sanchayanmaity/gem5
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
gem5/configs/common/O3_ARM_v7a.py

191 lines
6.1 KiB
Python
Raw Normal View History

configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
# Copyright (c) 2012 The Regents of The University of Michigan
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met: redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer;
# redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution;
# neither the name of the copyright holders nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#
# Authors: Ron Dreslinski
from m5.objects import *
# Simple ALU Instructions have a latency of 1
class O3_ARM_v7a_Simple_Int(FUDesc):
opList = [ OpDesc(opClass='IntAlu', opLat=1) ]
count = 2
# Complex ALU instructions have a variable latencies
class O3_ARM_v7a_Complex_Int(FUDesc):
cpu: o3: replace issueLatency with bool pipelined Currently, each op class has a parameter issueLat that denotes the cycles after which another op of the same class can be issued. As of now, this latency can either be one cycle (fully pipelined) or same as execution latency of the op (not at all pipelined). The fact that issueLat is a parameter of type Cycles makes one believe that it can be set to any value. To avoid the confusion, the parameter is being renamed as 'pipelined' with type boolean. If set to true, the op would execute in a fully pipelined fashion. Otherwise, it would execute in an unpipelined fashion.
2015-04-30 05:35:22 +02:00
opList = [ OpDesc(opClass='IntMult', opLat=3, pipelined=True),
OpDesc(opClass='IntDiv', opLat=12, pipelined=False),
OpDesc(opClass='IprAccess', opLat=3, pipelined=True) ]
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
count = 1
arm: make the bi-mode predictor the default for O3_ARM_v7a_BP the branch predictor used in the Cortex-A15 is a bi-mode style predictor, see: http://arm.com/files/pdf/at-exploring_the_design_of_the_cortex-a15.pdf and http://nvidia.com/docs/IO/116757/NVIDIA_Quad_a15_whitepaper_FINALv2.pdf this patch makes the bi-mode predictor the default for the ARM O3 CPU.
2014-06-30 19:50:01 +02:00
# Floating point and SIMD instructions
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
class O3_ARM_v7a_FP(FUDesc):
opList = [ OpDesc(opClass='SimdAdd', opLat=4),
OpDesc(opClass='SimdAddAcc', opLat=4),
OpDesc(opClass='SimdAlu', opLat=4),
OpDesc(opClass='SimdCmp', opLat=4),
OpDesc(opClass='SimdCvt', opLat=3),
OpDesc(opClass='SimdMisc', opLat=3),
OpDesc(opClass='SimdMult',opLat=5),
OpDesc(opClass='SimdMultAcc',opLat=5),
OpDesc(opClass='SimdShift',opLat=3),
OpDesc(opClass='SimdShiftAcc', opLat=3),
OpDesc(opClass='SimdSqrt', opLat=9),
OpDesc(opClass='SimdFloatAdd',opLat=5),
OpDesc(opClass='SimdFloatAlu',opLat=5),
OpDesc(opClass='SimdFloatCmp', opLat=3),
OpDesc(opClass='SimdFloatCvt', opLat=3),
OpDesc(opClass='SimdFloatDiv', opLat=3),
OpDesc(opClass='SimdFloatMisc', opLat=3),
OpDesc(opClass='SimdFloatMult', opLat=3),
OpDesc(opClass='SimdFloatMultAcc',opLat=1),
OpDesc(opClass='SimdFloatSqrt', opLat=9),
OpDesc(opClass='FloatAdd', opLat=5),
OpDesc(opClass='FloatCmp', opLat=5),
OpDesc(opClass='FloatCvt', opLat=5),
cpu: o3: replace issueLatency with bool pipelined Currently, each op class has a parameter issueLat that denotes the cycles after which another op of the same class can be issued. As of now, this latency can either be one cycle (fully pipelined) or same as execution latency of the op (not at all pipelined). The fact that issueLat is a parameter of type Cycles makes one believe that it can be set to any value. To avoid the confusion, the parameter is being renamed as 'pipelined' with type boolean. If set to true, the op would execute in a fully pipelined fashion. Otherwise, it would execute in an unpipelined fashion.
2015-04-30 05:35:22 +02:00
OpDesc(opClass='FloatDiv', opLat=9, pipelined=False),
OpDesc(opClass='FloatSqrt', opLat=33, pipelined=False),
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
OpDesc(opClass='FloatMult', opLat=4) ]
count = 2
# Load/Store Units
class O3_ARM_v7a_Load(FUDesc):
opList = [ OpDesc(opClass='MemRead',opLat=2) ]
count = 1
class O3_ARM_v7a_Store(FUDesc):
opList = [OpDesc(opClass='MemWrite',opLat=2) ]
count = 1
# Functional Units for this CPU
class O3_ARM_v7a_FUP(FUPool):
FUList = [O3_ARM_v7a_Simple_Int(), O3_ARM_v7a_Complex_Int(),
O3_ARM_v7a_Load(), O3_ARM_v7a_Store(), O3_ARM_v7a_FP()]
arm: make the bi-mode predictor the default for O3_ARM_v7a_BP the branch predictor used in the Cortex-A15 is a bi-mode style predictor, see: http://arm.com/files/pdf/at-exploring_the_design_of_the_cortex-a15.pdf and http://nvidia.com/docs/IO/116757/NVIDIA_Quad_a15_whitepaper_FINALv2.pdf this patch makes the bi-mode predictor the default for the ARM O3 CPU.
2014-06-30 19:50:01 +02:00
# Bi-Mode Branch Predictor
cpu: re-organizes the branch predictor structure. Committed by: Nilay Vaish <nilay@cs.wisc.edu>
2015-04-14 00:33:57 +02:00
class O3_ARM_v7a_BP(BiModeBP):
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
globalPredictorSize = 8192
globalCtrBits = 2
choicePredictorSize = 8192
choiceCtrBits = 2
BTBEntries = 2048
BTBTagSize = 18
RASSize = 16
instShiftAmt = 2
branch predictor: move out of o3 and inorder cpus This patch moves the branch predictor files in the o3 and inorder directories to src/cpu/pred. This allows sharing the branch predictor across different cpu models. This patch was originally posted by Timothy Jones in July 2010 but never made it to the repository. --HG-- rename : src/cpu/o3/bpred_unit.cc => src/cpu/pred/bpred_unit.cc rename : src/cpu/o3/bpred_unit.hh => src/cpu/pred/bpred_unit.hh rename : src/cpu/o3/bpred_unit_impl.hh => src/cpu/pred/bpred_unit_impl.hh rename : src/cpu/o3/sat_counter.hh => src/cpu/pred/sat_counter.hh
2013-01-24 19:28:51 +01:00
class O3_ARM_v7a_3(DerivO3CPU):
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
LQEntries = 16
SQEntries = 16
LSQDepCheckShift = 0
LFSTSize = 1024
SSITSize = 1024
decodeToFetchDelay = 1
renameToFetchDelay = 1
iewToFetchDelay = 1
commitToFetchDelay = 1
renameToDecodeDelay = 1
iewToDecodeDelay = 1
commitToDecodeDelay = 1
iewToRenameDelay = 1
commitToRenameDelay = 1
commitToIEWDelay = 1
fetchWidth = 3
cpu: allow the fetch buffer to be smaller than a cache line the current implementation of the fetch buffer in the o3 cpu is only allowed to be the size of a cache line. some architectures, e.g., ARM, have fetch buffers smaller than a cache line, see slide 22 at: http://www.arm.com/files/pdf/at-exploring_the_design_of_the_cortex-a15.pdf this patch allows the fetch buffer to be set to values smaller than a cache line.
2013-11-15 19:21:15 +01:00
fetchBufferSize = 16
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
fetchToDecodeDelay = 3
decodeWidth = 3
decodeToRenameDelay = 2
renameWidth = 3
renameToIEWDelay = 1
issueToExecuteDelay = 1
dispatchWidth = 6
issueWidth = 8
wbWidth = 8
fuPool = O3_ARM_v7a_FUP()
iewToCommitDelay = 1
renameToROBDelay = 1
commitWidth = 8
squashWidth = 8
trapLatency = 13
backComSize = 5
forwardComSize = 5
numPhysIntRegs = 128
arm: Add support for ARMv8 (AArch64 & AArch32) Note: AArch64 and AArch32 interworking is not supported. If you use an AArch64 kernel you are restricted to AArch64 user-mode binaries. This will be addressed in a later patch. Note: Virtualization is only supported in AArch32 mode. This will also be fixed in a later patch. Contributors: Giacomo Gabrielli (TrustZone, LPAE, system-level AArch64, AArch64 NEON, validation) Thomas Grocutt (AArch32 Virtualization, AArch64 FP, validation) Mbou Eyole (AArch64 NEON, validation) Ali Saidi (AArch64 Linux support, code integration, validation) Edmund Grimley-Evans (AArch64 FP) William Wang (AArch64 Linux support) Rene De Jong (AArch64 Linux support, performance opt.) Matt Horsnell (AArch64 MP, validation) Matt Evans (device models, code integration, validation) Chris Adeniyi-Jones (AArch64 syscall-emulation) Prakash Ramrakhyani (validation) Dam Sunwoo (validation) Chander Sudanthi (validation) Stephan Diestelhorst (validation) Andreas Hansson (code integration, performance opt.) Eric Van Hensbergen (performance opt.) Gabe Black
2014-01-24 22:29:34 +01:00
numPhysFloatRegs = 192
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
numIQEntries = 32
numROBEntries = 40
cpu: Rename defer_registration->switched_out The defer_registration parameter is used to prevent a CPU from initializing at startup, leaving it in the "switched out" mode. The name of this parameter (and the help string) is confusing. This patch renames it to switched_out, which should be more descriptive.
2013-01-07 19:05:45 +01:00
switched_out = False
branch predictor: move out of o3 and inorder cpus This patch moves the branch predictor files in the o3 and inorder directories to src/cpu/pred. This allows sharing the branch predictor across different cpu models. This patch was originally posted by Timothy Jones in July 2010 but never made it to the repository. --HG-- rename : src/cpu/o3/bpred_unit.cc => src/cpu/pred/bpred_unit.cc rename : src/cpu/o3/bpred_unit.hh => src/cpu/pred/bpred_unit.hh rename : src/cpu/o3/bpred_unit_impl.hh => src/cpu/pred/bpred_unit_impl.hh rename : src/cpu/o3/sat_counter.hh => src/cpu/pred/sat_counter.hh
2013-01-24 19:28:51 +01:00
branchPred = O3_ARM_v7a_BP()
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
# Instruction Cache
class O3_ARM_v7a_ICache(BaseCache):
Mem: Use cycles to express cache-related latencies This patch changes the cache-related latencies from an absolute time expressed in Ticks, to a number of cycles that can be scaled with the clock period of the caches. Ultimately this patch serves to enable future work that involves dynamic frequency scaling. As an immediate benefit it also makes it more convenient to specify cache performance without implicitly assuming a specific CPU core operating frequency. The stat blocked_cycles that actually counter in ticks is now updated to count in cycles. As the timing is now rounded to the clock edges of the cache, there are some regressions that change. Plenty of them have very minor changes, whereas some regressions with a short run-time are perturbed quite significantly. A follow-on patch updates all the statistics for the regressions.
2012-10-15 14:10:54 +02:00
hit_latency = 1
response_latency = 1
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
mshrs = 2
tgts_per_mshr = 8
size = '32kB'
assoc = 2
arm, configs: Do not forward snoops from I cache This fix simply tells the I cache to not forward snoops to the fetch unit (since there is really no reason to do so).
2015-03-27 09:56:10 +01:00
forward_snoops = False
mem: Allow read-only caches and check compliance This patch adds a parameter to the BaseCache to enable a read-only cache, for example for the instruction cache, or table-walker cache (not for x86). A number of checks are put in place in the code to ensure a read-only cache does not end up with dirty data. A follow-on patch adds suitable read requests to allow a read-only cache to explicitly ask for clean data.
2015-07-03 16:14:39 +02:00
is_read_only = True
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
# Data Cache
class O3_ARM_v7a_DCache(BaseCache):
Mem: Use cycles to express cache-related latencies This patch changes the cache-related latencies from an absolute time expressed in Ticks, to a number of cycles that can be scaled with the clock period of the caches. Ultimately this patch serves to enable future work that involves dynamic frequency scaling. As an immediate benefit it also makes it more convenient to specify cache performance without implicitly assuming a specific CPU core operating frequency. The stat blocked_cycles that actually counter in ticks is now updated to count in cycles. As the timing is now rounded to the clock edges of the cache, there are some regressions that change. Plenty of them have very minor changes, whereas some regressions with a short run-time are perturbed quite significantly. A follow-on patch updates all the statistics for the regressions.
2012-10-15 14:10:54 +02:00
hit_latency = 2
response_latency = 2
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
mshrs = 6
tgts_per_mshr = 8
size = '32kB'
assoc = 2
write_buffers = 16
arm: make the bi-mode predictor the default for O3_ARM_v7a_BP the branch predictor used in the Cortex-A15 is a bi-mode style predictor, see: http://arm.com/files/pdf/at-exploring_the_design_of_the_cortex-a15.pdf and http://nvidia.com/docs/IO/116757/NVIDIA_Quad_a15_whitepaper_FINALv2.pdf this patch makes the bi-mode predictor the default for the ARM O3 CPU.
2014-06-30 19:50:01 +02:00
# TLB Cache
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
# Use a cache as a L2 TLB
class O3_ARM_v7aWalkCache(BaseCache):
Mem: Use cycles to express cache-related latencies This patch changes the cache-related latencies from an absolute time expressed in Ticks, to a number of cycles that can be scaled with the clock period of the caches. Ultimately this patch serves to enable future work that involves dynamic frequency scaling. As an immediate benefit it also makes it more convenient to specify cache performance without implicitly assuming a specific CPU core operating frequency. The stat blocked_cycles that actually counter in ticks is now updated to count in cycles. As the timing is now rounded to the clock edges of the cache, there are some regressions that change. Plenty of them have very minor changes, whereas some regressions with a short run-time are perturbed quite significantly. A follow-on patch updates all the statistics for the regressions.
2012-10-15 14:10:54 +02:00
hit_latency = 4
response_latency = 4
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
mshrs = 6
tgts_per_mshr = 8
size = '1kB'
assoc = 8
write_buffers = 16
arch, cpu: Do not forward snoops to table walker This patch simplifies the overall CPU by changing the TLB caches such that they do not forward snoops to the table walker port(s). Note that only ARM and X86 are affected. There is no reason for the ports to snoop as they do not actually take any action, and from a performance point of view we are better of not snooping more than we have to. Should it at a later point be required to snoop for a particular TLB design it is easy enough to add it back.
2015-05-05 09:22:27 +02:00
forward_snoops = False
mem: Allow read-only caches and check compliance This patch adds a parameter to the BaseCache to enable a read-only cache, for example for the instruction cache, or table-walker cache (not for x86). A number of checks are put in place in the code to ensure a read-only cache does not end up with dirty data. A follow-on patch adds suitable read requests to allow a read-only cache to explicitly ask for clean data.
2015-07-03 16:14:39 +02:00
is_read_only = True
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
# L2 Cache
class O3_ARM_v7aL2(BaseCache):
Mem: Use cycles to express cache-related latencies This patch changes the cache-related latencies from an absolute time expressed in Ticks, to a number of cycles that can be scaled with the clock period of the caches. Ultimately this patch serves to enable future work that involves dynamic frequency scaling. As an immediate benefit it also makes it more convenient to specify cache performance without implicitly assuming a specific CPU core operating frequency. The stat blocked_cycles that actually counter in ticks is now updated to count in cycles. As the timing is now rounded to the clock edges of the cache, there are some regressions that change. Plenty of them have very minor changes, whereas some regressions with a short run-time are perturbed quite significantly. A follow-on patch updates all the statistics for the regressions.
2012-10-15 14:10:54 +02:00
hit_latency = 12
response_latency = 12
configs: actually add ARMv7a-like cpu/cache file
2012-01-26 22:44:43 +01:00
mshrs = 16
tgts_per_mshr = 8
size = '1MB'
assoc = 16
write_buffers = 8
arm, configs: Do not forward snoops from I cache This fix simply tells the I cache to not forward snoops to the fetch unit (since there is really no reason to do so).
2015-03-27 09:56:10 +01:00
prefetch_on_access = True
prefetcher: Make prefetcher a sim object instead of it being a parameter on cache
2012-02-12 23:07:38 +01:00
# Simple stride prefetcher
Mem: Use cycles to express cache-related latencies This patch changes the cache-related latencies from an absolute time expressed in Ticks, to a number of cycles that can be scaled with the clock period of the caches. Ultimately this patch serves to enable future work that involves dynamic frequency scaling. As an immediate benefit it also makes it more convenient to specify cache performance without implicitly assuming a specific CPU core operating frequency. The stat blocked_cycles that actually counter in ticks is now updated to count in cycles. As the timing is now rounded to the clock edges of the cache, there are some regressions that change. Plenty of them have very minor changes, whereas some regressions with a short run-time are perturbed quite significantly. A follow-on patch updates all the statistics for the regressions.
2012-10-15 14:10:54 +02:00
prefetcher = StridePrefetcher(degree=8, latency = 1)
arm: make the PseudoLRU tags the default for the O3_ARM_v7aL2 the Cortex-A15 has a random replacement policy for its L2 cache. see the Cortex-A15 Technical Reference Manual 1.7 About the L2 memory system. this patch makes the PseudoLRU tags the default for the ARM O3 CPU's L2 cache.
2014-07-28 18:22:00 +02:00
tags = RandomRepl()
Reference in a new issue Copy permalink