summaryrefslogtreecommitdiff
path: root/src/mem/protocol
AgeCommit message (Collapse)Author
2019-03-20invisispec-1.0 sourceIru Cai
2018-03-12mem-ruby: Fix RubyPrefetcher support in MESI_Two_LevelRico Amslinger
Only a small quantity of prefetches were issued, as the positive feedback mechanism was not implemented. This commit adds a new action po_observeHit, which notifies the RubyPrefetcher of successful prefetches and resets the prefetch flag. When a cache line was replaced by a prefetch, the wrong queue could be stalled. This commit adds a new event PF_L1_Replacement, which stalls the correct queue. The behavior when receiving a prefetch or instruction fetch while in PF_IS_I (prefetch caused GETs, but got invalidated before the response was received) was undefined. This was changed to drop the prefetch request or change the state to non-prefetch, respectively. This behavior is analogous to IS_I (non-prefetch caused GETs, but got invalidated before the response was received) and the data case, respectively. In my local branch a major (20+%) performance increase can be observed in SPEC2006 gobmk and leslie3d when enabling the prefetcher. Some other benchmarks like bwaves, GemsFDTD, sphinx and wrf show smaller (~10%) performance increases. Unfortunately, the performance in most other SPEC benchmarks is still poor, most likely as the prefetcher does not detect strides fast/often enough. In order to push the change timely (most benchmarks have runtimes in the order of days on my machine even with the smallest parameters) after checkout, I have only run gobmk with the base repository + this commit. The results match those of my local branch. Change-Id: I9903a2fcd02060ea5e619b409f31f7d6fac47ae8 Reviewed-on: https://gem5-review.googlesource.com/8801 Reviewed-by: Jason Lowe-Power <jason@lowepower.com> Reviewed-by: Swapnil Haria <swapnilster@gmail.com> Maintainer: Jason Lowe-Power <jason@lowepower.com>
2018-01-11mem-ruby: Fix wakeup timeouts for the MOESI_CMP_token protocolNikos Nikoleris
This changeset fixes a bug that was affecting the MOESI_CMP_token protocol where setting the next timeout required an absolute tick in the future. Change-Id: Ibfdb59354e13c7e552cb3389e71bda010f333249 Reviewed-by: Andreas Sandberg <andreas.sandberg@arm.com> Reviewed-on: https://gem5-review.googlesource.com/7163 Reviewed-by: Jason Lowe-Power <jason@lowepower.com> Maintainer: Jason Lowe-Power <jason@lowepower.com>
2018-01-11mem-ruby: Remove function that maps responses to a DMA engineNikos Nikoleris
The function map_Address_to_DMA was used to route responses to the first (and assumed to be the only) DMA engine in the system. This function is now unused as protocols handle responses and route them to the right DMA engine. Change-Id: I2fba913cf2f12321d1a1e38e7ee85bdf26b8a47a Reviewed-by: Andreas Sandberg <andreas.sandberg@arm.com> Reviewed-on: https://gem5-review.googlesource.com/7162 Reviewed-by: Jason Lowe-Power <jason@lowepower.com> Maintainer: Jason Lowe-Power <jason@lowepower.com>
2018-01-11mem-ruby: Add support for multiple DMA engines in MESI_Two_LevelNikos Nikoleris
Previously the MESI_Two_Level protocol supported systems with a single DMA engine and responses from the directory to DMA requests were routed back to the only DMA engine. This changeset adds support for multiple DMA engines in the system by routing the response to the DMA engine that originally sent the request. Change-Id: I10ceda682ea29746636862ec8ef2a9c4220ca045 Reviewed-by: Andreas Sandberg <andreas.sandberg@arm.com> Reviewed-on: https://gem5-review.googlesource.com/7161 Reviewed-by: Jason Lowe-Power <jason@lowepower.com> Maintainer: Jason Lowe-Power <jason@lowepower.com>
2017-11-10scons: Move Transform and termcap functionality into their own files.Gabe Black
Change-Id: Ica08e93f3873a7eafd02fe7d44c3bdbf0ce7f6b7 Reviewed-on: https://gem5-review.googlesource.com/5565 Reviewed-by: Gabe Black <gabeblack@google.com> Maintainer: Gabe Black <gabeblack@google.com>
2017-06-13ruby: Add support for address ranges in the directoryNikos Nikoleris
Previously the directory covered a flat address range that always started from address 0. This change adds a vector of address ranges with interleaving and hashing that each directory keeps track of and the necessary flexibility to support systems with non continuous memory ranges. Change-Id: I6ea1c629bdf4c5137b7d9c89dbaf6c826adfd977 Reviewed-by: Andreas Sandberg <andreas.sandberg@arm.com> Reviewed-on: https://gem5-review.googlesource.com/2903 Reviewed-by: Bradford Beckmann <brad.beckmann@amd.com> Reviewed-by: Jason Lowe-Power <jason@lowepower.com> Maintainer: Jason Lowe-Power <jason@lowepower.com>
2017-04-05ruby: Fix MOESI_CMP_directory for new DMA status changes.Javier Cano-Cano
Multiple outstanding DMA requests introduced new DMA states that didn't be considered into slicc code. This patch implements the missed DMA state changes on MOESI_CMP_directory protocol. Change-Id: I700d441d76556b7e77e0d507904af6ec6ba59cc2 Signed-off-by: Michael LeBeane <michael.lebeane@amd.com> Reviewed-on: https://gem5-review.googlesource.com/2380 Reviewed-by: Jason Lowe-Power <jason@lowepower.com> Reviewed-by: Michael LeBeane <Michael.Lebeane@amd.com> Maintainer: Jason Lowe-Power <jason@lowepower.com>
2016-11-19ruby: init MessageSizeType of SequencerMsg to Request_ControlSooraj Puthoor
SequencerMsg is autogenerated by slicc scripts and the MessageSizeType is initialized to the max enume value by default. The DMASequencer pushes this message to the mandatory queue and since the MessageSizeType is unitialized, string_to_MessageSizeType() function used by traces to print the message fails with a panic. This patch avoids this problem by initializing MessageSizeType of SequencerMsg to Request_Control.
2016-10-26ruby: Allow multiple outstanding DMA requestsMichael LeBeane
DMA sequencers and protocols can currently only issue one DMA access at a time. This patch implements the necessary functionality to support multiple outstanding DMA requests in Ruby.
2016-10-06ruby: rename ALPHA_Network_test protocol to Garnet_standalone.Tushar Krishna
Over the past 6 years, we realized that the protocol is essentially used to run the garnet network in a standalone manner, and feed standard synthetic traffic patterns through it.
2016-04-26ruby: Rename pkt to m_pkt so it may be accessed via SLICCMatthew Poremba
Allow usage of packet class in ruby for convenience purposes. This may be used to access members of the packet/request class (e.g., via helper functions) and/or push protocol specific information to the packets SenderState without needing to modify SLICC types and protocols in multiple locations.
2016-01-22ruby: removed Write_Only AccessPermissionBrad Beckmann
2016-01-19gpu-compute: AMD's baseline GPU modelTony Gutierrez
2016-01-19mem: write combining for ruby protocolsTony Gutierrez
This patch adds support for write-combining in ruby.
2016-01-19* * *Tony Gutierrez
mem: support for gpu-style RMWs in ruby This patch adds support for GPU-style read-modify-write (RMW) operations in ruby. Such atomic operations are traditionally executed at the memory controller (instead of through an L1 cache using cache-line locking). Currently, this patch works by propogating operation functors through the memory system.
2015-07-20mem: misc flags for AMD gpu modelBlake Hechtman
This patch add support to mark memory requests/packets with attributes defined in HSA, such as memory order and scope.
2015-07-20ruby: slicc: have a static MachineTypeTony Gutierrez
This patch is imported from reviewboard patch 2551 by Nilay. This patch moves from a dynamically defined MachineType to a statically defined one. The need for this patch was felt since a dynamically defined type prevents us from having types for which no machine definition may exist. The following changes have been made: i. each machine definition now uses a type from the MachineType enumeration instead of any random identifier. This required changing the grammar and the *.sm files. ii. MachineType enumeration defined statically in RubySlicc_Exports.sm. * * * normal protocol fixes for nilay's parser machine type fix
2015-07-20ruby: slicc: remove support for single machine, multiple typesTony Gutierrez
This patch is imported from reviewboard patch 2550 by Nilay. It was possible to specify multiple machine types with a single state machine. This seems unnecessary and is being removed.
2015-11-13slicc: fixes for the Address to Addr changeset (11025)Tony Gutierrez
misc changes now that Address has become Addr including int to address util function
2015-11-13ruby: add BoolVecJoe Gross
The BoolVec typedef and insertion operator overload function simplify usage of vectors of type bool
2015-09-23ruby: bloom filters: refactor codeNilay Vaish
2015-09-18ruby: print addresses in hexNilay Vaish
Changeset 4872dbdea907 replaced Address by Addr, but did not make changes to print statements. So the addresses which were being printed in hex earlier along with their line address, were now being printed in decimals. This patch adds a function printAddress(Addr) that can be used to print the address in hex along with the lines address. This function has been put to use in some of the places. At other places, change has been made to print just the address in hex.
2015-09-16ruby: Add missing block deallocations in MOESI_hammerLena Olson
Some blocks in MOESI hammer were not getting deallocated when they were set to an idle state (e.g. by invalidate or other_getx/s messages). While functionally correct, this caused some bad effects on performance, such as blocks in I in the L1s getting sent to the L2 upon eviction, in turn evicting valid blocks. Also, if a valid block was in LRU, that block could be evicted rather than a block in I. This patch adds in the missing deallocations. Committed by: Nilay Vaish<nilay@cs.wisc.edu>
2015-09-16ruby: message buffer, timer table: significant changesNilay Vaish
This patch changes MessageBuffer and TimerTable, two structures used for buffering messages by components in ruby. These structures would no longer maintain pointers to clock objects. Functions in these structures have been changed to take as input current time in Tick. Similarly, these structures will not operate on Cycle valued latencies for different operations. The corresponding functions would need to be provided with these latencies by components invoking the relevant functions. These latencies should also be in Ticks. I felt the need for these changes while trying to speed up ruby. The ultimate aim is to eliminate Consumer class and replace it with an EventManager object in the MessageBuffer and TimerTable classes. This object would be used for scheduling events. The event itself would contain information on the object and function to be invoked. In hindsight, it seems I should have done this while I was moving away from use of a single global clock in the memory system. That change led to introduction of clock objects that replaced the global clock object. It never crossed my mind that having clock object pointers is not a good design. And now I really don't like the fact that we have separate consumer, receiver and sender pointers in message buffers.
2015-09-16slicc: export uint64_t instead of uint64Anthony Gutierrez
2015-09-05ruby: call setMRU from L1 controllers, not from sequencerNilay Vaish
Currently the sequencer calls the function setMRU that updates the replacement policy structures with the first level caches. While functionally this is correct, the problem is that this requires calling findTagInSet() which is an expensive function. This patch removes the calls to setMRU from the sequencer. All controllers should now update the replacement policy on their own. The set and the way index for a given cache entry can be found within the AbstractCacheEntry structure. Use these indicies to update the replacement policy structures.
2015-09-05ruby: declare all protocol message buffers as parametersNilay Vaish
MessageBuffer is a SimObject now. There were protocols that still declared some of the message buffers are variables of the controller, but not as input parameters. Special handling was required for these variables in the SLICC compiler. This patch changes this. Now all message buffers are declared as input parameters.
2015-08-19ruby: reverts to changeset: bf82f1f7b040Nilay Vaish
2015-08-14ruby: add accessor functions to SLICC def of MachineIDNilay Vaish
2015-08-14ruby: call setMRU from L1 controllers, not from sequencerNilay Vaish
Currently the sequencer calls the function setMRU that updates the replacement policy structures with the first level caches. While functionally this is correct, the problem is that this requires calling findTagInSet() which is an expensive function. This patch removes the calls to setMRU from the sequencer. All controllers should now update the replacement policy on their own. The set and the way index for a given cache entry can be found within the AbstractCacheEntry structure. Use these indicies to update the replacement policy structures.
2015-08-14ruby: drop the [] notation for lookup function.Nilay Vaish
This is in preparation for adding a second arugment to the lookup function for the CacheMemory class. The change to *.sm files was made using the following sed command: sed -i 's/\[\([0-9A-Za-z._()]*\)\]/.lookup(\1)/' src/mem/protocol/*.sm
2015-08-14ruby: replace Address by AddrNilay Vaish
This patch eliminates the type Address defined by the ruby memory system. This memory system would now use the type Addr that is in use by the rest of the system.
2015-08-14ruby: rename variables Addr to addrNilay Vaish
Avoid clash between type Addr and variable name Addr.
2015-08-14ruby: Protocol changes for SimObject MessageBuffersJoel Hestness
2015-08-14ruby: Expose MessageBuffers as SimObjectsJoel Hestness
Expose MessageBuffers from SLICC controllers as SimObjects that can be manipulated in Python. This patch has numerous benefits: 1) First and foremost, it exposes MessageBuffers as SimObjects that can be manipulated in Python code. This allows parameters to be set and checked in Python code to avoid obfuscating parameters within protocol files. Further, now as SimObjects, MessageBuffer parameters are printed to config output files as a way to track parameters across simulations (e.g. buffer sizes) 2) Cleans up special-case code for responseFromMemory buffers, and aligns their instantiation and use with mandatoryQueue buffers. These two special buffers are the only MessageBuffers that are exposed to components outside of SLICC controllers, and they're both slave ends of these buffers. They should be exposed outside of SLICC in the same way, and this patch does it. 3) Distinguishes buffer-specific parameters from buffer-to-network parameters. Specifically, buffer size, randomization, ordering, recycle latency, and ports are all specific to a MessageBuffer, while the virtual network ID and type are intrinsics of how the buffer is connected to network ports. The former are specified in the Python object, while the latter are specified in the controller *.sm files. Unlike buffer-specific parameters, which may need to change depending on the simulated system structure, buffer-to-network parameters can be specified statically for most or all different simulated systems.
2015-08-14ruby: Change PerfectCacheMemory::lookup to return pointerJoel Hestness
CacheMemory and DirectoryMemory lookup functions return pointers to entries stored in the memory. Bring PerfectCacheMemory in line with this convention, and clean up SLICC code generation that was in place solely to handle references like that which was returned by PerfectCacheMemory::lookup.
2015-08-03ruby: mesi three level: multiple corrections to the protocolNilay Vaish
1. Eliminate state NP in L0 and L1 Caches: The two states 'NP' and 'I' both mean that the cache block is not present in the cache. 'I' also means that the cache entry has been allocated. This causes problems when we do not correctly initialize the cache entry when it is re-used. Hence, this patch eliminates the state NP altogether. Everytime a new block comes into the cache, a cache entry is allocated. Everytime a block leaves, the corresponding entry is deallocated. 2. Separate transient state for instruction fetches: purely for accouting purposes. 3. Drop state IS_I in L1 Cache and the message type STALE_DATA: when invalidation is received for a block in IS, the block used to be moved to IS_I. This meant that the data that would arrive in future would be used but not stored since the controller lost the permissions after gaining them. This state is being dropped and now invalidation messages would not processed till the data has arrived. This also means that STALE_DATA type is not longer required.
2015-08-03ruby: mesi two,three level: copy data only when dirtyNilay Vaish
The level 2 controller has a bug. In one particular action, the data block was copied from a message irrespective whether the block is dirty or not. In cases when L1 sends no data, the data value copied was incorrect.
2015-07-20slicc: enable overloading in functions not in classesBrad Beckmann
For many years the slicc symbol table has supported overloaded functions in external classes. This patch extends that support to functions that are not part of classes (a.k.a. no parent). For example, this support allows slicc to understand that mapAddressToRange is overloaded and the NodeID is an optional parameter.
2015-07-20ruby: adds size and empty apis to the msg buffer stallmapDavid Hashe
2015-07-20ruby: fix deadlock bug in banked array resource checksDavid Hashe
The Ruby banked array resource checks (initiated from SLICC) did a check and allocate at the same time. If a transition needs more than one resource, then it might check/allocate resource #1, then fail to get resource #2. Another transition might then try to get the same resources, but in reverse order. Deadlock. This patch separates resource checking and resource reservation into two steps to avoid deadlock.
2015-07-20mem: add request types for acquire and releaseDavid Hashe
Add support for acquire and release requests. These synchronization operations are commonly supported by several modern instruction sets.
2015-07-20ruby: allocate a block in CacheMemory without updating LRU stateDavid Hashe
2015-07-20ruby: speed up function used for cache walksDavid Hashe
This patch adds a few helpful functions that allow .sm files to directly invalidate all cache blocks using a trigger queue rather than rely on each individual cache block to be invalidated via requests from the mandatory queue.
2015-07-20ruby: give access to cache tag/data latencies from SLICCDavid Hashe
This patch exposes the tag and data array latencies to the SLICC state machines so that it can be used to determine the correct enqueue latency for response messages.
2015-07-20slicc: support for multiple message types on the same bufferDavid Hashe
This patch allows SLICC protocols to use more than one message type with a message buffer. For example, you can declare two in ports as such: in_port(ResponseQueue_in, ResponseMsg, responseFromDir, rank=3) { ... } in_port(tgtResponseQueue_in, TgtResponseMsg, responseFromDir, rank=2) { ... }
2015-07-20ruby: re-added the addressToInt slicc interface functionBrad Beckmann
This helper function is very useful converting address offsets to integers that can be used for protocol specific destination mapping.
2015-07-04ruby: drop NetworkMessage classNilay Vaish
This patch drops the NetworkMessage class. The relevant data members and functions have been moved to the Message class, which was the parent of NetworkMessage.
2015-07-04ruby: mesi three level: name change to avoid clashNilay Vaish
The accessor function getDestination() for Destination variable in the coherence message clashes with the getDestination() that is part of the Message class. Hence the name change.