diff options
45 files changed, 755 insertions, 3586 deletions
diff --git a/SConstruct b/SConstruct index 50ca2d6e4..7e8c6c2f0 100644 --- a/SConstruct +++ b/SConstruct @@ -66,6 +66,7 @@ # Python library imports import sys import os +from os.path import join as joinpath # Check for recent-enough Python and SCons versions. If your system's # default installation of Python is not recent enough, you can use a @@ -89,11 +90,11 @@ except: # The absolute path to the current directory (where this file lives). ROOT = Dir('.').abspath -# Paths to the M5 and external source trees. -SRCDIR = os.path.join(ROOT, 'src') +# Path to the M5 source tree. +SRCDIR = joinpath(ROOT, 'src') # tell python where to find m5 python code -sys.path.append(os.path.join(ROOT, 'src/python')) +sys.path.append(joinpath(ROOT, 'src/python')) ################################################### # @@ -105,13 +106,6 @@ sys.path.append(os.path.join(ROOT, 'src/python')) # Find default configuration & binary. Default(os.environ.get('M5_DEFAULT_BINARY', 'build/ALPHA_SE/m5.debug')) -# Ask SCons which directory it was invoked from. -launch_dir = GetLaunchDir() - -# Make targets relative to invocation directory -abs_targets = map(lambda x: os.path.normpath(os.path.join(launch_dir, str(x))), - BUILD_TARGETS) - # helper function: find last occurrence of element in list def rfind(l, elt, offs = -1): for i in range(len(l)+offs, 0, -1): @@ -141,6 +135,19 @@ def compare_versions(v1, v2): # recognize that ALPHA_SE specifies the configuration because it # follow 'build' in the bulid path. +# Generate absolute paths to targets so we can see where the build dir is +if COMMAND_LINE_TARGETS: + # Ask SCons which directory it was invoked from + launch_dir = GetLaunchDir() + # Make targets relative to invocation directory + abs_targets = map(lambda x: os.path.normpath(joinpath(launch_dir, str(x))), + COMMAND_LINE_TARGETS) +else: + # Default targets are relative to root of tree + abs_targets = map(lambda x: os.path.normpath(joinpath(ROOT, str(x))), + DEFAULT_TARGETS) + + # Generate a list of the unique build roots and configs that the # collected targets reference. build_paths = [] @@ -152,7 +159,7 @@ for t in abs_targets: except: print "Error: no non-leaf 'build' dir found on target path", t Exit(1) - this_build_root = os.path.join('/',*path_dirs[:build_top+1]) + this_build_root = joinpath('/',*path_dirs[:build_top+1]) if not build_root: build_root = this_build_root else: @@ -160,7 +167,7 @@ for t in abs_targets: print "Error: build targets not under same build root\n"\ " %s\n %s" % (build_root, this_build_root) Exit(1) - build_path = os.path.join('/',*path_dirs[:build_top+2]) + build_path = joinpath('/',*path_dirs[:build_top+2]) if build_path not in build_paths: build_paths.append(build_path) @@ -183,7 +190,7 @@ if ARGUMENTS.get('CC', None): if ARGUMENTS.get('CXX', None): env['CXX'] = ARGUMENTS.get('CXX') -env.SConsignFile(os.path.join(build_root,"sconsign")) +env.SConsignFile(joinpath(build_root,"sconsign")) # Default duplicate option is to use hard links, but this messes up # when you use emacs to edit a file in the target dir, as emacs moves @@ -243,8 +250,8 @@ env.Append(SCANNERS = swig_scanner) # Platform-specific configuration. Note again that we assume that all # builds under a given build root run on the same host platform. conf = Configure(env, - conf_dir = os.path.join(build_root, '.scons_config'), - log_file = os.path.join(build_root, 'scons_config.log')) + conf_dir = joinpath(build_root, '.scons_config'), + log_file = joinpath(build_root, 'scons_config.log')) # Find Python include and library directories for embedding the # interpreter. For consistency, we will use the same Python @@ -257,7 +264,7 @@ conf = Configure(env, py_version_name = 'python' + sys.version[:3] # include path, e.g. /usr/local/include/python2.4 -py_header_path = os.path.join(sys.exec_prefix, 'include', py_version_name) +py_header_path = joinpath(sys.exec_prefix, 'include', py_version_name) env.Append(CPPPATH = py_header_path) # verify that it works if not conf.CheckHeader('Python.h', '<>'): @@ -267,7 +274,7 @@ if not conf.CheckHeader('Python.h', '<>'): # add library path too if it's not in the default place py_lib_path = None if sys.exec_prefix != '/usr': - py_lib_path = os.path.join(sys.exec_prefix, 'lib') + py_lib_path = joinpath(sys.exec_prefix, 'lib') elif sys.platform == 'cygwin': # cygwin puts the .dll in /bin for some reason py_lib_path = '/bin' @@ -330,7 +337,7 @@ env['ALL_ISA_LIST'] = ['alpha', 'sparc', 'mips'] env['ALL_CPU_LIST'] = ['AtomicSimpleCPU', 'TimingSimpleCPU', 'O3CPU', 'OzoneCPU'] -if os.path.isdir(os.path.join(SRCDIR, 'src/encumbered/cpu/full')): +if os.path.isdir(joinpath(SRCDIR, 'encumbered/cpu/full')): env['ALL_CPU_LIST'] += ['FullCPU'] # Sticky options get saved in the options file so they persist from @@ -417,7 +424,7 @@ def config_emitter(target, source, env): # extract option name from Builder arg option = str(target[0]) # True target is config header file - target = os.path.join('config', option.lower() + '.hh') + target = joinpath('config', option.lower() + '.hh') val = env[option] if isinstance(val, bool): # Force value to 0/1 @@ -466,7 +473,7 @@ Usage: scons [scons options] [build options] [target(s)] # libelf build is shared across all configs in the build root. env.SConscript('ext/libelf/SConscript', - build_dir = os.path.join(build_root, 'libelf'), + build_dir = joinpath(build_root, 'libelf'), exports = 'env') ################################################### @@ -531,7 +538,7 @@ for build_path in build_paths: # Options for $BUILD_ROOT/$BUILD_DIR are stored in # $BUILD_ROOT/options/$BUILD_DIR so you can nuke # $BUILD_ROOT/$BUILD_DIR without losing your options settings. - current_opts_file = os.path.join(build_root, 'options', build_dir) + current_opts_file = joinpath(build_root, 'options', build_dir) if os.path.isfile(current_opts_file): sticky_opts.files.append(current_opts_file) print "Using saved options file %s" % current_opts_file @@ -546,8 +553,8 @@ for build_path in build_paths: # Get default build options from source tree. Options are # normally determined by name of $BUILD_DIR, but can be # overriden by 'default=' arg on command line. - default_opts_file = os.path.join('build_opts', - ARGUMENTS.get('default', build_dir)) + default_opts_file = joinpath('build_opts', + ARGUMENTS.get('default', build_dir)) if os.path.isfile(default_opts_file): sticky_opts.files.append(default_opts_file) print "Options file %s not found,\n using defaults in %s" \ @@ -611,7 +618,7 @@ for build_path in build_paths: # Set up the regression tests for each build. for e in envList: SConscript('tests/SConscript', - build_dir = os.path.join(build_path, 'tests', e.Label), + build_dir = joinpath(build_path, 'tests', e.Label), exports = { 'env' : e }, duplicate = False) Help(help_text) diff --git a/src/SConscript b/src/SConscript index 429e1bee1..f54e1de0d 100644 --- a/src/SConscript +++ b/src/SConscript @@ -108,6 +108,7 @@ base_sources = Split(''' mem/cache/coherence/coherence_protocol.cc mem/cache/coherence/uni_coherence.cc mem/cache/miss/blocking_buffer.cc + mem/cache/miss/miss_buffer.cc mem/cache/miss/miss_queue.cc mem/cache/miss/mshr.cc mem/cache/miss/mshr_queue.cc diff --git a/src/arch/isa_parser.py b/src/arch/isa_parser.py index 6504c7b32..a96622d4a 100755 --- a/src/arch/isa_parser.py +++ b/src/arch/isa_parser.py @@ -1180,15 +1180,16 @@ class IntRegOperand(Operand): if (self.ctype == 'float' or self.ctype == 'double'): error(0, 'Attempt to read integer register as FP') if (self.size == self.dflt_size): - return '%s = xc->readIntReg(this, %d);\n' % \ + return '%s = xc->readIntRegOperand(this, %d);\n' % \ (self.base_name, self.src_reg_idx) elif (self.size > self.dflt_size): - int_reg_val = 'xc->readIntReg(this, %d)' % (self.src_reg_idx) + int_reg_val = 'xc->readIntRegOperand(this, %d)' % \ + (self.src_reg_idx) if (self.is_signed): int_reg_val = 'sext<%d>(%s)' % (self.dflt_size, int_reg_val) return '%s = %s;\n' % (self.base_name, int_reg_val) else: - return '%s = bits(xc->readIntReg(this, %d), %d, 0);\n' % \ + return '%s = bits(xc->readIntRegOperand(this, %d), %d, 0);\n' % \ (self.base_name, self.src_reg_idx, self.size-1) def makeWrite(self): @@ -1201,7 +1202,7 @@ class IntRegOperand(Operand): wb = ''' { %s final_val = %s; - xc->setIntReg(this, %d, final_val);\n + xc->setIntRegOperand(this, %d, final_val);\n if (traceData) { traceData->setData(final_val); } }''' % (self.dflt_ctype, final_val, self.dest_reg_idx) return wb @@ -1227,13 +1228,13 @@ class FloatRegOperand(Operand): bit_select = 0 width = 0; if (self.ctype == 'float'): - func = 'readFloatReg' + func = 'readFloatRegOperand' width = 32; elif (self.ctype == 'double'): - func = 'readFloatReg' + func = 'readFloatRegOperand' width = 64; else: - func = 'readFloatRegBits' + func = 'readFloatRegOperandBits' if (self.ctype == 'uint32_t'): width = 32; elif (self.ctype == 'uint64_t'): @@ -1259,18 +1260,18 @@ class FloatRegOperand(Operand): width = 0 if (self.ctype == 'float'): width = 32 - func = 'setFloatReg' + func = 'setFloatRegOperand' elif (self.ctype == 'double'): width = 64 - func = 'setFloatReg' + func = 'setFloatRegOperand' elif (self.ctype == 'uint32_t'): - func = 'setFloatRegBits' + func = 'setFloatRegOperandBits' width = 32 elif (self.ctype == 'uint64_t'): - func = 'setFloatRegBits' + func = 'setFloatRegOperandBits' width = 64 else: - func = 'setFloatRegBits' + func = 'setFloatRegOperandBits' final_ctype = 'uint%d_t' % self.dflt_size if (self.size != self.dflt_size and self.is_signed): final_val = 'sext<%d>(%s)' % (self.size, self.base_name) diff --git a/src/arch/mips/isa/formats/fp.isa b/src/arch/mips/isa/formats/fp.isa index cdb892b3f..153f3f949 100644 --- a/src/arch/mips/isa/formats/fp.isa +++ b/src/arch/mips/isa/formats/fp.isa @@ -99,7 +99,7 @@ output exec {{ int size = sizeof(src_op) * 8; for (int i = 0; i < inst->numSrcRegs(); i++) { - uint64_t src_bits = xc->readFloatRegBits(inst, 0, size); + uint64_t src_bits = xc->readFloatRegOperandBits(inst, 0, size); if (isNan(&src_bits, size) ) { if (isSnan(&src_bits, size)) { @@ -113,7 +113,7 @@ output exec {{ mips_nan = src_bits; } - xc->setFloatRegBits(inst, 0, mips_nan, size); + xc->setFloatRegOperandBits(inst, 0, mips_nan, size); if (traceData) { traceData->setData(mips_nan); } return true; } @@ -139,7 +139,7 @@ output exec {{ } //Set value to QNAN - cpu->setFloatRegBits(inst, 0, mips_nan, size); + cpu->setFloatRegOperandBits(inst, 0, mips_nan, size); //Read FCSR from FloatRegFile uint32_t fcsr_bits = cpu->tcBase()->readFloatRegBits(FCSR); diff --git a/src/base/compression/base.hh b/src/base/compression/base.hh new file mode 100644 index 000000000..8d1f8d6b5 --- /dev/null +++ b/src/base/compression/base.hh @@ -0,0 +1,71 @@ +/* + * Copyright (c) 2003-2005 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: Erik Hallnor + * Nathan Binkert + */ + +#ifndef __BASE_COMPRESSION_BASE_HH__ +#define __BASE_COMPRESSION_BASE_HH__ + +/** + * @file + * This file defines a base (abstract virtual) compression algorithm object. + */ + +#include <inttypes.h> + +/** + * Abstract virtual compression algorithm object. + */ +class CompressionAlgorithm +{ + public: + virtual ~CompressionAlgorithm() {} + + /** + * Uncompress the data, causes a fatal since no data should be compressed. + * @param dest The output buffer. + * @param src The compressed data. + * @param size The number of bytes in src. + * + * @retval The size of the uncompressed data. + */ + virtual int uncompress(uint8_t * dest, uint8_t *src, int size) = 0; + + /** + * Compress the data, just returns the source data. + * @param dest The output buffer. + * @param src The data to be compressed. + * @param size The number of bytes in src. + * + * @retval The size of the compressed data. + */ + virtual int compress(uint8_t *dest, uint8_t *src, int size) = 0; +}; + +#endif //__BASE_COMPRESSION_BASE_HH__ diff --git a/src/base/compression/lzss_compression.hh b/src/base/compression/lzss_compression.hh index 35e4dcb3f..a23d07ffd 100644 --- a/src/base/compression/lzss_compression.hh +++ b/src/base/compression/lzss_compression.hh @@ -35,12 +35,12 @@ * LZSSCompression declarations. */ -#include "sim/host.hh" // for uint8_t +#include "base/compression/base.hh" /** * Simple LZSS compression scheme. */ -class LZSSCompression +class LZSSCompression : public CompressionAlgorithm { /** * Finds the longest substring for the given offset. diff --git a/src/base/compression/null_compression.hh b/src/base/compression/null_compression.hh index 5a582d828..ff110807a 100644 --- a/src/base/compression/null_compression.hh +++ b/src/base/compression/null_compression.hh @@ -38,41 +38,23 @@ */ #include "base/misc.hh" // for fatal() -#include "sim/host.hh" +#include "base/compression/base.hh" /** * A dummy compression class to use when no data compression is desired. */ -class NullCompression +class NullCompression : public CompressionAlgorithm { public: - /** - * Uncompress the data, causes a fatal since no data should be compressed. - * @param dest The output buffer. - * @param src The compressed data. - * @param size The number of bytes in src. - * - * @retval The size of the uncompressed data. - */ - static int uncompress(uint8_t * dest, uint8_t *src, int size) + int uncompress(uint8_t * dest, uint8_t *src, int size) { fatal("Can't uncompress data"); } - /** - * Compress the data, just returns the source data. - * @param dest The output buffer. - * @param src The data to be compressed. - * @param size The number of bytes in src. - * - * @retval The size of the compressed data. - */ - - static int compress(uint8_t *dest, uint8_t *src, int size) + int compress(uint8_t *dest, uint8_t *src, int size) { - memcpy(dest,src,size); - return size; + fatal("Can't compress data"); } }; diff --git a/src/cpu/base_dyn_inst.hh b/src/cpu/base_dyn_inst.hh index 0f7976631..17e121dbf 100644 --- a/src/cpu/base_dyn_inst.hh +++ b/src/cpu/base_dyn_inst.hh @@ -501,14 +501,15 @@ class BaseDynInst : public FastAlloc, public RefCounted double readDoubleResult() { return instResult.dbl; } /** Records an integer register being set to a value. */ - void setIntReg(const StaticInst *si, int idx, uint64_t val) + void setIntRegOperand(const StaticInst *si, int idx, uint64_t val) { if (recordResult) instResult.integer = val; } /** Records an fp register being set to a value. */ - void setFloatReg(const StaticInst *si, int idx, FloatReg val, int width) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val, + int width) { if (recordResult) { if (width == 32) @@ -521,21 +522,22 @@ class BaseDynInst : public FastAlloc, public RefCounted } /** Records an fp register being set to a value. */ - void setFloatReg(const StaticInst *si, int idx, FloatReg val) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val) { if (recordResult) instResult.dbl = (double)val; } /** Records an fp register being set to an integer value. */ - void setFloatRegBits(const StaticInst *si, int idx, uint64_t val, int width) + void setFloatRegOperandBits(const StaticInst *si, int idx, uint64_t val, + int width) { if (recordResult) instResult.integer = val; } /** Records an fp register being set to an integer value. */ - void setFloatRegBits(const StaticInst *si, int idx, uint64_t val) + void setFloatRegOperandBits(const StaticInst *si, int idx, uint64_t val) { if (recordResult) instResult.integer = val; diff --git a/src/cpu/checker/cpu.hh b/src/cpu/checker/cpu.hh index 9be54529f..3e08193ee 100644 --- a/src/cpu/checker/cpu.hh +++ b/src/cpu/checker/cpu.hh @@ -216,42 +216,44 @@ class CheckerCPU : public BaseCPU // storage (which is pretty hard to imagine they would have reason // to do). - uint64_t readIntReg(const StaticInst *si, int idx) + uint64_t readIntRegOperand(const StaticInst *si, int idx) { return thread->readIntReg(si->srcRegIdx(idx)); } - FloatReg readFloatReg(const StaticInst *si, int idx, int width) + FloatReg readFloatRegOperand(const StaticInst *si, int idx, int width) { int reg_idx = si->srcRegIdx(idx) - TheISA::FP_Base_DepTag; return thread->readFloatReg(reg_idx, width); } - FloatReg readFloatReg(const StaticInst *si, int idx) + FloatReg readFloatRegOperand(const StaticInst *si, int idx) { int reg_idx = si->srcRegIdx(idx) - TheISA::FP_Base_DepTag; return thread->readFloatReg(reg_idx); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx, int width) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx, + int width) { int reg_idx = si->srcRegIdx(idx) - TheISA::FP_Base_DepTag; return thread->readFloatRegBits(reg_idx, width); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx) { int reg_idx = si->srcRegIdx(idx) - TheISA::FP_Base_DepTag; return thread->readFloatRegBits(reg_idx); } - void setIntReg(const StaticInst *si, int idx, uint64_t val) + void setIntRegOperand(const StaticInst *si, int idx, uint64_t val) { thread->setIntReg(si->destRegIdx(idx), val); result.integer = val; } - void setFloatReg(const StaticInst *si, int idx, FloatReg val, int width) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val, + int width) { int reg_idx = si->destRegIdx(idx) - TheISA::FP_Base_DepTag; thread->setFloatReg(reg_idx, val, width); @@ -265,22 +267,23 @@ class CheckerCPU : public BaseCPU }; } - void setFloatReg(const StaticInst *si, int idx, FloatReg val) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val) { int reg_idx = si->destRegIdx(idx) - TheISA::FP_Base_DepTag; thread->setFloatReg(reg_idx, val); result.dbl = (double)val; } - void setFloatRegBits(const StaticInst *si, int idx, FloatRegBits val, - int width) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val, int width) { int reg_idx = si->destRegIdx(idx) - TheISA::FP_Base_DepTag; thread->setFloatRegBits(reg_idx, val, width); result.integer = val; } - void setFloatRegBits(const StaticInst *si, int idx, FloatRegBits val) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val) { int reg_idx = si->destRegIdx(idx) - TheISA::FP_Base_DepTag; thread->setFloatRegBits(reg_idx, val); diff --git a/src/cpu/exec_context.hh b/src/cpu/exec_context.hh index 13f70fa79..edccd747f 100644 --- a/src/cpu/exec_context.hh +++ b/src/cpu/exec_context.hh @@ -48,39 +48,42 @@ class ExecContext { // to do). /** Reads an integer register. */ - uint64_t readIntReg(const StaticInst *si, int idx); + uint64_t readIntRegOperand(const StaticInst *si, int idx); /** Reads a floating point register of a specific width. */ - FloatReg readFloatReg(const StaticInst *si, int idx, int width); + FloatReg readFloatRegOperand(const StaticInst *si, int idx, int width); /** Reads a floating point register of single register width. */ - FloatReg readFloatReg(const StaticInst *si, int idx); + FloatReg readFloatRegOperand(const StaticInst *si, int idx); /** Reads a floating point register of a specific width in its * binary format, instead of by value. */ - FloatRegBits readFloatRegBits(const StaticInst *si, int idx, int width); + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx, + int width); /** Reads a floating point register in its binary format, instead * of by value. */ - FloatRegBits readFloatRegBits(const StaticInst *si, int idx); + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx); /** Sets an integer register to a value. */ - void setIntReg(const StaticInst *si, int idx, uint64_t val); + void setIntRegOperand(const StaticInst *si, int idx, uint64_t val); /** Sets a floating point register of a specific width to a value. */ - void setFloatReg(const StaticInst *si, int idx, FloatReg val, int width); + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val, + int width); /** Sets a floating point register of single width to a value. */ - void setFloatReg(const StaticInst *si, int idx, FloatReg val); + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val); /** Sets the bits of a floating point register of a specific width * to a binary value. */ - void setFloatRegBits(const StaticInst *si, int idx, - FloatRegBits val, int width); + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val, int width); /** Sets the bits of a floating point register of single width * to a binary value. */ - void setFloatRegBits(const StaticInst *si, int idx, FloatRegBits val); + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val); /** Reads the PC. */ uint64_t readPC(); diff --git a/src/cpu/o3/alpha/dyn_inst.hh b/src/cpu/o3/alpha/dyn_inst.hh index 4e2ce63df..c340a806a 100644 --- a/src/cpu/o3/alpha/dyn_inst.hh +++ b/src/cpu/o3/alpha/dyn_inst.hh @@ -147,27 +147,28 @@ class AlphaDynInst : public BaseDynInst<Impl> // storage (which is pretty hard to imagine they would have reason // to do). - uint64_t readIntReg(const StaticInst *si, int idx) + uint64_t readIntRegOperand(const StaticInst *si, int idx) { return this->cpu->readIntReg(this->_srcRegIdx[idx]); } - FloatReg readFloatReg(const StaticInst *si, int idx, int width) + FloatReg readFloatRegOperand(const StaticInst *si, int idx, int width) { return this->cpu->readFloatReg(this->_srcRegIdx[idx], width); } - FloatReg readFloatReg(const StaticInst *si, int idx) + FloatReg readFloatRegOperand(const StaticInst *si, int idx) { return this->cpu->readFloatReg(this->_srcRegIdx[idx]); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx, int width) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx, + int width) { return this->cpu->readFloatRegBits(this->_srcRegIdx[idx], width); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx) { return this->cpu->readFloatRegBits(this->_srcRegIdx[idx]); } @@ -175,35 +176,37 @@ class AlphaDynInst : public BaseDynInst<Impl> /** @todo: Make results into arrays so they can handle multiple dest * registers. */ - void setIntReg(const StaticInst *si, int idx, uint64_t val) + void setIntRegOperand(const StaticInst *si, int idx, uint64_t val) { this->cpu->setIntReg(this->_destRegIdx[idx], val); - BaseDynInst<Impl>::setIntReg(si, idx, val); + BaseDynInst<Impl>::setIntRegOperand(si, idx, val); } - void setFloatReg(const StaticInst *si, int idx, FloatReg val, int width) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val, + int width) { this->cpu->setFloatReg(this->_destRegIdx[idx], val, width); - BaseDynInst<Impl>::setFloatReg(si, idx, val, width); + BaseDynInst<Impl>::setFloatRegOperand(si, idx, val, width); } - void setFloatReg(const StaticInst *si, int idx, FloatReg val) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val) { this->cpu->setFloatReg(this->_destRegIdx[idx], val); - BaseDynInst<Impl>::setFloatReg(si, idx, val); + BaseDynInst<Impl>::setFloatRegOperand(si, idx, val); } - void setFloatRegBits(const StaticInst *si, int idx, - FloatRegBits val, int width) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val, int width) { this->cpu->setFloatRegBits(this->_destRegIdx[idx], val, width); - BaseDynInst<Impl>::setFloatRegBits(si, idx, val); + BaseDynInst<Impl>::setFloatRegOperandBits(si, idx, val); } - void setFloatRegBits(const StaticInst *si, int idx, FloatRegBits val) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val) { this->cpu->setFloatRegBits(this->_destRegIdx[idx], val); - BaseDynInst<Impl>::setFloatRegBits(si, idx, val); + BaseDynInst<Impl>::setFloatRegOperandBits(si, idx, val); } public: diff --git a/src/cpu/o3/commit_impl.hh b/src/cpu/o3/commit_impl.hh index 1527efe1a..3178410a8 100644 --- a/src/cpu/o3/commit_impl.hh +++ b/src/cpu/o3/commit_impl.hh @@ -1180,16 +1180,16 @@ DefaultCommit<Impl>::getInsts() rename_idx < fromRename->size; rename_idx++) { DynInstPtr inst = fromRename->insts[rename_idx]; - int tid = inst->threadNumber; if (!inst->isSquashed()) { DPRINTF(Commit, "Inserting PC %#x [sn:%i] [tid:%i] into ", - "skidBuffer.\n", inst->readPC(), inst->seqNum, tid); + "skidBuffer.\n", inst->readPC(), inst->seqNum, + inst->threadNumber); skidBuffer.push(inst); } else { DPRINTF(Commit, "Instruction PC %#x [sn:%i] [tid:%i] was " "squashed, skipping.\n", - inst->readPC(), inst->seqNum, tid); + inst->readPC(), inst->seqNum, inst->threadNumber); } } } diff --git a/src/cpu/o3/fetch_impl.hh b/src/cpu/o3/fetch_impl.hh index f9e90614a..6cff52429 100644 --- a/src/cpu/o3/fetch_impl.hh +++ b/src/cpu/o3/fetch_impl.hh @@ -151,36 +151,6 @@ DefaultFetch<Impl>::DefaultFetch(Params *params) " RoundRobin,LSQcount,IQcount}\n"); } - // Size of cache block. - cacheBlkSize = 64; - - // Create mask to get rid of offset bits. - cacheBlkMask = (cacheBlkSize - 1); - - for (int tid=0; tid < numThreads; tid++) { - - fetchStatus[tid] = Running; - - priorityList.push_back(tid); - - memReq[tid] = NULL; - - // Create space to store a cache line. - cacheData[tid] = new uint8_t[cacheBlkSize]; - cacheDataPC[tid] = 0; - cacheDataValid[tid] = false; - - delaySlotInfo[tid].branchSeqNum = -1; - delaySlotInfo[tid].numInsts = 0; - delaySlotInfo[tid].targetAddr = 0; - delaySlotInfo[tid].targetReady = false; - - stalls[tid].decode = false; - stalls[tid].rename = false; - stalls[tid].iew = false; - stalls[tid].commit = false; - } - // Get the size of an instruction. instSize = sizeof(TheISA::MachInst); } @@ -353,6 +323,36 @@ DefaultFetch<Impl>::initStage() nextNPC[tid] = cpu->readNextNPC(tid); #endif } + + // Size of cache block. + cacheBlkSize = icachePort->peerBlockSize(); + + // Create mask to get rid of offset bits. + cacheBlkMask = (cacheBlkSize - 1); + + for (int tid=0; tid < numThreads; tid++) { + + fetchStatus[tid] = Running; + + priorityList.push_back(tid); + + memReq[tid] = NULL; + + // Create space to store a cache line. + cacheData[tid] = new uint8_t[cacheBlkSize]; + cacheDataPC[tid] = 0; + cacheDataValid[tid] = false; + + delaySlotInfo[tid].branchSeqNum = -1; + delaySlotInfo[tid].numInsts = 0; + delaySlotInfo[tid].targetAddr = 0; + delaySlotInfo[tid].targetReady = false; + + stalls[tid].decode = false; + stalls[tid].rename = false; + stalls[tid].iew = false; + stalls[tid].commit = false; + } } template<class Impl> diff --git a/src/cpu/o3/lsq_unit_impl.hh b/src/cpu/o3/lsq_unit_impl.hh index 379724166..7ce3dc9cb 100644 --- a/src/cpu/o3/lsq_unit_impl.hh +++ b/src/cpu/o3/lsq_unit_impl.hh @@ -418,7 +418,8 @@ LSQUnit<Impl>::executeLoad(DynInstPtr &inst) // realizes there is activity. // Mark it as executed unless it is an uncached load that // needs to hit the head of commit. - if (!(inst->req->isUncacheable()) || inst->isAtCommit()) { + if (!(inst->req && inst->req->isUncacheable()) || + inst->isAtCommit()) { inst->setExecuted(); } iewStage->instToCommit(inst); diff --git a/src/cpu/o3/mips/dyn_inst.hh b/src/cpu/o3/mips/dyn_inst.hh index bf82168ce..009a08012 100755 --- a/src/cpu/o3/mips/dyn_inst.hh +++ b/src/cpu/o3/mips/dyn_inst.hh @@ -136,27 +136,28 @@ class MipsDynInst : public BaseDynInst<Impl> // storage (which is pretty hard to imagine they would have reason // to do). - uint64_t readIntReg(const StaticInst *si, int idx) + uint64_t readIntRegOperand(const StaticInst *si, int idx) { return this->cpu->readIntReg(this->_srcRegIdx[idx]); } - FloatReg readFloatReg(const StaticInst *si, int idx, int width) + FloatReg readFloatRegOperand(const StaticInst *si, int idx, int width) { return this->cpu->readFloatReg(this->_srcRegIdx[idx], width); } - FloatReg readFloatReg(const StaticInst *si, int idx) + FloatReg readFloatRegOperand(const StaticInst *si, int idx) { return this->cpu->readFloatReg(this->_srcRegIdx[idx]); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx, int width) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx, + int width) { return this->cpu->readFloatRegBits(this->_srcRegIdx[idx], width); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx) { return this->cpu->readFloatRegBits(this->_srcRegIdx[idx]); } @@ -164,35 +165,37 @@ class MipsDynInst : public BaseDynInst<Impl> /** @todo: Make results into arrays so they can handle multiple dest * registers. */ - void setIntReg(const StaticInst *si, int idx, uint64_t val) + void setIntRegOperand(const StaticInst *si, int idx, uint64_t val) { this->cpu->setIntReg(this->_destRegIdx[idx], val); - BaseDynInst<Impl>::setIntReg(si, idx, val); + BaseDynInst<Impl>::setIntRegOperand(si, idx, val); } - void setFloatReg(const StaticInst *si, int idx, FloatReg val, int width) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val, + int width) { this->cpu->setFloatReg(this->_destRegIdx[idx], val, width); - BaseDynInst<Impl>::setFloatReg(si, idx, val, width); + BaseDynInst<Impl>::setFloatRegOperand(si, idx, val, width); } - void setFloatReg(const StaticInst *si, int idx, FloatReg val) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val) { this->cpu->setFloatReg(this->_destRegIdx[idx], val); - BaseDynInst<Impl>::setFloatReg(si, idx, val); + BaseDynInst<Impl>::setFloatRegOperand(si, idx, val); } - void setFloatRegBits(const StaticInst *si, int idx, - FloatRegBits val, int width) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val, int width) { this->cpu->setFloatRegBits(this->_destRegIdx[idx], val, width); - BaseDynInst<Impl>::setFloatRegBits(si, idx, val); + BaseDynInst<Impl>::setFloatRegOperandBits(si, idx, val); } - void setFloatRegBits(const StaticInst *si, int idx, FloatRegBits val) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val) { this->cpu->setFloatRegBits(this->_destRegIdx[idx], val); - BaseDynInst<Impl>::setFloatRegBits(si, idx, val); + BaseDynInst<Impl>::setFloatRegOperandBits(si, idx, val); } public: diff --git a/src/cpu/ozone/cpu.hh b/src/cpu/ozone/cpu.hh index c1373944d..0da446c9c 100644 --- a/src/cpu/ozone/cpu.hh +++ b/src/cpu/ozone/cpu.hh @@ -448,30 +448,6 @@ class OzoneCPU : public BaseCPU } #endif - /** Old CPU read from memory function. No longer used. */ - template <class T> - Fault read(Request *req, T &data) - { -#if 0 -#if FULL_SYSTEM && defined(TARGET_ALPHA) - if (req->isLocked()) { - req->xc->setMiscReg(TheISA::Lock_Addr_DepTag, req->paddr); - req->xc->setMiscReg(TheISA::Lock_Flag_DepTag, true); - } -#endif - if (req->isLocked()) { - lockAddrList.insert(req->paddr); - lockFlag = true; - } -#endif - Fault error; - - error = this->mem->read(req, data); - data = gtoh(data); - return error; - } - - /** CPU read function, forwards read to LSQ. */ template <class T> Fault read(Request *req, T &data, int load_idx) @@ -479,81 +455,6 @@ class OzoneCPU : public BaseCPU return backEnd->read(req, data, load_idx); } - /** Old CPU write to memory function. No longer used. */ - template <class T> - Fault write(Request *req, T &data) - { -#if 0 -#if FULL_SYSTEM && defined(TARGET_ALPHA) - ExecContext *xc; - - // If this is a store conditional, act appropriately - if (req->isLocked()) { - xc = req->xc; - - if (req->isUncacheable()) { - // Don't update result register (see stq_c in isa_desc) - req->result = 2; - xc->setStCondFailures(0);//Needed? [RGD] - } else { - bool lock_flag = xc->readMiscReg(TheISA::Lock_Flag_DepTag); - Addr lock_addr = xc->readMiscReg(TheISA::Lock_Addr_DepTag); - req->result = lock_flag; - if (!lock_flag || - ((lock_addr & ~0xf) != (req->paddr & ~0xf))) { - xc->setMiscReg(TheISA::Lock_Flag_DepTag, false); - xc->setStCondFailures(xc->readStCondFailures() + 1); - if (((xc->readStCondFailures()) % 100000) == 0) { - std::cerr << "Warning: " - << xc->readStCondFailures() - << " consecutive store conditional failures " - << "on cpu " << req->xc->readCpuId() - << std::endl; - } - return NoFault; - } - else xc->setStCondFailures(0); - } - } - - // Need to clear any locked flags on other proccessors for - // this address. Only do this for succsful Store Conditionals - // and all other stores (WH64?). Unsuccessful Store - // Conditionals would have returned above, and wouldn't fall - // through. - for (int i = 0; i < this->system->threadContexts.size(); i++){ - xc = this->system->threadContexts[i]; - if ((xc->readMiscReg(TheISA::Lock_Addr_DepTag) & ~0xf) == - (req->paddr & ~0xf)) { - xc->setMiscReg(TheISA::Lock_Flag_DepTag, false); - } - } - -#endif - - if (req->isLocked()) { - if (req->isUncacheable()) { - req->result = 2; - } else { - if (this->lockFlag) { - if (lockAddrList.find(req->paddr) != - lockAddrList.end()) { - req->result = 1; - } else { - req->result = 0; - return NoFault; - } - } else { - req->result = 0; - return NoFault; - } - } - } -#endif - - return this->mem->write(req, (T)htog(data)); - } - /** CPU write function, forwards write to LSQ. */ template <class T> Fault write(Request *req, T &data, int store_idx) diff --git a/src/cpu/ozone/dyn_inst.hh b/src/cpu/ozone/dyn_inst.hh index 9445a5309..88f96b14b 100644 --- a/src/cpu/ozone/dyn_inst.hh +++ b/src/cpu/ozone/dyn_inst.hh @@ -146,12 +146,12 @@ class OzoneDynInst : public BaseDynInst<Impl> // storage (which is pretty hard to imagine they would have reason // to do). - uint64_t readIntReg(const StaticInst *si, int idx) + uint64_t readIntRegOperand(const StaticInst *si, int idx) { return srcInsts[idx]->readIntResult(); } - FloatReg readFloatReg(const StaticInst *si, int idx, int width) + FloatReg readFloatRegOperand(const StaticInst *si, int idx, int width) { switch(width) { case 32: @@ -164,17 +164,18 @@ class OzoneDynInst : public BaseDynInst<Impl> } } - FloatReg readFloatReg(const StaticInst *si, int idx) + FloatReg readFloatRegOperand(const StaticInst *si, int idx) { return srcInsts[idx]->readFloatResult(); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx, int width) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx, + int width) { return srcInsts[idx]->readIntResult(); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx) { return srcInsts[idx]->readIntResult(); } @@ -182,28 +183,30 @@ class OzoneDynInst : public BaseDynInst<Impl> /** @todo: Make results into arrays so they can handle multiple dest * registers. */ - void setIntReg(const StaticInst *si, int idx, uint64_t val) + void setIntRegOperand(const StaticInst *si, int idx, uint64_t val) { BaseDynInst<Impl>::setIntReg(si, idx, val); } - void setFloatReg(const StaticInst *si, int idx, FloatReg val, int width) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val, + int width) { BaseDynInst<Impl>::setFloatReg(si, idx, val, width); } - void setFloatReg(const StaticInst *si, int idx, FloatReg val) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val) { BaseDynInst<Impl>::setFloatReg(si, idx, val); } - void setFloatRegBits(const StaticInst *si, int idx, - FloatRegBits val, int width) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val, int width) { BaseDynInst<Impl>::setFloatRegBits(si, idx, val); } - void setFloatRegBits(const StaticInst *si, int idx, FloatRegBits val) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val) { BaseDynInst<Impl>::setFloatRegBits(si, idx, val); } diff --git a/src/cpu/ozone/inorder_back_end.hh b/src/cpu/ozone/inorder_back_end.hh index b2522bdc8..4fd8e02f8 100644 --- a/src/cpu/ozone/inorder_back_end.hh +++ b/src/cpu/ozone/inorder_back_end.hh @@ -236,25 +236,6 @@ InorderBackEnd<Impl>::read(Addr addr, T &data, unsigned flags) */ return fault; } -#if 0 -template <class Impl> -template <class T> -Fault -InorderBackEnd<Impl>::read(MemReqPtr &req, T &data) -{ -#if FULL_SYSTEM && defined(TARGET_ALPHA) - if (req->isLocked()) { - req->xc->setMiscReg(TheISA::Lock_Addr_DepTag, req->paddr); - req->xc->setMiscReg(TheISA::Lock_Flag_DepTag, true); - } -#endif - - Fault error; - error = thread->mem->read(req, data); - data = LittleEndianGuest::gtoh(data); - return error; -} -#endif template <class Impl> template <class T> @@ -296,61 +277,6 @@ InorderBackEnd<Impl>::write(T data, Addr addr, unsigned flags, uint64_t *res) */ return fault; } -#if 0 -template <class Impl> -template <class T> -Fault -InorderBackEnd<Impl>::write(MemReqPtr &req, T &data) -{ -#if FULL_SYSTEM && defined(TARGET_ALPHA) - ExecContext *xc; - - // If this is a store conditional, act appropriately - if (req->isLocked()) { - xc = req->xc; - - if (req->isUncacheable()) { - // Don't update result register (see stq_c in isa_desc) - req->result = 2; - xc->setStCondFailures(0);//Needed? [RGD] - } else { - bool lock_flag = xc->readMiscReg(TheISA::Lock_Flag_DepTag); - Addr lock_addr = xc->readMiscReg(TheISA::Lock_Addr_DepTag); - req->result = lock_flag; - if (!lock_flag || - ((lock_addr & ~0xf) != (req->paddr & ~0xf))) { - xc->setMiscReg(TheISA::Lock_Flag_DepTag, false); - xc->setStCondFailures(xc->readStCondFailures() + 1); - if (((xc->readStCondFailures()) % 100000) == 0) { - std::cerr << "Warning: " - << xc->readStCondFailures() - << " consecutive store conditional failures " - << "on cpu " << req->xc->readCpuId() - << std::endl; - } - return NoFault; - } - else xc->setStCondFailures(0); - } - } - - // Need to clear any locked flags on other proccessors for - // this address. Only do this for succsful Store Conditionals - // and all other stores (WH64?). Unsuccessful Store - // Conditionals would have returned above, and wouldn't fall - // through. - for (int i = 0; i < cpu->system->execContexts.size(); i++){ - xc = cpu->system->execContexts[i]; - if ((xc->readMiscReg(TheISA::Lock_Addr_DepTag) & ~0xf) == - (req->paddr & ~0xf)) { - xc->setMiscReg(TheISA::Lock_Flag_DepTag, false); - } - } - -#endif - return thread->mem->write(req, (T)LittleEndianGuest::htog(data)); -} -#endif template <class Impl> template <class T> diff --git a/src/cpu/simple/base.hh b/src/cpu/simple/base.hh index efb884325..c39bfa9cd 100644 --- a/src/cpu/simple/base.hh +++ b/src/cpu/simple/base.hh @@ -213,60 +213,63 @@ class BaseSimpleCPU : public BaseCPU // storage (which is pretty hard to imagine they would have reason // to do). - uint64_t readIntReg(const StaticInst *si, int idx) + uint64_t readIntRegOperand(const StaticInst *si, int idx) { return thread->readIntReg(si->srcRegIdx(idx)); } - FloatReg readFloatReg(const StaticInst *si, int idx, int width) + FloatReg readFloatRegOperand(const StaticInst *si, int idx, int width) { int reg_idx = si->srcRegIdx(idx) - TheISA::FP_Base_DepTag; return thread->readFloatReg(reg_idx, width); } - FloatReg readFloatReg(const StaticInst *si, int idx) + FloatReg readFloatRegOperand(const StaticInst *si, int idx) { int reg_idx = si->srcRegIdx(idx) - TheISA::FP_Base_DepTag; return thread->readFloatReg(reg_idx); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx, int width) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx, + int width) { int reg_idx = si->srcRegIdx(idx) - TheISA::FP_Base_DepTag; return thread->readFloatRegBits(reg_idx, width); } - FloatRegBits readFloatRegBits(const StaticInst *si, int idx) + FloatRegBits readFloatRegOperandBits(const StaticInst *si, int idx) { int reg_idx = si->srcRegIdx(idx) - TheISA::FP_Base_DepTag; return thread->readFloatRegBits(reg_idx); } - void setIntReg(const StaticInst *si, int idx, uint64_t val) + void setIntRegOperand(const StaticInst *si, int idx, uint64_t val) { thread->setIntReg(si->destRegIdx(idx), val); } - void setFloatReg(const StaticInst *si, int idx, FloatReg val, int width) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val, + int width) { int reg_idx = si->destRegIdx(idx) - TheISA::FP_Base_DepTag; thread->setFloatReg(reg_idx, val, width); } - void setFloatReg(const StaticInst *si, int idx, FloatReg val) + void setFloatRegOperand(const StaticInst *si, int idx, FloatReg val) { int reg_idx = si->destRegIdx(idx) - TheISA::FP_Base_DepTag; thread->setFloatReg(reg_idx, val); } - void setFloatRegBits(const StaticInst *si, int idx, - FloatRegBits val, int width) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val, int width) { int reg_idx = si->destRegIdx(idx) - TheISA::FP_Base_DepTag; thread->setFloatRegBits(reg_idx, val, width); } - void setFloatRegBits(const StaticInst *si, int idx, FloatRegBits val) + void setFloatRegOperandBits(const StaticInst *si, int idx, + FloatRegBits val) { int reg_idx = si->destRegIdx(idx) - TheISA::FP_Base_DepTag; thread->setFloatRegBits(reg_idx, val); diff --git a/src/cpu/simple_thread.hh b/src/cpu/simple_thread.hh index acbefeb67..f2f79c070 100644 --- a/src/cpu/simple_thread.hh +++ b/src/cpu/simple_thread.hh @@ -234,75 +234,6 @@ class SimpleThread : public ThreadState /// Set the status to Halted. void halt(); -/* - template <class T> - Fault read(RequestPtr &req, T &data) - { -#if FULL_SYSTEM && THE_ISA == ALPHA_ISA - if (req->isLocked()) { - req->xc->setMiscReg(TheISA::Lock_Addr_DepTag, req->paddr); - req->xc->setMiscReg(TheISA::Lock_Flag_DepTag, true); - } -#endif - - Fault error; - error = mem->prot_read(req->paddr, data, req->size); - data = LittleEndianGuest::gtoh(data); - return error; - } - - template <class T> - Fault write(RequestPtr &req, T &data) - { -#if FULL_SYSTEM && THE_ISA == ALPHA_ISA - ExecContext *xc; - - // If this is a store conditional, act appropriately - if (req->isLocked()) { - xc = req->xc; - - if (req->isUncacheable()) { - // Don't update result register (see stq_c in isa_desc) - req->result = 2; - xc->setStCondFailures(0);//Needed? [RGD] - } else { - bool lock_flag = xc->readMiscReg(TheISA::Lock_Flag_DepTag); - Addr lock_addr = xc->readMiscReg(TheISA::Lock_Addr_DepTag); - req->result = lock_flag; - if (!lock_flag || - ((lock_addr & ~0xf) != (req->paddr & ~0xf))) { - xc->setMiscReg(TheISA::Lock_Flag_DepTag, false); - xc->setStCondFailures(xc->readStCondFailures() + 1); - if (((xc->readStCondFailures()) % 100000) == 0) { - std::cerr << "Warning: " - << xc->readStCondFailures() - << " consecutive store conditional failures " - << "on cpu " << req->xc->readCpuId() - << std::endl; - } - return NoFault; - } - else xc->setStCondFailures(0); - } - } - - // Need to clear any locked flags on other proccessors for - // this address. Only do this for succsful Store Conditionals - // and all other stores (WH64?). Unsuccessful Store - // Conditionals would have returned above, and wouldn't fall - // through. - for (int i = 0; i < system->execContexts.size(); i++){ - xc = system->execContexts[i]; - if ((xc->readMiscReg(TheISA::Lock_Addr_DepTag) & ~0xf) == - (req->paddr & ~0xf)) { - xc->setMiscReg(TheISA::Lock_Flag_DepTag, false); - } - } - -#endif - return mem->prot_write(req->paddr, (T)htog(data), req->size); - } -*/ virtual bool misspeculating(); Fault instRead(RequestPtr &req) diff --git a/src/mem/cache/cache.cc b/src/mem/cache/cache.cc index db66c096e..29ca09060 100644 --- a/src/mem/cache/cache.cc +++ b/src/mem/cache/cache.cc @@ -37,7 +37,6 @@ */ #include "mem/config/cache.hh" -#include "mem/config/compression.hh" #include "mem/cache/tags/cache_tags.hh" @@ -61,11 +60,6 @@ #include "mem/cache/tags/split_lifo.hh" #endif -#include "base/compression/null_compression.hh" -#if defined(USE_LZSS_COMPRESSION) -#include "base/compression/lzss_compression.hh" -#endif - #include "mem/cache/miss/miss_queue.hh" #include "mem/cache/miss/blocking_buffer.hh" @@ -79,68 +73,28 @@ #if defined(USE_CACHE_FALRU) -template class Cache<CacheTags<FALRU,NullCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<FALRU,NullCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<FALRU,NullCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<FALRU,NullCompression>, MissQueue, UniCoherence>; -#if defined(USE_LZSS_COMPRESSION) -template class Cache<CacheTags<FALRU,LZSSCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<FALRU,LZSSCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<FALRU,LZSSCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<FALRU,LZSSCompression>, MissQueue, UniCoherence>; -#endif +template class Cache<CacheTags<FALRU>, SimpleCoherence>; +template class Cache<CacheTags<FALRU>, UniCoherence>; #endif #if defined(USE_CACHE_IIC) -template class Cache<CacheTags<IIC,NullCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<IIC,NullCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<IIC,NullCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<IIC,NullCompression>, MissQueue, UniCoherence>; -#if defined(USE_LZSS_COMPRESSION) -template class Cache<CacheTags<IIC,LZSSCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<IIC,LZSSCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<IIC,LZSSCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<IIC,LZSSCompression>, MissQueue, UniCoherence>; -#endif +template class Cache<CacheTags<IIC>, SimpleCoherence>; +template class Cache<CacheTags<IIC>, UniCoherence>; #endif #if defined(USE_CACHE_LRU) -template class Cache<CacheTags<LRU,NullCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<LRU,NullCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<LRU,NullCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<LRU,NullCompression>, MissQueue, UniCoherence>; -#if defined(USE_LZSS_COMPRESSION) -template class Cache<CacheTags<LRU,LZSSCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<LRU,LZSSCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<LRU,LZSSCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<LRU,LZSSCompression>, MissQueue, UniCoherence>; -#endif +template class Cache<CacheTags<LRU>, SimpleCoherence>; +template class Cache<CacheTags<LRU>, UniCoherence>; #endif #if defined(USE_CACHE_SPLIT) -template class Cache<CacheTags<Split,NullCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<Split,NullCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<Split,NullCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<Split,NullCompression>, MissQueue, UniCoherence>; -#if defined(USE_LZSS_COMPRESSION) -template class Cache<CacheTags<Split,LZSSCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<Split,LZSSCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<Split,LZSSCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<Split,LZSSCompression>, MissQueue, UniCoherence>; -#endif +template class Cache<CacheTags<Split>, SimpleCoherence>; +template class Cache<CacheTags<Split>, UniCoherence>; #endif #if defined(USE_CACHE_SPLIT_LIFO) -template class Cache<CacheTags<SplitLIFO,NullCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<SplitLIFO,NullCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<SplitLIFO,NullCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<SplitLIFO,NullCompression>, MissQueue, UniCoherence>; -#if defined(USE_LZSS_COMPRESSION) -template class Cache<CacheTags<SplitLIFO,LZSSCompression>, BlockingBuffer, SimpleCoherence>; -template class Cache<CacheTags<SplitLIFO,LZSSCompression>, BlockingBuffer, UniCoherence>; -template class Cache<CacheTags<SplitLIFO,LZSSCompression>, MissQueue, SimpleCoherence>; -template class Cache<CacheTags<SplitLIFO,LZSSCompression>, MissQueue, UniCoherence>; -#endif +template class Cache<CacheTags<SplitLIFO>, SimpleCoherence>; +template class Cache<CacheTags<SplitLIFO>, UniCoherence>; #endif #endif //DOXYGEN_SHOULD_SKIP_THIS diff --git a/src/mem/cache/cache.hh b/src/mem/cache/cache.hh index 1f3b087bb..29502042c 100644 --- a/src/mem/cache/cache.hh +++ b/src/mem/cache/cache.hh @@ -42,6 +42,7 @@ #include "cpu/smt.hh" // SMT_MAX_THREADS #include "mem/cache/base_cache.hh" +#include "mem/cache/miss/miss_buffer.hh" #include "mem/cache/prefetch/prefetcher.hh" //Forward decleration @@ -55,7 +56,7 @@ class MSHR; * @sa MissQueue. Coherence handles all coherence policy details @sa * UniCoherence, SimpleMultiCoherence. */ -template <class TagStore, class Buffering, class Coherence> +template <class TagStore, class Coherence> class Cache : public BaseCache { public: @@ -68,12 +69,12 @@ class Cache : public BaseCache /** Tag and data Storage */ TagStore *tags; /** Miss and Writeback handler */ - Buffering *missQueue; + MissBuffer *missQueue; /** Coherence protocol. */ Coherence *coherence; /** Prefetcher */ - Prefetcher<TagStore, Buffering> *prefetcher; + Prefetcher<TagStore> *prefetcher; /** * The clock ratio of the outgoing bus. @@ -105,16 +106,16 @@ class Cache : public BaseCache { public: TagStore *tags; - Buffering *missQueue; + MissBuffer *missQueue; Coherence *coherence; BaseCache::Params baseParams; - Prefetcher<TagStore, Buffering> *prefetcher; + Prefetcher<TagStore> *prefetcher; bool prefetchAccess; int hitLatency; - Params(TagStore *_tags, Buffering *mq, Coherence *coh, + Params(TagStore *_tags, MissBuffer *mq, Coherence *coh, BaseCache::Params params, - Prefetcher<TagStore, Buffering> *_prefetcher, + Prefetcher<TagStore> *_prefetcher, bool prefetch_access, int hit_latency) : tags(_tags), missQueue(mq), coherence(coh), baseParams(params), diff --git a/src/mem/cache/cache_builder.cc b/src/mem/cache/cache_builder.cc index 03646ec2a..4b1b19718 100644 --- a/src/mem/cache/cache_builder.cc +++ b/src/mem/cache/cache_builder.cc @@ -37,7 +37,6 @@ // Must be included first to determine which caches we want #include "mem/config/cache.hh" -#include "mem/config/compression.hh" #include "mem/config/prefetch.hh" #include "mem/cache/base_cache.hh" @@ -69,9 +68,7 @@ // Compression Templates #include "base/compression/null_compression.hh" -#if defined(USE_LZSS_COMPRESSION) #include "base/compression/lzss_compression.hh" -#endif // CacheTags Templates #include "mem/cache/tags/cache_tags.hh" @@ -211,156 +208,127 @@ BEGIN_INIT_SIM_OBJECT_PARAMS(BaseCache) END_INIT_SIM_OBJECT_PARAMS(BaseCache) -#define BUILD_CACHE(t, comp, b, c) do { \ - Prefetcher<CacheTags<t, comp>, b> *pf; \ +#define BUILD_CACHE(t, c) do { \ + Prefetcher<CacheTags<t> > *pf; \ if (pf_policy == "tagged") { \ - BUILD_TAGGED_PREFETCHER(t, comp, b); \ + BUILD_TAGGED_PREFETCHER(t); \ } \ else if (pf_policy == "stride") { \ - BUILD_STRIDED_PREFETCHER(t, comp, b); \ + BUILD_STRIDED_PREFETCHER(t); \ } \ else if (pf_policy == "ghb") { \ - BUILD_GHB_PREFETCHER(t, comp, b); \ + BUILD_GHB_PREFETCHER(t); \ } \ else { \ - BUILD_NULL_PREFETCHER(t, comp, b); \ + BUILD_NULL_PREFETCHER(t); \ } \ - Cache<CacheTags<t, comp>, b, c>::Params params(tagStore, mq, coh, \ + Cache<CacheTags<t>, c>::Params params(tagStore, mq, coh, \ base_params, \ - /*in_bus, out_bus,*/ pf, \ + pf, \ prefetch_access, hit_latency); \ - Cache<CacheTags<t, comp>, b, c> *retval = \ - new Cache<CacheTags<t, comp>, b, c>(getInstanceName(), /*hier,*/ \ - params); \ -/* if (in_bus == NULL) { \ - retval->setSlaveInterface(new MemoryInterface<Cache<CacheTags<t, comp>, b, c> >(getInstanceName(), hier, retval, mem_trace)); \ - } else { \ - retval->setSlaveInterface(new SlaveInterface<Cache<CacheTags<t, comp>, b, c>, Bus>(getInstanceName(), hier, retval, in_bus, mem_trace)); \ - } \ - retval->setMasterInterface(new MasterInterface<Cache<CacheTags<t, comp>, b, c>, Bus>(getInstanceName(), hier, retval, out_bus)); \ - out_bus->rangeChange(); \ - return retval; \ -*/return retval; \ + Cache<CacheTags<t>, c> *retval = \ + new Cache<CacheTags<t>, c>(getInstanceName(), params); \ +return retval; \ } while (0) #define BUILD_CACHE_PANIC(x) do { \ panic("%s not compiled into M5", x); \ } while (0) -#if defined(USE_LZSS_COMPRESSION) -#define BUILD_COMPRESSED_CACHE(TAGS, tags, b, c) do { \ - if (compressed_bus || store_compressed){ \ - CacheTags<TAGS, LZSSCompression> *tagStore = \ - new CacheTags<TAGS, LZSSCompression>(tags, \ - compression_latency, \ - true, store_compressed, \ - adaptive_compression, \ - prefetch_miss); \ - BUILD_CACHE(TAGS, LZSSCompression, b, c); \ - } else { \ - CacheTags<TAGS, NullCompression> *tagStore = \ - new CacheTags<TAGS, NullCompression>(tags, \ - compression_latency, \ - true, store_compressed, \ - adaptive_compression, \ - prefetch_miss); \ - BUILD_CACHE(TAGS, NullCompression, b, c); \ - } \ - } while (0) -#else -#define BUILD_COMPRESSED_CACHE(TAGS, tags, b, c) do { \ - if (compressed_bus || store_compressed){ \ - BUILD_CACHE_PANIC("compressed caches"); \ +#define BUILD_COMPRESSED_CACHE(TAGS, tags, c) \ + do { \ + CompressionAlgorithm *compAlg; \ + if (compressed_bus || store_compressed) { \ + compAlg = new LZSSCompression(); \ } else { \ - CacheTags<TAGS, NullCompression> *tagStore = \ - new CacheTags<TAGS, NullCompression>(tags, \ - compression_latency, \ - true, store_compressed, \ - adaptive_compression \ - prefetch_miss); \ - BUILD_CACHE(TAGS, NullCompression, b, c); \ + compAlg = new NullCompression(); \ } \ + CacheTags<TAGS> *tagStore = \ + new CacheTags<TAGS>(tags, compression_latency, true, \ + store_compressed, adaptive_compression, \ + compressed_bus, \ + compAlg, prefetch_miss); \ + BUILD_CACHE(TAGS, c); \ } while (0) -#endif #if defined(USE_CACHE_FALRU) -#define BUILD_FALRU_CACHE(b,c) do { \ +#define BUILD_FALRU_CACHE(c) do { \ FALRU *tags = new FALRU(block_size, size, latency); \ - BUILD_COMPRESSED_CACHE(FALRU, tags, b, c); \ + BUILD_COMPRESSED_CACHE(FALRU, tags, c); \ } while (0) #else -#define BUILD_FALRU_CACHE(b, c) BUILD_CACHE_PANIC("falru cache") +#define BUILD_FALRU_CACHE(c) BUILD_CACHE_PANIC("falru cache") #endif #if defined(USE_CACHE_LRU) -#define BUILD_LRU_CACHE(b, c) do { \ +#define BUILD_LRU_CACHE(c) do { \ LRU *tags = new LRU(numSets, block_size, assoc, latency); \ - BUILD_COMPRESSED_CACHE(LRU, tags, b, c); \ + BUILD_COMPRESSED_CACHE(LRU, tags, c); \ } while (0) #else -#define BUILD_LRU_CACHE(b, c) BUILD_CACHE_PANIC("lru cache") +#define BUILD_LRU_CACHE(c) BUILD_CACHE_PANIC("lru cache") #endif #if defined(USE_CACHE_SPLIT) -#define BUILD_SPLIT_CACHE(b, c) do { \ +#define BUILD_SPLIT_CACHE(c) do { \ Split *tags = new Split(numSets, block_size, assoc, split_size, lifo, \ two_queue, latency); \ - BUILD_COMPRESSED_CACHE(Split, tags, b, c); \ + BUILD_COMPRESSED_CACHE(Split, tags, c); \ } while (0) #else -#define BUILD_SPLIT_CACHE(b, c) BUILD_CACHE_PANIC("split cache") +#define BUILD_SPLIT_CACHE(c) BUILD_CACHE_PANIC("split cache") #endif #if defined(USE_CACHE_SPLIT_LIFO) -#define BUILD_SPLIT_LIFO_CACHE(b, c) do { \ +#define BUILD_SPLIT_LIFO_CACHE(c) do { \ SplitLIFO *tags = new SplitLIFO(block_size, size, assoc, \ latency, two_queue, -1); \ - BUILD_COMPRESSED_CACHE(SplitLIFO, tags, b, c); \ + BUILD_COMPRESSED_CACHE(SplitLIFO, tags, c); \ } while (0) #else -#define BUILD_SPLIT_LIFO_CACHE(b, c) BUILD_CACHE_PANIC("lifo cache") +#define BUILD_SPLIT_LIFO_CACHE(c) BUILD_CACHE_PANIC("lifo cache") #endif #if defined(USE_CACHE_IIC) -#define BUILD_IIC_CACHE(b ,c) do { \ +#define BUILD_IIC_CACHE(c) do { \ IIC *tags = new IIC(iic_params); \ - BUILD_COMPRESSED_CACHE(IIC, tags, b, c); \ + BUILD_COMPRESSED_CACHE(IIC, tags, c); \ } while (0) #else -#define BUILD_IIC_CACHE(b, c) BUILD_CACHE_PANIC("iic") +#define BUILD_IIC_CACHE(c) BUILD_CACHE_PANIC("iic") #endif -#define BUILD_CACHES(b, c) do { \ +#define BUILD_CACHES(c) do { \ if (repl == NULL) { \ if (numSets == 1) { \ - BUILD_FALRU_CACHE(b, c); \ + BUILD_FALRU_CACHE(c); \ } else { \ if (split == true) { \ - BUILD_SPLIT_CACHE(b, c); \ + BUILD_SPLIT_CACHE(c); \ } else if (lifo == true) { \ - BUILD_SPLIT_LIFO_CACHE(b, c); \ + BUILD_SPLIT_LIFO_CACHE(c); \ } else { \ - BUILD_LRU_CACHE(b, c); \ + BUILD_LRU_CACHE(c); \ } \ } \ } else { \ - BUILD_IIC_CACHE(b, c); \ + BUILD_IIC_CACHE(c); \ } \ } while (0) #define BUILD_COHERENCE(b) do { \ if (protocol == NULL) { \ UniCoherence *coh = new UniCoherence(); \ - BUILD_CACHES(b, UniCoherence); \ + BUILD_CACHES(UniCoherence); \ } else { \ SimpleCoherence *coh = new SimpleCoherence(protocol); \ - BUILD_CACHES(b, SimpleCoherence); \ + BUILD_CACHES(SimpleCoherence); \ } \ } while (0) #if defined(USE_TAGGED) -#define BUILD_TAGGED_PREFETCHER(t, comp, b) pf = new \ - TaggedPrefetcher<CacheTags<t, comp>, b>(prefetcher_size, \ +#define BUILD_TAGGED_PREFETCHER(t) pf = new \ + TaggedPrefetcher<CacheTags<t> >(prefetcher_size, \ !prefetch_past_page, \ prefetch_serial_squash, \ prefetch_cache_check_push, \ @@ -368,12 +336,12 @@ END_INIT_SIM_OBJECT_PARAMS(BaseCache) prefetch_latency, \ prefetch_degree) #else -#define BUILD_TAGGED_PREFETCHER(t, comp, b) BUILD_CACHE_PANIC("Tagged Prefetcher") +#define BUILD_TAGGED_PREFETCHER(t) BUILD_CACHE_PANIC("Tagged Prefetcher") #endif #if defined(USE_STRIDED) -#define BUILD_STRIDED_PREFETCHER(t, comp, b) pf = new \ - StridePrefetcher<CacheTags<t, comp>, b>(prefetcher_size, \ +#define BUILD_STRIDED_PREFETCHER(t) pf = new \ + StridePrefetcher<CacheTags<t> >(prefetcher_size, \ !prefetch_past_page, \ prefetch_serial_squash, \ prefetch_cache_check_push, \ @@ -382,12 +350,12 @@ END_INIT_SIM_OBJECT_PARAMS(BaseCache) prefetch_degree, \ prefetch_use_cpu_id) #else -#define BUILD_STRIDED_PREFETCHER(t, comp, b) BUILD_CACHE_PANIC("Stride Prefetcher") +#define BUILD_STRIDED_PREFETCHER(t) BUILD_CACHE_PANIC("Stride Prefetcher") #endif #if defined(USE_GHB) -#define BUILD_GHB_PREFETCHER(t, comp, b) pf = new \ - GHBPrefetcher<CacheTags<t, comp>, b>(prefetcher_size, \ +#define BUILD_GHB_PREFETCHER(t) pf = new \ + GHBPrefetcher<CacheTags<t> >(prefetcher_size, \ !prefetch_past_page, \ prefetch_serial_squash, \ prefetch_cache_check_push, \ @@ -396,12 +364,12 @@ END_INIT_SIM_OBJECT_PARAMS(BaseCache) prefetch_degree, \ prefetch_use_cpu_id) #else -#define BUILD_GHB_PREFETCHER(t, comp, b) BUILD_CACHE_PANIC("GHB Prefetcher") +#define BUILD_GHB_PREFETCHER(t) BUILD_CACHE_PANIC("GHB Prefetcher") #endif #if defined(USE_TAGGED) -#define BUILD_NULL_PREFETCHER(t, comp, b) pf = new \ - TaggedPrefetcher<CacheTags<t, comp>, b>(prefetcher_size, \ +#define BUILD_NULL_PREFETCHER(t) pf = new \ + TaggedPrefetcher<CacheTags<t> >(prefetcher_size, \ !prefetch_past_page, \ prefetch_serial_squash, \ prefetch_cache_check_push, \ @@ -409,7 +377,7 @@ END_INIT_SIM_OBJECT_PARAMS(BaseCache) prefetch_latency, \ prefetch_degree) #else -#define BUILD_NULL_PREFETCHER(t, comp, b) BUILD_CACHE_PANIC("NULL Prefetcher (uses Tagged)") +#define BUILD_NULL_PREFETCHER(t) BUILD_CACHE_PANIC("NULL Prefetcher (uses Tagged)") #endif CREATE_SIM_OBJECT(BaseCache) diff --git a/src/mem/cache/cache_impl.hh b/src/mem/cache/cache_impl.hh index 3a681bc52..6742d5892 100644 --- a/src/mem/cache/cache_impl.hh +++ b/src/mem/cache/cache_impl.hh @@ -55,9 +55,9 @@ bool SIGNAL_NACK_HACK; -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> bool -Cache<TagStore,Buffering,Coherence>:: +Cache<TagStore,Coherence>:: doTimingAccess(PacketPtr pkt, CachePort *cachePort, bool isCpuSide) { if (isCpuSide) @@ -81,9 +81,9 @@ doTimingAccess(PacketPtr pkt, CachePort *cachePort, bool isCpuSide) return true; } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> Tick -Cache<TagStore,Buffering,Coherence>:: +Cache<TagStore,Coherence>:: doAtomicAccess(PacketPtr pkt, bool isCpuSide) { if (isCpuSide) @@ -103,9 +103,9 @@ doAtomicAccess(PacketPtr pkt, bool isCpuSide) return hitLatency; } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> void -Cache<TagStore,Buffering,Coherence>:: +Cache<TagStore,Coherence>:: doFunctionalAccess(PacketPtr pkt, bool isCpuSide) { if (isCpuSide) @@ -123,19 +123,19 @@ doFunctionalAccess(PacketPtr pkt, bool isCpuSide) } } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> void -Cache<TagStore,Buffering,Coherence>:: +Cache<TagStore,Coherence>:: recvStatusChange(Port::Status status, bool isCpuSide) { } -template<class TagStore, class Buffering, class Coherence> -Cache<TagStore,Buffering,Coherence>:: +template<class TagStore, class Coherence> +Cache<TagStore,Coherence>:: Cache(const std::string &_name, - Cache<TagStore,Buffering,Coherence>::Params ¶ms) + Cache<TagStore,Coherence>::Params ¶ms) : BaseCache(_name, params.baseParams), prefetchAccess(params.prefetchAccess), tags(params.tags), missQueue(params.missQueue), @@ -154,9 +154,9 @@ Cache(const std::string &_name, invalidatePkt = new Packet(invalidateReq, Packet::InvalidateReq, 0); } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> void -Cache<TagStore,Buffering,Coherence>::regStats() +Cache<TagStore,Coherence>::regStats() { BaseCache::regStats(); tags->regStats(name()); @@ -165,9 +165,9 @@ Cache<TagStore,Buffering,Coherence>::regStats() prefetcher->regStats(name()); } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> bool -Cache<TagStore,Buffering,Coherence>::access(PacketPtr &pkt) +Cache<TagStore,Coherence>::access(PacketPtr &pkt) { //@todo Add back in MemDebug Calls // MemDebug::cacheAccess(pkt); @@ -253,9 +253,9 @@ Cache<TagStore,Buffering,Coherence>::access(PacketPtr &pkt) } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> PacketPtr -Cache<TagStore,Buffering,Coherence>::getPacket() +Cache<TagStore,Coherence>::getPacket() { assert(missQueue->havePending()); PacketPtr pkt = missQueue->getPacket(); @@ -276,9 +276,9 @@ Cache<TagStore,Buffering,Coherence>::getPacket() return pkt; } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> void -Cache<TagStore,Buffering,Coherence>::sendResult(PacketPtr &pkt, MSHR* mshr, +Cache<TagStore,Coherence>::sendResult(PacketPtr &pkt, MSHR* mshr, bool success) { if (success && !(SIGNAL_NACK_HACK)) { @@ -319,9 +319,9 @@ Cache<TagStore,Buffering,Coherence>::sendResult(PacketPtr &pkt, MSHR* mshr, } } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> void -Cache<TagStore,Buffering,Coherence>::handleResponse(PacketPtr &pkt) +Cache<TagStore,Coherence>::handleResponse(PacketPtr &pkt) { BlkType *blk = NULL; if (pkt->senderState) { @@ -363,16 +363,16 @@ Cache<TagStore,Buffering,Coherence>::handleResponse(PacketPtr &pkt) } } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> PacketPtr -Cache<TagStore,Buffering,Coherence>::getCoherencePacket() +Cache<TagStore,Coherence>::getCoherencePacket() { return coherence->getPacket(); } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> void -Cache<TagStore,Buffering,Coherence>::sendCoherenceResult(PacketPtr &pkt, +Cache<TagStore,Coherence>::sendCoherenceResult(PacketPtr &pkt, MSHR *cshr, bool success) { @@ -380,9 +380,9 @@ Cache<TagStore,Buffering,Coherence>::sendCoherenceResult(PacketPtr &pkt, } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> void -Cache<TagStore,Buffering,Coherence>::snoop(PacketPtr &pkt) +Cache<TagStore,Coherence>::snoop(PacketPtr &pkt) { if (pkt->req->isUncacheable()) { //Can't get a hit on an uncacheable address @@ -514,9 +514,9 @@ Cache<TagStore,Buffering,Coherence>::snoop(PacketPtr &pkt) tags->handleSnoop(blk, new_state); } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> void -Cache<TagStore,Buffering,Coherence>::snoopResponse(PacketPtr &pkt) +Cache<TagStore,Coherence>::snoopResponse(PacketPtr &pkt) { //Need to handle the response, if NACKED if (pkt->flags & NACKED_LINE) { @@ -533,9 +533,9 @@ Cache<TagStore,Buffering,Coherence>::snoopResponse(PacketPtr &pkt) } } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> void -Cache<TagStore,Buffering,Coherence>::invalidateBlk(Addr addr) +Cache<TagStore,Coherence>::invalidateBlk(Addr addr) { tags->invalidateBlk(addr); } @@ -544,9 +544,9 @@ Cache<TagStore,Buffering,Coherence>::invalidateBlk(Addr addr) /** * @todo Fix to not assume write allocate */ -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> Tick -Cache<TagStore,Buffering,Coherence>::probe(PacketPtr &pkt, bool update, +Cache<TagStore,Coherence>::probe(PacketPtr &pkt, bool update, CachePort* otherSidePort) { // MemDebug::cacheProbe(pkt); @@ -694,9 +694,9 @@ return 0; return 0; } -template<class TagStore, class Buffering, class Coherence> +template<class TagStore, class Coherence> Tick -Cache<TagStore,Buffering,Coherence>::snoopProbe(PacketPtr &pkt) +Cache<TagStore,Coherence>::snoopProbe(PacketPtr &pkt) { //Send a atomic (false) invalidate up if the protocol calls for it if (coherence->propogateInvalidate(pkt, false)) { diff --git a/src/mem/cache/miss/blocking_buffer.cc b/src/mem/cache/miss/blocking_buffer.cc index bf741e547..4a431d82d 100644 --- a/src/mem/cache/miss/blocking_buffer.cc +++ b/src/mem/cache/miss/blocking_buffer.cc @@ -33,11 +33,9 @@ * Definitions of a simple buffer for a blocking cache. */ -#include "cpu/smt.hh" //for maxThreadsPerCPU #include "mem/cache/base_cache.hh" #include "mem/cache/miss/blocking_buffer.hh" #include "mem/cache/prefetch/base_prefetcher.hh" -#include "sim/eventq.hh" // for Event declaration. #include "mem/request.hh" /** @@ -46,28 +44,11 @@ void BlockingBuffer::regStats(const std::string &name) { - using namespace Stats; - writebacks - .init(maxThreadsPerCPU) - .name(name + ".writebacks") - .desc("number of writebacks") - .flags(total) - ; + MissBuffer::regStats(name); } -void -BlockingBuffer::setCache(BaseCache *_cache) -{ - cache = _cache; - blkSize = cache->getBlockSize(); -} void -BlockingBuffer::setPrefetcher(BasePrefetcher *_prefetcher) -{ - prefetcher = _prefetcher; -} -void BlockingBuffer::handleMiss(PacketPtr &pkt, int blk_size, Tick time) { Addr blk_addr = pkt->getAddr() & ~(Addr)(blk_size - 1); @@ -241,3 +222,23 @@ BlockingBuffer::doWriteback(PacketPtr &pkt) cache->setBlocked(Blocked_NoWBBuffers); cache->setMasterRequest(Request_WB, curTick); } + + +MSHR * +BlockingBuffer::findMSHR(Addr addr) +{ + if (miss.addr == addr && miss.pkt) + return &miss; + return NULL; +} + + +bool +BlockingBuffer::findWrites(Addr addr, std::vector<MSHR*>& writes) +{ + if (wb.addr == addr && wb.pkt) { + writes.push_back(&wb); + return true; + } + return false; +} diff --git a/src/mem/cache/miss/blocking_buffer.hh b/src/mem/cache/miss/blocking_buffer.hh index 934a843a6..205068a8c 100644 --- a/src/mem/cache/miss/blocking_buffer.hh +++ b/src/mem/cache/miss/blocking_buffer.hh @@ -38,16 +38,14 @@ #include <vector> +#include "base/misc.hh" // for fatal() +#include "mem/cache/miss/miss_buffer.hh" #include "mem/cache/miss/mshr.hh" -#include "base/statistics.hh" - -class BaseCache; -class BasePrefetcher; /** * Miss and writeback storage for a blocking cache. */ -class BlockingBuffer +class BlockingBuffer : public MissBuffer { protected: /** Miss storage. */ @@ -55,38 +53,13 @@ protected: /** WB storage. */ MSHR wb; - //Params - - /** Allocate on write misses. */ - const bool writeAllocate; - - /** Pointer to the parent cache. */ - BaseCache* cache; - - BasePrefetcher* prefetcher; - - /** Block size of the parent cache. */ - int blkSize; - - // Statistics - /** - * @addtogroup CacheStatistics - * @{ - */ - /** Number of blocks written back per thread. */ - Stats::Vector<> writebacks; - - /** - * @} - */ - public: /** * Builds and initializes this buffer. * @param write_allocate If true, treat write misses the same as reads. */ BlockingBuffer(bool write_allocate) - : writeAllocate(write_allocate) + : MissBuffer(write_allocate) { } @@ -97,14 +70,6 @@ public: void regStats(const std::string &name); /** - * Called by the parent cache to set the back pointer. - * @param _cache A pointer to the parent cache. - */ - void setCache(BaseCache *_cache); - - void setPrefetcher(BasePrefetcher *_prefetcher); - - /** * Handle a cache miss properly. Requests the bus and marks the cache as * blocked. * @param pkt The request that missed in the cache. @@ -183,12 +148,7 @@ public: * @param asid The address space id. * @return A pointer to miss if it matches. */ - MSHR* findMSHR(Addr addr) - { - if (miss.addr == addr && miss.pkt) - return &miss; - return NULL; - } + MSHR* findMSHR(Addr addr); /** * Searches for the supplied address in the write buffer. @@ -197,16 +157,7 @@ public: * @param writes List of pointers to the matching writes. * @return True if there is a matching write. */ - bool findWrites(Addr addr, std::vector<MSHR*>& writes) - { - if (wb.addr == addr && wb.pkt) { - writes.push_back(&wb); - return true; - } - return false; - } - - + bool findWrites(Addr addr, std::vector<MSHR*>& writes); /** * Perform a writeback of dirty data to the given address. diff --git a/src/mem/cache/miss/miss_buffer.cc b/src/mem/cache/miss/miss_buffer.cc new file mode 100644 index 000000000..4d9cd0958 --- /dev/null +++ b/src/mem/cache/miss/miss_buffer.cc @@ -0,0 +1,62 @@ +/* + * Copyright (c) 2003-2006 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: Erik Hallnor + */ + +#include "cpu/smt.hh" //for maxThreadsPerCPU +#include "mem/cache/base_cache.hh" +#include "mem/cache/miss/miss_buffer.hh" +#include "mem/cache/prefetch/base_prefetcher.hh" + +/** + * @todo Move writebacks into shared BaseBuffer class. + */ +void +MissBuffer::regStats(const std::string &name) +{ + using namespace Stats; + writebacks + .init(maxThreadsPerCPU) + .name(name + ".writebacks") + .desc("number of writebacks") + .flags(total) + ; +} + +void +MissBuffer::setCache(BaseCache *_cache) +{ + cache = _cache; + blkSize = cache->getBlockSize(); +} + +void +MissBuffer::setPrefetcher(BasePrefetcher *_prefetcher) +{ + prefetcher = _prefetcher; +} diff --git a/src/mem/cache/miss/miss_buffer.hh b/src/mem/cache/miss/miss_buffer.hh new file mode 100644 index 000000000..3e3080578 --- /dev/null +++ b/src/mem/cache/miss/miss_buffer.hh @@ -0,0 +1,223 @@ +/* + * Copyright (c) 2003-2006 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: Steve Reinhardt + */ + +/** + * @file + * MissBuffer declaration. + */ + +#ifndef __MISS_BUFFER_HH__ +#define __MISS_BUFFER_HH__ + +class BaseCache; +class BasePrefetcher; +class MSHR; + +/** + * Abstract base class for cache miss buffering. + */ +class MissBuffer +{ + protected: + /** True if the cache should allocate on a write miss. */ + const bool writeAllocate; + + /** Pointer to the parent cache. */ + BaseCache *cache; + + /** The Prefetcher */ + BasePrefetcher *prefetcher; + + /** Block size of the parent cache. */ + int blkSize; + + // Statistics + /** + * @addtogroup CacheStatistics + * @{ + */ + /** Number of blocks written back per thread. */ + Stats::Vector<> writebacks; + + /** + * @} + */ + + public: + MissBuffer(bool write_allocate) + : writeAllocate(write_allocate) + { + } + + virtual ~MissBuffer() {} + + /** + * Called by the parent cache to set the back pointer. + * @param _cache A pointer to the parent cache. + */ + void setCache(BaseCache *_cache); + + void setPrefetcher(BasePrefetcher *_prefetcher); + + /** + * Register statistics for this object. + * @param name The name of the parent cache. + */ + virtual void regStats(const std::string &name); + + /** + * Handle a cache miss properly. Either allocate an MSHR for the request, + * or forward it through the write buffer. + * @param pkt The request that missed in the cache. + * @param blk_size The block size of the cache. + * @param time The time the miss is detected. + */ + virtual void handleMiss(PacketPtr &pkt, int blk_size, Tick time) = 0; + + /** + * Fetch the block for the given address and buffer the given target. + * @param addr The address to fetch. + * @param asid The address space of the address. + * @param blk_size The block size of the cache. + * @param time The time the miss is detected. + * @param target The target for the fetch. + */ + virtual MSHR *fetchBlock(Addr addr, int blk_size, Tick time, + PacketPtr &target) = 0; + + /** + * Selects a outstanding request to service. + * @return The request to service, NULL if none found. + */ + virtual PacketPtr getPacket() = 0; + + /** + * Set the command to the given bus command. + * @param pkt The request to update. + * @param cmd The bus command to use. + */ + virtual void setBusCmd(PacketPtr &pkt, Packet::Command cmd) = 0; + + /** + * Restore the original command in case of a bus transmission error. + * @param pkt The request to reset. + */ + virtual void restoreOrigCmd(PacketPtr &pkt) = 0; + + /** + * Marks a request as in service (sent on the bus). This can have side + * effect since storage for no response commands is deallocated once they + * are successfully sent. + * @param pkt The request that was sent on the bus. + */ + virtual void markInService(PacketPtr &pkt, MSHR* mshr) = 0; + + /** + * Collect statistics and free resources of a satisfied request. + * @param pkt The request that has been satisfied. + * @param time The time when the request is satisfied. + */ + virtual void handleResponse(PacketPtr &pkt, Tick time) = 0; + + /** + * Removes all outstanding requests for a given thread number. If a request + * has been sent to the bus, this function removes all of its targets. + * @param threadNum The thread number of the requests to squash. + */ + virtual void squash(int threadNum) = 0; + + /** + * Return the current number of outstanding misses. + * @return the number of outstanding misses. + */ + virtual int getMisses() = 0; + + /** + * Searches for the supplied address in the miss queue. + * @param addr The address to look for. + * @param asid The address space id. + * @return The MSHR that contains the address, NULL if not found. + * @warning Currently only searches the miss queue. If non write allocate + * might need to search the write buffer for coherence. + */ + virtual MSHR* findMSHR(Addr addr) = 0; + + /** + * Searches for the supplied address in the write buffer. + * @param addr The address to look for. + * @param asid The address space id. + * @param writes The list of writes that match the address. + * @return True if any writes are found + */ + virtual bool findWrites(Addr addr, std::vector<MSHR*>& writes) = 0; + + /** + * Perform a writeback of dirty data to the given address. + * @param addr The address to write to. + * @param asid The address space id. + * @param xc The execution context of the address space. + * @param size The number of bytes to write. + * @param data The data to write, can be NULL. + * @param compressed True if the data is compressed. + */ + virtual void doWriteback(Addr addr, int size, uint8_t *data, + bool compressed) = 0; + + /** + * Perform the given writeback request. + * @param pkt The writeback request. + */ + virtual void doWriteback(PacketPtr &pkt) = 0; + + /** + * Returns true if there are outstanding requests. + * @return True if there are outstanding requests. + */ + virtual bool havePending() = 0; + + /** + * Add a target to the given MSHR. This assumes it is in the miss queue. + * @param mshr The mshr to add a target to. + * @param pkt The target to add. + */ + virtual void addTarget(MSHR *mshr, PacketPtr &pkt) = 0; + + /** + * Allocate a MSHR to hold a list of targets to a block involved in a copy. + * If the block is marked done then the MSHR already holds the data to + * fill the block. Otherwise the block needs to be fetched. + * @param addr The address to buffer. + * @param asid The address space ID. + * @return A pointer to the allocated MSHR. + */ + virtual MSHR* allocateTargetList(Addr addr) = 0; +}; + +#endif //__MISS_BUFFER_HH__ diff --git a/src/mem/cache/miss/miss_queue.cc b/src/mem/cache/miss/miss_queue.cc index 3c4586272..1d3e22326 100644 --- a/src/mem/cache/miss/miss_queue.cc +++ b/src/mem/cache/miss/miss_queue.cc @@ -48,16 +48,25 @@ using namespace std; */ MissQueue::MissQueue(int numMSHRs, int numTargets, int write_buffers, bool write_allocate, bool prefetch_miss) - : mq(numMSHRs, 4), wb(write_buffers,numMSHRs+1000), numMSHR(numMSHRs), + : MissBuffer(write_allocate), + mq(numMSHRs, 4), wb(write_buffers,numMSHRs+1000), numMSHR(numMSHRs), numTarget(numTargets), writeBuffers(write_buffers), - writeAllocate(write_allocate), order(0), prefetchMiss(prefetch_miss) + order(0), prefetchMiss(prefetch_miss) { noTargetMSHR = NULL; } + +MissQueue::~MissQueue() +{ +} + + void MissQueue::regStats(const string &name) { + MissBuffer::regStats(name); + Request temp_req((Addr) NULL, 4, 0); Packet::Command temp_cmd = Packet::ReadReq; Packet temp_pkt(&temp_req, temp_cmd, 0); //@todo FIx command strings so this isn't neccessary @@ -65,13 +74,6 @@ MissQueue::regStats(const string &name) using namespace Stats; - writebacks - .init(maxThreadsPerCPU) - .name(name + ".writebacks") - .desc("number of writebacks") - .flags(total) - ; - // MSHR hit statistics for (int access_idx = 0; access_idx < NUM_MEM_CMDS; ++access_idx) { Packet::Command cmd = (Packet::Command)access_idx; @@ -336,18 +338,6 @@ MissQueue::regStats(const string &name) } -void -MissQueue::setCache(BaseCache *_cache) -{ - cache = _cache; - blkSize = cache->getBlockSize(); -} - -void -MissQueue::setPrefetcher(BasePrefetcher *_prefetcher) -{ - prefetcher = _prefetcher; -} MSHR* MissQueue::allocateMiss(PacketPtr &pkt, int size, Tick time) @@ -706,13 +696,13 @@ MissQueue::squash(int threadNum) } MSHR* -MissQueue::findMSHR(Addr addr) const +MissQueue::findMSHR(Addr addr) { return mq.findMatch(addr); } bool -MissQueue::findWrites(Addr addr, vector<MSHR*> &writes) const +MissQueue::findWrites(Addr addr, vector<MSHR*> &writes) { return wb.findMatches(addr,writes); } diff --git a/src/mem/cache/miss/miss_queue.hh b/src/mem/cache/miss/miss_queue.hh index b67a896f4..1f9bb1e0c 100644 --- a/src/mem/cache/miss/miss_queue.hh +++ b/src/mem/cache/miss/miss_queue.hh @@ -38,19 +38,18 @@ #include <vector> +#include "mem/cache/miss/miss_buffer.hh" #include "mem/cache/miss/mshr.hh" #include "mem/cache/miss/mshr_queue.hh" #include "base/statistics.hh" -class BaseCache; -class BasePrefetcher; /** * Manages cache misses and writebacks. Contains MSHRs to store miss data * and the writebuffer for writes/writebacks. * @todo need to handle data on writes better (encapsulate). * @todo need to make replacements/writebacks happen in Cache::access */ -class MissQueue +class MissQueue : public MissBuffer { protected: /** The MSHRs. */ @@ -66,16 +65,6 @@ class MissQueue const int numTarget; /** The number of write buffers. */ const int writeBuffers; - /** True if the cache should allocate on a write miss. */ - const bool writeAllocate; - /** Pointer to the parent cache. */ - BaseCache* cache; - - /** The Prefetcher */ - BasePrefetcher *prefetcher; - - /** The block size of the parent cache. */ - int blkSize; /** Increasing order number assigned to each incoming request. */ uint64_t order; @@ -87,9 +76,6 @@ class MissQueue * @addtogroup CacheStatistics * @{ */ - /** Number of blocks written back per thread. */ - Stats::Vector<> writebacks; - /** Number of misses that hit in the MSHRs per command and thread. */ Stats::Vector<> mshr_hits[NUM_MEM_CMDS]; /** Demand misses that hit in the MSHRs. */ @@ -204,14 +190,6 @@ class MissQueue void regStats(const std::string &name); /** - * Called by the parent cache to set the back pointer. - * @param _cache A pointer to the parent cache. - */ - void setCache(BaseCache *_cache); - - void setPrefetcher(BasePrefetcher *_prefetcher); - - /** * Handle a cache miss properly. Either allocate an MSHR for the request, * or forward it through the write buffer. * @param pkt The request that missed in the cache. @@ -289,7 +267,7 @@ class MissQueue * @warning Currently only searches the miss queue. If non write allocate * might need to search the write buffer for coherence. */ - MSHR* findMSHR(Addr addr) const; + MSHR* findMSHR(Addr addr); /** * Searches for the supplied address in the write buffer. @@ -298,7 +276,7 @@ class MissQueue * @param writes The list of writes that match the address. * @return True if any writes are found */ - bool findWrites(Addr addr, std::vector<MSHR*>& writes) const; + bool findWrites(Addr addr, std::vector<MSHR*>& writes); /** * Perform a writeback of dirty data to the given address. diff --git a/src/mem/cache/prefetch/ghb_prefetcher.cc b/src/mem/cache/prefetch/ghb_prefetcher.cc index 247ec6e8b..a6c419113 100644 --- a/src/mem/cache/prefetch/ghb_prefetcher.cc +++ b/src/mem/cache/prefetch/ghb_prefetcher.cc @@ -38,17 +38,11 @@ #include "mem/cache/tags/lru.hh" -#include "base/compression/null_compression.hh" - -#include "mem/cache/miss/miss_queue.hh" -#include "mem/cache/miss/blocking_buffer.hh" - #include "mem/cache/prefetch/ghb_prefetcher.hh" // Template Instantiations #ifndef DOXYGEN_SHOULD_SKIP_THIS -template class GHBPrefetcher<CacheTags<LRU,NullCompression>, MissQueue>; -template class GHBPrefetcher<CacheTags<LRU,NullCompression>, BlockingBuffer>; +template class GHBPrefetcher<CacheTags<LRU> >; #endif //DOXYGEN_SHOULD_SKIP_THIS diff --git a/src/mem/cache/prefetch/ghb_prefetcher.hh b/src/mem/cache/prefetch/ghb_prefetcher.hh index 14f5747df..c558a3e64 100644 --- a/src/mem/cache/prefetch/ghb_prefetcher.hh +++ b/src/mem/cache/prefetch/ghb_prefetcher.hh @@ -43,16 +43,16 @@ /** * A template-policy based cache. The behavior of the cache can be altered by * supplying different template policies. TagStore handles all tag and data - * storage @sa TagStore. Buffering handles all misses and writes/writebacks + * storage @sa TagStore. MissBuffer handles all misses and writes/writebacks * @sa MissQueue. Coherence handles all coherence policy details @sa * UniCoherence, SimpleMultiCoherence. */ -template <class TagStore, class Buffering> -class GHBPrefetcher : public Prefetcher<TagStore, Buffering> +template <class TagStore> +class GHBPrefetcher : public Prefetcher<TagStore> { protected: - Buffering* mq; + MissBuffer* mq; TagStore* tags; Addr second_last_miss_addr[64/*MAX_CPUS*/]; @@ -67,7 +67,7 @@ class GHBPrefetcher : public Prefetcher<TagStore, Buffering> GHBPrefetcher(int size, bool pageStop, bool serialSquash, bool cacheCheckPush, bool onlyData, Tick latency, int degree, bool useCPUId) - :Prefetcher<TagStore, Buffering>(size, pageStop, serialSquash, + :Prefetcher<TagStore>(size, pageStop, serialSquash, cacheCheckPush, onlyData), latency(latency), degree(degree), useCPUId(useCPUId) { diff --git a/src/mem/cache/prefetch/stride_prefetcher.cc b/src/mem/cache/prefetch/stride_prefetcher.cc index 93a096468..2204871cc 100644 --- a/src/mem/cache/prefetch/stride_prefetcher.cc +++ b/src/mem/cache/prefetch/stride_prefetcher.cc @@ -38,17 +38,11 @@ #include "mem/cache/tags/lru.hh" -#include "base/compression/null_compression.hh" - -#include "mem/cache/miss/miss_queue.hh" -#include "mem/cache/miss/blocking_buffer.hh" - #include "mem/cache/prefetch/stride_prefetcher.hh" // Template Instantiations #ifndef DOXYGEN_SHOULD_SKIP_THIS -template class StridePrefetcher<CacheTags<LRU,NullCompression>, MissQueue>; -template class StridePrefetcher<CacheTags<LRU,NullCompression>, BlockingBuffer>; +template class StridePrefetcher<CacheTags<LRU> >; #endif //DOXYGEN_SHOULD_SKIP_THIS diff --git a/src/mem/cache/prefetch/stride_prefetcher.hh b/src/mem/cache/prefetch/stride_prefetcher.hh index d6fb8ab66..57e430400 100644 --- a/src/mem/cache/prefetch/stride_prefetcher.hh +++ b/src/mem/cache/prefetch/stride_prefetcher.hh @@ -43,16 +43,16 @@ /** * A template-policy based cache. The behavior of the cache can be altered by * supplying different template policies. TagStore handles all tag and data - * storage @sa TagStore. Buffering handles all misses and writes/writebacks + * storage @sa TagStore. MissBuffer handles all misses and writes/writebacks * @sa MissQueue. Coherence handles all coherence policy details @sa * UniCoherence, SimpleMultiCoherence. */ -template <class TagStore, class Buffering> -class StridePrefetcher : public Prefetcher<TagStore, Buffering> +template <class TagStore> +class StridePrefetcher : public Prefetcher<TagStore> { protected: - Buffering* mq; + MissBuffer* mq; TagStore* tags; class strideEntry @@ -84,7 +84,7 @@ class StridePrefetcher : public Prefetcher<TagStore, Buffering> StridePrefetcher(int size, bool pageStop, bool serialSquash, bool cacheCheckPush, bool onlyData, Tick latency, int degree, bool useCPUId) - :Prefetcher<TagStore, Buffering>(size, pageStop, serialSquash, + :Prefetcher<TagStore>(size, pageStop, serialSquash, cacheCheckPush, onlyData), latency(latency), degree(degree), useCPUId(useCPUId) { diff --git a/src/mem/cache/prefetch/tagged_prefetcher.hh b/src/mem/cache/prefetch/tagged_prefetcher.hh index b61e57dcc..dc2aaec50 100644 --- a/src/mem/cache/prefetch/tagged_prefetcher.hh +++ b/src/mem/cache/prefetch/tagged_prefetcher.hh @@ -41,16 +41,16 @@ /** * A template-policy based cache. The behavior of the cache can be altered by * supplying different template policies. TagStore handles all tag and data - * storage @sa TagStore. Buffering handles all misses and writes/writebacks + * storage @sa TagStore. MissBuffer handles all misses and writes/writebacks * @sa MissQueue. Coherence handles all coherence policy details @sa * UniCoherence, SimpleMultiCoherence. */ -template <class TagStore, class Buffering> -class TaggedPrefetcher : public Prefetcher<TagStore, Buffering> +template <class TagStore> +class TaggedPrefetcher : public Prefetcher<TagStore> { protected: - Buffering* mq; + MissBuffer* mq; TagStore* tags; Tick latency; diff --git a/src/mem/cache/prefetch/tagged_prefetcher_impl.hh b/src/mem/cache/prefetch/tagged_prefetcher_impl.hh index a18de4571..b3d4284c7 100644 --- a/src/mem/cache/prefetch/tagged_prefetcher_impl.hh +++ b/src/mem/cache/prefetch/tagged_prefetcher_impl.hh @@ -36,20 +36,20 @@ #include "arch/isa_traits.hh" #include "mem/cache/prefetch/tagged_prefetcher.hh" -template <class TagStore, class Buffering> -TaggedPrefetcher<TagStore, Buffering>:: +template <class TagStore> +TaggedPrefetcher<TagStore>:: TaggedPrefetcher(int size, bool pageStop, bool serialSquash, bool cacheCheckPush, bool onlyData, Tick latency, int degree) - :Prefetcher<TagStore, Buffering>(size, pageStop, serialSquash, + :Prefetcher<TagStore>(size, pageStop, serialSquash, cacheCheckPush, onlyData), latency(latency), degree(degree) { } -template <class TagStore, class Buffering> +template <class TagStore> void -TaggedPrefetcher<TagStore, Buffering>:: +TaggedPrefetcher<TagStore>:: calculatePrefetch(PacketPtr &pkt, std::list<Addr> &addresses, std::list<Tick> &delays) { diff --git a/src/python/m5/objects/Repl.py b/src/python/m5/objects/Repl.py index 10892cf6f..b76aa1d6e 100644 --- a/src/python/m5/objects/Repl.py +++ b/src/python/m5/objects/Repl.py @@ -6,6 +6,6 @@ class Repl(SimObject): class GenRepl(Repl): type = 'GenRepl' - fresh_res = Param.Int("associativity") - num_pools = Param.Int("capacity in bytes") - pool_res = Param.Int("block size in bytes") + fresh_res = Param.Int("Fresh pool residency time") + num_pools = Param.Int("Number of priority pools") + pool_res = Param.Int("Pool residency time") diff --git a/src/python/m5/params.py b/src/python/m5/params.py index 9e5f985c3..d83d5f73f 100644 --- a/src/python/m5/params.py +++ b/src/python/m5/params.py @@ -237,6 +237,12 @@ class NumericParamValue(ParamValue): def __float__(self): return float(self.value) + def __long__(self): + return long(self.value) + + def __int__(self): + return int(self.value) + # hook for bounds checking def _check(self): return @@ -308,8 +314,11 @@ class CheckedInt(NumericParamValue): def __init__(self, value): if isinstance(value, str): self.value = convert.toInteger(value) - elif isinstance(value, (int, long, float)): + elif isinstance(value, (int, long, float, NumericParamValue)): self.value = long(value) + else: + raise TypeError, "Can't convert object of type %s to CheckedInt" \ + % type(value).__name__ self._check() class Int(CheckedInt): cxx_type = 'int'; size = 32; unsigned = False diff --git a/tests/long/20.parser/ref/alpha/linux/NOTE b/tests/long/20.parser/ref/alpha/linux/NOTE new file mode 100644 index 000000000..5e7d8c358 --- /dev/null +++ b/tests/long/20.parser/ref/alpha/linux/NOTE @@ -0,0 +1,6 @@ +I removed the reference outputs for this program because it's taking +way too long... over an hour for simple-atomic and over 19 hrs for +o3-timing. We need to find a shorter input if we want to keep this +in the regressions. + +Steve diff --git a/tests/long/30.eon/ref/alpha/linux/o3-timing/config.ini b/tests/long/30.eon/ref/alpha/linux/o3-timing/config.ini index c3a59fbce..e69de29bb 100644 --- a/tests/long/30.eon/ref/alpha/linux/o3-timing/config.ini +++ b/tests/long/30.eon/ref/alpha/linux/o3-timing/config.ini @@ -1,417 +0,0 @@ -[root] -type=Root -children=system -checkpoint= -clock=1000000000000 -max_tick=0 -output_file=cout -progress_interval=0 - -[debug] -break_cycles= - -[exetrace] -intel_format=false -pc_symbol=true -print_cpseq=false -print_cycle=true -print_data=true -print_effaddr=true -print_fetchseq=false -print_iregs=false -print_opclass=true -print_thread=true -speculative=true -trace_system=client - -[serialize] -count=10 -cycle=0 -dir=cpt.%012d -period=0 - -[stats] -descriptions=true -dump_cycle=0 -dump_period=0 -dump_reset=false -ignore_events= -mysql_db= -mysql_host= -mysql_password= -mysql_user= -project_name=test -simulation_name=test -simulation_sample=0 -text_compat=true -text_file=m5stats.txt - -[system] -type=System -children=cpu membus physmem -mem_mode=atomic -physmem=system.physmem - -[system.cpu] -type=DerivO3CPU -children=dcache fuPool icache l2cache toL2Bus workload -BTBEntries=4096 -BTBTagSize=16 -LFSTSize=1024 -LQEntries=32 -RASSize=16 -SQEntries=32 -SSITSize=1024 -activity=0 -backComSize=5 -choiceCtrBits=2 -choicePredictorSize=8192 -clock=1 -commitToDecodeDelay=1 -commitToFetchDelay=1 -commitToIEWDelay=1 -commitToRenameDelay=1 -commitWidth=8 -decodeToFetchDelay=1 -decodeToRenameDelay=1 -decodeWidth=8 -defer_registration=false -dispatchWidth=8 -fetchToDecodeDelay=1 -fetchTrapLatency=1 -fetchWidth=8 -forwardComSize=5 -fuPool=system.cpu.fuPool -function_trace=false -function_trace_start=0 -globalCtrBits=2 -globalHistoryBits=13 -globalPredictorSize=8192 -iewToCommitDelay=1 -iewToDecodeDelay=1 -iewToFetchDelay=1 -iewToRenameDelay=1 -instShiftAmt=2 -issueToExecuteDelay=1 -issueWidth=8 -localCtrBits=2 -localHistoryBits=11 -localHistoryTableSize=2048 -localPredictorSize=2048 -max_insts_all_threads=0 -max_insts_any_thread=0 -max_loads_all_threads=0 -max_loads_any_thread=0 -mem=system.cpu.dcache -numIQEntries=64 -numPhysFloatRegs=256 -numPhysIntRegs=256 -numROBEntries=192 -numRobs=1 -numThreads=1 -predType=tournament -renameToDecodeDelay=1 -renameToFetchDelay=1 -renameToIEWDelay=2 -renameToROBDelay=1 -renameWidth=8 -squashWidth=8 -system=system -trapLatency=13 -wbDepth=1 -wbWidth=8 -workload=system.cpu.workload -dcache_port=system.cpu.dcache.cpu_side -icache_port=system.cpu.icache.cpu_side - -[system.cpu.dcache] -type=BaseCache -adaptive_compression=false -assoc=2 -block_size=64 -compressed_bus=false -compression_latency=0 -do_copy=false -hash_delay=1 -hit_latency=1 -latency=1 -lifo=false -max_miss_count=0 -mshrs=10 -prefetch_access=false -prefetch_cache_check_push=true -prefetch_data_accesses_only=false -prefetch_degree=1 -prefetch_latency=10 -prefetch_miss=false -prefetch_past_page=false -prefetch_policy=none -prefetch_serial_squash=false -prefetch_use_cpu_id=true -prefetcher_size=100 -prioritizeRequests=false -protocol=Null -repl=Null -size=262144 -split=false -split_size=0 -store_compressed=false -subblock_size=0 -tgts_per_mshr=5 -trace_addr=0 -two_queue=false -write_buffers=8 -cpu_side=system.cpu.dcache_port -mem_side=system.cpu.toL2Bus.port[1] - -[system.cpu.fuPool] -type=FUPool -children=FUList0 FUList1 FUList2 FUList3 FUList4 FUList5 FUList6 FUList7 -FUList=system.cpu.fuPool.FUList0 system.cpu.fuPool.FUList1 system.cpu.fuPool.FUList2 system.cpu.fuPool.FUList3 system.cpu.fuPool.FUList4 system.cpu.fuPool.FUList5 system.cpu.fuPool.FUList6 system.cpu.fuPool.FUList7 - -[system.cpu.fuPool.FUList0] -type=FUDesc -children=opList0 -count=6 -opList=system.cpu.fuPool.FUList0.opList0 - -[system.cpu.fuPool.FUList0.opList0] -type=OpDesc -issueLat=1 -opClass=IntAlu -opLat=1 - -[system.cpu.fuPool.FUList1] -type=FUDesc -children=opList0 opList1 -count=2 -opList=system.cpu.fuPool.FUList1.opList0 system.cpu.fuPool.FUList1.opList1 - -[system.cpu.fuPool.FUList1.opList0] -type=OpDesc -issueLat=1 -opClass=IntMult -opLat=3 - -[system.cpu.fuPool.FUList1.opList1] -type=OpDesc -issueLat=19 -opClass=IntDiv -opLat=20 - -[system.cpu.fuPool.FUList2] -type=FUDesc -children=opList0 opList1 opList2 -count=4 -opList=system.cpu.fuPool.FUList2.opList0 system.cpu.fuPool.FUList2.opList1 system.cpu.fuPool.FUList2.opList2 - -[system.cpu.fuPool.FUList2.opList0] -type=OpDesc -issueLat=1 -opClass=FloatAdd -opLat=2 - -[system.cpu.fuPool.FUList2.opList1] -type=OpDesc -issueLat=1 -opClass=FloatCmp -opLat=2 - -[system.cpu.fuPool.FUList2.opList2] -type=OpDesc -issueLat=1 -opClass=FloatCvt -opLat=2 - -[system.cpu.fuPool.FUList3] -type=FUDesc -children=opList0 opList1 opList2 -count=2 -opList=system.cpu.fuPool.FUList3.opList0 system.cpu.fuPool.FUList3.opList1 system.cpu.fuPool.FUList3.opList2 - -[system.cpu.fuPool.FUList3.opList0] -type=OpDesc -issueLat=1 -opClass=FloatMult -opLat=4 - -[system.cpu.fuPool.FUList3.opList1] -type=OpDesc -issueLat=12 -opClass=FloatDiv -opLat=12 - -[system.cpu.fuPool.FUList3.opList2] -type=OpDesc -issueLat=24 -opClass=FloatSqrt -opLat=24 - -[system.cpu.fuPool.FUList4] -type=FUDesc -children=opList0 -count=0 -opList=system.cpu.fuPool.FUList4.opList0 - -[system.cpu.fuPool.FUList4.opList0] -type=OpDesc -issueLat=1 -opClass=MemRead -opLat=1 - -[system.cpu.fuPool.FUList5] -type=FUDesc -children=opList0 -count=0 -opList=system.cpu.fuPool.FUList5.opList0 - -[system.cpu.fuPool.FUList5.opList0] -type=OpDesc -issueLat=1 -opClass=MemWrite -opLat=1 - -[system.cpu.fuPool.FUList6] -type=FUDesc -children=opList0 opList1 -count=4 -opList=system.cpu.fuPool.FUList6.opList0 system.cpu.fuPool.FUList6.opList1 - -[system.cpu.fuPool.FUList6.opList0] -type=OpDesc -issueLat=1 -opClass=MemRead -opLat=1 - -[system.cpu.fuPool.FUList6.opList1] -type=OpDesc -issueLat=1 -opClass=MemWrite -opLat=1 - -[system.cpu.fuPool.FUList7] -type=FUDesc -children=opList0 -count=1 -opList=system.cpu.fuPool.FUList7.opList0 - -[system.cpu.fuPool.FUList7.opList0] -type=OpDesc -issueLat=3 -opClass=IprAccess -opLat=3 - -[system.cpu.icache] -type=BaseCache -adaptive_compression=false -assoc=2 -block_size=64 -compressed_bus=false -compression_latency=0 -do_copy=false -hash_delay=1 -hit_latency=1 -latency=1 -lifo=false -max_miss_count=0 -mshrs=10 -prefetch_access=false -prefetch_cache_check_push=true -prefetch_data_accesses_only=false -prefetch_degree=1 -prefetch_latency=10 -prefetch_miss=false -prefetch_past_page=false -prefetch_policy=none -prefetch_serial_squash=false -prefetch_use_cpu_id=true -prefetcher_size=100 -prioritizeRequests=false -protocol=Null -repl=Null -size=131072 -split=false -split_size=0 -store_compressed=false -subblock_size=0 -tgts_per_mshr=5 -trace_addr=0 -two_queue=false -write_buffers=8 -cpu_side=system.cpu.icache_port -mem_side=system.cpu.toL2Bus.port[0] - -[system.cpu.l2cache] -type=BaseCache -adaptive_compression=false -assoc=2 -block_size=64 -compressed_bus=false -compression_latency=0 -do_copy=false -hash_delay=1 -hit_latency=1 -latency=1 -lifo=false -max_miss_count=0 -mshrs=10 -prefetch_access=false -prefetch_cache_check_push=true -prefetch_data_accesses_only=false -prefetch_degree=1 -prefetch_latency=10 -prefetch_miss=false -prefetch_past_page=false -prefetch_policy=none -prefetch_serial_squash=false -prefetch_use_cpu_id=true -prefetcher_size=100 -prioritizeRequests=false -protocol=Null -repl=Null -size=2097152 -split=false -split_size=0 -store_compressed=false -subblock_size=0 -tgts_per_mshr=5 -trace_addr=0 -two_queue=false -write_buffers=8 -cpu_side=system.cpu.toL2Bus.port[2] -mem_side=system.membus.port[1] - -[system.cpu.toL2Bus] -type=Bus -bus_id=0 -port=system.cpu.icache.mem_side system.cpu.dcache.mem_side system.cpu.l2cache.cpu_side - -[system.cpu.workload] -type=LiveProcess -cmd=hello -env= -executable=tests/test-progs/hello/bin/alpha/linux/hello -input=cin -output=cout -system=system - -[system.membus] -type=Bus -bus_id=0 -port=system.physmem.port system.cpu.l2cache.mem_side - -[system.physmem] -type=PhysicalMemory -file= -latency=1 -range=0:134217727 -port=system.membus.port[0] - -[trace] -bufsize=0 -dump_on_exit=false -file=cout -flags= -ignore= -start=0 - diff --git a/tests/long/30.eon/ref/alpha/linux/o3-timing/config.out b/tests/long/30.eon/ref/alpha/linux/o3-timing/config.out index f491a3081..e69de29bb 100644 --- a/tests/long/30.eon/ref/alpha/linux/o3-timing/config.out +++ b/tests/long/30.eon/ref/alpha/linux/o3-timing/config.out @@ -1,403 +0,0 @@ -[root] -type=Root -clock=1000000000000 -max_tick=0 -progress_interval=0 -output_file=cout - -[system.physmem] -type=PhysicalMemory -file= -range=[0,134217727] -latency=1 - -[system] -type=System -physmem=system.physmem -mem_mode=atomic - -[system.membus] -type=Bus -bus_id=0 - -[system.cpu.workload] -type=LiveProcess -cmd=hello -executable=tests/test-progs/hello/bin/alpha/linux/hello -input=cin -output=cout -env= -system=system - -[system.cpu.dcache] -type=BaseCache -size=262144 -assoc=2 -block_size=64 -latency=1 -mshrs=10 -tgts_per_mshr=5 -write_buffers=8 -prioritizeRequests=false -do_copy=false -protocol=null -trace_addr=0 -hash_delay=1 -repl=null -compressed_bus=false -store_compressed=false -adaptive_compression=false -compression_latency=0 -block_size=64 -max_miss_count=0 -addr_range=[0,18446744073709551615] -split=false -split_size=0 -lifo=false -two_queue=false -prefetch_miss=false -prefetch_access=false -prefetcher_size=100 -prefetch_past_page=false -prefetch_serial_squash=false -prefetch_latency=10 -prefetch_degree=1 -prefetch_policy=none -prefetch_cache_check_push=true -prefetch_use_cpu_id=true -prefetch_data_accesses_only=false -hit_latency=1 - -[system.cpu.fuPool.FUList0.opList0] -type=OpDesc -opClass=IntAlu -opLat=1 -issueLat=1 - -[system.cpu.fuPool.FUList0] -type=FUDesc -opList=system.cpu.fuPool.FUList0.opList0 -count=6 - -[system.cpu.fuPool.FUList1.opList0] -type=OpDesc -opClass=IntMult -opLat=3 -issueLat=1 - -[system.cpu.fuPool.FUList1.opList1] -type=OpDesc -opClass=IntDiv -opLat=20 -issueLat=19 - -[system.cpu.fuPool.FUList1] -type=FUDesc -opList=system.cpu.fuPool.FUList1.opList0 system.cpu.fuPool.FUList1.opList1 -count=2 - -[system.cpu.fuPool.FUList2.opList0] -type=OpDesc -opClass=FloatAdd -opLat=2 -issueLat=1 - -[system.cpu.fuPool.FUList2.opList1] -type=OpDesc -opClass=FloatCmp -opLat=2 -issueLat=1 - -[system.cpu.fuPool.FUList2.opList2] -type=OpDesc -opClass=FloatCvt -opLat=2 -issueLat=1 - -[system.cpu.fuPool.FUList2] -type=FUDesc -opList=system.cpu.fuPool.FUList2.opList0 system.cpu.fuPool.FUList2.opList1 system.cpu.fuPool.FUList2.opList2 -count=4 - -[system.cpu.fuPool.FUList3.opList0] -type=OpDesc -opClass=FloatMult -opLat=4 -issueLat=1 - -[system.cpu.fuPool.FUList3.opList1] -type=OpDesc -opClass=FloatDiv -opLat=12 -issueLat=12 - -[system.cpu.fuPool.FUList3.opList2] -type=OpDesc -opClass=FloatSqrt -opLat=24 -issueLat=24 - -[system.cpu.fuPool.FUList3] -type=FUDesc -opList=system.cpu.fuPool.FUList3.opList0 system.cpu.fuPool.FUList3.opList1 system.cpu.fuPool.FUList3.opList2 -count=2 - -[system.cpu.fuPool.FUList4.opList0] -type=OpDesc -opClass=MemRead -opLat=1 -issueLat=1 - -[system.cpu.fuPool.FUList4] -type=FUDesc -opList=system.cpu.fuPool.FUList4.opList0 -count=0 - -[system.cpu.fuPool.FUList5.opList0] -type=OpDesc -opClass=MemWrite -opLat=1 -issueLat=1 - -[system.cpu.fuPool.FUList5] -type=FUDesc -opList=system.cpu.fuPool.FUList5.opList0 -count=0 - -[system.cpu.fuPool.FUList6.opList0] -type=OpDesc -opClass=MemRead -opLat=1 -issueLat=1 - -[system.cpu.fuPool.FUList6.opList1] -type=OpDesc -opClass=MemWrite -opLat=1 -issueLat=1 - -[system.cpu.fuPool.FUList6] -type=FUDesc -opList=system.cpu.fuPool.FUList6.opList0 system.cpu.fuPool.FUList6.opList1 -count=4 - -[system.cpu.fuPool.FUList7.opList0] -type=OpDesc -opClass=IprAccess -opLat=3 -issueLat=3 - -[system.cpu.fuPool.FUList7] -type=FUDesc -opList=system.cpu.fuPool.FUList7.opList0 -count=1 - -[system.cpu.fuPool] -type=FUPool -FUList=system.cpu.fuPool.FUList0 system.cpu.fuPool.FUList1 system.cpu.fuPool.FUList2 system.cpu.fuPool.FUList3 system.cpu.fuPool.FUList4 system.cpu.fuPool.FUList5 system.cpu.fuPool.FUList6 system.cpu.fuPool.FUList7 - -[system.cpu] -type=DerivO3CPU -clock=1 -numThreads=1 -activity=0 -workload=system.cpu.workload -mem=system.cpu.dcache -checker=null -max_insts_any_thread=0 -max_insts_all_threads=0 -max_loads_any_thread=0 -max_loads_all_threads=0 -cachePorts=200 -decodeToFetchDelay=1 -renameToFetchDelay=1 -iewToFetchDelay=1 -commitToFetchDelay=1 -fetchWidth=8 -renameToDecodeDelay=1 -iewToDecodeDelay=1 -commitToDecodeDelay=1 -fetchToDecodeDelay=1 -decodeWidth=8 -iewToRenameDelay=1 -commitToRenameDelay=1 -decodeToRenameDelay=1 -renameWidth=8 -commitToIEWDelay=1 -renameToIEWDelay=2 -issueToExecuteDelay=1 -dispatchWidth=8 -issueWidth=8 -wbWidth=8 -wbDepth=1 -fuPool=system.cpu.fuPool -iewToCommitDelay=1 -renameToROBDelay=1 -commitWidth=8 -squashWidth=8 -trapLatency=13 -backComSize=5 -forwardComSize=5 -predType=tournament -localPredictorSize=2048 -localCtrBits=2 -localHistoryTableSize=2048 -localHistoryBits=11 -globalPredictorSize=8192 -globalCtrBits=2 -globalHistoryBits=13 -choicePredictorSize=8192 -choiceCtrBits=2 -BTBEntries=4096 -BTBTagSize=16 -RASSize=16 -LQEntries=32 -SQEntries=32 -LFSTSize=1024 -SSITSize=1024 -numPhysIntRegs=256 -numPhysFloatRegs=256 -numIQEntries=64 -numROBEntries=192 -smtNumFetchingThreads=1 -smtFetchPolicy=SingleThread -smtLSQPolicy=Partitioned -smtLSQThreshold=100 -smtIQPolicy=Partitioned -smtIQThreshold=100 -smtROBPolicy=Partitioned -smtROBThreshold=100 -smtCommitPolicy=RoundRobin -instShiftAmt=2 -defer_registration=false -function_trace=false -function_trace_start=0 - -[system.cpu.icache] -type=BaseCache -size=131072 -assoc=2 -block_size=64 -latency=1 -mshrs=10 -tgts_per_mshr=5 -write_buffers=8 -prioritizeRequests=false -do_copy=false -protocol=null -trace_addr=0 -hash_delay=1 -repl=null -compressed_bus=false -store_compressed=false -adaptive_compression=false -compression_latency=0 -block_size=64 -max_miss_count=0 -addr_range=[0,18446744073709551615] -split=false -split_size=0 -lifo=false -two_queue=false -prefetch_miss=false -prefetch_access=false -prefetcher_size=100 -prefetch_past_page=false -prefetch_serial_squash=false -prefetch_latency=10 -prefetch_degree=1 -prefetch_policy=none -prefetch_cache_check_push=true -prefetch_use_cpu_id=true -prefetch_data_accesses_only=false -hit_latency=1 - -[system.cpu.l2cache] -type=BaseCache -size=2097152 -assoc=2 -block_size=64 -latency=1 -mshrs=10 -tgts_per_mshr=5 -write_buffers=8 -prioritizeRequests=false -do_copy=false -protocol=null -trace_addr=0 -hash_delay=1 -repl=null -compressed_bus=false -store_compressed=false -adaptive_compression=false -compression_latency=0 -block_size=64 -max_miss_count=0 -addr_range=[0,18446744073709551615] -split=false -split_size=0 -lifo=false -two_queue=false -prefetch_miss=false -prefetch_access=false -prefetcher_size=100 -prefetch_past_page=false -prefetch_serial_squash=false -prefetch_latency=10 -prefetch_degree=1 -prefetch_policy=none -prefetch_cache_check_push=true -prefetch_use_cpu_id=true -prefetch_data_accesses_only=false -hit_latency=1 - -[system.cpu.toL2Bus] -type=Bus -bus_id=0 - -[trace] -flags= -start=0 -bufsize=0 -file=cout -dump_on_exit=false -ignore= - -[stats] -descriptions=true -project_name=test -simulation_name=test -simulation_sample=0 -text_file=m5stats.txt -text_compat=true -mysql_db= -mysql_user= -mysql_password= -mysql_host= -events_start=-1 -dump_reset=false -dump_cycle=0 -dump_period=0 -ignore_events= - -[random] -seed=1 - -[exetrace] -speculative=true -print_cycle=true -print_opclass=true -print_thread=true -print_effaddr=true -print_data=true -print_iregs=false -print_fetchseq=false -print_cpseq=false -print_reg_delta=false -pc_symbol=true -intel_format=false -trace_system=client - -[debug] -break_cycles= - diff --git a/tests/long/30.eon/ref/alpha/linux/o3-timing/m5stats.txt b/tests/long/30.eon/ref/alpha/linux/o3-timing/m5stats.txt index 5d4f9235a..e69de29bb 100644 --- a/tests/long/30.eon/ref/alpha/linux/o3-timing/m5stats.txt +++ b/tests/long/30.eon/ref/alpha/linux/o3-timing/m5stats.txt @@ -1,1974 +0,0 @@ - ----------- Begin Simulation Statistics ---------- -global.BPredUnit.BTBCorrect 0 # Number of correct BTB predictions (this stat may not work properly. -global.BPredUnit.BTBHits 542 # Number of BTB hits -global.BPredUnit.BTBLookups 1938 # Number of BTB lookups -global.BPredUnit.RASInCorrect 48 # Number of incorrect RAS predictions. -global.BPredUnit.condIncorrect 420 # Number of conditional branches incorrect -global.BPredUnit.condPredicted 1304 # Number of conditional branches predicted -global.BPredUnit.lookups 2256 # Number of BP lookups -global.BPredUnit.usedRAS 291 # Number of times the RAS was used to get a target. -host_inst_rate 41797 # Simulator instruction rate (inst/s) -host_mem_usage 160344 # Number of bytes of host memory used -host_seconds 0.13 # Real time elapsed on the host -host_tick_rate 50948 # Simulator tick rate (ticks/s) -memdepunit.memDep.conflictingLoads 12 # Number of conflicting loads. -memdepunit.memDep.conflictingStores 259 # Number of conflicting stores. -memdepunit.memDep.insertedLoads 2050 # Number of loads inserted to the mem dependence unit. -memdepunit.memDep.insertedStores 1221 # Number of stores inserted to the mem dependence unit. -sim_freq 1000000000000 # Frequency of simulated ticks -sim_insts 5623 # Number of instructions simulated -sim_seconds 0.000000 # Number of seconds simulated -sim_ticks 6870 # Number of ticks simulated -system.cpu.commit.COM:branches 862 # Number of branches committed -system.cpu.commit.COM:bw_lim_events 74 # number cycles where commit BW limit reached -system.cpu.commit.COM:bw_limited 0 # number of insts not committed due to BW limits -system.cpu.commit.COM:committed_per_cycle.start_dist # Number of insts commited each cycle -system.cpu.commit.COM:committed_per_cycle.samples 6116 -system.cpu.commit.COM:committed_per_cycle.min_value 0 - 0 3908 6389.80% - 1 1064 1739.70% - 2 389 636.04% - 3 210 343.36% - 4 153 250.16% - 5 93 152.06% - 6 76 124.26% - 7 149 243.62% - 8 74 120.99% -system.cpu.commit.COM:committed_per_cycle.max_value 8 -system.cpu.commit.COM:committed_per_cycle.end_dist - -system.cpu.commit.COM:count 5640 # Number of instructions committed -system.cpu.commit.COM:loads 979 # Number of loads committed -system.cpu.commit.COM:membars 0 # Number of memory barriers committed -system.cpu.commit.COM:refs 1791 # Number of memory references committed -system.cpu.commit.COM:swp_count 0 # Number of s/w prefetches committed -system.cpu.commit.branchMispredicts 337 # The number of times a branch was mispredicted -system.cpu.commit.commitCommittedInsts 5640 # The number of committed instructions -system.cpu.commit.commitNonSpecStalls 17 # The number of times commit has been forced to stall to communicate backwards -system.cpu.commit.commitSquashedInsts 4350 # The number of squashed insts skipped by commit -system.cpu.committedInsts 5623 # Number of Instructions Simulated -system.cpu.committedInsts_total 5623 # Number of Instructions Simulated -system.cpu.cpi 1.221768 # CPI: Cycles Per Instruction -system.cpu.cpi_total 1.221768 # CPI: Total CPI of All Threads -system.cpu.dcache.ReadReq_accesses 1538 # number of ReadReq accesses(hits+misses) -system.cpu.dcache.ReadReq_avg_miss_latency 3.072000 # average ReadReq miss latency -system.cpu.dcache.ReadReq_avg_mshr_miss_latency 2.240000 # average ReadReq mshr miss latency -system.cpu.dcache.ReadReq_hits 1413 # number of ReadReq hits -system.cpu.dcache.ReadReq_miss_latency 384 # number of ReadReq miss cycles -system.cpu.dcache.ReadReq_miss_rate 0.081274 # miss rate for ReadReq accesses -system.cpu.dcache.ReadReq_misses 125 # number of ReadReq misses -system.cpu.dcache.ReadReq_mshr_hits 25 # number of ReadReq MSHR hits -system.cpu.dcache.ReadReq_mshr_miss_latency 224 # number of ReadReq MSHR miss cycles -system.cpu.dcache.ReadReq_mshr_miss_rate 0.065020 # mshr miss rate for ReadReq accesses -system.cpu.dcache.ReadReq_mshr_misses 100 # number of ReadReq MSHR misses -system.cpu.dcache.WriteReq_accesses 821 # number of WriteReq accesses(hits+misses) -system.cpu.dcache.WriteReq_avg_miss_latency 2.467742 # average WriteReq miss latency -system.cpu.dcache.WriteReq_avg_mshr_miss_latency 2.140845 # average WriteReq mshr miss latency -system.cpu.dcache.WriteReq_hits 635 # number of WriteReq hits -system.cpu.dcache.WriteReq_miss_latency 459 # number of WriteReq miss cycles -system.cpu.dcache.WriteReq_miss_rate 0.226553 # miss rate for WriteReq accesses -system.cpu.dcache.WriteReq_misses 186 # number of WriteReq misses -system.cpu.dcache.WriteReq_mshr_hits 108 # number of WriteReq MSHR hits -system.cpu.dcache.WriteReq_mshr_miss_latency 152 # number of WriteReq MSHR miss cycles -system.cpu.dcache.WriteReq_mshr_miss_rate 0.086480 # mshr miss rate for WriteReq accesses -system.cpu.dcache.WriteReq_mshr_misses 71 # number of WriteReq MSHR misses -system.cpu.dcache.avg_blocked_cycles_no_mshrs <err: div-0> # average number of cycles each access was blocked -system.cpu.dcache.avg_blocked_cycles_no_targets 0.800000 # average number of cycles each access was blocked -system.cpu.dcache.avg_refs 11.505618 # Average number of references to valid blocks. -system.cpu.dcache.blocked_no_mshrs 0 # number of cycles access was blocked -system.cpu.dcache.blocked_no_targets 5 # number of cycles access was blocked -system.cpu.dcache.blocked_cycles_no_mshrs 0 # number of cycles access was blocked -system.cpu.dcache.blocked_cycles_no_targets 4 # number of cycles access was blocked -system.cpu.dcache.cache_copies 0 # number of cache copies performed -system.cpu.dcache.demand_accesses 2359 # number of demand (read+write) accesses -system.cpu.dcache.demand_avg_miss_latency 2.710611 # average overall miss latency -system.cpu.dcache.demand_avg_mshr_miss_latency 2.198830 # average overall mshr miss latency -system.cpu.dcache.demand_hits 2048 # number of demand (read+write) hits -system.cpu.dcache.demand_miss_latency 843 # number of demand (read+write) miss cycles -system.cpu.dcache.demand_miss_rate 0.131836 # miss rate for demand accesses -system.cpu.dcache.demand_misses 311 # number of demand (read+write) misses -system.cpu.dcache.demand_mshr_hits 133 # number of demand (read+write) MSHR hits -system.cpu.dcache.demand_mshr_miss_latency 376 # number of demand (read+write) MSHR miss cycles -system.cpu.dcache.demand_mshr_miss_rate 0.072488 # mshr miss rate for demand accesses -system.cpu.dcache.demand_mshr_misses 171 # number of demand (read+write) MSHR misses -system.cpu.dcache.fast_writes 0 # number of fast writes performed -system.cpu.dcache.mshr_cap_events 0 # number of times MSHR cap was activated -system.cpu.dcache.no_allocate_misses 0 # Number of misses that were no-allocate -system.cpu.dcache.overall_accesses 2359 # number of overall (read+write) accesses -system.cpu.dcache.overall_avg_miss_latency 2.710611 # average overall miss latency -system.cpu.dcache.overall_avg_mshr_miss_latency 2.198830 # average overall mshr miss latency -system.cpu.dcache.overall_avg_mshr_uncacheable_latency <err: div-0> # average overall mshr uncacheable latency -system.cpu.dcache.overall_hits 2048 # number of overall hits -system.cpu.dcache.overall_miss_latency 843 # number of overall miss cycles -system.cpu.dcache.overall_miss_rate 0.131836 # miss rate for overall accesses -system.cpu.dcache.overall_misses 311 # number of overall misses -system.cpu.dcache.overall_mshr_hits 133 # number of overall MSHR hits -system.cpu.dcache.overall_mshr_miss_latency 376 # number of overall MSHR miss cycles -system.cpu.dcache.overall_mshr_miss_rate 0.072488 # mshr miss rate for overall accesses -system.cpu.dcache.overall_mshr_misses 171 # number of overall MSHR misses -system.cpu.dcache.overall_mshr_uncacheable_latency 0 # number of overall MSHR uncacheable cycles -system.cpu.dcache.overall_mshr_uncacheable_misses 0 # number of overall MSHR uncacheable misses -system.cpu.dcache.prefetcher.num_hwpf_already_in_cache 0 # number of hwpf that were already in the cache -system.cpu.dcache.prefetcher.num_hwpf_already_in_mshr 0 # number of hwpf that were already in mshr -system.cpu.dcache.prefetcher.num_hwpf_already_in_prefetcher 0 # number of hwpf that were already in the prefetch queue -system.cpu.dcache.prefetcher.num_hwpf_evicted 0 # number of hwpf removed due to no buffer left -system.cpu.dcache.prefetcher.num_hwpf_identified 0 # number of hwpf identified -system.cpu.dcache.prefetcher.num_hwpf_issued 0 # number of hwpf issued -system.cpu.dcache.prefetcher.num_hwpf_removed_MSHR_hit 0 # number of hwpf removed because MSHR allocated -system.cpu.dcache.prefetcher.num_hwpf_span_page 0 # number of hwpf spanning a virtual page -system.cpu.dcache.prefetcher.num_hwpf_squashed_from_miss 0 # number of hwpf that got squashed due to a miss aborting calculation time -system.cpu.dcache.replacements 0 # number of replacements -system.cpu.dcache.sampled_refs 178 # Sample count of references to valid blocks. -system.cpu.dcache.soft_prefetch_mshr_full 0 # number of mshr full events for SW prefetching instrutions -system.cpu.dcache.tagsinuse 119.831029 # Cycle average of tags in use -system.cpu.dcache.total_refs 2048 # Total number of references to valid blocks. -system.cpu.dcache.warmup_cycle 0 # Cycle when the warmup percentage was hit. -system.cpu.dcache.writebacks 0 # number of writebacks -system.cpu.decode.DECODE:BlockedCycles 387 # Number of cycles decode is blocked -system.cpu.decode.DECODE:BranchMispred 93 # Number of times decode detected a branch misprediction -system.cpu.decode.DECODE:BranchResolved 185 # Number of times decode resolved a branch -system.cpu.decode.DECODE:DecodedInsts 12349 # Number of instructions handled by decode -system.cpu.decode.DECODE:IdleCycles 3542 # Number of cycles decode is idle -system.cpu.decode.DECODE:RunCycles 2158 # Number of cycles decode is running -system.cpu.decode.DECODE:SquashCycles 754 # Number of cycles decode is squashing -system.cpu.decode.DECODE:SquashedInsts 286 # Number of squashed instructions handled by decode -system.cpu.decode.DECODE:UnblockCycles 30 # Number of cycles decode is unblocking -system.cpu.fetch.Branches 2256 # Number of branches that fetch encountered -system.cpu.fetch.CacheLines 1582 # Number of cache lines fetched -system.cpu.fetch.Cycles 3905 # Number of cycles fetch has run and was not squashing or blocked -system.cpu.fetch.IcacheSquashes 148 # Number of outstanding Icache misses that were squashed -system.cpu.fetch.Insts 13707 # Number of instructions fetch has processed -system.cpu.fetch.SquashCycles 456 # Number of cycles fetch has spent squashing -system.cpu.fetch.branchRate 0.328336 # Number of branch fetches per cycle -system.cpu.fetch.icacheStallCycles 1582 # Number of cycles fetch is stalled on an Icache miss -system.cpu.fetch.predictedBranches 833 # Number of branches that fetch has predicted taken -system.cpu.fetch.rate 1.994906 # Number of inst fetches per cycle -system.cpu.fetch.rateDist.start_dist # Number of instructions fetched each cycle (Total) -system.cpu.fetch.rateDist.samples 6871 -system.cpu.fetch.rateDist.min_value 0 - 0 4549 6620.58% - 1 174 253.24% - 2 186 270.70% - 3 157 228.50% - 4 211 307.09% - 5 153 222.68% - 6 171 248.87% - 7 105 152.82% - 8 1165 1695.53% -system.cpu.fetch.rateDist.max_value 8 -system.cpu.fetch.rateDist.end_dist - -system.cpu.icache.ReadReq_accesses 1582 # number of ReadReq accesses(hits+misses) -system.cpu.icache.ReadReq_avg_miss_latency 2.960245 # average ReadReq miss latency -system.cpu.icache.ReadReq_avg_mshr_miss_latency 1.996885 # average ReadReq mshr miss latency -system.cpu.icache.ReadReq_hits 1255 # number of ReadReq hits -system.cpu.icache.ReadReq_miss_latency 968 # number of ReadReq miss cycles -system.cpu.icache.ReadReq_miss_rate 0.206700 # miss rate for ReadReq accesses -system.cpu.icache.ReadReq_misses 327 # number of ReadReq misses -system.cpu.icache.ReadReq_mshr_hits 6 # number of ReadReq MSHR hits -system.cpu.icache.ReadReq_mshr_miss_latency 641 # number of ReadReq MSHR miss cycles -system.cpu.icache.ReadReq_mshr_miss_rate 0.202908 # mshr miss rate for ReadReq accesses -system.cpu.icache.ReadReq_mshr_misses 321 # number of ReadReq MSHR misses -system.cpu.icache.avg_blocked_cycles_no_mshrs no value # average number of cycles each access was blocked -system.cpu.icache.avg_blocked_cycles_no_targets no value # average number of cycles each access was blocked -system.cpu.icache.avg_refs 3.909657 # Average number of references to valid blocks. -system.cpu.icache.blocked_no_mshrs 0 # number of cycles access was blocked -system.cpu.icache.blocked_no_targets 0 # number of cycles access was blocked -system.cpu.icache.blocked_cycles_no_mshrs 0 # number of cycles access was blocked -system.cpu.icache.blocked_cycles_no_targets 0 # number of cycles access was blocked -system.cpu.icache.cache_copies 0 # number of cache copies performed -system.cpu.icache.demand_accesses 1582 # number of demand (read+write) accesses -system.cpu.icache.demand_avg_miss_latency 2.960245 # average overall miss latency -system.cpu.icache.demand_avg_mshr_miss_latency 1.996885 # average overall mshr miss latency -system.cpu.icache.demand_hits 1255 # number of demand (read+write) hits -system.cpu.icache.demand_miss_latency 968 # number of demand (read+write) miss cycles -system.cpu.icache.demand_miss_rate 0.206700 # miss rate for demand accesses -system.cpu.icache.demand_misses 327 # number of demand (read+write) misses -system.cpu.icache.demand_mshr_hits 6 # number of demand (read+write) MSHR hits -system.cpu.icache.demand_mshr_miss_latency 641 # number of demand (read+write) MSHR miss cycles -system.cpu.icache.demand_mshr_miss_rate 0.202908 # mshr miss rate for demand accesses -system.cpu.icache.demand_mshr_misses 321 # number of demand (read+write) MSHR misses -system.cpu.icache.fast_writes 0 # number of fast writes performed -system.cpu.icache.mshr_cap_events 0 # number of times MSHR cap was activated -system.cpu.icache.no_allocate_misses 0 # Number of misses that were no-allocate -system.cpu.icache.overall_accesses 1582 # number of overall (read+write) accesses -system.cpu.icache.overall_avg_miss_latency 2.960245 # average overall miss latency -system.cpu.icache.overall_avg_mshr_miss_latency 1.996885 # average overall mshr miss latency -system.cpu.icache.overall_avg_mshr_uncacheable_latency <err: div-0> # average overall mshr uncacheable latency -system.cpu.icache.overall_hits 1255 # number of overall hits -system.cpu.icache.overall_miss_latency 968 # number of overall miss cycles -system.cpu.icache.overall_miss_rate 0.206700 # miss rate for overall accesses -system.cpu.icache.overall_misses 327 # number of overall misses -system.cpu.icache.overall_mshr_hits 6 # number of overall MSHR hits -system.cpu.icache.overall_mshr_miss_latency 641 # number of overall MSHR miss cycles -system.cpu.icache.overall_mshr_miss_rate 0.202908 # mshr miss rate for overall accesses -system.cpu.icache.overall_mshr_misses 321 # number of overall MSHR misses -system.cpu.icache.overall_mshr_uncacheable_latency 0 # number of overall MSHR uncacheable cycles -system.cpu.icache.overall_mshr_uncacheable_misses 0 # number of overall MSHR uncacheable misses -system.cpu.icache.prefetcher.num_hwpf_already_in_cache 0 # number of hwpf that were already in the cache -system.cpu.icache.prefetcher.num_hwpf_already_in_mshr 0 # number of hwpf that were already in mshr -system.cpu.icache.prefetcher.num_hwpf_already_in_prefetcher 0 # number of hwpf that were already in the prefetch queue -system.cpu.icache.prefetcher.num_hwpf_evicted 0 # number of hwpf removed due to no buffer left -system.cpu.icache.prefetcher.num_hwpf_identified 0 # number of hwpf identified -system.cpu.icache.prefetcher.num_hwpf_issued 0 # number of hwpf issued -system.cpu.icache.prefetcher.num_hwpf_removed_MSHR_hit 0 # number of hwpf removed because MSHR allocated -system.cpu.icache.prefetcher.num_hwpf_span_page 0 # number of hwpf spanning a virtual page -system.cpu.icache.prefetcher.num_hwpf_squashed_from_miss 0 # number of hwpf that got squashed due to a miss aborting calculation time -system.cpu.icache.replacements 0 # number of replacements -system.cpu.icache.sampled_refs 321 # Sample count of references to valid blocks. -system.cpu.icache.soft_prefetch_mshr_full 0 # number of mshr full events for SW prefetching instrutions -system.cpu.icache.tagsinuse 176.393247 # Cycle average of tags in use -system.cpu.icache.total_refs 1255 # Total number of references to valid blocks. -system.cpu.icache.warmup_cycle 0 # Cycle when the warmup percentage was hit. -system.cpu.icache.writebacks 0 # number of writebacks -system.cpu.iew.EXEC:branches 1206 # Number of branches executed -system.cpu.iew.EXEC:insts 7969 # Number of executed instructions -system.cpu.iew.EXEC:loads 1610 # Number of load instructions executed -system.cpu.iew.EXEC:nop 37 # number of nop insts executed -system.cpu.iew.EXEC:rate 1.159802 # Inst execution rate -system.cpu.iew.EXEC:refs 2599 # number of memory reference insts executed -system.cpu.iew.EXEC:squashedInsts 419 # Number of squashed instructions skipped in execute -system.cpu.iew.EXEC:stores 989 # Number of stores executed -system.cpu.iew.EXEC:swp 0 # number of swp insts executed -system.cpu.iew.WB:consumers 5438 # num instructions consuming a value -system.cpu.iew.WB:count 7722 # cumulative count of insts written-back -system.cpu.iew.WB:fanout 0.744575 # average fanout of values written-back -system.cpu.iew.WB:penalized 0 # number of instrctions required to write to 'other' IQ -system.cpu.iew.WB:penalized_rate 0 # fraction of instructions written-back that wrote to 'other' IQ -system.cpu.iew.WB:producers 4049 # num instructions producing a value -system.cpu.iew.WB:rate 1.123854 # insts written-back per cycle -system.cpu.iew.WB:sent 7762 # cumulative count of insts sent to commit -system.cpu.iew.branchMispredicts 393 # Number of branch mispredicts detected at execute -system.cpu.iew.iewBlockCycles 4 # Number of cycles IEW is blocking -system.cpu.iew.iewDispLoadInsts 2050 # Number of dispatched load instructions -system.cpu.iew.iewDispNonSpecInsts 21 # Number of dispatched non-speculative instructions -system.cpu.iew.iewDispSquashedInsts 272 # Number of squashed instructions skipped by dispatch -system.cpu.iew.iewDispStoreInsts 1221 # Number of dispatched store instructions -system.cpu.iew.iewDispatchedInsts 9990 # Number of instructions dispatched to IQ -system.cpu.iew.iewIQFullEvents 0 # Number of times the IQ has become full, causing a stall -system.cpu.iew.iewIdleCycles 0 # Number of cycles IEW is idle -system.cpu.iew.iewLSQFullEvents 0 # Number of times the LSQ has become full, causing a stall -system.cpu.iew.iewSquashCycles 754 # Number of cycles IEW is squashing -system.cpu.iew.iewUnblockCycles 0 # Number of cycles IEW is unblocking -system.cpu.iew.lsq.thread.0.blockedLoads 1 # Number of blocked loads due to partial load-store forwarding -system.cpu.iew.lsq.thread.0.cacheBlocked 5 # Number of times an access to memory failed due to the cache being blocked -system.cpu.iew.lsq.thread.0.forwLoads 55 # Number of loads that had data forwarded from stores -system.cpu.iew.lsq.thread.0.ignoredResponses 5 # Number of memory responses ignored because the instruction is squashed -system.cpu.iew.lsq.thread.0.invAddrLoads 0 # Number of loads ignored due to an invalid address -system.cpu.iew.lsq.thread.0.invAddrSwpfs 0 # Number of software prefetches ignored due to an invalid address -system.cpu.iew.lsq.thread.0.rescheduledLoads 1 # Number of loads that were rescheduled -system.cpu.iew.lsq.thread.0.squashedLoads 1071 # Number of loads squashed -system.cpu.iew.lsq.thread.0.squashedStores 409 # Number of stores squashed -system.cpu.iew.memOrderViolationEvents 41 # Number of memory order violations -system.cpu.iew.predictedNotTakenIncorrect 296 # Number of branches that were predicted not taken incorrectly -system.cpu.iew.predictedTakenIncorrect 97 # Number of branches that were predicted taken incorrectly -system.cpu.ipc 0.818486 # IPC: Instructions Per Cycle -system.cpu.ipc_total 0.818486 # IPC: Total IPC of All Threads -system.cpu.iq.IQ:residence:(null).start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:(null).samples 0 -system.cpu.iq.IQ:residence:(null).min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:(null).max_value 0 -system.cpu.iq.IQ:residence:(null).end_dist - -system.cpu.iq.IQ:residence:IntAlu.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:IntAlu.samples 0 -system.cpu.iq.IQ:residence:IntAlu.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:IntAlu.max_value 0 -system.cpu.iq.IQ:residence:IntAlu.end_dist - -system.cpu.iq.IQ:residence:IntMult.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:IntMult.samples 0 -system.cpu.iq.IQ:residence:IntMult.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:IntMult.max_value 0 -system.cpu.iq.IQ:residence:IntMult.end_dist - -system.cpu.iq.IQ:residence:IntDiv.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:IntDiv.samples 0 -system.cpu.iq.IQ:residence:IntDiv.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:IntDiv.max_value 0 -system.cpu.iq.IQ:residence:IntDiv.end_dist - -system.cpu.iq.IQ:residence:FloatAdd.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:FloatAdd.samples 0 -system.cpu.iq.IQ:residence:FloatAdd.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:FloatAdd.max_value 0 -system.cpu.iq.IQ:residence:FloatAdd.end_dist - -system.cpu.iq.IQ:residence:FloatCmp.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:FloatCmp.samples 0 -system.cpu.iq.IQ:residence:FloatCmp.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:FloatCmp.max_value 0 -system.cpu.iq.IQ:residence:FloatCmp.end_dist - -system.cpu.iq.IQ:residence:FloatCvt.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:FloatCvt.samples 0 -system.cpu.iq.IQ:residence:FloatCvt.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:FloatCvt.max_value 0 -system.cpu.iq.IQ:residence:FloatCvt.end_dist - -system.cpu.iq.IQ:residence:FloatMult.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:FloatMult.samples 0 -system.cpu.iq.IQ:residence:FloatMult.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:FloatMult.max_value 0 -system.cpu.iq.IQ:residence:FloatMult.end_dist - -system.cpu.iq.IQ:residence:FloatDiv.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:FloatDiv.samples 0 -system.cpu.iq.IQ:residence:FloatDiv.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:FloatDiv.max_value 0 -system.cpu.iq.IQ:residence:FloatDiv.end_dist - -system.cpu.iq.IQ:residence:FloatSqrt.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:FloatSqrt.samples 0 -system.cpu.iq.IQ:residence:FloatSqrt.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:FloatSqrt.max_value 0 -system.cpu.iq.IQ:residence:FloatSqrt.end_dist - -system.cpu.iq.IQ:residence:MemRead.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:MemRead.samples 0 -system.cpu.iq.IQ:residence:MemRead.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:MemRead.max_value 0 -system.cpu.iq.IQ:residence:MemRead.end_dist - -system.cpu.iq.IQ:residence:MemWrite.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:MemWrite.samples 0 -system.cpu.iq.IQ:residence:MemWrite.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:MemWrite.max_value 0 -system.cpu.iq.IQ:residence:MemWrite.end_dist - -system.cpu.iq.IQ:residence:IprAccess.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:IprAccess.samples 0 -system.cpu.iq.IQ:residence:IprAccess.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:IprAccess.max_value 0 -system.cpu.iq.IQ:residence:IprAccess.end_dist - -system.cpu.iq.IQ:residence:InstPrefetch.start_dist # cycles from dispatch to issue -system.cpu.iq.IQ:residence:InstPrefetch.samples 0 -system.cpu.iq.IQ:residence:InstPrefetch.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.IQ:residence:InstPrefetch.max_value 0 -system.cpu.iq.IQ:residence:InstPrefetch.end_dist - -system.cpu.iq.ISSUE:(null)_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:(null)_delay.samples 0 -system.cpu.iq.ISSUE:(null)_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:(null)_delay.max_value 0 -system.cpu.iq.ISSUE:(null)_delay.end_dist - -system.cpu.iq.ISSUE:IntAlu_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:IntAlu_delay.samples 0 -system.cpu.iq.ISSUE:IntAlu_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:IntAlu_delay.max_value 0 -system.cpu.iq.ISSUE:IntAlu_delay.end_dist - -system.cpu.iq.ISSUE:IntMult_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:IntMult_delay.samples 0 -system.cpu.iq.ISSUE:IntMult_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:IntMult_delay.max_value 0 -system.cpu.iq.ISSUE:IntMult_delay.end_dist - -system.cpu.iq.ISSUE:IntDiv_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:IntDiv_delay.samples 0 -system.cpu.iq.ISSUE:IntDiv_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:IntDiv_delay.max_value 0 -system.cpu.iq.ISSUE:IntDiv_delay.end_dist - -system.cpu.iq.ISSUE:FloatAdd_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:FloatAdd_delay.samples 0 -system.cpu.iq.ISSUE:FloatAdd_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:FloatAdd_delay.max_value 0 -system.cpu.iq.ISSUE:FloatAdd_delay.end_dist - -system.cpu.iq.ISSUE:FloatCmp_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:FloatCmp_delay.samples 0 -system.cpu.iq.ISSUE:FloatCmp_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:FloatCmp_delay.max_value 0 -system.cpu.iq.ISSUE:FloatCmp_delay.end_dist - -system.cpu.iq.ISSUE:FloatCvt_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:FloatCvt_delay.samples 0 -system.cpu.iq.ISSUE:FloatCvt_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:FloatCvt_delay.max_value 0 -system.cpu.iq.ISSUE:FloatCvt_delay.end_dist - -system.cpu.iq.ISSUE:FloatMult_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:FloatMult_delay.samples 0 -system.cpu.iq.ISSUE:FloatMult_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:FloatMult_delay.max_value 0 -system.cpu.iq.ISSUE:FloatMult_delay.end_dist - -system.cpu.iq.ISSUE:FloatDiv_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:FloatDiv_delay.samples 0 -system.cpu.iq.ISSUE:FloatDiv_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:FloatDiv_delay.max_value 0 -system.cpu.iq.ISSUE:FloatDiv_delay.end_dist - -system.cpu.iq.ISSUE:FloatSqrt_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:FloatSqrt_delay.samples 0 -system.cpu.iq.ISSUE:FloatSqrt_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:FloatSqrt_delay.max_value 0 -system.cpu.iq.ISSUE:FloatSqrt_delay.end_dist - -system.cpu.iq.ISSUE:MemRead_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:MemRead_delay.samples 0 -system.cpu.iq.ISSUE:MemRead_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:MemRead_delay.max_value 0 -system.cpu.iq.ISSUE:MemRead_delay.end_dist - -system.cpu.iq.ISSUE:MemWrite_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:MemWrite_delay.samples 0 -system.cpu.iq.ISSUE:MemWrite_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:MemWrite_delay.max_value 0 -system.cpu.iq.ISSUE:MemWrite_delay.end_dist - -system.cpu.iq.ISSUE:IprAccess_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:IprAccess_delay.samples 0 -system.cpu.iq.ISSUE:IprAccess_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:IprAccess_delay.max_value 0 -system.cpu.iq.ISSUE:IprAccess_delay.end_dist - -system.cpu.iq.ISSUE:InstPrefetch_delay.start_dist # cycles from operands ready to issue -system.cpu.iq.ISSUE:InstPrefetch_delay.samples 0 -system.cpu.iq.ISSUE:InstPrefetch_delay.min_value 0 - 0 0 - 2 0 - 4 0 - 6 0 - 8 0 - 10 0 - 12 0 - 14 0 - 16 0 - 18 0 - 20 0 - 22 0 - 24 0 - 26 0 - 28 0 - 30 0 - 32 0 - 34 0 - 36 0 - 38 0 - 40 0 - 42 0 - 44 0 - 46 0 - 48 0 - 50 0 - 52 0 - 54 0 - 56 0 - 58 0 - 60 0 - 62 0 - 64 0 - 66 0 - 68 0 - 70 0 - 72 0 - 74 0 - 76 0 - 78 0 - 80 0 - 82 0 - 84 0 - 86 0 - 88 0 - 90 0 - 92 0 - 94 0 - 96 0 - 98 0 -system.cpu.iq.ISSUE:InstPrefetch_delay.max_value 0 -system.cpu.iq.ISSUE:InstPrefetch_delay.end_dist - -system.cpu.iq.ISSUE:FU_type_0 8388 # Type of FU issued -system.cpu.iq.ISSUE:FU_type_0.start_dist - (null) 2 0.02% # Type of FU issued - IntAlu 5594 66.69% # Type of FU issued - IntMult 1 0.01% # Type of FU issued - IntDiv 0 0.00% # Type of FU issued - FloatAdd 2 0.02% # Type of FU issued - FloatCmp 0 0.00% # Type of FU issued - FloatCvt 0 0.00% # Type of FU issued - FloatMult 0 0.00% # Type of FU issued - FloatDiv 0 0.00% # Type of FU issued - FloatSqrt 0 0.00% # Type of FU issued - MemRead 1757 20.95% # Type of FU issued - MemWrite 1032 12.30% # Type of FU issued - IprAccess 0 0.00% # Type of FU issued - InstPrefetch 0 0.00% # Type of FU issued -system.cpu.iq.ISSUE:FU_type_0.end_dist -system.cpu.iq.ISSUE:fu_busy_cnt 115 # FU busy when requested -system.cpu.iq.ISSUE:fu_busy_rate 0.013710 # FU busy rate (busy events/executed inst) -system.cpu.iq.ISSUE:fu_full.start_dist - (null) 0 0.00% # attempts to use FU when none available - IntAlu 1 0.87% # attempts to use FU when none available - IntMult 0 0.00% # attempts to use FU when none available - IntDiv 0 0.00% # attempts to use FU when none available - FloatAdd 0 0.00% # attempts to use FU when none available - FloatCmp 0 0.00% # attempts to use FU when none available - FloatCvt 0 0.00% # attempts to use FU when none available - FloatMult 0 0.00% # attempts to use FU when none available - FloatDiv 0 0.00% # attempts to use FU when none available - FloatSqrt 0 0.00% # attempts to use FU when none available - MemRead 76 66.09% # attempts to use FU when none available - MemWrite 38 33.04% # attempts to use FU when none available - IprAccess 0 0.00% # attempts to use FU when none available - InstPrefetch 0 0.00% # attempts to use FU when none available -system.cpu.iq.ISSUE:fu_full.end_dist -system.cpu.iq.ISSUE:issued_per_cycle.start_dist # Number of insts issued each cycle -system.cpu.iq.ISSUE:issued_per_cycle.samples 6871 -system.cpu.iq.ISSUE:issued_per_cycle.min_value 0 - 0 3753 5462.09% - 1 894 1301.12% - 2 723 1052.25% - 3 614 893.61% - 4 451 656.38% - 5 279 406.05% - 6 104 151.36% - 7 41 59.67% - 8 12 17.46% -system.cpu.iq.ISSUE:issued_per_cycle.max_value 8 -system.cpu.iq.ISSUE:issued_per_cycle.end_dist - -system.cpu.iq.ISSUE:rate 1.220783 # Inst issue rate -system.cpu.iq.iqInstsAdded 9932 # Number of instructions added to the IQ (excludes non-spec) -system.cpu.iq.iqInstsIssued 8388 # Number of instructions issued -system.cpu.iq.iqNonSpecInstsAdded 21 # Number of non-speculative instructions added to the IQ -system.cpu.iq.iqSquashedInstsExamined 3990 # Number of squashed instructions iterated over during squash; mainly for profiling -system.cpu.iq.iqSquashedInstsIssued 21 # Number of squashed instructions issued -system.cpu.iq.iqSquashedNonSpecRemoved 4 # Number of squashed non-spec instructions that were removed -system.cpu.iq.iqSquashedOperandsExamined 2486 # Number of squashed operands that are examined and possibly removed from graph -system.cpu.l2cache.ReadReq_accesses 499 # number of ReadReq accesses(hits+misses) -system.cpu.l2cache.ReadReq_avg_miss_latency 2.042254 # average ReadReq miss latency -system.cpu.l2cache.ReadReq_avg_mshr_miss_latency 1 # average ReadReq mshr miss latency -system.cpu.l2cache.ReadReq_hits 2 # number of ReadReq hits -system.cpu.l2cache.ReadReq_miss_latency 1015 # number of ReadReq miss cycles -system.cpu.l2cache.ReadReq_miss_rate 0.995992 # miss rate for ReadReq accesses -system.cpu.l2cache.ReadReq_misses 497 # number of ReadReq misses -system.cpu.l2cache.ReadReq_mshr_miss_latency 490 # number of ReadReq MSHR miss cycles -system.cpu.l2cache.ReadReq_mshr_miss_rate 0.981964 # mshr miss rate for ReadReq accesses -system.cpu.l2cache.ReadReq_mshr_misses 490 # number of ReadReq MSHR misses -system.cpu.l2cache.avg_blocked_cycles_no_mshrs <err: div-0> # average number of cycles each access was blocked -system.cpu.l2cache.avg_blocked_cycles_no_targets <err: div-0> # average number of cycles each access was blocked -system.cpu.l2cache.avg_refs 0.004024 # Average number of references to valid blocks. -system.cpu.l2cache.blocked_no_mshrs 0 # number of cycles access was blocked -system.cpu.l2cache.blocked_no_targets 0 # number of cycles access was blocked -system.cpu.l2cache.blocked_cycles_no_mshrs 0 # number of cycles access was blocked -system.cpu.l2cache.blocked_cycles_no_targets 0 # number of cycles access was blocked -system.cpu.l2cache.cache_copies 0 # number of cache copies performed -system.cpu.l2cache.demand_accesses 499 # number of demand (read+write) accesses -system.cpu.l2cache.demand_avg_miss_latency 2.042254 # average overall miss latency -system.cpu.l2cache.demand_avg_mshr_miss_latency 1 # average overall mshr miss latency -system.cpu.l2cache.demand_hits 2 # number of demand (read+write) hits -system.cpu.l2cache.demand_miss_latency 1015 # number of demand (read+write) miss cycles -system.cpu.l2cache.demand_miss_rate 0.995992 # miss rate for demand accesses -system.cpu.l2cache.demand_misses 497 # number of demand (read+write) misses -system.cpu.l2cache.demand_mshr_hits 0 # number of demand (read+write) MSHR hits -system.cpu.l2cache.demand_mshr_miss_latency 490 # number of demand (read+write) MSHR miss cycles -system.cpu.l2cache.demand_mshr_miss_rate 0.981964 # mshr miss rate for demand accesses -system.cpu.l2cache.demand_mshr_misses 490 # number of demand (read+write) MSHR misses -system.cpu.l2cache.fast_writes 0 # number of fast writes performed -system.cpu.l2cache.mshr_cap_events 0 # number of times MSHR cap was activated -system.cpu.l2cache.no_allocate_misses 0 # Number of misses that were no-allocate -system.cpu.l2cache.overall_accesses 499 # number of overall (read+write) accesses -system.cpu.l2cache.overall_avg_miss_latency 2.042254 # average overall miss latency -system.cpu.l2cache.overall_avg_mshr_miss_latency 1 # average overall mshr miss latency -system.cpu.l2cache.overall_avg_mshr_uncacheable_latency <err: div-0> # average overall mshr uncacheable latency -system.cpu.l2cache.overall_hits 2 # number of overall hits -system.cpu.l2cache.overall_miss_latency 1015 # number of overall miss cycles -system.cpu.l2cache.overall_miss_rate 0.995992 # miss rate for overall accesses -system.cpu.l2cache.overall_misses 497 # number of overall misses -system.cpu.l2cache.overall_mshr_hits 0 # number of overall MSHR hits -system.cpu.l2cache.overall_mshr_miss_latency 490 # number of overall MSHR miss cycles -system.cpu.l2cache.overall_mshr_miss_rate 0.981964 # mshr miss rate for overall accesses -system.cpu.l2cache.overall_mshr_misses 490 # number of overall MSHR misses -system.cpu.l2cache.overall_mshr_uncacheable_latency 0 # number of overall MSHR uncacheable cycles -system.cpu.l2cache.overall_mshr_uncacheable_misses 0 # number of overall MSHR uncacheable misses -system.cpu.l2cache.prefetcher.num_hwpf_already_in_cache 0 # number of hwpf that were already in the cache -system.cpu.l2cache.prefetcher.num_hwpf_already_in_mshr 0 # number of hwpf that were already in mshr -system.cpu.l2cache.prefetcher.num_hwpf_already_in_prefetcher 0 # number of hwpf that were already in the prefetch queue -system.cpu.l2cache.prefetcher.num_hwpf_evicted 0 # number of hwpf removed due to no buffer left -system.cpu.l2cache.prefetcher.num_hwpf_identified 0 # number of hwpf identified -system.cpu.l2cache.prefetcher.num_hwpf_issued 0 # number of hwpf issued -system.cpu.l2cache.prefetcher.num_hwpf_removed_MSHR_hit 0 # number of hwpf removed because MSHR allocated -system.cpu.l2cache.prefetcher.num_hwpf_span_page 0 # number of hwpf spanning a virtual page -system.cpu.l2cache.prefetcher.num_hwpf_squashed_from_miss 0 # number of hwpf that got squashed due to a miss aborting calculation time -system.cpu.l2cache.replacements 0 # number of replacements -system.cpu.l2cache.sampled_refs 497 # Sample count of references to valid blocks. -system.cpu.l2cache.soft_prefetch_mshr_full 0 # number of mshr full events for SW prefetching instrutions -system.cpu.l2cache.tagsinuse 295.773395 # Cycle average of tags in use -system.cpu.l2cache.total_refs 2 # Total number of references to valid blocks. -system.cpu.l2cache.warmup_cycle 0 # Cycle when the warmup percentage was hit. -system.cpu.l2cache.writebacks 0 # number of writebacks -system.cpu.numCycles 6871 # number of cpu cycles simulated -system.cpu.rename.RENAME:BlockCycles 4 # Number of cycles rename is blocking -system.cpu.rename.RENAME:CommittedMaps 4051 # Number of HB maps that are committed -system.cpu.rename.RENAME:IdleCycles 3758 # Number of cycles rename is idle -system.cpu.rename.RENAME:LSQFullEvents 62 # Number of times rename has blocked due to LSQ full -system.cpu.rename.RENAME:RenameLookups 14786 # Number of register rename lookups that rename has made -system.cpu.rename.RENAME:RenamedInsts 11555 # Number of instructions processed by rename -system.cpu.rename.RENAME:RenamedOperands 8634 # Number of destination operands rename has renamed -system.cpu.rename.RENAME:RunCycles 1975 # Number of cycles rename is running -system.cpu.rename.RENAME:SquashCycles 754 # Number of cycles rename is squashing -system.cpu.rename.RENAME:UnblockCycles 111 # Number of cycles rename is unblocking -system.cpu.rename.RENAME:UndoneMaps 4583 # Number of HB maps that are undone due to squashing -system.cpu.rename.RENAME:serializeStallCycles 269 # count of cycles rename stalled for serializing inst -system.cpu.rename.RENAME:serializingInsts 26 # count of serializing insts renamed -system.cpu.rename.RENAME:skidInsts 408 # count of insts added to the skid buffer -system.cpu.rename.RENAME:tempSerializingInsts 21 # count of temporary serializing insts renamed -system.cpu.workload.PROG:num_syscalls 17 # Number of system calls - ----------- End Simulation Statistics ---------- diff --git a/tests/long/30.eon/ref/alpha/linux/o3-timing/stderr b/tests/long/30.eon/ref/alpha/linux/o3-timing/stderr index 8893caac8..e69de29bb 100644 --- a/tests/long/30.eon/ref/alpha/linux/o3-timing/stderr +++ b/tests/long/30.eon/ref/alpha/linux/o3-timing/stderr @@ -1,3 +0,0 @@ -warn: Entering event queue @ 0. Starting simulation... -warn: cycle 0: fault (page_table_fault) detected @ PC 0x000000 -warn: Increasing stack 0x11ff92000:0x11ff9b000 to 0x11ff90000:0x11ff9b000 because of access to 0x11ff91ff0 diff --git a/tests/long/30.eon/ref/alpha/linux/o3-timing/stdout b/tests/long/30.eon/ref/alpha/linux/o3-timing/stdout index fbb329a2f..e69de29bb 100644 --- a/tests/long/30.eon/ref/alpha/linux/o3-timing/stdout +++ b/tests/long/30.eon/ref/alpha/linux/o3-timing/stdout @@ -1,13 +0,0 @@ -Hello world! -M5 Simulator System - -Copyright (c) 2001-2006 -The Regents of The University of Michigan -All Rights Reserved - - -M5 compiled Sep 5 2006 15:28:48 -M5 started Tue Sep 5 15:42:12 2006 -M5 executing on zizzer.eecs.umich.edu -command line: build/ALPHA_SE/m5.opt -d build/ALPHA_SE/tests/opt/quick/00.hello/alpha/linux/o3-timing tests/run.py quick/00.hello/alpha/linux/o3-timing -Exiting @ tick 6870 because target called exit() diff --git a/util/regress b/util/regress index 8e60b6fee..5d6f321f2 100755 --- a/util/regress +++ b/util/regress @@ -43,11 +43,13 @@ optparser.add_option('--builds', dest='builds', help='comma-separated list of build targets to test ' " (default: '%default')" ) optparser.add_option('--variants', dest='variants', - default='opt', + default='fast', help='comma-separated list of build variants to test ' " (default: '%default')" ) optparser.add_option('--scons-opts', dest='scons_opts', default='', help='scons options', metavar='OPTS') +optparser.add_option('-j', '--jobs', type='int', default=1, + help='number of parallel jobs to use') (options, tests) = optparser.parse_args() @@ -75,10 +77,11 @@ def shellquote(s): try: if not tests: - print "No tests specified." - sys.exit(1) - - if 'all' in tests: + print "No tests specified, just building binaries." + targets = ['build/%s/m5.%s' % (build, variant) + for build in builds + for variant in variants] + elif 'all' in tests: targets = ['build/%s/tests/%s' % (build, variant) for build in builds for variant in variants] @@ -88,7 +91,11 @@ try: for variant in variants for test in tests] - system('scons %s %s' % (options.scons_opts, ' '.join(targets))) + scons_opts = options.scons_opts + if options.jobs != 1: + scons_opts += ' -j %d' % options.jobs + + system('scons %s %s' % (scons_opts, ' '.join(targets))) sys.exit(0) |