Nathan Binkert [Tue, 30 Mar 2010 00:39:02 +0000 (20:39 -0400)]
style: cleanup the Ruby Tester
Korey Sewell [Sat, 27 Mar 2010 06:23:00 +0000 (02:23 -0400)]
m5: merge inorder updates
Korey Sewell [Sat, 27 Mar 2010 06:21:22 +0000 (02:21 -0400)]
inorder: update twolf/vortex regressions
Korey Sewell [Sat, 27 Mar 2010 05:40:05 +0000 (01:40 -0400)]
inorder: write-hints bug fix
make sure to only read 1 src reg. for write-hint and any other similar
'store' instruction. Reading the source reg when its not necessary
can cause the simulator to read from uninitialized values
Timothy M. Jones [Thu, 25 Mar 2010 12:43:52 +0000 (12:43 +0000)]
CPU: Added comments to address translation classes.
Nathan Binkert [Wed, 24 Mar 2010 05:49:43 +0000 (22:49 -0700)]
ruby: continue style pass
Nathan Binkert [Tue, 23 Mar 2010 23:31:47 +0000 (16:31 -0700)]
regress: add some new options
add -n/--no-exec which doesn't execute scons, but just prints the command line
add -j0 which tries to calculate how many cpus you have
add -D/--build-dir to specify a build directory other than ./build
Steve Reinhardt [Tue, 23 Mar 2010 15:50:59 +0000 (08:50 -0700)]
cpu: get rid of uncached access "events"
These recordEvent() calls could cause crashes since they
access the req pointer after it's potentially been
deleted during a failed translation call. (Similar
problem to the traceData bug fixed in the previous cset.)
Moving them above the translation call (as was done
recentlyi in cset
8b2b8e5e7d35) avoids the crash
but doesn't work, since at that point we don't know if
the access is uncached or not.
It's not clear why these calls are there, and no one
seems to use them, so we'll just delete them. If they
are needed, they should be moved to somewhere that's
guaranteed to be after the translation completes but
before the request is possibly deleted, e.g., in
finishTranslation().
Steve Reinhardt [Tue, 23 Mar 2010 15:50:57 +0000 (08:50 -0700)]
cpu: fix exec tracing memory corruption bug
Accessing traceData (to call setAddress() and/or setData())
after initiating a timing translation was causing crashes,
since a failed translation could delete the traceData
object before returning.
It turns out that there was never a need to access traceData
after initiating the translation, as the traced data was
always available earlier; this ordering was merely
historical. Furthermore, traceData->setAddress() and
traceData->setData() were being called both from the CPU
model and the ISA definition, often redundantly.
This patch standardizes all setAddress and setData calls
for memory instructions to be in the CPU models and not
in the ISA definition. It also moves those calls above
the translation calls to eliminate the crashes.
Korey Sewell [Tue, 23 Mar 2010 04:29:10 +0000 (00:29 -0400)]
m5merge(2): another merge of regression stats
Korey Sewell [Tue, 23 Mar 2010 04:26:53 +0000 (00:26 -0400)]
inorder: update hello world for alpha and mips
Korey Sewell [Tue, 23 Mar 2010 04:21:19 +0000 (00:21 -0400)]
m5merge: ruby + inorder
automerge of updated inorder regressions and ruby style pass
Korey Sewell [Tue, 23 Mar 2010 04:14:52 +0000 (00:14 -0400)]
inorder: update twolf regression
Korey Sewell [Tue, 23 Mar 2010 03:39:23 +0000 (23:39 -0400)]
inorder: update vortex regression
Nathan Binkert [Tue, 23 Mar 2010 01:43:53 +0000 (18:43 -0700)]
ruby: style pass
Korey Sewell [Mon, 22 Mar 2010 21:19:48 +0000 (17:19 -0400)]
inorder: import name for addtl. bpred stats
Maximilien Breughe [Mon, 22 Mar 2010 20:59:12 +0000 (16:59 -0400)]
inorder: fix squash bug in branch predictor
Korey Sewell [Mon, 22 Mar 2010 19:38:28 +0000 (15:38 -0400)]
inorder: fix address list bug
Brad Beckmann [Mon, 22 Mar 2010 18:19:17 +0000 (11:19 -0700)]
ruby: improved isReadWrite fix me comment
Brad Beckmann [Mon, 22 Mar 2010 04:22:22 +0000 (21:22 -0700)]
ruby: Regression updates for new ruby config locations
Brad Beckmann [Mon, 22 Mar 2010 04:22:22 +0000 (21:22 -0700)]
ruby: Removed the unnecessary MachineType message fields
Brad Beckmann [Mon, 22 Mar 2010 04:22:22 +0000 (21:22 -0700)]
ruby: Reorganized Ruby topology and protocol files
--HG--
rename : configs/ruby/MESI_CMP_directory.py => configs/ruby/protocols/MESI_CMP_directory.py
rename : configs/ruby/MI_example.py => configs/ruby/protocols/MI_example.py
rename : configs/ruby/MOESI_CMP_directory.py => configs/ruby/protocols/MOESI_CMP_directory.py
rename : configs/ruby/MOESI_CMP_token.py => configs/ruby/protocols/MOESI_CMP_token.py
rename : configs/ruby/MOESI_hammer.py => configs/ruby/protocols/MOESI_hammer.py
rename : configs/ruby/networks/MeshDirCorners.py => src/mem/ruby/network/topologies/MeshDirCorners.py
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
added sparse memory support to hammer
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Disable adaptive routing by for faster simulation perf.
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Changed the default set size to 1
Previously, the set size was set to 4. This was mostly do to the fact that a
crazy graduate student use to create networks with 256 l2 cache banks. Now it
is far more likely that users will create systems with less than 64 of any
particular controller type. Therefore Ruby should be optimized for a set size
of 1.
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Reordered protocol buffers
Reordered vnet priorities to agree with PerfectSwitch for protocols MI_example,
MOESI_CMP_token, and MOESI_hammer
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Adds configurable bit selection for numa mapping
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Added flag to disable mem_vec allocation
The RubySystem flag no_mem_vec will disable Ruby from allocating it's memory
data array.
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Ruby support for sparse memory
The patch includes direct support for the MI example protocol.
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Finally removed bash code cira. 2001ish!
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Ruby support for LLSC
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Minor dma latency initialization fix
Tushar Krishna [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Fix multiple wakeups in Ruby Eventqueue
Fix bug in Ruby Event queue to avoid multiple wakeups of same consumer in
same cycle
Brad Beckmann [Mon, 22 Mar 2010 04:22:21 +0000 (21:22 -0700)]
ruby: Removed the obsolete file specified network files
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
ruby: Added copyright to many Ruby *.py files
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
ruby: removed ruby.config from configs/example
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
ruby: Fixed small data msg bug in MOESI_hammer-dir
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
TimingSimpleCPU: Fixed uncacacheable request read bug
Previously the recording of an uncached read occurred after the request was
possibly deleted within the translateTiming function.
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
ruby: Python config files now sets a unique id for each sequencer
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
ruby: fixed how ruby_fs creates phsyical memory
Now ruby_fs creates physical memory of the right size.
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
ruby: Removed the no longer used rubymem files
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
ruby: Fix MOESI_hammer cache profiler calls for L2 misses
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
ruby: Added a mesh network with directories at the corners
The patch creates a specific mesh network where directories are at the corners.
The patch is a good example of how to create an arbitrary network, similar to
the old file specified network, while leveraging scripts and loops when
possible.
Brad Beckmann [Mon, 22 Mar 2010 04:22:20 +0000 (21:22 -0700)]
ruby: Removed deprecated stats from the main profiler
Nathan Binkert [Tue, 16 Mar 2010 15:15:16 +0000 (08:15 -0700)]
orion: Make declarations match definition
Nathan Binkert [Mon, 15 Mar 2010 03:58:45 +0000 (20:58 -0700)]
ruby: Fix copyrights on files
Mostly files missed during import or screwed up during import
Nathan Binkert [Sat, 13 Mar 2010 02:42:56 +0000 (18:42 -0800)]
slicc: Change the code generation so that the generated code is easier to read
Nathan Binkert [Sat, 13 Mar 2010 01:31:08 +0000 (17:31 -0800)]
packet: add a method to set the size
Nathan Binkert [Sat, 13 Mar 2010 01:31:04 +0000 (17:31 -0800)]
eventq: rearrange a little bit so I can add some stuff
Nathan Binkert [Sat, 13 Mar 2010 01:31:02 +0000 (17:31 -0800)]
eventq: remove some unused includes
Nathan Binkert [Fri, 12 Mar 2010 23:11:09 +0000 (15:11 -0800)]
bugfix: since pow() causes a bug don't use it
It's a power of two anyway, so why use it in the first place.
Nathan Binkert [Thu, 11 Mar 2010 02:33:11 +0000 (18:33 -0800)]
ruby: get rid of std-includes.hh
Do not use "using namespace std;" in headers
Include header files as needed
Nathan Binkert [Thu, 11 Mar 2010 00:22:27 +0000 (16:22 -0800)]
ruby: remove calc_host.diff since we don't use it
Nathan Binkert [Thu, 11 Mar 2010 00:22:26 +0000 (16:22 -0800)]
ruby: get rid of the ioutil stuff since it isn't used anymore
Nathan Binkert [Thu, 11 Mar 2010 00:22:26 +0000 (16:22 -0800)]
slicc: have a central mechanism for creating a code_formatter.
This makes it easier to add global variables like protocol
Nathan Binkert [Wed, 10 Mar 2010 23:39:34 +0000 (15:39 -0800)]
scons: import ply to work around scons sys.path weirdness
Nathan Binkert [Mon, 1 Mar 2010 03:28:09 +0000 (19:28 -0800)]
SmartDict: Make SmartDict an attrdict
Nathan Binkert [Mon, 1 Mar 2010 03:28:09 +0000 (19:28 -0800)]
uart: use integer versions of time instead of messing around with floats
Gabe Black [Sun, 28 Feb 2010 02:52:57 +0000 (18:52 -0800)]
Config: Fix fs.py's call to CacheConfig.config_cache.
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
cpu_models: get rid of cpu_models.py and move the stuff into SCons
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: Make SCons import the isa_parser
this is instead of forking a new interpreter
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: move the operand map stuff into the ISAParser class.
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: move more support functions into the ISAParser class
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: move more stuff into the ISAParser class
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: move the formatMap and exportContext into the ISAParser class
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: Make stack objects class members instead of globals
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: add a debug variable that changes how errors are reported.
This allows us to get tracebacks in certain cases where they're more
useful than our error message.
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: Use an exception to flag error
This allows the error to propagate more easily
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: Move more stuff into the ISAParser class
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: move code around to prepare for putting more stuff in the class
Nathan Binkert [Sat, 27 Feb 2010 02:14:48 +0000 (18:14 -0800)]
isa_parser: simple fixes, formatting and style
Nathan Binkert [Sat, 27 Feb 2010 02:09:41 +0000 (18:09 -0800)]
events: Give EventWrapped a default name and description
Lisa Hsu [Thu, 25 Feb 2010 18:13:40 +0000 (10:13 -0800)]
configs: pull out cache configuration code from se.py and fs.py.
Most of these frontend configurations share cache configuration code, pull it out so that
changes to caches don't have to require changing multiple config files.
Lisa Hsu [Thu, 25 Feb 2010 18:08:41 +0000 (10:08 -0800)]
stats: update stats for the changes I pushed re: shared cache occupancy
Lisa Hsu [Wed, 24 Feb 2010 21:46:55 +0000 (13:46 -0800)]
cache stats: account for writebacks and/or device occupancy in the cache.
Plus, a minor bugfix that neglects to update blk->contextSrc in certain cases on a cache insert.
Lisa Hsu [Tue, 23 Feb 2010 17:34:22 +0000 (09:34 -0800)]
cache: Make caches sharing aware and add occupancy stats.
On the config end, if a shared L2 is created for the system, it is
parameterized to have n sharers as defined by option.num_cpus. In addition to
making the cache sharing aware so that discriminating tag policies can make use
of context_ids to make decisions, I added an occupancy AverageStat and an occ %
stat to each cache so that you could know which contexts are occupying how much
cache on average, both in terms of blocks and percentage. Note that since
devices have context_id -1, having an array of occ stats that correspond to
each context_id will break here, so in FS mode I add an extra bucket for device
blocks. This bucket is explicitly not added in SE mode in order to not only
avoid ugliness in the stats.txt file, but to avoid broken stats (some formulas
break when a bucket is 0).
Lisa Hsu [Tue, 23 Feb 2010 17:33:18 +0000 (09:33 -0800)]
stats: this makes some fixes to AverageStat and AverageVector.
Also, make Formulas work on AverageVector. First, Stat::Average (and thus
Stats::AverageVector) was broken when coming out of a checkpoint and on resets,
this fixes that. Formulas also didn't work with AverageVector, but added
support for that.
Lisa Hsu [Tue, 23 Feb 2010 17:33:09 +0000 (09:33 -0800)]
cache: pull CacheSet out of LRU so that other tags can use associative sets.
Timothy M. Jones [Sat, 20 Feb 2010 20:11:58 +0000 (20:11 +0000)]
BaseDynInst: Preserve the faults returned from read and write.
When implementing timing address translations instead of atomic, I
forgot to preserve the faults that are returned from the read and
write calls. This patch reinstates them.
Timothy M. Jones [Fri, 12 Feb 2010 19:53:20 +0000 (19:53 +0000)]
O3PCU: Split loads and stores that cross cache line boundaries.
When each load or store is sent to the LSQ, we check whether it will cross a
cache line boundary and, if so, split it in two. This creates two TLB
translations and two memory requests. Care has to be taken if the first
packet of a split load is sent but the second blocks the cache. Similarly,
for a store, if the first packet cannot be sent, we must store the second
one somewhere to retry later.
This modifies the LSQSenderState class to record both packets in a split
load or store.
Finally, a new const variable, HasUnalignedMemAcc, is added to each ISA
to indicate whether unaligned memory accesses are allowed. This is used
throughout the changed code so that compiler can optimise away code dealing
with split requests for ISAs that don't need them.
Timothy M. Jones [Fri, 12 Feb 2010 19:53:19 +0000 (19:53 +0000)]
BaseDynInst: Make the TLB translation timing instead of atomic.
This initiates a timing translation and passes the read or write on to the
processor before waiting for it to finish. Once the translation is finished,
the instruction's state is updated via the 'finish' function. A new
DataTranslation class is created to handle this.
The idea is taken from the implementation of timing translations in
TimingSimpleCPU by Gabe Black. This patch also separates out the timing
translations from this CPU and uses the new DataTranslation class.
Timothy M. Jones [Fri, 12 Feb 2010 19:53:19 +0000 (19:53 +0000)]
Power ISA: Add an alignment fault to Power ISA and check alignment in TLB.
Brad Beckmann [Thu, 11 Feb 2010 00:40:54 +0000 (16:40 -0800)]
ruby: fixed data block assignment fix
Fixed data block assignment to not delete if not internally allocated.
Brad Beckmann [Thu, 11 Feb 2010 00:40:54 +0000 (16:40 -0800)]
ruby: Initialize sender in MI_example-dir
Brad Beckmann [Thu, 11 Feb 2010 00:40:54 +0000 (16:40 -0800)]
ruby: Fixed slicc to initialize the m_is_blocking flag
Brad Beckmann [Mon, 1 Feb 2010 22:27:16 +0000 (14:27 -0800)]
ruby: Added FS support to the simple mesh topology
Added full-system support to the simple mesh toplogy by allowing dma contrllers
to be attached to router zero in the network.
Brad Beckmann [Mon, 1 Feb 2010 19:07:38 +0000 (11:07 -0800)]
ruby: Set default protocol back to MI_example
Brad Beckmann [Mon, 1 Feb 2010 06:28:13 +0000 (22:28 -0800)]
merge
Brad Beckmann [Mon, 1 Feb 2010 06:21:01 +0000 (22:21 -0800)]
m5: Added PROTOCOL default for regress fix
Korey Sewell [Sun, 31 Jan 2010 23:31:28 +0000 (18:31 -0500)]
inorder: update hello world mips
Korey Sewell [Sun, 31 Jan 2010 23:31:20 +0000 (18:31 -0500)]
inorder: vortex alpha regression
Korey Sewell [Sun, 31 Jan 2010 23:31:14 +0000 (18:31 -0500)]
inorder: twolf alpha regression
Korey Sewell [Sun, 31 Jan 2010 23:31:09 +0000 (18:31 -0500)]
inorder: update hello world alpha
Korey Sewell [Sun, 31 Jan 2010 23:30:59 +0000 (18:30 -0500)]
inorder: double delete inst bug
Make sure that instructions are dereferenced/deleted twice by marking they are
on the remove list
Korey Sewell [Sun, 31 Jan 2010 23:30:48 +0000 (18:30 -0500)]
inorder: inst count mgmt
Korey Sewell [Sun, 31 Jan 2010 23:30:43 +0000 (18:30 -0500)]
inorder: implement split stores
Korey Sewell [Sun, 31 Jan 2010 23:30:35 +0000 (18:30 -0500)]
inorder: implement split loads
Korey Sewell [Sun, 31 Jan 2010 23:30:24 +0000 (18:30 -0500)]
inorder: add activity stats
Korey Sewell [Sun, 31 Jan 2010 23:30:08 +0000 (18:30 -0500)]
inorder: object cleanup in destructors
Korey Sewell [Sun, 31 Jan 2010 23:29:59 +0000 (18:29 -0500)]
inorder: user per-thread dummy insts/reqs