Certain instructions (some atomics and buffer_wbinvl1_vol) deadlock
in the coalescer, where sendTimingReq fails, fails a retry, and then
never retries again.
This fix sets m_cache_inv_pkt to null before calling
completeHitCallback(), as that allows the failed packets to be retried
again.
Change-Id: I4a51c741360f385f8b4c3f2a31a9410f18e095d9
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/37477
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Reviewed-by: Matthew Poremba <matthew.poremba@amd.com>
Maintainer: Matt Sinclair <mattdsinclair@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
The variable 'm_allow_zero_latency' was only used in an assert message in
`src/mem/ruby/network/MessageBuffer.cc`. This assert is stripped when
compiling to gem5.fast, resulting in the compilation failing with an
unused variable error.
This assert is better as a panic_if, which will not be stripped out
during the .fast compilation.
Change-Id: I5de74982fa42b3291899ddcf73f7140079e1ec3f
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/36697
Maintainer: Bobby R. Bruce <bbruce@ucdavis.edu>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
There is a flow of packets as so:
WriteResp -> WriteReq -> WriteCompleteResp
These packets share some variables, in particular senderState and a
status vector.
One issue was the WriteResp packet decremented the status vector, which
was used by the WriteCompleteResp packets to determine when to handle
the global memory response. This could lead to multiple
WriteCompleteResp packets attempting to handle the global memory
response.
Because of that, the WriteCompleteResp packets needed to handle the
status vector. this patch moves WriteCompleteResp packet handling back
into ComputeUnit::DataPort::processMemRespEvent from
ComputeUnit::DataPort::recvTimingResp. This helps remove some redundant
code.
This patch has the WriteResp packet return without doing any status
vector handling, and without deleting the senderState, which had
previously caused a segfault.
Another issue was WriteCompleteResp packets weren't being issued for
each active lane, as the coalesced request was being issued too early.
In order to fix that, we have to ensure every active lane puts their
request into their applicable coalesced request before issuing the
coalesced request. Because of that change, we change the issuing of
CoalescedRequests from GPUCoalescer::coalescePacket to
GPUCoalescer::completeIssue.
That change involves adding a new variable to store the
CoalescedRequests that are created in the calls to coalescePacket. This
variable is a map from instruction sequence number to coalesced
requests.
Additionally, the WriteCompleteResp packet was attempting to access
physical memory in hitCallback while not having any data, which
caused a crash. This can be resolved either by not allowing
WriteCompleteResp packets to access memory, or by copying the data
from the WriteReq packet. This patch denies WriteCompleteResp packets
memory access in hitCallback.
Finally, in VIPERCoalescer::writeCompleteCallback there was a map
that held the WriteComplete packets, but no packets were ever being
removed. This patch removes packets that match the address that was
passed in to the function.
Change-Id: I9a064a0def2bf6c513f5295596c56b1b652b0ca4
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/33656
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Reviewed-by: Anthony Gutierrez <anthony.gutierrez@amd.com>
Maintainer: Matt Sinclair <mattdsinclair@gmail.com>
Maintainer: Anthony Gutierrez <anthony.gutierrez@amd.com>
Tested-by: kokoro <noreply+kokoro@google.com>
The create() method on Params structs usually instantiate SimObjects
using a constructor which takes the Params struct as a parameter
somehow. There has been a lot of needless variation in how that was
done, making it annoying to pass Params down to base classes. Some of
the different forms were:
const Params &
Params &
Params *
const Params *
Params const*
This change goes through and fixes up every constructor and every
create() method to use the const Params & form. We use a reference
because the Params struct should never be null. We use const because
neither the create method nor the consuming object should modify the
record of the parameters as they came in from the config. That would
make consuming them not idempotent, and make it impossible to tell what
the actual simulation configuration was since it would change from any
user visible form (config script, config.ini, dot pdf output).
Change-Id: I77453cba52fdcfd5f4eec92dfb0bddb5a9945f31
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/35938
Reviewed-by: Gabe Black <gabeblack@google.com>
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Maintainer: Gabe Black <gabeblack@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Added mapAddressToDownstreamMachine that may be used by the protocols
to map an address to different target donwstream controller of the same
type.
These functions do not use the global mapping provided by the network
and map addresses to one of the controllers specified in the
downstream_destinations parameter.
This change facilitates reusing the same cache state-machine/controllers
to model different levels of the cache hierarchy.
Change-Id: I9a202e9461e0d2f16ed232ff8b60bbde2d15570d
Signed-off-by: Tiago Mück <tiago.muck@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/31415
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Returns the offset of an address with respect to a base address.
Looks unnecessary, but SLICC doesn't support casting and the '-'
operator for Addr types, so the alternative to this would be to add
more some helpers like 'addrToUint64' and 'uint64ToInt'.
Change-Id: I90480cec4c8b2e6bb9706f8b94ed33abe3c93e78
Signed-off-by: Tiago Mück <tiago.muck@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/31270
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Add incomingTransactionStart/End and outgoingTransactionStart/End
functions that can be called from the protocol to profile events
that initiate a transaction locally (e.g. an incoming request) and
remotely (e.g. outgoing requests). The generated stats will include
histograms of the latency for completing each type of transaction.
This assumes assumes the protocol uses different trigger events for
initiating incoming and outgoing transactions.
Change-Id: Ib528641b9676c68907b5989b6a09bfe91373f9c9
Signed-off-by: Tiago Mück <tiago.muck@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/31421
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Reviewed-by: John Alsop <johnathan.alsop@amd.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
There are cases in which we need to prevent randomization for a
specific buffer when enabled at the RubySystem level (e.g. a internal
trigger queue that requires zero latency enqueue, while other buffers
can be randomized).
This changes the randomization parameter to support enabling and
disabling randomization regardless of the RubySystem setting.
Change-Id: If7520153cc5864897fa42e8911a6f8acbcf01db5
Signed-off-by: Tiago Mück <tiago.muck@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/31419
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Ruby prefetcher's non-unit filter is a circular queue, so use the class
created for this functionality.
This changes the behavior, since previously iterating through the
filter was completely arbitrary, and now it iterates from the
beginning of the queue to the end when accessing and updating
the filter's contents.
Change-Id: I3148efcbef00da0c8f6cf2dee7fb86f6c2ddb27d
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/24533
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Ruby prefetcher's unit filter is a circular queue, so use the class
created for this functionality.
This changes the behavior, since previously iterating through the
filter was completely arbitrary, and now it iterates from the
beginning of the queue to the end when accessing and updating
the filter's contents.
Change-Id: I834be88a33580d5857c38e9bae8b289c5a6250b9
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/24532
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Messages may be enqueued and be ready in the same cycle.
Using this feature may introduce nondeterminism in the protocol and
should be used in specific cases. A case study is to avoid needing an
additional cycle for internal protocol triggers (e.g. the All_Acks
event in src/mem/ruby/protocol/MOESI_CMP_directory-L2cache.sm).
To mitigate modeling mistakes, the 'allow_zero_latency' parameter must
be set for a MessageBuffer where this behavior is acceptable.
This changes also updates the Consumer to schedule events according to
this new behavior. The original implementation would not schedule a new
wakeup event if the wakeup for the Consumer had already been executed
in that cycle.
Additional authors:
- Tuan Ta <tuan.ta2@arm.com>
Change-Id: Ib194e7b4b4ee4b06da1baea17c0eb743f650dfdd
Signed-off-by: Tiago Mück <tiago.muck@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/31255
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
The is a bug in the GPUCoalescer which occurs in the following
situation:
1) An instruction crosses a page boundary causing multiple TLB requests
to be sent.
2) The TLB responses arrive at different times, causing the vector
memory requests to be sent at different times.
3) The first vector memory request completes before the second vector
memory request arrives at the coalescer.
This caused the coalescer to consider the instruction sequence number
done and return its token. Then the second request would arrive and
complete sending back another token. Eventually this increases the token
count beyond the maximum tripping an assert.
This change keeps track of the number of per-lane requests which are
expected to be sent in the vector memory request by looking at the exec
mask of the instruction. The token is not returned until the expected
number of per-lane requests have been coalesced. This fixes "#7" in the
list of issues in JIRA-300. There are also style fixes for local
variables in code nearby the changes in this CL.
Change-Id: I152fd9397920ad82ba6079112908387e71ff3cce
JIRA: https://gem5.atlassian.net/browse/GEM5-300
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/35176
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Reviewed-by: Kyle Roarty <kyleroarty1716@gmail.com>
Maintainer: Matt Sinclair <mattdsinclair@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This change replaces the __attribute__ syntax with the now standard [[]]
syntax. It also reorganizes compiler.hh so that all special macros have
some explanatory text saying what they do, and each attribute which has a
standard version can use that if available and what version of c++ it's
standard in is put in a comment.
Also, the requirements as far as where you put [[]] style attributes are
a little more strict than the old school __attribute__ style. The use of
the attribute macros was updated to fit these new, more strict
requirements.
Change-Id: Iace44306a534111f1c38b9856dc9e88cd9b49d2a
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/35219
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Maintainer: Gabe Black <gabeblack@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
These were including instruction class definitions from x86 for some
reason. There was no code in those .cc files which actually used
anything from them, as evidenced by the fact that the GCN3_X86 build
still works. No other code in the file was conditionally compiled as of
today.
Change-Id: I3cef8348fb601dd7af67665cf64bbf514c91c3db
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/34577
Reviewed-by: Matthew Poremba <matthew.poremba@amd.com>
Maintainer: Bobby R. Bruce <bbruce@ucdavis.edu>
Tested-by: kokoro <noreply+kokoro@google.com>
This version of garnet includes HeteroGarnet which
supports heterogenous interconnect systems, flexible
router and link configurations, and better debugging
resources.
This patch changes the garnet directory structure
to not include the version number. The user will be
informed about the garnet version being used.
Change-Id: Id4763421528305193ae0cd10c159b385a9513553
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/34259
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Initializing the network bridge with NULL causes it to have
an class error when instatiating a link. The bridge is only
needed whne either a CDC or SerDes is enabled. This is handled
later during construction of the GarnetLink.
Change-Id: If19a21a6d9bf49449b9c390467d08d3422ae991a
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/34257
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This patch augments the MESI_Three_Level Ruby protocol with hardware
transactional memory support.
The HTM implementation relies on buffering of speculative memory updates.
The core notifies the L0 cache controller that a new transaction has
started and the controller in turn places itself in transactional state
(htmTransactionalState := true).
When operating in transactional state, the usual MESI protocol changes
slightly. Lines loaded or stored are marked as part of a transaction's
read and write set respectively. If there is an invalidation request to
cache line in the read/write set, the transaction is marked as failed.
Similarly, if there is a read request by another core to a speculatively
written cache line, i.e. in the write set, the transaction is marked as
failed. If failed, all subsequent loads and stores from the core are
made benign, i.e. made into NOPS at the cache controller, and responses
are marked to indicate that the transactional state has failed. When the
core receives these marked responses, it generates a HtmFailureFault
with the reason for the transaction failure. Servicing this fault does
two things--
(a) Restores the architectural checkpoint
(b) Sends an HTM abort signal to the cache controller
The restoration includes all registers in the checkpoint as well as the
program counter of the instruction before the transaction started.
The abort signal is sent to the L0 cache controller and resets the
failed transactional state. It resets the transactional read and write
sets and invalidates any speculatively written cache lines. It also
exits the transactional state so that the MESI protocol operates as
usual.
Alternatively, if the instructions within a transaction complete without
triggering a HtmFailureFault, the transaction can be committed. The core
is responsible for notifying the cache controller that the transaction
is complete and the cache controller makes all speculative writes
visible to the rest of the system and exits the transactional state.
Notifting the cache controller is done through HtmCmd Requests which are
a subtype of Load Requests.
KUDOS:
The code is based on a previous pull request by Pradip Vallathol who
developed HTM and TSX support in Gem5 as part of his master’s thesis:
http://reviews.gem5.org/r/2308/index.html
JIRA: https://gem5.atlassian.net/browse/GEM5-587
Change-Id: Icc328df93363486e923b8bd54f4d77741d8f5650
Signed-off-by: Giacomo Travaglini <giacomo.travaglini@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/30319
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Currently there are independent round robin arbiter at each
input port and output port. Every time a VC is selected for
output allocation round robin is incremented irrespective of
if it is selected by its output port or not. This leads to
unfair arbitration at input port and is well known[1]. This
patch fixes it to increment only if the output port also
selects it.
[1] D. U. Becker and W. J. Dally, "Allocator implementations
for network-on-chip routers," Proceedings of the Conference
on High Performance Computing Networking, Storage and
Analysis, Portland, OR, 2009, pp. 1-12
Change-Id: I65963fb8082c51c0e3c6e031a8b87b4f5c3626e1
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/32601
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Currently the Switch Allocator takes up most of the simulation
wall clock time. This function checks for all VCs to see if it
should wakeup next. The input units which are simulated before
the switch allocator could have scheduled it already. This patch
adds a check for it.
Change-Id: I8609d4e7f925aa5e97198f6cd07466530f6fcf4c
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/32600
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
This change allows configuring each router with a certain number
of VCs for each VNET. This is beneficial when dealing with
heterogenous link widths in a system. Configuring VCs
for each router allows one to ensure equal throughput
within the network while avoiding head-of-line blocking.
Changing a router's VCs number can be done in topology files
using the vcs_per_vnet value argument of router.
Change-Id: Icf4f510248128429a1a11f19f9802ee96f340611
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/32599
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This upgrades the garnet model to support HeteroGarnet
1) Static and dynamic multi-freq domains in network
2) Support for CDC
3) Separate links for each message class
4) Separate linkwidth for each message class
5) Support for SerDes
Change-Id: I6d00e3b5cb3745e849d221066cb46b2138c47871
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/32597
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
