Commit graph

3278 commits

Author SHA1 Message Date
Kevin Lim
a50e83c134 Fix assertion. I haven't tested it fully (I can't reproduce Lisa's error) but I believe it should fix what she's running into (which was definitely a bug).
src/cpu/o3/fetch_impl.hh:
    Move assertion to area where it should really always be true.  Sometimes you might recvRetry and not necessarily be blocked (if there was a squash).

--HG--
extra : convert_revision : 76ad35357e7f4c44fa544ffed071096a62053018
2006-10-13 17:35:23 -04:00
Ali Saidi
ca4063ac00 fix a bug in CopyStringOut. dprintk appears to work again.
--HG--
extra : convert_revision : cd0d13a85ddc7599308db8604a8f63a48679cc05
2006-10-13 14:28:46 -04:00
Lisa Hsu
339b1f8516 Merge zizzer:/bk/newmem
into  zed.eecs.umich.edu:/z/hsul/work/m5/newmem

src/cpu/simple/timing.cc:
    hand merge

--HG--
extra : convert_revision : 083bf102249ad9bc63c447dbf85d3863f935f647
2006-10-12 18:56:57 -04:00
Ali Saidi
3d2764acf3 replace functional code in tport with fixPacket().
fixPacket() should be used anywhere a functional packet and timing packet are found to have the same address.

--HG--
extra : convert_revision : 783ec438271b24ddb0ae742b4efd1ed7d6be93f3
2006-10-12 15:30:30 -04:00
Korey Sewell
e11c0d380a Merge zizzer:/bk/newmem
into  zooks.eecs.umich.edu:/y/ksewell/research/m5-sim/cleanrepo

--HG--
extra : convert_revision : f02fd56fc1ec658fe2a81d0e0b0d053b7606f7f2
2006-10-12 15:05:29 -04:00
Korey Sewell
790e746bd9 config file updates
--HG--
extra : convert_revision : b873e0b436d71d86b87b8d9df7115bcc7ceb2b50
2006-10-12 15:04:14 -04:00
Ron Dreslinski
eddbb6801f Fix CSHR retrys
--HG--
extra : convert_revision : caa7664f6c945396fa38ce62fbda018ebed4eaa6
2006-10-12 15:02:56 -04:00
Ali Saidi
4a96779350 Merge zizzer:/bk/newmem
into  zeep.pool:/z/saidi/work/m5.newmem.head

--HG--
extra : convert_revision : 0e184a0784100112db5841c587bd3dd638f8bdc0
2006-10-12 15:02:50 -04:00
Ali Saidi
0615d92d33 small bus updates for functional accesses
--HG--
extra : convert_revision : c7a6b199c74ed4b4ffab14bbffb51e72d75b7742
2006-10-12 15:02:25 -04:00
Korey Sewell
1ebe4b0dec Merge zizzer:/bk/newmem
into  zooks.eecs.umich.edu:/y/ksewell/research/m5-sim/cleanrepo

--HG--
extra : convert_revision : 86abc4dabb7c72edc90fae951314d6a6b5c73705
2006-10-12 14:36:41 -04:00
Korey Sewell
665901de8a Add test binary & inputs
--HG--
extra : convert_revision : efc0249ae6f51b9f4f49741aafb2cad21e1fb11e
2006-10-12 14:36:21 -04:00
Ron Dreslinski
3aaa3456dc Merge zizzer:/bk/newmem
into  zazzer.eecs.umich.edu:/z/rdreslin/m5bk/newmemcleanest

--HG--
extra : convert_revision : fa5b2cfa79d87a0612b8116d407a8b2959d9095a
2006-10-12 14:31:31 -04:00
Ron Dreslinski
fe230ddb8f Remove bus and top level parameters from cache
src/mem/cache/base_cache.hh:
    Remove top level param from cache
src/mem/cache/coherence/uni_coherence.cc:
    Remove top level parameters from the cache

--HG--
extra : convert_revision : 4437aeedc20866869de7f9ab123dfa7baeebedf0
2006-10-12 14:21:25 -04:00
Ali Saidi
2c9d506f46 Merge zizzer:/bk/newmem
into  zeep.pool:/z/saidi/work/m5.newmem.head

src/mem/packet.hh:
    hand merge

--HG--
extra : convert_revision : 3f77707360235dc98c6b12a0367ca64a401313df
2006-10-12 14:18:42 -04:00
Ali Saidi
3ba2ed6aef add a traceflag for functional accesses
implement fix packet and add the ability to print a packet to a ostream
remove tabs in packet.hh (Could people stop inserting them??!?!?!)
mark const functions in packet.hh as such

src/base/traceflags.py:
    add a traceflag for functional accesses
src/mem/packet.cc:
    implement fix packet and add the ability to print a packet to a ostream
src/mem/packet.hh:
    add the ability to print a packet to an ostream
    remove tabs in file
    mark const functions as such

--HG--
extra : convert_revision : 4297bce5e1d3abbab48be5bd9eb9e982b751fc7c
2006-10-12 14:15:09 -04:00
Ron Dreslinski
f89b56b61a Check the response queue on functional accesses.
The response queue is not tying up an MSHR, should we change that or assume infinite storage for responses?

src/mem/cache/base_cache.cc:
src/mem/tport.cc:
    Add in functional check of retry queued packets.

--HG--
extra : convert_revision : 0cb40b3a96d37a5e9eec95312d660ec6a9ce526a
2006-10-12 13:59:03 -04:00
Ron Dreslinski
6ffdc7b4d7 Another memleak in the memtester (need [] with the delete)
src/cpu/memtest/memtest.cc:
    Another memleak in the memtester

--HG--
extra : convert_revision : f7ab079e90d578fb6b9d1ff238d049fcce55b01b
2006-10-12 13:45:28 -04:00
Ron Dreslinski
dd18ffe51d Fix a memory leak in the memtester
--HG--
extra : convert_revision : 93062b0f1a3ba7a5210e2f27099f20ae8f66522b
2006-10-12 13:43:12 -04:00
Ron Dreslinski
ba4c224c39 Fix problems with unCacheable addresses in timing-coherence
src/base/traceflags.py:
src/mem/physical.cc:
    Add debug falgs fro physical memory accesses
src/mem/cache/cache_impl.hh:
    Snoops to uncacheable blocks should not happen
src/mem/cache/miss/miss_queue.cc:
    Set the size properly on unCacheable accesses

--HG--
extra : convert_revision : fc78192863afb11fc7c591fba169021b9e127d16
2006-10-12 13:33:21 -04:00
Ron Dreslinski
78aec04b66 Merge zizzer:/bk/newmem
into  zazzer.eecs.umich.edu:/z/rdreslin/m5bk/newmemcleanest

--HG--
extra : convert_revision : b6224624ea493531e5ecc161bede7433f96d8a0f
2006-10-11 21:04:11 -04:00
Ron Dreslinski
388d484269 Make default ID unique (not broadcast)
Fix a segfault associated with DefaultId

src/mem/bus.cc:
    Handle a segfault in the bus when DefaultPort was being used
src/mem/bus.hh:
    Make the Default ID more unique (it overlapped with Broadcast ID)

--HG--
extra : convert_revision : 9182805c5cf4d9fe004e6c5be8547a8f41ed7bfe
2006-10-11 20:54:06 -04:00
Ron Dreslinski
14c8e8b227 Forgot to mark myself as on the retry list
--HG--
extra : convert_revision : c20170320a284a7bf143a929e4d3aa1475a8bfe0
2006-10-11 19:47:11 -04:00
Korey Sewell
652a09a512 add spec2k tests
--HG--
extra : convert_revision : 619d6af7013b597ade6440f8cf84d6d099e31763
2006-10-11 19:26:56 -04:00
Ron Dreslinski
3c7e0ec752 Fix bus in FS mode.
src/mem/bus.cc:
    Add debugging statement
src/mem/bus.hh:
    Fix implementation of bus for subsequent recvTimings while handling a retry request.
src/mem/tport.cc:
    Rework timing port to retry properly

--HG--
extra : convert_revision : fbfb5e8b4a625e49c6cd764da1df46a4f336b1b2
2006-10-11 19:25:48 -04:00
Korey Sewell
7442bd2664 add bzip test-prog
--HG--
extra : convert_revision : f5197172b094d59ced84423eb34c31ddf23689ee
2006-10-11 19:07:10 -04:00
Lisa Hsu
ba795552f5 System not global object, need to preface it with objects.
--HG--
extra : convert_revision : 5e105d7082a8c103fb5d5383c3093734bfd590f5
2006-10-11 18:54:31 -04:00
Lisa Hsu
8acecfef9b since memoryMode was put into the System (from SimObject), things got broken - this fixes it so that changeToTiming/changeToAtomic works.
src/python/m5/SimObject.py:
    now that setMemoryMode is a method in System, need to convert the SimObject * _ccObject into a system ptr to call setMemoryMode.
src/sim/main.cc:
    need this conversion now.
src/sim/sim_object.hh:
    put the enum back into SimObject.
src/sim/system.hh:
    memoryMode is now a part of SimObject, need the ::'s

--HG--
extra : convert_revision : 0ade06957fa57b497798e1f50c237ca1badc821d
2006-10-11 18:53:50 -04:00
Lisa Hsu
024b33a1ef some drain changes in timing (kevin's) and some memory mode assertion changes so that when you come out of resume, you only assert if you're really wrong.
src/cpu/simple/atomic.cc:
    memory mode assertion change so that it only goes off if it's supposed to.
src/cpu/simple/timing.cc:
    some drain changes (kevin's) and some changes to memoryMode assertions so that they don't go off when they're not supposed to.

--HG--
extra : convert_revision : 007d8610f097e08f01367b905ada49f93cf37ca3
2006-10-11 18:44:48 -04:00
Ron Dreslinski
567afbf6ce More cache fixes. Atomic coherence now works as well.
src/cpu/memtest/memtest.cc:
src/cpu/memtest/memtest.hh:
    Make Memtester able to test atomic as well
src/mem/bus.cc:
src/mem/bus.hh:
    Handle atomic snoops properly for cache->cache transfers
src/mem/cache/cache_impl.hh:
    Debug output.
    Clean up memleak in atomic mode.
    Set hitLatency.
    Still need to send back reasonable number for atomic return value.
src/mem/packet.cc:
    Add command strings for new commands
src/python/m5/objects/MemTest.py:
    Add param to test atomic memory.

--HG--
extra : convert_revision : 43f880e29215776167c16ea90793ebf8122c785b
2006-10-11 18:28:33 -04:00
Ron Dreslinski
03c42ea590 Update for Atomic Coherece with Gabes bus
--HG--
extra : convert_revision : 6a23052056d1c61cba0a4c77f1030cee419c6fa3
2006-10-11 01:59:38 -04:00
Ron Dreslinski
60252f8e63 Interesting memtest finally.
Get over 500,000 reads on each of 8 testers before memory leak becomes large.

tests/configs/memtest.py:
    Update test to be more interesting

--HG--
extra : convert_revision : 4258b798fbeeed2a376f1bfac100a109eb05620e
2006-10-11 01:18:20 -04:00
Ron Dreslinski
4e35c5656f Merge zizzer:/n/wexford/x/gblack/m5/newmem_bus
into  zazzer.eecs.umich.edu:/z/rdreslin/m5bk/newmemcleanest

--HG--
extra : convert_revision : 70187b8f04d0f8424512f64bdade05bf1aca85a3
2006-10-11 01:02:18 -04:00
Ron Dreslinski
c2012601e9 Use bus response time paramteres
Fix bug with deadlocking

src/mem/cache/base_cache.cc:
    Make sure to not wait anymore

--HG--
extra : convert_revision : 5f7b44a1c475820b9862275a0d6113ec2991735d
2006-10-11 01:01:40 -04:00
Gabe Black
7767f5af73 Don't call recvRetry if the bus is busy anyway. This takes care of a corner case as well when dealing with grants that aren't used.
--HG--
extra : convert_revision : 38f7ef1b41477fb2a2438387ef3a81cccd3e7a8a
2006-10-11 00:54:47 -04:00
Ron Dreslinski
07dad71f6f Merge zizzer:/n/wexford/x/gblack/m5/newmem_bus
into  zazzer.eecs.umich.edu:/z/rdreslin/m5bk/newmemcleanest

--HG--
extra : convert_revision : d2d19b27533f35c6570ee84c6c83b2919f27b97f
2006-10-11 00:31:40 -04:00
Gabe Black
a139e4394d Make the bus work if the other sides recvRetry doesn't call sendTiming for some reason.
--HG--
extra : convert_revision : e722ddb0354a5c021dc7c44a3e2f0a64e962442b
2006-10-11 00:26:21 -04:00
Ron Dreslinski
04f71f1226 When turning asserts into if's don't forget to invert.
src/mem/cache/base_cache.cc:
    When turning asserts into if's don't forget to invert.
    Must be too sleepy.

--HG--
extra : convert_revision : ea38d5a4b4ddde7b5266b3b2c83bbc256218af9a
2006-10-11 00:19:31 -04:00
Ron Dreslinski
23bbd14426 Writebacks can be pulled out from under the BusRequest when snoops of uprgades to owned blocks hit in the WB buffer
--HG--
extra : convert_revision : f0502836a79ce303150daa7e571badb0bce3a97a
2006-10-11 00:13:53 -04:00
Ron Dreslinski
c9102b08fa Only issue responses if we aren;t already blocked
--HG--
extra : convert_revision : 511c0bcd44b93d5499eefa8399f36ef8b6607311
2006-10-10 23:53:10 -04:00
Ron Dreslinski
ca694ca7b1 Merge zizzer:/n/wexford/x/gblack/m5/newmem_bus
into  zazzer.eecs.umich.edu:/z/rdreslin/m5bk/newmemcleanest

src/mem/bus.cc:
    SCCS merged

--HG--
extra : convert_revision : 18608114350c466a56ab499ae523b01fcb2f6ef2
2006-10-10 23:37:14 -04:00
Gabe Black
8353b1e21f Make the bus is occupied for none broadcast packets as well.
--HG--
extra : convert_revision : aef3c625172e92be8f29c4c57077fefee43046bb
2006-10-10 23:28:33 -04:00
Ron Dreslinski
477a3b0b61 Merge zizzer:/n/wexford/x/gblack/m5/newmem_bus
into  zazzer.eecs.umich.edu:/z/rdreslin/m5bk/newmemcleanest

src/mem/bus.cc:
    SCCS merged

--HG--
extra : convert_revision : eaae105025635c37af06cf72bb061ce82def9dc9
2006-10-10 22:52:52 -04:00
Ron Dreslinski
1de8eae43a Debugging info
src/base/traceflags.py:
    Add new flags for cacheport
src/mem/bus.cc:
    Add debugging info
src/mem/cache/base_cache.cc:
    Add debuggin info

--HG--
extra : convert_revision : a6c4b452466a8e0b50a86e886833cb6e29edc748
2006-10-10 22:50:36 -04:00
Gabe Black
59dd317cb5 Put in an accounting mechanism and an assert to make sure something doesn't try to send another packet while it's still waiting for the bus.
--HG--
extra : convert_revision : 4a2b83111e49f71ca27e05c98b55bc3bac8d9f53
2006-10-10 22:10:08 -04:00
Gabe Black
404b2a951d Fixed a corner case and simplified the logic in Packet::intersect.
--HG--
extra : convert_revision : b57c31ca7c220e701d34e02bb07ce392370e4428
2006-10-10 17:49:31 -04:00
Ron Dreslinski
27c59dc370 Merge zizzer:/n/wexford/x/gblack/m5/newmem_bus
into  zazzer.eecs.umich.edu:/z/rdreslin/m5bk/newmemcleanest

--HG--
extra : convert_revision : 7b7a1b03ffed36bce49595962ea57c08d1d1a4ad
2006-10-10 17:32:24 -04:00
Ron Dreslinski
aff3d92c00 Some more code cleanup
src/mem/cache/base_cache.cc:
    Add sanity checks
src/mem/cache/base_cache.hh:
    Fix for retry mechanism

--HG--
extra : convert_revision : 9298e32e64194b1ef3fe51242595eaa56dcbbcfd
2006-10-10 17:25:50 -04:00
Gabe Black
549412b333 Changed the bus to use a bool to keep track of retries rather than a pointer
src/mem/tport.cc:
    minor formatting tweak

--HG--
extra : convert_revision : 7391d142815c5876fcc0f991bd053e6a1781c101
2006-10-10 17:24:03 -04:00
Gabe Black
5f9aca531d Merge zizzer.eecs.umich.edu:/bk/newmem
into  zeep.eecs.umich.edu:/home/gblack/m5/newmem_bus

--HG--
extra : convert_revision : aa59d3169d84bcd13b8c97f22b52aeef43dc33c3
2006-10-10 17:18:09 -04:00
Ron Dreslinski
995146ead7 Fix some more mem leaks, still some left
Update retry mechanism

src/mem/cache/base_cache.cc:
    Rework the retry mechanism
src/mem/cache/base_cache.hh:
    Rework the retry mechanism
    Try to fix memory bug
src/mem/cache/cache_impl.hh:
    Rework upgrades to not be blocked by slave
src/mem/cache/miss/mshr_queue.cc:
    Fix mem leak on writebacks

--HG--
extra : convert_revision : 3cec234ee441edf398ec8d0f51a0c5d7ada1e2be
2006-10-10 17:10:56 -04:00