There are several parts to this PR to work towards #1349 .
(1) Make RubySystem::getBlockSizeBytes non-static by providing ways to
access the block size or passing the block size explicitly to classes.
The main changes are:
- DataBlocks must be explicitly allocated. A default ctor still exists
to avoid needing to heavily modify SLICC. The size can be set using a
realloc function, operator=, or copy ctor. This is handled completely
transparently meaning no protocol or config changes are required.
- WriteMask now requires block size to be set. This is also handled
transparently by modifying the SLICC parser to identify WriteMask
types and call setBlockSize().
- AbstractCacheEntry and TBE classes now require block size to be set.
This is handled transparently by modifying the SLICC parser to
identify these classes and call initBlockSize() which calls
setBlockSize() for any DataBlock or WriteMask.
- All AbstractControllers now have a pointer to RubySystem. This is
assigned in SLICC generated code and requires no changes to protocol
or configs.
- The Ruby Message class now requires block size in all constructors.
This is added to the argument list automatically by the SLICC parser.
(2) Relax dependence on common functions in
src/mem/ruby/common/Address.hh
so that RubySystem::getBlockSizeBits is no longer static. Many classes
already have a way to get block size from the previous commit, so they
simply multiple by 8 to get the number of bits. For handling SLICC and
reducing the number of changes, define makeCacheLine, getOffset, etc. in
RubyPort and AbstractController. The only protocol changes required are
to change any "RubySystem::foo()" calls with "m_ruby_system->foo()".
For classes which do not have a way to get access to block size but
still used makeLineAddress, getOffset, etc., the block size must be
passed to that class. This requires some changes to the SimObject
interface for two commonly used classes: DirectoryMemory and
RubyPrefecther, resulting in user-facing API changes
User-facing API changes:
- DirectoryMemory and RubyPrefetcher now require the cache line size as
a non-optional argument.
- RubySequencer SimObjects now require RubySystem as a non-optional
argument.
- TesterThread in the GPU ruby tester now requires the cache line size
as a non-optional argument.
(3) Removes static member variables in RubySystem which control
randomization, cooldown, and warmup. These are mostly used by the Ruby
Network. The network classes are modified to take these former static
variables as parameters which are passed to the corresponding method
(e.g., enqueue, delayHead, etc.) rather than needing a RubySystem object
at all.
Change-Id: Ia63c2ad5cf0bf9d1cbdffba5d3a679bb4d3b1220
(4) There are two major SLICC generated static methods:
getNumControllers()
on each cache controller which returns the number of controllers created
by the configs at run time and the functions which access this method,
which are MachineType_base_count and MachineType_base_number. These need
to be removed to create multiple RubySystem objects otherwise NetDest,
version value, and other objects are incorrect.
To remove the static requirement, MachineType_base_count and
MachineType_base_number are moved to RubySystem. Any class which needs
to call these methods must now have a pointer to a RubySystem. To enable
that, several changes are made:
- RubyRequest and Message now require a RubySystem pointer in the
constructor. The pointer is passed to fields in the Message class
which require a RubySystem pointer (e.g., NetDest). SLICC is modified
to do this automatically.
- SLICC structures may now optionally take an "implicit constructor"
which can be used to call a non-default constructor for locally
defined variables (e.g., temporary variables within SLICC actions). A
statement such as "NetDest bcast_dest;" in SLICC will implicitly
append a call to the NetDest constructor taking RubySystem, for
example.
- RubySystem gets passed to Ruby network objects (Network, Topology).
This PR changes memory and cache sizes in various parts of the gem5
codebase to use binary units (e.g. KiB) instead of metric units (e.g.
kB). This makes the codebase more consistent, as gem5 automatically
converts memory and cache sizes that are in metric units to binary
units.
This PR also adds a warning message to let users know when an
auto-conversion from base 10 to base 2 units occurs.
There were a few places in configs and in the comments of various files
where I didn't change the metric units, as I couldn't figure out where
the parameters with those units were being used.
The Vega ISA's s_memtime instruction is used to obtain a cycle value
from the GPU. Previously, this was implemented to obtain the cycle count
when the memtime instruction reached the execute stage of the GPU
pipeline. However, from microbenchmarking we have found that this under
reports the latency for memtime instructions relative to real hardware.
Thus, we changed its behavior to go through the scalar memory pipeline
and obtain a latency value from the the SQC (L1 I$). This mirrors the
suggestion of the AMD Vega ISA manual that s_memtime should be treated
like a s_load_dwordx2.
The default latency was set based on microbenchmarking.
Change-Id: I5e251dde28c06fe1c492aea4abf9f34f05784420
This commit contains the rest of the base 2 vs base 10 cache/memory
size clarifications. It also changes the warning message to use
warn(). With these changes, the warning message should now no
longer show up during a fresh compilation of gem5.
Change-Id: Ia63f841bdf045b76473437f41548fab27dc19631
Rather than adding the options to *every* config that might be using
GPU_VIPER.py, just change the Ruby config to check if the option is
available before trying to use it. Otherwise, reverts to what was the
default on stable.
Change-Id: Ia6f1d0827d489ee2a35c598b644461cbff59e247
Adding RP_choose function to change replacement policies among
TreePLRU, LRU, FIFO, LFU, LIP, MRU, NRU, RRIP, SecondChance AND ShiPMem replacement
policies for TCC, TCP and SQC caches for GPU
Adding RP_choose functions to change replacement policies among
TreePLRU, LRU, FIFO, LFU, LIP, MRU, NRU, RRIP, SecondChance AND ShiPMem replacement
policies for TCC, TCP and SQC caches for GPU
Adding RP_choose functions to change replacement policies among
TreePLRU, LRU, FIFO, LFU, LIP, MRU, NRU, RRIP, SecondChance AND ShiPMem replacement
policies for TCC, TCP and SQC caches for GPU
The scalar cache is not being invalidated which causes stale data to be
left in the scalar cache between GPU kernels. This commit sends
invalidates to the scalar cache when the SQC is invalidated. This is a
sufficient baseline for simulation.
Since the number of invalidates might be larger than the mandatory queue
can hold and no flash invalidate mechanism exists in the VIPER protocol,
the command line option for the mandatory queue size is removed, which
is the same behavior as the SQC.
Change-Id: I1723f224711b04caa4c88beccfa8fb73ccf56572
Added a resource constraint, AtomicALUOperation, to GLC atomics
performed in the TCC.
The resource constraint uses a new class, ALUFreeList array. The class
assumes the following:
- There are a fixed number of atomic ALU pipelines
- While a new cache line can be processed in each pipeline each cycle,
if a cache line is currently going through a pipeline, it can't be
processed again until it's finished
Two configuration parameters have been used to tune this behavior:
- tcc-num-atomic-alus corresponds to the number of atomic ALU pipelines
- atomic-alu-latency corresponds to the latency of atomic ALU pipelines
Change-Id: I25bdde7dafc3877590bb6536efdf57b8c540a939
PR #367 adds an option to configs/ruby/GPU_VIPER.py that was not added
to the corresponding dGPU equal for GPUFS and thus all GPUFS runs are
failing. Fixed in this patch.
Added checks to ensure that atomics are not performed in the TCC when it
is configured as a write-through cache. Also added SLC bit overwrite to
ensure directory preforms atomics when there is a write-through TCC.
Change-Id: I4514e6c8022aeb7785f2c59871cd9acec8161ed8
Previously, the L1, L2 number of banks and L2 latencies were not
configurable through command line arguments. This commit adds support to
configure them through the arguments '--tcp-num-banks' for number of
banks in L1, '--tcc-num-banks' for number of banks in L2, and
'--tcc-tag-access-latency', and '--tcc-data-access-latency'
Change-Id: Ie3b713ead16865fd7120e2d809ebfa56b69bc4a1
Added a GLC atomic latency parameter (glc-atomic-latency) used when
enqueueing response messages regarding atomics directly performed in
the TCC. This latency is added in addition to the L2 response latency
(TCC_latency). This represents the latency of performing an atomic
within the L2.
With this change, the TCC response queue will receive enqueues with
varying latencies as GLC atomic responses will have this added GLC
atomic latency while data responses will not. To accommodate this in
light of the queue having strict FIFO ordering (which would be violated
here), this change also adds an optional parameter bypassStrictFIFO to
the SLICC enqueue function which allows overriding strict FIFO
requirements for individual messages on a case-by-case basis. This
parameter is only being used in the TCC's atomic response enqueue call.
Change-Id: Iabd52cbd2c0cc385c1fb3fe7bcd0cc64bdb40aac
Update the TCP_latency input arg to reflect what it does -- in
combination with the number of banks, it determines the number of
accesses that can happen in the L1 (TCP) in a given cycle. It does
not directly affect the L1 latency as the name implies. Instead,
the mandatory_queue_latency does this.
Change-Id: Ib6cbc8367ce2b1f30005d137384f53650a403b49
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/61309
Reviewed-by: Bobby Bruce <bbruce@ucdavis.edu>
Maintainer: Bobby Bruce <bbruce@ucdavis.edu>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Matthew Poremba <matthew.poremba@amd.com>
The VIPER configuration uses the MOESI_AMD_Base protocol's directory.
This protocol does not wait for memory ACKs. As a result, this can lead
to read requests being pulled out of the MessageBuffer between the
directory and DRAMCtrl before a write request to the same address. This
leads to inconsistent data. To fix this, make the MessageBuffers
ordered. Since these MessageBuffers are essentially just an interface
between SLICC and DRAMCtrl, and DRAMCtrl can reorder requests properly,
this should not cause any large impact on performance due to the
constraint.
Also remove the duplicate instantiation of these MessageBuffers.
Change-Id: I59653717cc79884e733af3958adfc14941703958
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/57411
Reviewed-by: Matthew Poremba <matthew.poremba@amd.com>
Maintainer: Matthew Poremba <matthew.poremba@amd.com>
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Maintainer: Matt Sinclair <mattdsinclair@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Remove the line "For use for simulation and test purposes only" in files
were AMD is the only copyright holder listed in the header. This happens
to be the case for all files where this line exists, removing it
completely from gem5.
Change-Id: I623f266b002f564301b28774f49081099cfc60fd
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/53943
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
In order to have more fine grained control over which SLICC controllers
are part of which Ruby network in a disjoint configuration, the
create_system function in GPU_VIPER is broken up into multiple construct
calls for each SLICC machine type in the protocol. By default this does
not change anything functionally. A future config will use the construct
calls to explicitly set which network (CPU or GPU) the controller is in.
Change-Id: Ic038b300c5c3732e96992ef4bfe14e43fa0ea824
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/51847
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Viper is checking for the dma's type before making the port assignment.
In FullSystem mode the IDE device is a PortRef and does not have an
attribute 'type.' This handles the various types a bit better and
ensures that IDE device, the protocol tester, and upcoming DMA devices
related to FullSystem can be added.
Change-Id: I6879b25c6aabbbc22b0ee8dc9cbfec6399f70daa
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/44806
Reviewed-by: Matthew Poremba <matthew.poremba@amd.com>
Maintainer: Matthew Poremba <matthew.poremba@amd.com>
Tested-by: kokoro <noreply+kokoro@google.com>
There was a missing option (--buffers-size) used to set the mandatory
queue size for the scalar controllers. This patch renames the option to
be more clear, and adds it to the argument parser.
Default of 128 taken from the implementation on the GCN staging branch
Change-Id: I58b6b57be07498cdf6e39c0bb85982674ec4caa6
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/32676
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Maintainer: Anthony Gutierrez <anthony.gutierrez@amd.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This was originally from the GCN staging branch, which only had
GPU_VIPER.py, but the other GPU_VIPER configs had DirMem as well, so I
applied this change to all of them.
The patch replaces the Directory in DirCntrl from DirMem to
RubyDirectoryMemory. This fixes errors that DirMem caused relating to
setting class variables. It also generates and sets addr_ranges in
DirCntrl as RubyDirectoryMemory uses the parent object's addr_ranges
in its code
The style checker complained about a line length in GPU_VIPER_Region,
so the patch also fixes that
Change-Id: Icec96777a51d8a826b576fc752fae0f7f15427bc
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/32674
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Reviewed-by: Bradford Beckmann <brad.beckmann@amd.com>
Maintainer: Bradford Beckmann <brad.beckmann@amd.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This changeset adds the necessary changes for running
GCN3 ISA with VIPER in apu_se.py.
Changes to the VIPER protocol configs are made to add support
for DMA and scalar caches.
hsaTopology is added to help the pseudo FS create the files
needed by ROCm to understand the device on which the SW is
being run.
Change-Id: I0f47a6a36bb241a26972c0faafafcf332a7d7d1f
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/30274
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Maintainer: Bradford Beckmann <brad.beckmann@amd.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Calls to queueMemoryRead and queueMemoryWrite do not consider the size
of the queue between ruby directories and DRAMCtrl which causes infinite
buffering in the queued port between the two. This adds a MessageBuffer
in between which uses enqueues in SLICC and is therefore size checked
before any SLICC transaction pushing to the buffer can occur, removing
the infinite buffering between the two.
Change-Id: Iedb9070844e4f6c8532a9c914d126105ec98d0bc
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/27427
Tested-by: Gem5 Cloud Project GCB service account <345032938727@cloudbuild.gserviceaccount.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Bradford Beckmann <brad.beckmann@amd.com>
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Maintainer: Bradford Beckmann <brad.beckmann@amd.com>
Add support in Ruby to use all replacement policies in Classic.
Furthermore, if new replacement policies are added to the
Classic system, the Ruby system will recognize new policies
without any other changes in Ruby system. The following list
all the major changes:
* Make Ruby cache entries (AbstractCacheEntry) inherit from
Classic cache entries (ReplaceableEntry). By doing this,
replacement policies can use cache entries from Ruby caches.
AccessPermission and print function are moved from
AbstractEntry to AbstractCacheEntry, so AbstractEntry is no
longer needed.
* DirectoryMemory and all SLICC files are changed to use
AbstractCacheEntry as their cache entry interface. So do the
python files in mem/slicc/ast which check the entry
interface.
* "main='false'" argument is added to the protocol files where
the DirectoryEntry is defined. This change helps
differentiate DirectoryEntry from CacheEntry because they are
both the instances of AbstractCacheEntry now.
* Use BaseReplacementPolicy in Ruby caches instead of
AbstractReplacementPolicy so that Ruby caches will recognize
the replacement policies from Classic.
* Add getLastAccess() and useOccupancy() function to Classic
system so that Ruby caches can use them. Move lastTouchTick
to ReplacementData struct because it's needed by
getLastAccess() to return the correct value.
* Add a 2-dimensional array of ReplacementData in Ruby caches
to store information for different replacement policies. Note
that, unlike Classic caches, where policy information is
stored in cache entries, the policy information needs to be
stored in a new 2-dimensional array. This is due to Ruby
caches deleting the cache entry every time the corresponding
cache line get evicted.
Change-Id: Idff6fdd2102a552c103e9d5f31f779aae052943f
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/20879
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Maintainer: Jason Lowe-Power <jason@lowepower.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This patch updates the FileSystemConfig so it works with more kinds of
config scripts (e.g., the Learning gem5 scripts).
There are 4 main changes:
- Added system as a parameter to the config_filesystem function so the
function can search the system for the number of CPUs instead of relying
on options from Options.py
- Instead of calling redirect_paths everywhere config_filesystem is
used, now it is implicitly called.
- Cleaned up the Ruby scripts a bit to remove redundant calls to
config_filesystem
- Added a config_filesystem call to the Ruby Learning gem5 script
(currently the only Learning gem5 script that requires it).
In the future, I think it would be better to move the config_filesystem
call into simulate.py, probably into the instantiate function. I tried to
use the per-CPU configuration parameters instead of options from
Options.py, but that's not possible until after the SimObject params
have been finalized in instantiate.
Change-Id: Ie6501a7435cfb3ac9d2b45be3722388b34063b1e
Signed-off-by: Jason Lowe-Power <jason@lowepower.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/18848
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Tested-by: kokoro <noreply+kokoro@google.com>
Python 2.7 used to return lists for operations such as map and range,
this has changed in Python 3. To make the configs Python 3 compliant,
add explicit conversions from iterators to lists where needed, replace
xrange with range, and fix changes to exec syntax.
This change doesn't fix import paths since that might require us to
restructure the configs slightly.
Change-Id: Idcea8482b286779fc98b4e144ca8f54069c08024
Signed-off-by: Andreas Sandberg <andreas.sandberg@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/16002
Reviewed-by: Gabe Black <gabeblack@google.com>
