gem5.git
12 years agoSLICC: Put functions of a controller in its .cc file
Nilay Vaish [Thu, 28 Jul 2011 01:20:53 +0000 (20:20 -0500)]
SLICC: Put functions of a controller in its .cc file
Currently, functions associated with a controller go into separate files.
This patch puts all the functions in the controller's .cc file. This should
hopefully take away some time from compilation.

12 years agoRuby: Fix instantiations of DMA controller and sequencer
Nilay Vaish [Tue, 26 Jul 2011 17:20:22 +0000 (12:20 -0500)]
Ruby: Fix instantiations of DMA controller and sequencer
The patch on Ruby functional accesses made changes to the process of
instantiating controllers and sequencers. The DMA controller and
sequencer was not updated, hence this patch.

12 years agoMerged with Gabe's changeset.
Nilay Vaish [Mon, 25 Jul 2011 23:31:30 +0000 (18:31 -0500)]
Merged with Gabe's changeset.

12 years agoRuby: Fix dma controller configs/ruby/MI_example.py
Nilay Vaish [Mon, 25 Jul 2011 23:18:31 +0000 (18:18 -0500)]
Ruby: Fix dma controller configs/ruby/MI_example.py
The dma controller in configs/ruby/MI_example.py was not being set correctly.
This patch fixes it.

12 years agoSCons: Only print all the SConsopts being read if verbose is turned on.
Gabe Black [Tue, 19 Jul 2011 09:56:02 +0000 (02:56 -0700)]
SCons: Only print all the SConsopts being read if verbose is turned on.

12 years agoinorder-fs: temp. regression removal
Korey Sewell [Sat, 16 Jul 2011 01:26:18 +0000 (21:26 -0400)]
inorder-fs: temp. regression removal
remove this regression till the fix for the hwrei instruction is put in

12 years agoMem: Fix issue with prefetches originating at non-L1 caches getting stale data
Ali Saidi [Fri, 15 Jul 2011 16:53:35 +0000 (11:53 -0500)]
Mem: Fix issue with prefetches originating at non-L1 caches getting stale data

Prefetch requests issued from the L2 or below wouldn't check if valid data is
present higher in the system. If a prefetch into the L2 occured at the same
time as writeback from a higher-level cache the dirty data could be replaced
in by unmodified data in memory.

12 years agoO3: Create a pipeline activity viewer for the O3 CPU model.
Giacomo Gabrielli [Fri, 15 Jul 2011 16:53:35 +0000 (11:53 -0500)]
O3: Create a pipeline activity viewer for the O3 CPU model.

Implemented a pipeline activity viewer as a python script (util/o3-pipeview.py)
and modified O3 code base to support an extra trace flag (O3PipeView) for
generating traces to be used as inputs by the tool.

12 years agoARM: Update stats for better miscreg support for MP configurations.
Ali Saidi [Fri, 15 Jul 2011 16:53:35 +0000 (11:53 -0500)]
ARM: Update stats for better miscreg support for MP configurations.

12 years agoARM: Fix SWP/SWPB undefined instruction behavior
Wade Walker [Fri, 15 Jul 2011 16:53:34 +0000 (11:53 -0500)]
ARM: Fix SWP/SWPB undefined instruction behavior

SWP and SWPB now throw an undefined instruction exception if
SCTLR.SW == 0. This also required the MIDR to be changed
slightly so programs can correctly determine that gem5 supports
the ARM v7 behavior of SWP/SWPB (in ARM v6, SWP/SWPB were
deprecated, but not disabled at CPU startup).

12 years agoARM: Add two unimplemented miscellaneous registers.
Wade Walker [Fri, 15 Jul 2011 16:53:34 +0000 (11:53 -0500)]
ARM: Add two unimplemented miscellaneous registers.

Adds MISCREG_ID_MMFR2 and removes break on access to MISCREG_CLIDR. Both
registers now return values that are consistent with current ARM
implementations.

13 years agose.py: Fixes the way ruby's options are added
Nilay Vaish [Tue, 12 Jul 2011 00:57:10 +0000 (19:57 -0500)]
se.py: Fixes the way ruby's options are added

13 years agoX86: implements copyRegs() function
Nilay Vaish [Mon, 11 Jul 2011 21:52:52 +0000 (16:52 -0500)]
X86: implements copyRegs() function
This patch implements the copyRegs() function for the x86 architecture.
The patch assumes that no side effects other than TLB invalidation need
to be considered while copying the registers. This may not hold true in
future.

13 years agoISA: Get rid of the unused mem_acc_type template parameter.
Gabe Black [Mon, 11 Jul 2011 11:47:06 +0000 (04:47 -0700)]
ISA: Get rid of the unused mem_acc_type template parameter.

13 years agoO3: Update stats for fetch and bp changes.
Ali Saidi [Sun, 10 Jul 2011 17:56:09 +0000 (12:56 -0500)]
O3: Update stats for fetch and bp changes.

13 years agoBranch predictor: Fixes the tournament branch predictor.
Mrinmoy Ghosh [Sun, 10 Jul 2011 17:56:08 +0000 (12:56 -0500)]
Branch predictor: Fixes the tournament branch predictor.

Branch predictor could not predict a branch in a nested loop because:
 1. The global history was not updated after a mispredict squash.
 2. The global history was updated in the fetch stage. The choice predictors
    that were updated  used the changed global history. This is incorrect, as
    it incorporates the state of global history after the branch in
    encountered. Fixed update to choice predictor using the global history
    state before the branch happened.
 3. The global predictor table was also updated using the global history state
    before the branch happened as above.

Additionally, parameters to initialize ctr and history size were reversed.

13 years agoO3: Fix up pipelining icache accesses in fetch stage to function properly
Geoffrey Blake [Sun, 10 Jul 2011 17:56:08 +0000 (12:56 -0500)]
O3: Fix up pipelining icache accesses in fetch stage to function properly

Fixed up the patch from Yasuko Watanabe that enabled pipelining of fetch accessess to
icache to work with recent changes to main repository.
Also added in ability for fetch stage to delay issuing the fault carrying
nop when a pipeline fetch causes a fault and no fetch bandwidth is available
until the next cycle.

13 years agoIO: Handle case where ISA Fake device is being used as a fake memory.
Ali Saidi [Sun, 10 Jul 2011 17:56:08 +0000 (12:56 -0500)]
IO: Handle case where ISA Fake device is being used as a fake memory.

13 years agoO3: Make sure fetch doesn't go off into the weeds during speculation.
Ali Saidi [Sun, 10 Jul 2011 17:56:08 +0000 (12:56 -0500)]
O3: Make sure fetch doesn't go off into the weeds during speculation.

13 years agoConfig: Add support for a Self.all proxy object
Ali Saidi [Sun, 10 Jul 2011 17:56:08 +0000 (12:56 -0500)]
Config: Add support for a Self.all proxy object

13 years agoARM: Fix mp interrupt bug in GIC.
Daniel Johnson [Sun, 10 Jul 2011 17:56:08 +0000 (12:56 -0500)]
ARM: Fix mp interrupt bug in GIC.

Missing "!" made multiprocessor interrupts operate incorrectly.

13 years agoalpha:hwrei:rollback for o3
Korey Sewell [Fri, 8 Jul 2011 01:32:49 +0000 (21:32 -0400)]
alpha:hwrei:rollback for o3
change hwrei back to being a non-control instruction so O3-FS mode will work
add squash in inorder that will catch a hwrei (or any other genric instruction)
that isnt a control inst but changes the PC. Additional testing still needs to be done
for inorder-FS mode but this change will free O3 development back up in the interim

13 years agoruby: added generic dma machine
Brad Beckmann [Thu, 7 Jul 2011 01:45:15 +0000 (18:45 -0700)]
ruby: added generic dma machine

13 years agoMOESI_hammer: Fixed uniprocessor DMA bug
Brad Beckmann [Thu, 7 Jul 2011 01:44:42 +0000 (18:44 -0700)]
MOESI_hammer: Fixed uniprocessor DMA bug

13 years agoslicc: add a protocol statement and an include statement
Nathan Binkert [Wed, 6 Jul 2011 01:30:05 +0000 (18:30 -0700)]
slicc: add a protocol statement and an include statement
All protocols must specify their name
The include statement allows any file to include another file.

13 years agoslicc: cleanup slicc code and make it less verbose
Nathan Binkert [Wed, 6 Jul 2011 01:30:05 +0000 (18:30 -0700)]
slicc: cleanup slicc code and make it less verbose

13 years agogrammar: better encapsulation of a grammar and parsing
Nathan Binkert [Wed, 6 Jul 2011 01:30:04 +0000 (18:30 -0700)]
grammar: better encapsulation of a grammar and parsing
This makes it possible to use the grammar multiple times and use the multiple
instances concurrently.  This makes implementing an include statement as part
of a grammar possible.

13 years agoX86: Add a config for an FS regression on O3.
Gabe Black [Wed, 6 Jul 2011 00:46:46 +0000 (17:46 -0700)]
X86: Add a config for an FS regression on O3.

13 years agoISAs: Streamline some spots where Mem is used in the ISA descriptions.
Gabe Black [Tue, 5 Jul 2011 23:52:57 +0000 (16:52 -0700)]
ISAs: Streamline some spots where Mem is used in the ISA descriptions.

13 years agoISA parser: Define operand types with a ctype directly.
Gabe Black [Tue, 5 Jul 2011 23:52:15 +0000 (16:52 -0700)]
ISA parser: Define operand types with a ctype directly.

13 years agoISA parser: Simplify operand type handling.
Gabe Black [Tue, 5 Jul 2011 23:48:18 +0000 (16:48 -0700)]
ISA parser: Simplify operand type handling.

This change simplifies the code surrounding operand type handling and makes it
depend only on the ctype that goes with each operand type. Future changes will
allow defining operand types by their ctypes directly, convert the ISAs over
to that style of definition, and then remove support for the old style. These
changes are to make it easier to use non-builtin types like classes or
structures as the type for operands.

13 years agoMerged with Gabe's recent changes.
Nilay Vaish [Sun, 3 Jul 2011 16:38:25 +0000 (11:38 -0500)]
Merged with Gabe's recent changes.

13 years agoNetwork_test: Conform it with functional access changes in Ruby
Nilay Vaish [Sun, 3 Jul 2011 16:33:46 +0000 (11:33 -0500)]
Network_test: Conform it with functional access changes in Ruby
Addition of functional access support to Ruby necessitated some changes to
the way coherence protocols are written. I had forgotten to update the
Network_test protocol. This patch makes those updates.

13 years agotracediff: Check for --debug-flags now instead of --trace-flags.
Gabe Black [Sun, 3 Jul 2011 05:52:26 +0000 (22:52 -0700)]
tracediff: Check for --debug-flags now instead of --trace-flags.

13 years agoExecContext: Rename the readBytes/writeBytes functions to readMem and writeMem.
Gabe Black [Sun, 3 Jul 2011 05:35:04 +0000 (22:35 -0700)]
ExecContext: Rename the readBytes/writeBytes functions to readMem and writeMem.

readBytes and writeBytes had the word "bytes" in their names because they
accessed blobs of bytes. This distinguished them from the read and write
functions which handled higher level data types. Because those functions don't
exist any more, this change renames readBytes and writeBytes to more general
names, readMem and writeMem, which reflect the fact that they are how you read
and write memory. This also makes their names more consistent with the
register reading/writing functions, although those are still read and set for
some reason.

13 years agoExecContext: Get rid of the now unused read/write templated functions.
Gabe Black [Sun, 3 Jul 2011 05:34:58 +0000 (22:34 -0700)]
ExecContext: Get rid of the now unused read/write templated functions.

13 years agoISA: Use readBytes/writeBytes for all instruction level memory operations.
Gabe Black [Sun, 3 Jul 2011 05:34:29 +0000 (22:34 -0700)]
ISA: Use readBytes/writeBytes for all instruction level memory operations.

13 years agoStats: Update stats for the x86 store fault fix.
Gabe Black [Sun, 3 Jul 2011 05:31:42 +0000 (22:31 -0700)]
Stats: Update stats for the x86 store fault fix.

13 years agoX86: Fix store microops so they don't drop faults in timing mode.
Gabe Black [Sun, 3 Jul 2011 05:31:22 +0000 (22:31 -0700)]
X86: Fix store microops so they don't drop faults in timing mode.

If a fault was returned by the CPU when a store initiated it's write, the
store instruction would ignore the fault. This change fixes that.

13 years agoRuby: Commit files missing from previous commit
Nilay Vaish [Fri, 1 Jul 2011 21:29:33 +0000 (16:29 -0500)]
Ruby: Commit files missing from previous commit
The previous commit on functional access support in Ruby did not have
some of the files required. This patch adds those files to the repository.

13 years agoRegression: Updates regression outputs for Ruby memtest
Brad Beckmann [Fri, 1 Jul 2011 00:57:26 +0000 (19:57 -0500)]
Regression: Updates regression outputs for Ruby memtest
This patch updates the regression outputs for Ruby memtest. This was
required because of the changes carried out by the addition of functional
access support to Ruby.

13 years agoconfig: removed unnecessary slashes
Brad Beckmann [Fri, 1 Jul 2011 00:54:02 +0000 (19:54 -0500)]
config: removed unnecessary slashes
This patch removes unnecessary slashes from a couple of python scripts.

13 years agoRuby: Add support for functional accesses
Brad Beckmann ext:(%2C%20Nilay%20Vaish%20%3Cnilay%40cs.wisc.edu%3E) [Fri, 1 Jul 2011 00:49:26 +0000 (19:49 -0500)]
Ruby: Add support for functional accesses
This patch rpovides functional access support in Ruby. Currently only
the M5Port of RubyPort supports functional accesses. The support for
functional through the PioPort will be added as a separate patch.

13 years agoarch: print next upc correctly
Nilay Vaish [Tue, 28 Jun 2011 23:27:38 +0000 (18:27 -0500)]
arch: print next upc correctly
The patch corrects the print statement which prints the current and
the next pc. Instead of the next upc, the next pc was being printed.

13 years agoRuby: remove unused functions in CacheMemory: get/setMemoryValue
Joel Hestness [Fri, 24 Jun 2011 20:47:35 +0000 (15:47 -0500)]
Ruby: remove unused functions in CacheMemory: get/setMemoryValue

13 years agomips: fix nmsub and nmadd definitions
Deyaun Guo [Thu, 23 Jun 2011 03:35:21 +0000 (23:35 -0400)]
mips: fix nmsub and nmadd definitions
the -/+ signs were flipped for nmsub_s, nmsub_d, and nmadd_d

13 years agoX86: Eliminate an unused argument for building store microops.
Gabe Black [Wed, 22 Jun 2011 02:28:14 +0000 (19:28 -0700)]
X86: Eliminate an unused argument for building store microops.

13 years agoinorder: sparc: add 02.insttest regression
Korey Sewell [Tue, 21 Jun 2011 02:44:24 +0000 (22:44 -0400)]
inorder: sparc: add 02.insttest regression

13 years agoinorder: sparc: add hello world regression
Korey Sewell [Tue, 21 Jun 2011 02:44:22 +0000 (22:44 -0400)]
inorder: sparc: add hello world regression
- add InOrderCPU compile option to SPARC
- add hello regression for SPARC

13 years agomerge regression updates
Korey Sewell [Mon, 20 Jun 2011 22:58:31 +0000 (18:58 -0400)]
merge regression updates

13 years agoalpha:o3:simple: update simout/err files
Korey Sewell [Mon, 20 Jun 2011 22:57:14 +0000 (18:57 -0400)]
alpha:o3:simple: update simout/err files
A few prior changesets have changed the gem5 output in a way that wont cause
errors but may be confusing for someone trying to debug the regressions. Ones that I caught
were:
- no more "warn: <hash address"
- typo in the ALPHA Prefetch unimplemented warning

Additionaly, the last updated stats changes rearrange the ordering of the stats output even though
they are still correct stats (gem5 is smart enough to detect this). All the regressions pass
w/the same stats even though it looks like they are being changed.

13 years agoinorder: alpha-hello regression update
Korey Sewell [Mon, 20 Jun 2011 16:21:10 +0000 (12:21 -0400)]
inorder: alpha-hello regression update

13 years agoinorder: merge gabes compile fix
Korey Sewell [Mon, 20 Jun 2011 12:37:25 +0000 (08:37 -0400)]
inorder: merge gabes compile fix

13 years agoInOder: Fix a compile error.
Gabe Black [Mon, 20 Jun 2011 09:29:14 +0000 (02:29 -0700)]
InOder: Fix a compile error.

13 years agoinorder: gem5.opt compile
Korey Sewell [Mon, 20 Jun 2011 03:26:36 +0000 (23:26 -0400)]
inorder: gem5.opt compile
variable name typo.

13 years agoinorder: update eon regr w/eon info
Korey Sewell [Mon, 20 Jun 2011 01:54:53 +0000 (21:54 -0400)]
inorder: update eon regr w/eon info
previous commit copied over O3 stats, this one puts the inorder ones in the right place

13 years agoinorder: add 10.linux-boot regression
Korey Sewell [Mon, 20 Jun 2011 01:43:43 +0000 (21:43 -0400)]
inorder: add 10.linux-boot regression

13 years agoinorder: add eon regression
Korey Sewell [Mon, 20 Jun 2011 01:43:43 +0000 (21:43 -0400)]
inorder: add eon regression

13 years agoinorder: update SE regressions
Korey Sewell [Mon, 20 Jun 2011 01:43:42 +0000 (21:43 -0400)]
inorder: update SE regressions

13 years agoinorder: clear reg. dep entry after removing from list
Korey Sewell [Mon, 20 Jun 2011 01:43:42 +0000 (21:43 -0400)]
inorder: clear reg. dep entry after removing from list
this will safeguard future code from trying to remove
from the list twice. That code wouldnt break but would
waste time.

13 years agoinorder: se: squash after syscalls
Korey Sewell [Mon, 20 Jun 2011 01:43:42 +0000 (21:43 -0400)]
inorder: se: squash after syscalls

13 years agoinorder: cleanup dprintfs in cache unit
Korey Sewell [Mon, 20 Jun 2011 01:43:42 +0000 (21:43 -0400)]
inorder: cleanup dprintfs in cache unit

13 years agoinorder: SE mode TLB faults
Korey Sewell [Mon, 20 Jun 2011 01:43:42 +0000 (21:43 -0400)]
inorder: SE mode TLB faults
handle them like we do in FS mode, by blocking the TLB until the fault
is handled by the fault->invoke()

13 years agoinorder:tracing: fix fault tracing bug
Korey Sewell [Mon, 20 Jun 2011 01:43:42 +0000 (21:43 -0400)]
inorder:tracing: fix fault tracing bug

13 years agoinorder: se compile fixes
Korey Sewell [Mon, 20 Jun 2011 01:43:42 +0000 (21:43 -0400)]
inorder: se compile fixes

13 years agoinorder: add necessary debug flag header files
Korey Sewell [Mon, 20 Jun 2011 01:43:41 +0000 (21:43 -0400)]
inorder: add necessary debug flag header files

13 years agomips: mark unaligned access flag as true
Korey Sewell [Mon, 20 Jun 2011 01:43:41 +0000 (21:43 -0400)]
mips: mark unaligned access flag as true

13 years agoinorder: clear fetchbuffer on traps
Korey Sewell [Mon, 20 Jun 2011 01:43:41 +0000 (21:43 -0400)]
inorder: clear fetchbuffer on traps
implement clearfetchbufferfunction
extend predecoder to use multiple threads and clear those on trap

13 years agoinorder: use separate float-reg bits function in dyninst
Korey Sewell [Mon, 20 Jun 2011 01:43:41 +0000 (21:43 -0400)]
inorder: use separate float-reg bits function in dyninst
this will make sure we get the correct view of a FP register

13 years agoinorder: use trapPending flag to manage traps
Korey Sewell [Mon, 20 Jun 2011 01:43:41 +0000 (21:43 -0400)]
inorder: use trapPending flag to manage traps

13 years agoinorder/dtb: make sure DTB translate correct address
Korey Sewell [Mon, 20 Jun 2011 01:43:41 +0000 (21:43 -0400)]
inorder/dtb: make sure DTB translate correct address
The DTB expects the correct PC in the ThreadContext
but how if the memory accesses are speculative? Shouldn't
we send along the requestor's PC to the translate functions?

13 years agoinorder: handle serializing instructions
Korey Sewell [Mon, 20 Jun 2011 01:43:41 +0000 (21:43 -0400)]
inorder: handle serializing instructions
including IPR accesses and store-conditionals. These class of instructions will not
execute correctly in a superscalar machine

13 years agoalpha: fix warn_once for prefetches
Korey Sewell [Mon, 20 Jun 2011 01:43:40 +0000 (21:43 -0400)]
alpha: fix warn_once for prefetches

13 years agoalpha: naming for dtb faults
Korey Sewell [Mon, 20 Jun 2011 01:43:40 +0000 (21:43 -0400)]
alpha: naming for dtb faults
Just "dfault" gets confusing while debugging. Why not
differentiate whether it's an access violation or page
fault

13 years agoinorder: dont handle multiple faults on same cycle
Korey Sewell [Mon, 20 Jun 2011 01:43:40 +0000 (21:43 -0400)]
inorder: dont handle multiple faults on same cycle
if a faulting instruction reaches an execution unit,
then ignore it and pass it through the pipeline.

Once we recognize the fault in the graduation unit,
dont allow a second fault to creep in on the same cycle.

13 years agoinorder: register ports for FS mode
Korey Sewell [Mon, 20 Jun 2011 01:43:40 +0000 (21:43 -0400)]
inorder: register ports for FS mode
handle "snoop" port registration as well as functional
port setup for FS mode

13 years agoinorder: check for interrupts each tick
Korey Sewell [Mon, 20 Jun 2011 01:43:40 +0000 (21:43 -0400)]
inorder: check for interrupts each tick
use a dummy instruction to facilitate the squash after
the interrupts trap

13 years agoinorder: explicit fault check
Korey Sewell [Mon, 20 Jun 2011 01:43:40 +0000 (21:43 -0400)]
inorder: explicit fault check
Before graduating an instruction, explicitly check fault
by making the fault check it's own separate command
that can be put on an instruction schedule.

13 years agoinorder: squash and trap behind a tlb fault
Korey Sewell [Mon, 20 Jun 2011 01:43:39 +0000 (21:43 -0400)]
inorder: squash and trap behind a tlb fault

13 years agoinorder: stall stores on store conditionals & compare/swaps
Korey Sewell [Mon, 20 Jun 2011 01:43:39 +0000 (21:43 -0400)]
inorder: stall stores on store conditionals & compare/swaps

13 years agoalpha: make hwrei a control inst
Korey Sewell [Mon, 20 Jun 2011 01:43:39 +0000 (21:43 -0400)]
alpha: make hwrei a control inst
this always changes the PC and is basically an impromptu branch instruction. why
not speculate on this instead of always be forced to mispredict/squash after the
hwrei gets resolved?

The InOrder model needs this marked as "isControl" so it knows to update the PC
after the ALU executes it. If this isnt marked as control, then it's going to
force the model to check the PC of every instruction at commit (what O3 does?),
and that would be a wasteful check for a very high percentage of instructions.

13 years agoinorder: make InOrder CPU FS compilable/visible
Korey Sewell [Mon, 20 Jun 2011 01:43:39 +0000 (21:43 -0400)]
inorder: make InOrder CPU FS compilable/visible
make syscall a SE mode only functionality
copy over basic FS functions (hwrei) to make FS compile

13 years agoinorder: remove memdep tracking for default pipeline
Korey Sewell [Mon, 20 Jun 2011 01:43:39 +0000 (21:43 -0400)]
inorder: remove memdep tracking for default pipeline
speculative load/store pipelines can reenable this

13 years agoinorder: fetchBuffer tracking
Korey Sewell [Mon, 20 Jun 2011 01:43:39 +0000 (21:43 -0400)]
inorder: fetchBuffer tracking
calculate blocks in use for the fetch buffer to figure out how many total blocks
are pending

13 years agoinorder: redefine DynInst FP result type
Korey Sewell [Mon, 20 Jun 2011 01:43:38 +0000 (21:43 -0400)]
inorder: redefine DynInst FP result type
Sharing the FP value w/the integer values was giving inconsistent results esp. when
their is a 32-bit integer register matched w/a 64-bit float value

13 years agoinorder: treat SE mode syscalls as a trapping instruction
Korey Sewell [Mon, 20 Jun 2011 01:43:38 +0000 (21:43 -0400)]
inorder: treat SE mode syscalls as a trapping instruction
define a syscallContext to schedule the syscall and then use syscall() to actually perform the action

13 years agoinorder: bug in mdu
Korey Sewell [Mon, 20 Jun 2011 01:43:38 +0000 (21:43 -0400)]
inorder: bug in mdu
segfault was caused by squashed multiply thats in the process of an event.
use isProcessing flag to handle this and cleanup the MDU code

13 years agoinorder: optionally track faulting instructions
Korey Sewell [Mon, 20 Jun 2011 01:43:38 +0000 (21:43 -0400)]
inorder: optionally track faulting instructions

13 years agoinorder: cleanup events in resource pool
Korey Sewell [Mon, 20 Jun 2011 01:43:38 +0000 (21:43 -0400)]
inorder: cleanup events in resource pool
remove events in the resource pool that can be called from the CPU event, since the CPU
event is scheduled at the same time at the resource pool event.
----
Also, match the resPool event function names to the cpu event function names
----

13 years agoinorder: don't stall after stores
Korey Sewell [Mon, 20 Jun 2011 01:43:38 +0000 (21:43 -0400)]
inorder: don't stall after stores
once a ST is sent off, it's OK to keep processing, however it's a little more
complicated to handle the packet acknowledging the store is completed

13 years agoinorder: don't stall after stores
Korey Sewell [Mon, 20 Jun 2011 01:43:37 +0000 (21:43 -0400)]
inorder: don't stall after stores
once a ST is sent off, it's OK to keep processing, however it's a little more
complicated to handle the packet acknowledging the store is completed

13 years agoinorder: remove decode squash
Korey Sewell [Mon, 20 Jun 2011 01:43:37 +0000 (21:43 -0400)]
inorder: remove decode squash
also, cleanup comments for gem5.fast compilation

13 years agoinorder: support for compare and swap insts
Korey Sewell [Mon, 20 Jun 2011 01:43:37 +0000 (21:43 -0400)]
inorder: support for compare and swap insts
dont treat read() and write() fields as mut. exclusive

13 years agoinorder: branch predictor update
Korey Sewell [Mon, 20 Jun 2011 01:43:37 +0000 (21:43 -0400)]
inorder: branch predictor update
only update BTB on a taken branch and update branch predictor w/pcstate from instruction
---
only pay attention to branch predictor updates if the the inst. is in fact a branch

13 years agoinorder: priority for grad/squash events
Korey Sewell [Mon, 20 Jun 2011 01:43:37 +0000 (21:43 -0400)]
inorder: priority for grad/squash events
define separate priority resource pool squash and graduate events

13 years agoinorder: remove stalls on trap squash
Korey Sewell [Mon, 20 Jun 2011 01:43:37 +0000 (21:43 -0400)]
inorder: remove stalls on trap squash

13 years agoinorder: no dep. tracking for zero reg
Korey Sewell [Mon, 20 Jun 2011 01:43:37 +0000 (21:43 -0400)]
inorder: no dep. tracking for zero reg
this causes forwarding a bad value register value

13 years agoimported patch recoverPCfromTrap
Korey Sewell [Mon, 20 Jun 2011 01:43:37 +0000 (21:43 -0400)]
imported patch recoverPCfromTrap

13 years agoimported patch squash_from_next_stage
Korey Sewell [Mon, 20 Jun 2011 01:43:36 +0000 (21:43 -0400)]
imported patch squash_from_next_stage

13 years agoinorder: add flatDestReg member to dyninst
Korey Sewell [Mon, 20 Jun 2011 01:43:36 +0000 (21:43 -0400)]
inorder: add flatDestReg member to dyninst
use it in reg. dep. tracking