Gabe Black [Fri, 11 Mar 2011 19:27:36 +0000 (11:27 -0800)]
SCons: Stop embedding the mercurial revision into the binary.
This causes a lot of rebuilds that could have otherwise possibly been
avoided, and, more annoyingly, a lot of unnecessary rerunning of the
regressions. The benefits of having the revision in the output haven't
materialized, so this change removes it.
Gabe Black [Fri, 11 Mar 2011 19:27:26 +0000 (11:27 -0800)]
Gems: Eliminate the now unused GEMS_ROOT scons variable.
Gabe Black [Fri, 11 Mar 2011 19:27:16 +0000 (11:27 -0800)]
Ruby: Get rid of the dead ruby tester.
None of the code in the ruby tester directory is compiled or referred to
outside of that directory. This change eliminates it. If it's needed in the
future, it can be revived from the history. In the mean time, this removes
clutter and the only use of the GEMS_ROOT scons variable.
Yi Xiang [Wed, 9 Mar 2011 05:43:11 +0000 (21:43 -0800)]
Alpha: Fix the datatypes of some values read from the simulated kernel.
Gabe Black [Fri, 4 Mar 2011 08:11:02 +0000 (00:11 -0800)]
SCons: Fix the polarity on the --ignore-style check.
Gabe Black [Fri, 4 Mar 2011 07:55:21 +0000 (23:55 -0800)]
SCons: Clean up some inconsistent capitalization in scons options.
Gabe Black [Fri, 4 Mar 2011 07:54:31 +0000 (23:54 -0800)]
SCons: Turn some scons variables into command line options.
Gabe Black [Fri, 4 Mar 2011 07:01:38 +0000 (23:01 -0800)]
Mips: MIPS_FS doesn't build currently, so delete it to avoid confusion.
MIPS_FS doesn't build and presumably doesn't work right now. Users might see
the MIP_FS file in build_opts and expect it to work. To avoid confusion, this
change deletes that file.
Gabe Black [Thu, 3 Mar 2011 06:53:11 +0000 (22:53 -0800)]
Statetrace: Stub out the missing i386 version of sendState.
Gabe Black [Thu, 3 Mar 2011 06:53:11 +0000 (22:53 -0800)]
Statetrace: Rename i386 to i686.
--HG--
rename : util/statetrace/arch/i386/tracechild.cc => util/statetrace/arch/i686/tracechild.cc
rename : util/statetrace/arch/i386/tracechild.hh => util/statetrace/arch/i686/tracechild.hh
Gabe Black [Thu, 3 Mar 2011 06:53:11 +0000 (22:53 -0800)]
Statetrace: Fix the i686 detection macro.
Gabe Black [Thu, 3 Mar 2011 06:53:11 +0000 (22:53 -0800)]
Statetrace: Use sys/user.h instead of linux/user.h.
Gabe Black [Thu, 3 Mar 2011 06:53:11 +0000 (22:53 -0800)]
Statetrace: Tweak the help for the -nt option.
Gabe Black [Thu, 3 Mar 2011 06:53:11 +0000 (22:53 -0800)]
Statetrace: Accomodate cross compiling statetrace with scons.
--HG--
rename : util/statetrace/arch/tracechild_amd64.cc => util/statetrace/arch/amd64/tracechild.cc
rename : util/statetrace/arch/tracechild_amd64.hh => util/statetrace/arch/amd64/tracechild.hh
rename : util/statetrace/arch/tracechild_arm.cc => util/statetrace/arch/arm/tracechild.cc
rename : util/statetrace/arch/tracechild_arm.hh => util/statetrace/arch/arm/tracechild.hh
rename : util/statetrace/arch/tracechild_i386.cc => util/statetrace/arch/i386/tracechild.cc
rename : util/statetrace/arch/tracechild_i386.hh => util/statetrace/arch/i386/tracechild.hh
rename : util/statetrace/arch/tracechild_sparc.cc => util/statetrace/arch/sparc/tracechild.cc
rename : util/statetrace/arch/tracechild_sparc.hh => util/statetrace/arch/sparc/tracechild.hh
rename : util/statetrace/tracechild_arch.cc => util/statetrace/base/arch_check.h
rename : util/statetrace/regstate.hh => util/statetrace/base/regstate.hh
rename : util/statetrace/statetrace.cc => util/statetrace/base/statetrace.cc
rename : util/statetrace/tracechild.cc => util/statetrace/base/tracechild.cc
rename : util/statetrace/tracechild.hh => util/statetrace/base/tracechild.hh
Gabe Black [Thu, 3 Mar 2011 06:53:11 +0000 (22:53 -0800)]
Statetrace: Convert the build to scons.
Gabe Black [Thu, 3 Mar 2011 06:53:11 +0000 (22:53 -0800)]
Statetrace: Allow the user to override CXX.
Gabe Black [Thu, 3 Mar 2011 06:53:10 +0000 (22:53 -0800)]
Statetrace: Get rid of explicit register name handling.
Gabe Black [Thu, 3 Mar 2011 06:53:10 +0000 (22:53 -0800)]
Statetrace: Kill the printer functionality in statetrace.
Gabe Black [Thu, 3 Mar 2011 06:53:10 +0000 (22:53 -0800)]
Statetrace: Clean up style.
Gabe Black [Wed, 2 Mar 2011 08:41:44 +0000 (00:41 -0800)]
X86: Use the npc as the pc when doing a nativetrace, not what M5 considers the pc.
Gabe Black [Wed, 2 Mar 2011 08:41:38 +0000 (00:41 -0800)]
X86: Decode the mysterious and elusive ffreep x87 instruction.
The internet says this instruction was created by accident when an Intel CPU
failed to decode x87 instructions properly. It's been documented on a few rare
occasions and has generally worked to ensure backwards compatability. One
source claims that the gcc toolchain is basically the only thing that emits
it, and that emulators/binary translators like qemu and bochs implement it.
We won't actually implement it here since we're hardly implementing any other
x87 instructions either. If we were to implement it, it would behave the same
as ffree but then also pop the register stack.
http://www.pagetable.com/?p=16
Gabe Black [Wed, 2 Mar 2011 07:18:47 +0000 (23:18 -0800)]
Spelling: Fix the a spelling error by changing mmaped to mmapped.
There may not be a formally correct spelling for the past tense of mmap, but
mmapped is the spelling Google doesn't try to autocorrect. This makes sense
because it mirrors the past tense of map->mapped and not the past tense of
cape->caped.
--HG--
rename : src/arch/alpha/mmaped_ipr.hh => src/arch/alpha/mmapped_ipr.hh
rename : src/arch/arm/mmaped_ipr.hh => src/arch/arm/mmapped_ipr.hh
rename : src/arch/mips/mmaped_ipr.hh => src/arch/mips/mmapped_ipr.hh
rename : src/arch/power/mmaped_ipr.hh => src/arch/power/mmapped_ipr.hh
rename : src/arch/sparc/mmaped_ipr.hh => src/arch/sparc/mmapped_ipr.hh
rename : src/arch/x86/mmaped_ipr.hh => src/arch/x86/mmapped_ipr.hh
Gabe Black [Wed, 2 Mar 2011 07:18:00 +0000 (23:18 -0800)]
X86: Update stats for the x86 o3 hello world regression.
Gabe Black [Wed, 2 Mar 2011 06:42:59 +0000 (22:42 -0800)]
X86: Mark IO reads and writes as non-speculative.
Gabe Black [Wed, 2 Mar 2011 06:42:18 +0000 (22:42 -0800)]
X86: Mark prefetches as such in their instruction and request flags.
Nilay Vaish [Tue, 1 Mar 2011 21:26:11 +0000 (15:26 -0600)]
Ruby: Fix DPRINTF bugs in PerfectSwitch and MessageBuffer
At a couple of places in PerfectSwitch.cc and MessageBuffer.cc, DPRINTF()
has not been provided with correct number of arguments. The patch fixes these
bugs.
Gabe Black [Tue, 1 Mar 2011 11:00:42 +0000 (03:00 -0800)]
SCons: Separately label the global non-sticky options.
The global sticky options were being printed with a heading, and then the
global nonsticky options were being printed immediately after them without a
heading. Because the two lists ran together and the first had its own heading,
it looked like -all- those options where sticky even though some of them
aren't. This change adds a label to the second list so it's clear they're
different.
Gabe Black [Tue, 1 Mar 2011 10:59:09 +0000 (02:59 -0800)]
Ruby: Mention that Ruby's bound checking option only applies to Ruby.
Gabe Black [Mon, 28 Feb 2011 00:25:06 +0000 (16:25 -0800)]
X86: If PCI config space is disabled, pass through to regular IO addresses.
Gabe Black [Mon, 28 Feb 2011 00:24:54 +0000 (16:24 -0800)]
X86: Update X86_FS stats.
Gabe Black [Mon, 28 Feb 2011 00:24:10 +0000 (16:24 -0800)]
X86: Use regular read requests in the walker instead of read exclusive.
Korey Sewell [Sun, 27 Feb 2011 19:17:26 +0000 (14:17 -0500)]
inorder: bzip2 regression update
Nathan Binkert [Sun, 27 Feb 2011 05:43:11 +0000 (21:43 -0800)]
getopt: Remove GPL code.
This code is unused and should never have been committed
Nilay Vaish [Fri, 25 Feb 2011 23:55:20 +0000 (17:55 -0600)]
Ruby: Remove store buffer
This patch removes the store buffer from Ruby. It is not in use currently.
Since libruby is being and store buffer makes calls to libruby, it is not
possible to maintain it until substantial changes are made.
Nilay Vaish [Fri, 25 Feb 2011 23:54:56 +0000 (17:54 -0600)]
Ruby: Remove libruby
This patch removes libruby_internal.hh, libruby.hh and libruby.cc. It moves
the contents to libruby.hh to RubyRequest.hh and RubyRequest.cc files.
Nilay Vaish [Fri, 25 Feb 2011 23:51:56 +0000 (17:51 -0600)]
Ruby: Make Address.hh independent of RubySystem
This patch changes Address.hh so that it is not dependent on RubySystem.
This dependence seems unecessary. All those functions that depend on
RubySystem have been moved to Address.cc file.
Nilay Vaish [Fri, 25 Feb 2011 23:51:02 +0000 (17:51 -0600)]
Ruby: Make DataBlock.hh independent of RubySystem
This patch changes DataBlock.hh so that it is not dependent on RubySystem.
This dependence seems unecessary. All those functions that depende on
RubySystem have been moved to DataBlock.cc file.
Timothy M. Jones [Fri, 25 Feb 2011 13:50:29 +0000 (13:50 +0000)]
O3CPU: Fix iqCount and lsqCount SMT fetch policies.
Fixes two of the SMT fetch policies in O3CPU that were returning the count
of instructions in the IQ or LSQ rather than the thread ID to fetch from.
Gabe Black [Thu, 24 Feb 2011 10:14:45 +0000 (02:14 -0800)]
Configs: Explicitly import env in Benchmarks.py
env was being implicitly imported into Benchmarks.py through SysPaths.py.
This change brings it in explicitly in the file where it's used.
Brad Beckmann [Thu, 24 Feb 2011 00:41:59 +0000 (16:41 -0800)]
regress: MOESI_hammer memtest updates
Brad Beckmann [Thu, 24 Feb 2011 00:41:59 +0000 (16:41 -0800)]
ruby: automate permission setting
This patch integrates permissions with cache and memory states, and then
automates the setting of permissions within the generated code. No longer
does one need to manually set the permissions within the setState funciton.
This patch will faciliate easier functional access support by always correctly
setting permissions for both cache and memory states.
--HG--
rename : src/mem/slicc/ast/EnumDeclAST.py => src/mem/slicc/ast/StateDeclAST.py
rename : src/mem/slicc/ast/TypeFieldEnumAST.py => src/mem/slicc/ast/TypeFieldStateAST.py
Brad Beckmann [Thu, 24 Feb 2011 00:41:58 +0000 (16:41 -0800)]
MOESI_hammer: cache probe address clean up
Brad Beckmann [Thu, 24 Feb 2011 00:41:58 +0000 (16:41 -0800)]
ruby: cleaned up access permission enum
Brad Beckmann [Thu, 24 Feb 2011 00:41:26 +0000 (16:41 -0800)]
ruby: removed unsupported protocol files
Korey Sewell [Wed, 23 Feb 2011 21:35:25 +0000 (16:35 -0500)]
inorder: add 00.gzip and 60.bzip2 regression tests
Korey Sewell [Wed, 23 Feb 2011 21:35:18 +0000 (16:35 -0500)]
inorder: InstSeqNum bug
Because int and not InstSeqNum was used in a couple of places, you can
overflow the int type and thus get wierd bugs when the sequence number
is negative (or some wierd value)
Korey Sewell [Wed, 23 Feb 2011 21:35:04 +0000 (16:35 -0500)]
inorder: dyn inst initialization
remove constructors that werent being used (it just gets confusing)
use initialization list for all the variables instead of relying on initVars()
function
Korey Sewell [Wed, 23 Feb 2011 21:30:45 +0000 (16:30 -0500)]
inorder: cache packet handling
-use a pointer to CacheReqPacket instead of PacketPtr so correct destructors
get called on packet deletion
- make sure to delete the packet if the cache blocks the sendTiming request
or for some reason we dont use the packet
- dont overwrite memory requests since in the worst case an instruction will
be replaying a request so no need to keep allocating a new request
- we dont use retryPkt so delete it
- fetch code was split out already, so just assert that this is a memory
reference inst. and that the staticInst is available
Ali Saidi [Wed, 23 Feb 2011 21:10:50 +0000 (15:10 -0600)]
ARM: Update regression tests for preceeding changes.
Ali Saidi [Wed, 23 Feb 2011 21:10:50 +0000 (15:10 -0600)]
Mem: Print out memory when access > 8 bytes
Ali Saidi [Wed, 23 Feb 2011 21:10:50 +0000 (15:10 -0600)]
ARM: Set ITSTATE correctly after FlushPipe
Ali Saidi [Wed, 23 Feb 2011 21:10:50 +0000 (15:10 -0600)]
ARM: This panic can be hit during misspeculation so it can't exist.
Ali Saidi [Wed, 23 Feb 2011 21:10:50 +0000 (15:10 -0600)]
ARM: Bad interworking warn way to noisy when running real code w/misspeculation.
Ali Saidi [Wed, 23 Feb 2011 21:10:50 +0000 (15:10 -0600)]
O3: When a prefetch causes a fault, don't record it in the inst
Giacomo Gabrielli [Wed, 23 Feb 2011 21:10:50 +0000 (15:10 -0600)]
ARM: NEON instruction templates modified to set the predicate flag to false when needed.
Ali Saidi [Wed, 23 Feb 2011 21:10:49 +0000 (15:10 -0600)]
O3: If there is an outstanding table walk don't let the inst queue sleep.
If there is an outstanding table walk and no other activity in the CPU
it can go to sleep and never wake up. This change makes the instruction
queue always active if the CPU is waiting for a store to translate.
If Gabe changes the way this code works then the below should be removed
as indicated by the todo.
Ali Saidi [Wed, 23 Feb 2011 21:10:49 +0000 (15:10 -0600)]
ARM: Squash state on FPSCR stride or len write.
Matt Horsnell [Wed, 23 Feb 2011 21:10:49 +0000 (15:10 -0600)]
ARM: Mark store conditionals as such.
Ali Saidi [Wed, 23 Feb 2011 21:10:49 +0000 (15:10 -0600)]
ARM: Do something for ISB, DSB, DMB
Ali Saidi [Wed, 23 Feb 2011 21:10:49 +0000 (15:10 -0600)]
ARM: Fix bug that let two table walks occur in parallel.
Ali Saidi [Wed, 23 Feb 2011 21:10:49 +0000 (15:10 -0600)]
Includes: Don't include isa_traits.hh and use the TheISA namespace unless really needed.
Ali Saidi [Wed, 23 Feb 2011 21:10:49 +0000 (15:10 -0600)]
ARM: Make Noop actually decode to a noop and set it's instflags.
Ali Saidi [Wed, 23 Feb 2011 21:10:49 +0000 (15:10 -0600)]
O3: Fix bug when a squash occurs right before TLB miss returns.
In this case we need to throw away the TLB miss, not assume it was the
one we were waiting for.
Ali Saidi [Wed, 23 Feb 2011 21:10:48 +0000 (15:10 -0600)]
ARM: Delete OABI syscall handling.
We only support EABI binaries, so there is no reason to support OABI syscalls.
The loader detects OABI calls and fatal() so there is no reason to even check
here.
Ali Saidi [Wed, 23 Feb 2011 21:10:48 +0000 (15:10 -0600)]
CLCD: Fix some serialization bugs with the clcd controller.
Ali Saidi [Wed, 23 Feb 2011 21:10:48 +0000 (15:10 -0600)]
ARM: Clarifies creation of Linux and baremetal ARM systems.
makeArmSystem creates both bare-metal and Linux systems more cleanly.
machine_type was never optional though listed as an optional argument; a system
such as "RealView_PBX" must now be explicitly specified. Now that it is a
required argument, the placement of the arguments has changed slightly
requiring some changes to calls that create ARM systems.
Ali Saidi [Wed, 23 Feb 2011 21:10:48 +0000 (15:10 -0600)]
ARM: Add support for read of 100MHz clock in system controller.
Ali Saidi [Wed, 23 Feb 2011 21:10:48 +0000 (15:10 -0600)]
ARM: Reset simulation statistics when pref counters are reset.
The ARM performance counters are not currently supported by the model.
This patch interprets a 'reset performance counters' command to mean 'reset
the simulator statistics' instead.
Ali Saidi [Wed, 23 Feb 2011 21:10:48 +0000 (15:10 -0600)]
ARM: Adds dummy support for a L2 latency miscreg.
Korey Sewell [Wed, 23 Feb 2011 19:26:55 +0000 (14:26 -0500)]
configs: cache: add cache line size option
Korey Sewell [Wed, 23 Feb 2011 06:01:46 +0000 (01:01 -0500)]
configs: set default cache params
It's confusing (especially to new users), when you are setting some standard
parameters (as defined in Options.py) and they aren't reflected in the simulations
so we might as well link the settings in CacheConfig.py to those in Options.py
Korey Sewell [Wed, 23 Feb 2011 05:58:42 +0000 (00:58 -0500)]
ruby: extend dprintfs for RubyGenerated TraceFlag
"executing" isnt a very descriptive debug message and in going through the
output you get multiple messages that say "executing" but nothing to help
you parse through the code/execution.
So instead, at least print out the name of the action that is taking
place in these functions.
Korey Sewell [Wed, 23 Feb 2011 05:58:40 +0000 (00:58 -0500)]
ruby: cleaning up RubyQueue and RubyNetwork dprintfs
Overall, continue to progress Ruby debug messages to more of the normal M5
debug message style
- add a name() to the Ruby Throttle & PerfectSwitch objects so that the debug output
isn't littered w/"global:" everywhere.
- clean up messages that print over multiple lines when possible
- clean up duplicate prints in the message buffer
Brad Beckmann [Tue, 22 Feb 2011 19:16:40 +0000 (11:16 -0800)]
m5: merged in hammer fix
Nilay Vaish [Sat, 19 Feb 2011 23:32:43 +0000 (17:32 -0600)]
Ruby: Machine Type missing in MOESI CMP directory protocol
In certain actions of the L1 cache controller, while creating an outgoing
message, the machine type was not being set. This results in a
segmentation fault when trace is collected. Joseph Pusudesris provided
his patch for fixing this issue.
Nilay Vaish [Sat, 19 Feb 2011 23:32:00 +0000 (17:32 -0600)]
Ruby: clean MOESI CMP directory protocol
The L1 cache controller file contains references to foo and goo queues, which
are not in use at all. These have been removed.
Korey Sewell [Fri, 18 Feb 2011 19:35:15 +0000 (14:35 -0500)]
m5: merge inorder/release-notes/make_release changes
Korey Sewell [Fri, 18 Feb 2011 19:31:37 +0000 (14:31 -0500)]
inorder: regr-update: reduce dynamic mem. use to speedup sims
previous changesets took a closer look at memory mgmt in the inorder model and sought to avoid
dynamic memory mgmt (for access to pipeline resources) as much as possible. For the regressions
that were run, the sims are about 2x speedup from changeset 7726 which is the last change
since the recent commits in Feb. (note: these regressions now are 4-issue CPUs instead of just 1-issue)
Korey Sewell [Fri, 18 Feb 2011 19:31:31 +0000 (14:31 -0500)]
inorder: add names and slot #s to res. dprints
Korey Sewell [Fri, 18 Feb 2011 19:30:38 +0000 (14:30 -0500)]
inorder: ignore nops in execution unit
Korey Sewell [Fri, 18 Feb 2011 19:30:05 +0000 (14:30 -0500)]
inorder: update graduation unit
make sure instructions are able to commit before writing back to the RF
do not commit more than 1 non-speculative instruction per cycle
Korey Sewell [Fri, 18 Feb 2011 19:29:48 +0000 (14:29 -0500)]
inorder: recognize isSerializeAfter flag
keep track of when an instruction needs the execution
behind it to be serialized. Without this, in SE Mode
instructions can execute behind a system call exit().
Korey Sewell [Fri, 18 Feb 2011 19:29:44 +0000 (14:29 -0500)]
inorder: update default thread size(=1)
a lot of structures get allocated based off that MaxThreads parameter so this is an
effort to not abuse it
Korey Sewell [Fri, 18 Feb 2011 19:29:40 +0000 (14:29 -0500)]
inorder: don't overuse getLatency()
resources don't need to call getLatency because the latency is already a member
in the class. If there is some type of special case where different instructions
impose a different latency inside a resource then we can revisit this and
add getLatency() back in
Korey Sewell [Fri, 18 Feb 2011 19:29:31 +0000 (14:29 -0500)]
inorder: update max. resource bandwidths
each resource has a certain # of requests it can take per cycle. update the #s here
to be more realistic based off of the pipeline width and if the resource needs to
be accessed on multiple cycles
Korey Sewell [Fri, 18 Feb 2011 19:29:26 +0000 (14:29 -0500)]
inorder: cleanup in destructors
cleanup hanging pointers and other cruft in the destructors
Korey Sewell [Fri, 18 Feb 2011 19:29:17 +0000 (14:29 -0500)]
inorder: fix cache/fetch unit memory leaks
---
need to delete the cache request's data on clearRequest() now that we are recycling
requests
---
fetch unit needs to deallocate the fetch buffer blocks when they are replaced or
squashed.
Korey Sewell [Fri, 18 Feb 2011 19:29:02 +0000 (14:29 -0500)]
inorder: remove events for zero-cycle resources
if a resource has a zero cycle latency (e.g. RegFile write), then dont allocate an event
for it to use
Korey Sewell [Fri, 18 Feb 2011 19:28:37 +0000 (14:28 -0500)]
inorder: update pipeline interface for handling finished resource reqs
formerly, to free up bandwidth in a resource, we could just change the pointer in that resource
but at the same time the pipeline stages had visibility to see what happened to a resource request.
Now that we are recycling these requests (to avoid too much dynamic allocation), we can't throw
away the request too early or the pipeline stage gets bad information. Instead, mark when a request
is done with the resource all together and then let the pipeline stage call back to the resource
that it's time to free up the bandwidth for more instructions
*** inteface notes ***
- When an instruction completes and is done in a resource for that cycle, call done()
- When an instruction fails and is done with a resource for that cycle, call done(false)
- When an instruction completes, but isnt finished with a resource, call completed()
- When an instruction fails, but isnt finished with a resource, call completed(false)
* * *
inorder: tlbmiss wakeup bug fix
Korey Sewell [Fri, 18 Feb 2011 19:28:30 +0000 (14:28 -0500)]
inorder: remove request map, use request vector
take away all instances of reqMap in the code and make all references use the built-in
request vectors inside of each resource. The request map was dynamically allocating
a request per instruction. The request vector just allocates N number of requests
during instantiation and then the surrounding code is fixed up to reuse those N requests
***
setRequest() and clearRequest() are the new accessors needed to define a new
request in a resource
Korey Sewell [Fri, 18 Feb 2011 19:28:22 +0000 (14:28 -0500)]
inorder: add valid bit for resource requests
this will allow us to reuse resource requests within a resource instead
of always dynamically allocating
Korey Sewell [Fri, 18 Feb 2011 19:28:10 +0000 (14:28 -0500)]
inorder: remove reqRemoveList
we are going to be getting away from creating new resource requests for every
instruction so no more need to keep track of a reqRemoveList and clean it up
every tick
Korey Sewell [Fri, 18 Feb 2011 19:27:52 +0000 (14:27 -0500)]
inorder: initialize res. req. vectors based on resource bandwidth
first change in an optimization that will stop InOrder from allocating new memory for every instruction's
request to a resource. This gets expensive since every instruction needs to access ~10 requests before
graduation. Instead, the plan is to allocate just enough resource request objects to satisfy each resource's
bandwidth (e.g. the execution unit would need to allocate 3 resource request objects for a 1-issue pipeline
since on any given cycle it could have 2 read requests and 1 write request) and then let the instructions
contend and reuse those allocated requests. The end result is a smaller memory footprint for the InOrder model
and increased simulation performance
Nathan Binkert [Wed, 16 Feb 2011 15:57:04 +0000 (10:57 -0500)]
merge alpha system files into tree
Gabe Black [Wed, 16 Feb 2011 07:22:32 +0000 (23:22 -0800)]
Util: Get rid of the make_release.py script.
Since we're not doing releases any more we don't really need this script. If
we need it in the future, we can resurrect it from the history.
Nathan Binkert [Wed, 16 Feb 2011 06:34:02 +0000 (00:34 -0600)]
Cleanup system directory to fit into modern M5 tree
Nathan Binkert [Wed, 16 Feb 2011 06:34:01 +0000 (00:34 -0600)]
copyright: update copyright on alpha system files
Gabe Black [Tue, 15 Feb 2011 23:58:16 +0000 (15:58 -0800)]
X86: Get rid of "inline" on the MicroPanic constructor in decoder.cc.
This was making certain versions of gcc omit the function from the object file
which would break the build.
Gabe Black [Tue, 15 Feb 2011 05:36:37 +0000 (21:36 -0800)]
Info: Clean up some info files.
Get rid of RELEASE_NOTES since we no longer do releases, update some of the
information in README, and update the date in LICENSE.
Nilay Vaish [Mon, 14 Feb 2011 22:14:54 +0000 (16:14 -0600)]
Ruby: Improve Change PerfectSwitch's wakeup function
Currently the wakeup function for the PerfectSwitch contains three loops -
loop on number of virtual networks
loop on number of incoming links
loop till all messages for this (link, network) have been routed
With an 8 processor mesh network and Hammer protocol, about 11-12% of the
was observed to have been spent in this function, which is the highest
amongst all the functions. It was found that the innermost loop is executed
about 45 times per invocation of the wakeup function, when each invocation
of the wakeup function processes just about one message.
The patch tries to do away with the redundant executions of the innermost
loop. Counters have been added for each virtual network that record the
number of messages that need to be routed for that virtual network. The
inner loops are only executed when the number of messages for that particular
virtual network > 0. This does away with almost 80% of the executions of the
innermost loop. The function now consumes about 5-6% of the total execution
time.