Brad Beckmann [Fri, 20 Aug 2010 18:46:12 +0000 (11:46 -0700)]
memtest: Memtester support for DMA
This patch adds DMA testing to the Memtester and is inherits many changes from
Polina's old tester_dma_extension patch. Since Ruby does not work in atomic
mode, the atomic mode options are removed.
Brad Beckmann [Fri, 20 Aug 2010 18:46:12 +0000 (11:46 -0700)]
ruby: Added ruby_request_type ostream def to libruby.hh
Brad Beckmann [Fri, 20 Aug 2010 18:46:12 +0000 (11:46 -0700)]
slicc: Consolidated the protocol stats printing
Created a separate ProfileDumper that consolidates the generated stats for
each controller of a certain type.
Brad Beckmann [Fri, 20 Aug 2010 18:46:11 +0000 (11:46 -0700)]
config: Improve ruby simobject names
This patch attaches ruby objects to the system before the topology is
created so that their simobject names read their meaningful variable
names instead of their topology name.
Brad Beckmann [Fri, 20 Aug 2010 18:46:11 +0000 (11:46 -0700)]
config: Added the topology description to m5 config.ini
Brad Beckmann [Fri, 20 Aug 2010 18:46:11 +0000 (11:46 -0700)]
ruby: added token broadcast config params to cmd options
Brad Beckmann [Fri, 20 Aug 2010 18:44:09 +0000 (11:44 -0700)]
config: reorganized how ruby specifies command-line options
Brad Beckmann [Fri, 20 Aug 2010 18:41:35 +0000 (11:41 -0700)]
ruby: Fixed printout when Sequencer detects a deadlock
Brad Beckmann [Fri, 20 Aug 2010 18:41:35 +0000 (11:41 -0700)]
MESI_CMP_directory: bug fix for old PUTX requests
Brad Beckmann [Fri, 20 Aug 2010 18:41:35 +0000 (11:41 -0700)]
config: moved python protocol config files
Moved the python protocol config files back to their original location to avoid
addToPath calls.
--HG--
rename : configs/ruby/protocols/MESI_CMP_directory.py => configs/ruby/MESI_CMP_directory.py
rename : configs/ruby/protocols/MI_example.py => configs/ruby/MI_example.py
rename : configs/ruby/protocols/MOESI_CMP_directory.py => configs/ruby/MOESI_CMP_directory.py
rename : configs/ruby/protocols/MOESI_CMP_token.py => configs/ruby/MOESI_CMP_token.py
rename : configs/ruby/protocols/MOESI_hammer.py => configs/ruby/MOESI_hammer.py
Steve Reinhardt [Tue, 17 Aug 2010 12:49:05 +0000 (05:49 -0700)]
misc: add some AMD copyright notices
Meant to add these with the previous batch of csets.
Steve Reinhardt [Tue, 17 Aug 2010 12:20:39 +0000 (05:20 -0700)]
x86: minor checkpointing bug fixes
Steve Reinhardt [Tue, 17 Aug 2010 12:17:06 +0000 (05:17 -0700)]
sim: revamp unserialization procedure
Replace direct call to unserialize() on each SimObject with a pair of
calls for better control over initialization in both ckpt and non-ckpt
cases.
If restoring from a checkpoint, loadState(ckpt) is called on each
SimObject. The default implementation simply calls unserialize() if
there is a corresponding checkpoint section, so we get backward
compatibility for existing objects. However, objects can override
loadState() to get other behaviors, e.g., doing other programmed
initializations after unserialize(), or complaining if no checkpoint
section is found. (Note that the default warning for a missing
checkpoint section is now gone.)
If not restoring from a checkpoint, we call the new initState() method
on each SimObject instead. This provides a hook for state
initializations that are only required when *not* restoring from a
checkpoint.
Given this new framework, do some cleanup of LiveProcess subclasses
and X86System, which were (in some cases) emulating initState()
behavior in startup via a local flag or (in other cases) erroneously
doing initializations in startup() that clobbered state loaded earlier
by unserialize().
Steve Reinhardt [Tue, 17 Aug 2010 12:17:06 +0000 (05:17 -0700)]
sim: fold checkpoint restore code into instantiate()
The separate restoreCheckpoint() call is gone; just pass
the checkpoint dir as an optional arg to instantiate().
This change is a precursor to some more extensive
reworking of the startup code.
Steve Reinhardt [Tue, 17 Aug 2010 12:17:06 +0000 (05:17 -0700)]
configs: clean up checkpoint code in Simulation.py
Small change to clean up some redundant code.
Should not have any functional impact.
Steve Reinhardt [Tue, 17 Aug 2010 12:14:03 +0000 (05:14 -0700)]
test: Update stats for python object iteration.
Small changes in tests with data races due to new object creation
order.
Steve Reinhardt [Tue, 17 Aug 2010 12:11:00 +0000 (05:11 -0700)]
sim: clean up child handling
The old code for handling SimObject children was kind of messy,
with children stored both in _values and _children, and
inconsistent and potentially buggy handling of SimObject
vectors. Now children are always stored in _children, and
SimObject vectors are consistently handled using the
SimObjectVector class.
Also, by deferring the parenting of SimObject-valued parameters
until the end (instead of doing it at assignment), we eliminate
the hole where one could assign a vector of SimObjects to a
parameter then append to that vector, with the appended objects
never getting parented properly.
This patch induces small stats changes in tests with data races
due to changes in the object creation & initialization order.
The new code does object vectors in order and so should be more
stable.
Steve Reinhardt [Tue, 17 Aug 2010 12:08:50 +0000 (05:08 -0700)]
sim: move iterating over SimObjects into Python.
Steve Reinhardt [Tue, 17 Aug 2010 12:06:22 +0000 (05:06 -0700)]
sim: fail on implicit creation of orphans via ports
Orphan SimObjects (not in the config hierarchy) could get
created implicitly if they have a port connection to a SimObject
that is in the hierarchy. This means that there are objects on
the C++ SimObject list (created via the C++ SimObject
constructor call) that are unknown to Python and will get
skipped if we walk the hierarchy from the Python side (as we are
about to do). This patch detects this situation and prints an
error message.
Also fix the rubytester config script which happened to rely on
this behavior.
Steve Reinhardt [Tue, 17 Aug 2010 12:06:22 +0000 (05:06 -0700)]
sim: make Python Root object a singleton
Enforce that the Python Root SimObject is instantiated only
once. The C++ Root object already panics if more than one is
created. This change avoids the need to track what the root
object is, since it's available from Root.getInstance() (if it
exists). It's now redundant to have the user pass the root
object to functions like instantiate(), checkpoint(), and
restoreCheckpoint(), so that arg is gone. Users who use
configs/common/Simulate.py should not notice.
Steve Reinhardt [Tue, 17 Aug 2010 12:06:22 +0000 (05:06 -0700)]
tests: update reference config.ini files for previous cset
Rename 'responder_set' to 'use_default_range'.
Steve Reinhardt [Tue, 17 Aug 2010 12:06:21 +0000 (05:06 -0700)]
bus: clean up default responder code.
Clean up some minor things left over from the default responder
change in rev
9af6fb59752f. Mostly renaming the 'responder_set'
param to 'use_default_range' to actually reflect what it does...
old name wasn't that descriptive in the first place, but now
it really doesn't make sense at all.
Also got rid of the bogus obsolete assignment to 'bus.responder'
which used to be a parameter but now is interpreted as an
implicit child assignment, and which was giving me problems in
the config restructuring to come. (A good argument for not
allowing implicit child assignments, IMO, but that's water under
the bridge, I'm afraid.)
Also moved the Bus constructor to the .cc file since that's
where it should have been all along.
Gabe Black [Sat, 14 Aug 2010 08:00:45 +0000 (01:00 -0700)]
Inorder: Fix compilation of m5.fast.
printMemData is only used in DPRINTFs. If those are removed by compiling
m5.fast, that function is unused, gcc generates a warning, that gets turned
into an error, and the build fails. This change surrounds the function
definition with #if TRACING_ON so it only gets compiled in if the DPRINTFs do
to.
Gabe Black [Fri, 13 Aug 2010 13:16:30 +0000 (06:16 -0700)]
Merge with head.
Gabe Black [Fri, 13 Aug 2010 13:16:02 +0000 (06:16 -0700)]
CPU: Add readBytes and writeBytes functions to the exec contexts.
Gabe Black [Fri, 13 Aug 2010 13:16:00 +0000 (06:16 -0700)]
InOrder: Clean up some DPRINTFs that print data sent to/from the cache.
Gabe Black [Fri, 13 Aug 2010 13:10:45 +0000 (06:10 -0700)]
CPU: Tidy up endianness handling for mmapped "IPR"s.
Joel Hestness [Fri, 13 Aug 2010 00:16:04 +0000 (17:16 -0700)]
util/m5/m5.c: ensure readfile() buffer pages are in page table
(and marked dirty, in case that matters) by touching them beforehand
Joel Hestness [Fri, 13 Aug 2010 00:16:02 +0000 (17:16 -0700)]
TimingSimpleCPU: fix NO_ACCESS memory op handling
When a request is NO_ACCESS (x86 CDA microinstruction), the memory op
doesn't go to the cache, so TimingSimpleCPU::completeDataAccess needs
to handle the case where the current status of the CPU is Running
and not DcacheWaitResponse or DTBWaitResponse
Nathan Binkert [Mon, 9 Aug 2010 05:57:16 +0000 (22:57 -0700)]
None, not none
Steve Reinhardt [Wed, 28 Jul 2010 03:00:38 +0000 (20:00 -0700)]
.hgignore: added src/doxygen
Ali Saidi [Tue, 27 Jul 2010 05:03:44 +0000 (01:03 -0400)]
ARM: Add regression tests
Timothy M. Jones [Thu, 22 Jul 2010 17:54:37 +0000 (18:54 +0100)]
Power: The condition register should be set or cleared upon a system call
return to indicate success or failure.
Timothy M. Jones [Thu, 22 Jul 2010 17:54:37 +0000 (18:54 +0100)]
LSQ Unit: After deleting part of a split request, set it to NULL so that it
isn't accidentally deleted again later (causing a segmentation fault).
Timothy M. Jones [Thu, 22 Jul 2010 17:54:37 +0000 (18:54 +0100)]
Port: Only indicate that a SimpleTimingPort is drained if its send event is
not scheduled, as well as the transmit list being empty.
Timothy M. Jones [Thu, 22 Jul 2010 17:52:02 +0000 (18:52 +0100)]
O3CPU: Fix a bug where stores in the cpu where never marked as split.
Timothy M. Jones [Thu, 22 Jul 2010 17:47:52 +0000 (18:47 +0100)]
Syscall: Don't close the simulator's standard file descriptors.
Timothy M. Jones [Thu, 22 Jul 2010 17:47:43 +0000 (18:47 +0100)]
O3CPU: O3's tick event gets squashed when it is switched out. When repeatedly
switching between O3 and another CPU, O3's tick event might still be scheduled
in the event queue (as squashed). Therefore, check for a squashed tick event
as well as a non-scheduled event when taking over from another CPU and deal
with it accordingly.
Timothy M. Jones [Thu, 22 Jul 2010 17:47:03 +0000 (18:47 +0100)]
Power: Provide a utility function to copy registers from one thread context
to another in the Power ISA.
Nathan Binkert [Thu, 22 Jul 2010 01:54:53 +0000 (18:54 -0700)]
stats: unify the two stats distribution type better
Nathan Binkert [Wed, 21 Jul 2010 22:53:53 +0000 (15:53 -0700)]
stats: cleanup a few small problems in stats
Nathan Binkert [Wed, 21 Jul 2010 22:53:53 +0000 (15:53 -0700)]
python: add a sorted dictionary class
It would be nice if python had a tree class that would do this for real,
but since we don't, we'll just keep a sorted list of keys and update
it on demand.
Nathan Binkert [Wed, 21 Jul 2010 22:53:52 +0000 (15:53 -0700)]
python: Add mechanism to override code compiled into the exectuable
If the user sets the environment variable M5_OVERRIDE_PY_SOURCE to
True, then imports that would normally find python code compiled into
the executable will instead first check in the absolute location where
the code was found during the build of the executable. This only
works for files in the src (or extras) directories, not automatically
generated files.
This is a developer feature!
Tushar Krishna [Wed, 21 Jul 2010 16:55:57 +0000 (09:55 -0700)]
Fix x86 XCHG macro-op to use locked micro-ops for all memory accesses
Steve Reinhardt [Sat, 17 Jul 2010 15:56:49 +0000 (08:56 -0700)]
SimObject: transparently forward Python attribute refs to C++.
This tidbit was pulled from a larger patch for Tim's sake, so
the comment reflects functions that haven't been exported yet.
I hope to commit them soon so it didn't seem worth cleaning up.
Gabe Black [Thu, 15 Jul 2010 09:11:56 +0000 (02:11 -0700)]
ARM: Make an SRS instruction with a bad mode cause an undefined instruction fault.
Gabe Black [Wed, 14 Jul 2010 05:41:47 +0000 (22:41 -0700)]
ARM: Adjust the FP_Base_DepTag to be larger than the largest int reg index.
Steve Reinhardt [Fri, 9 Jul 2010 00:56:13 +0000 (17:56 -0700)]
cache: fix bug in SC upgrade handling
This bug was introduced with the recent rework of SC
failure handling in cset
f97b62be544f.
Brad Beckmann [Thu, 8 Jul 2010 23:18:20 +0000 (16:18 -0700)]
garnet: Added topology print function to Garnet printStats
Tushar Krishna [Thu, 8 Jul 2010 23:18:20 +0000 (16:18 -0700)]
NetworkMessage copy constructor fix
Steve Reinhardt [Tue, 6 Jul 2010 05:57:23 +0000 (22:57 -0700)]
checkpointing: another small overload fix
On Nate's advice, overload 'char' as well as 'signed char'
and 'unsigned char'.
Steve Reinhardt [Tue, 6 Jul 2010 04:39:38 +0000 (21:39 -0700)]
sim: allow SimObject subclasses to define classmethods
(without requiring a leading underscore)
Also a little cleanup on type names in SimObject.py.
Steve Reinhardt [Tue, 6 Jul 2010 04:39:38 +0000 (21:39 -0700)]
sim: fold StartupCallback into SimObject
There used to be a reason to have StartupCallback
be a separate object, but not any more. Now
it's just confusing.
Steve Reinhardt [Tue, 6 Jul 2010 04:39:38 +0000 (21:39 -0700)]
checkpointing: minor cleanup.
Move some static checkpoint stuff into the
Checkpoint object namespace.
Steve Reinhardt [Tue, 6 Jul 2010 04:39:38 +0000 (21:39 -0700)]
checkpointing: fix minor bug
Somehow we now need to explicitly specialize on
'signed char' and not just 'char' to catch cases
like int8_t
Steve Reinhardt [Tue, 6 Jul 2010 04:39:38 +0000 (21:39 -0700)]
util: add a script for testing checkpointing
See comments in util/checkpoint-tester.py for details.
Steve Reinhardt [Tue, 6 Jul 2010 04:39:38 +0000 (21:39 -0700)]
process: get rid of some unused code & vars
Steve Reinhardt [Tue, 6 Jul 2010 04:39:38 +0000 (21:39 -0700)]
process: minor format/style cleanup
Tushar Krishna [Tue, 22 Jun 2010 22:36:07 +0000 (15:36 -0700)]
style: updated garnet to match M5 coding style
Korey Sewell [Mon, 28 Jun 2010 11:33:33 +0000 (07:33 -0400)]
inorder: remove another debug stat
Korey Sewell [Sat, 26 Jun 2010 13:41:39 +0000 (09:41 -0400)]
inorder: remove debugging stat
m5 doesnt do stats specific to binary and this resource request stat is probably only
useful for people who really know the ins/outs of the model anyway
Korey Sewell [Fri, 25 Jun 2010 21:42:55 +0000 (17:42 -0400)]
inorder: update regressions from RAS fix
Korey Sewell [Fri, 25 Jun 2010 21:42:35 +0000 (17:42 -0400)]
inorder: Return Address Stack bug
the nextPC was getting sent to the branch predictor not the current PC, so
the RAS was returning the wrong PC and mispredicting everything.
Korey Sewell [Fri, 25 Jun 2010 21:42:34 +0000 (17:42 -0400)]
inorder: resource scheduling backend
replace priority queue with vector of lists(1 list per stage) and place inside a class
so that we have more control of when an instruction uses a particular schedule entry
...
also, this is the 1st step toward making the InOrderCPU fully parameterizable. See the
wiki for details on this process
Gabe Black [Fri, 25 Jun 2010 07:21:48 +0000 (00:21 -0700)]
X86: Fix div2 flag calculation.
Korey Sewell [Thu, 24 Jun 2010 19:34:21 +0000 (15:34 -0400)]
inorder: update regressions
Korey Sewell [Thu, 24 Jun 2010 19:34:19 +0000 (15:34 -0400)]
inorder: cleanup virtual functions
remove the annotation 'virtual' from function declaration that isnt being derived from
Korey Sewell [Thu, 24 Jun 2010 19:34:12 +0000 (15:34 -0400)]
inorder: enforce 78-character rule
Korey Sewell [Thu, 24 Jun 2010 17:58:27 +0000 (13:58 -0400)]
inorder: exe_unit_stats for resolved branches
Korey Sewell [Thu, 24 Jun 2010 02:09:49 +0000 (22:09 -0400)]
inorder: squash from memory stall
this applies to multithreading models which would like to squash a thread on memory stall
Korey Sewell [Wed, 23 Jun 2010 22:21:44 +0000 (18:21 -0400)]
inorder: update regressions
Korey Sewell [Wed, 23 Jun 2010 22:21:12 +0000 (18:21 -0400)]
inorder: record load/store trace data
Korey Sewell [Wed, 23 Jun 2010 22:19:18 +0000 (18:19 -0400)]
inorder: update branch predictor
- use InOrderBPred instead of Resource for DPRINTFs
- account for DELAY SLOT in updating RAS and in squashing
- don't let squashed instructions update the predictor
- the BTB needs to use the ASID not the TID to work for multithreaded programs
- add stats for BTB hits
Korey Sewell [Wed, 23 Jun 2010 22:18:20 +0000 (18:18 -0400)]
inorder-stats: add instruction type stats
also, remove inst-req stats as default.good for debugging
but in terms of pure processor stats they aren't useful
Korey Sewell [Wed, 23 Jun 2010 22:15:23 +0000 (18:15 -0400)]
inorder: stall signal handling
remove stall only when necessary
add debugging printfs
Korey Sewell [Wed, 23 Jun 2010 22:14:59 +0000 (18:14 -0400)]
inorder: tick scheduling
use nextCycle to calculate ticks after addition
Steve Reinhardt [Wed, 23 Jun 2010 04:29:43 +0000 (21:29 -0700)]
cache: fix longstanding prefetcher bug
Thanks to Joe Gross for pointing this out (again?).
Apologies to anyone who pointed it out earlier and
we didn't listen.
Timothy M. Jones [Tue, 22 Jun 2010 23:53:17 +0000 (00:53 +0100)]
O3ThreadContext: When taking over from a previous context, only assert that
the system pointers match in Full System mode.
Steve Reinhardt [Wed, 16 Jun 2010 22:25:57 +0000 (15:25 -0700)]
stats: update stats for SC protocol change
Some subset of UpgradeReq messages shifted to the
new SCUpgradeReq type. Other than that there
are no significant differences.
Steve Reinhardt [Wed, 16 Jun 2010 22:25:57 +0000 (15:25 -0700)]
cache: fail store conditionals when upgrade loses race
Requires new "SCUpgradeReq" message that marks upgrades
for store conditionals, so downstream caches can fail
these when they run into invalidations.
See http://www.m5sim.org/flyspray/task/197
Steve Reinhardt [Wed, 16 Jun 2010 22:25:57 +0000 (15:25 -0700)]
cache: fix dirty bit setting
Only set the dirty bit when we actually write to a block
(not if we thought we might but didn't, as in a failed
SC or CAS). This requires makeing sure the dirty bit
stays set when we get an exclusive (writable) copy
in a cache-to-cache transfer from another owner, which
n turn requires copying the mem-inhibit flag from
timing-mode requests to their associated responses.
Nathan Binkert [Tue, 15 Jun 2010 21:00:41 +0000 (14:00 -0700)]
stats: rename print to display in the mysql code too...sorry
Nathan Binkert [Tue, 15 Jun 2010 15:34:19 +0000 (08:34 -0700)]
stats: rename print to display so it work in python
Nathan Binkert [Tue, 15 Jun 2010 08:18:36 +0000 (01:18 -0700)]
stats: only consider a formula initialized if there is a formula
Nathan Binkert [Tue, 15 Jun 2010 06:24:46 +0000 (23:24 -0700)]
stats: get rid of the never-really-used event stuff
Nathan Binkert [Tue, 15 Jun 2010 06:24:46 +0000 (23:24 -0700)]
util: clean up attrdict and import multiattrdict into m5.util
Nathan Binkert [Tue, 15 Jun 2010 06:24:46 +0000 (23:24 -0700)]
python: use ipython in --interactive if it is available
Nathan Binkert [Tue, 15 Jun 2010 06:24:45 +0000 (23:24 -0700)]
build: grab the right library if we're using a debug version of python
Nathan Binkert [Fri, 11 Jun 2010 06:17:07 +0000 (23:17 -0700)]
ruby: get rid of PrioHeap and use STL
One big difference is that PrioHeap puts the smallest element at the
top of the heap, whereas stl puts the largest element on top, so I
changed all comparisons so they did the right thing.
Some usage of PrioHeap was simply changed to a std::vector, using sort
at the right time, other usage had me just use the various heap functions
in the stl.
Nathan Binkert [Fri, 11 Jun 2010 06:17:07 +0000 (23:17 -0700)]
ruby: get rid of the Map class
Nathan Binkert [Fri, 11 Jun 2010 06:17:07 +0000 (23:17 -0700)]
ruby: get rid of Vector and use STL
add a couple of helper functions to base for deleteing all pointers in
a container and outputting containers to a stream
Nathan Binkert [Fri, 11 Jun 2010 06:17:06 +0000 (23:17 -0700)]
ruby: get rid of RefCnt and Allocator stuff use base/refcnt.hh
This was somewhat tricky because the RefCnt API was somewhat odd. The
biggest confusion was that the the RefCnt object's constructor that
took a TYPE& cloned the object. I created an explicit virtual clone()
function for things that took advantage of this version of the
constructor. I was conservative and used clone() when I was in doubt
of whether or not it was necessary. I still think that there are
probably too many instances of clone(), but hopefully not too many.
I converted several instances of const MsgPtr & to a simple MsgPtr.
If the function wants to avoid the overhead of creating another
reference, then it should just use a regular pointer instead of a ref
counting ptr.
There were a couple of instances where refcounted objects were created
on the stack. This seems pretty dangerous since if you ever
accidentally make a reference to that object with a ref counting
pointer, bad things are bound to happen.
Lisa Hsu [Wed, 9 Jun 2010 17:47:37 +0000 (10:47 -0700)]
flags: add comment to avoid future deletions since code appears redundant.
Lisa Hsu [Wed, 9 Jun 2010 00:16:36 +0000 (17:16 -0700)]
flags: Unserializing old checkpoints before the introduction
of the Initialized flag would break, set Initialized for events upon
unserialization.
Steve Reinhardt [Mon, 7 Jun 2010 16:19:59 +0000 (12:19 -0400)]
scons: make RUBY a regular (non-global) sticky var
and force it to True for builds that imply Ruby protocols
(else unexpected things happen when testing these builds
with RUBY=False).
m5test [Sun, 6 Jun 2010 22:39:10 +0000 (18:39 -0400)]
tests: Update O3 ref outputs to reflect Lisa's dist format change.
Steve Reinhardt [Sun, 6 Jun 2010 05:22:14 +0000 (01:22 -0400)]
stats: fix stats diff script
Previously the return value ignored missing/added stats,
making the regressions not tell you when you needed to
update the reference stats because of these changes.
Also stop filtering distributions when reporting these;
not sure why we did that in the first place.
Also get rid of obsolete hacks for the "fetch-loss" stats
that have been gone for a long time.
Steve Reinhardt [Fri, 4 Jun 2010 02:41:34 +0000 (19:41 -0700)]
More minor gdb-related cleanup.
Found several more stale includes and forward decls.
Steve Reinhardt [Thu, 3 Jun 2010 23:54:28 +0000 (16:54 -0700)]
Act like enabling CPUs is no big deal,
rather than a scary thing that might not work.
Steve Reinhardt [Thu, 3 Jun 2010 23:54:26 +0000 (16:54 -0700)]
Minor remote GDB cleanup.
Expand the help text on the --remote-gdb-port option so
people know you can use it to disable remote gdb without
reading the source code, and thus don't waste any time
trying to add a separate option to do that.
Clean up some gdb-related cruft I found while looking
for where one would add a gdb disable option, before
I found the comment that told me that I didn't need
to do that.