:00
Yunsup Lee
504eedbf89 fixes in in bit manipulation 2015-03-17 18:07:52 -07:00
Henry Cook
b08dced37c first cut at pending scoreboarding 2015-03-17 17:51:00 -07:00
Yunsup Lee
4fd01d82b8 don't block finish messages when grant'ing 2015-03-17 17:48:02 -07:00
Yunsup Lee
a52a729ab9 bugfix wmask handling 2015-03-17 15:54:21 -07:00
Henry Cook
fc0ae81a97 added finish counter 2015-03-17 15:04:30 -07:00
Henry Cook
5c2461c743 merge data wmask bugfix 2015-03-17 13:09:47 -07:00
Henry Cook
dc88614094 overlapping read/resps in l2 fix 2015-03-17 13:09:47 -07:00
Henry Cook
730a13abf2 pending read fix 2015-03-17 13:09:46 -07:00
Henry Cook
23f8033df5 turn off self probes again 2015-03-17 13:09:46 -07:00
Henry Cook
d9598d26f2 fix assert 2015-03-17 13:09:46 -07:00
Henry Cook
6d565d22e3 clean up acquire alloc in hub 2015-03-17 13:09:45 -07:00
Henry Cook
3f070eee1f first cut of merging puts/gets 2015-03-17 13:09:44 -07:00
Henry Cook
6af48e168a cleanup mergeData buffer 2015-03-17 13:08:14 -07:00
Henry Cook
9bedde9a8a re-merge mem resp queues 2015-03-17 12:22:57 -07:00
Henry Cook
1471d9debc fix assert 2015-03-17 05:40:05 -07:00
Yunsup Lee
0e51fef200 bugfix where an in-progress acquire can be blocked by another acquire tracker being free'd up in between 2015-03-17 05:37:56 -07:00
Henry Cook
ce9d4b6e70 further amo cleanups 2015-03-17 05:37:41 -07:00
Henry Cook
f35a6a574f Add a queue on released data coming in to L2 2015-03-16 13:25:01 -07:00
Henry Cook
b72230a9f0 PutBlock bugfix 2015-03-16 00:09:55 -07:00
Henry Cook
f6d1a2fb76 No more self-probes required 2015-03-16 00:09:38 -07:00
Henry Cook
23a6b007c1 Fix BroadcastHub AcquiteTracker allocation bug and clean up tracker wiring 2015-03-15 23:10:51 -07:00
Henry Cook
c03976896e separate queues for resp tag and data 2015-03-15 17:58:17 -07:00
Andrew Waterman
6e540825b2 Use entire 12-bit CSR address 2015-03-14 02:15:24 -07:00
Yunsup Lee
3a78ca210d bugfix in uncached TL to TL convertors 2015-03-12 16:33:41 -07:00
Henry Cook
8181262419 clean up incoherent and probe flags 2015-03-12 16:22:14 -07:00
Henry Cook
dcc84c4dd3 arbiter probe ready bugfix 2015-03-12 16:02:51 -07:00
Yunsup Lee
2c31ed6426 previous bug fix for meta data writeback wasn't quite right 2015-03-12 15:34:20 -07:00
Yunsup Lee
5e40c8ba77 write back meta data when cache miss even when coherence meta data is clean 2015-03-12 14:36:46 -07:00
Albert Ou
8f8022379c Fix AMO opcode extraction 2015-03-11 23:24:58 -07:00
Albert Ou
f75126c39c Require self probes for all built-in Acquire types 
This ensures that puts by the RoCC accelerator properly invalidates its
tile's L1 D$, with which it currently shares the same TileLink port.
 2015-03-11 23:24:58 -07:00
Henry Cook
1aff919c24 added prefetchAck Grant type 2015-03-11 17:32:06 -07:00
Henry Cook
059575c334 cleanup mergeData and prep for cleaner data_buffer in L2 2015-03-11 15:43:41 -07:00
Henry Cook
b4ed1d9121 Add builtin prefetch types to TileLink 2015-03-11 14:28:17 -07:00
Yunsup Lee
3ab1aca7de L2 subblock access bugfix 2015-03-11 01:56:47 -07:00
Henry Cook
17072a0041 L2 Writeback bugfix 2015-03-10 01:15:03 -07:00
Henry Cook
a1f04386f7 Headerless TileLinkIO and arbiters 2015-03-09 16:34:59 -07:00
Henry Cook
002f1a1b39 pin outer finish header 2015-03-09 12:40:37 -07:00
Henry Cook
df79e7ff8d secondary miss bug 2015-03-05 15:51:18 -08:00
Henry Cook
8e41fcf6fc reduce MemIFTag size, enable non pow2 HellaFLowQueue size 2015-03-05 15:51:02 -08:00
Henry Cook
1bed6ea498 New metadata-based coherence API 2015-02-28 17:32:03 -08:00
Henry Cook
0a8722e881 bugfix for indexing DataArray of of small L2 2015-02-17 00:37:40 -08:00
Henry Cook
0c66e70f14 cleanup of conflicts; allocation bugfix 2015-02-06 13:20:44 -08:00
Henry Cook
7b86ea17cf rename L2HellaCache to L2HellaCacheBank 2015-02-03 19:38:01 -08:00
Stephen Twigg
3b3250339a Explicitely convert results of Bits Muxes to UInt 
Chisel updated to emit SInt result instead of UInt so this commit addresses this change.
 2015-02-03 18:15:01 -08:00
Henry Cook
57340be72b doc update 2015-02-02 01:11:13 -08:00
Henry Cook
6141b3efc5 uncached -> builtin_type 2015-02-02 01:02:06 -08:00
Henry Cook
e6491d351f Offset AMOs within beat and return old value 2015-02-02 00:22:21 -08:00
Henry Cook
3aa030f960 Support for uncached sub-block reads and writes, major TileLink and CoherencePolicy refactor. 2015-02-01 20:37:16 -08:00
Henry Cook
7b4e9dd137 Block L2 transactions on the same set from proceeding in parallel 2015-02-01 20:29:23 -08:00
Henry Cook
973eb43128 state machine bug on uncached write hits 2015-02-01 20:29:23 -08:00
Henry Cook
f58f8bf385 Make L2 data array use a single Mem 2015-01-25 15:37:04 -08:00
Henry Cook
9ef00d187f %s/master/manager/g + better comments 2014-12-29 22:55:58 -08:00
Henry Cook
c76b4bc21d TileLink doc 2014-12-29 22:55:18 -08:00
Henry Cook
e62c71203e disconnect unused outer network headers 2014-12-22 18:50:37 -08:00
Henry Cook
2ef4357ca8 acquire allocation bugfix 2014-12-19 17:39:23 -08:00
Henry Cook
f234fe65ce Initial verison of L2WritebackUnit, passes MiT2 bmark tests 2014-12-19 03:03:53 -08:00
Henry Cook
d121af7f94 Simplify release handling 2014-12-18 17:12:29 -08:00
Henry Cook
bfcfc3fe18 refactor cache params 2014-12-17 14:28:14 -08:00
Henry Cook
ab39cbb15d cleanup DirectoryRepresentation and coherence params 2014-12-15 19:24:42 -08:00
Andrew Waterman
d04da83f96 Make data RAMs 1RW instead of 1R1W 2014-12-15 17:36:17 -08:00
Henry Cook
6a8b66231c Add uncached->cached tilelink converter 2014-12-12 17:06:03 -08:00
Henry Cook
424df2368f 1R/W L2 data array? 
Add TLDataBeats to new LLC; all bmarks pass
 2014-12-12 17:05:21 -08:00
Henry Cook
3026c46a9c Finish adding TLDataBeats to uncore & hub 2014-12-12 17:04:52 -08:00
Henry Cook
2f733a60db Begin adding TLDataBeats to uncore 2014-12-12 17:04:31 -08:00
Henry Cook
404773eb9f fix wb bug 2014-12-03 14:22:39 -08:00