Commit graph

776 commits

Author SHA1 Message Date
Polina Dudnik
de25decf37 Multi-line RMW handling 2009-08-14 14:24:15 -05:00
Polina Dudnik
4b924fd16c SMT atomics modifications:
don't allow enquing from other threads if servicing and atomic for a thread
2009-08-14 14:06:14 -05:00
Derek Hower
bcaf93d182 Automated merge with ssh://hg@m5sim.org/m5 2009-08-13 10:37:37 -05:00
Derek Hower
db40cb8f51 ruby: config bugfix 2009-08-13 10:37:09 -05:00
Tushar Krishna
35082a67b6 ruby/network data_msg_size bug fix with updated stats 2009-08-11 15:19:04 -07:00
Brad Beckmann
b89add1e3f merged Tushar's bug fix with public repository changes 2009-08-11 12:22:41 -07:00
Derek Hower
1a452d228b protocol: added recycle actions to MOESI DMA events 2009-08-09 13:58:40 -05:00
Tushar Krishna
b952eb19c1 bug fix for data_msg_size in network/Network.cc 2009-08-07 13:59:40 -07:00
Derek Hower
cbc52ef6c5 fixed MOESI_CMP_directory bug 2009-08-06 03:41:28 -05:00
Derek Hower
f5e0c56da2 protocol: fixed MOESI_CMP_directory bug 2009-08-06 01:15:55 -05:00
Derek Hower
a1b5a6320f ruby: better configuration assert message 2009-08-06 01:15:23 -05:00
Derek Hower
dff7c9eaa0 merge 2009-08-05 14:23:32 -05:00
Derek Hower
fbf7391bb0 ruby: configuration supports multiple runs in same session
These changes allow to run Ruby-gems multiple times from the same
ruby-lang script with different configurations
2009-08-05 14:20:32 -05:00
Derek Hower
1276df51e2 protocol: made MI_example dma mapping generic 2009-08-05 14:17:23 -05:00
Derek Hower
7f012ef8da ruby: made mapAddressToRange based off a bit count 2009-08-04 23:05:37 -05:00
Derek Hower
33b28fde7a slicc: added MOESI_CMP_directory, DMA SequencerMsg, parameterized controllers
This changeset contains a lot of different changes that are too
mingled to separate.  They are:

1.  Added MOESI_CMP_directory

I made the changes necessary to bring back MOESI_CMP_directory,
including adding a DMA controller.  I got rid of MOESI_CMP_directory_m
and made MOESI_CMP_directory use a memory controller. Added a new
configuration for two level protocols in general, and
MOESI_CMP_directory in particular.

2.  DMA Sequencer uses a generic SequencerMsg

I will eventually make the cache Sequencer use this type as well.  It
doesn't contain an offset field, just a physical address and a length.
MI_example has been updated to deal with this.

3. Parameterized Controllers

SLICC controllers can now take custom parameters to use for mapping,
latencies, etc.  Currently, only int parameters are supported.
2009-08-04 12:52:52 -05:00
Derek Hower
c1e0bd1df4 slicc: generate html by default 2009-08-04 12:42:45 -05:00
Nathan Binkert
bd7af84d5e slicc: better error messages when the python parser fails 2009-08-04 09:37:27 -07:00
Derek Hower
ac15e42c17 Automated merge with ssh://hg@m5sim.org/m5 2009-08-03 11:39:08 -05:00
Steve Reinhardt
a13a706a20 Fix setting of INST_FETCH flag for O3 CPU.
It's still broken in inorder.
Also enhance DPRINTFs in cache and physical memory so we
can see more easily whether it's getting set or not.
2009-08-01 22:50:14 -07:00
Steve Reinhardt
1c28004654 Clean up some inconsistencies with Request flags. 2009-08-01 22:50:13 -07:00
Steve Reinhardt
c0755e6085 Rename internal Request fields to start with '_'.
The inconsistency was causing a subtle bug with some of the
constructors where the params had the same name as the fields.
This is also a first step to switching the accessors over to
our new "standard", e.g., getVaddr() -> vaddr().
2009-08-01 22:50:10 -07:00
Derek Hower
d9ff3021ba ruby: fixed clearStats 2009-07-29 13:46:58 -05:00
Derek Hower
469256d823 ruby: removed unused/incorrect profiler state 2009-07-27 21:43:43 -05:00
Polina Dudnik
e7a3bda497 Fixed the licences plus minor fixes for compilation 2009-07-22 20:28:32 -05:00
Derek Hower
7f34ee36ec ruby: fixed sequencer RMW data bug 2009-07-21 19:42:09 -05:00
Derek Hower
80544cda8a ruby: libruby_init now takes parsed Ruby-lang config text
libruby_init now expects to get a file that contains the output of
running a ruby-lang configuration, opposed to the ruby-lang
configuration itself.
2009-07-21 18:33:05 -05:00
Derek Hower
e59d0e3e89 ruby: moved cache stats from Profiler to CacheMemory
Caches are now responsible for their own statistic gathering.  This
requires a direct callback from the protocol on misses, and so all
future protocols need to take this into account.
2009-07-20 09:40:43 -05:00
Derek Hower
308419b947 scons: removed RubyConfig from scons 2009-07-19 12:34:11 -05:00
Derek Hower
7cd2d8f687 ruby: removed all refs to old RubyConfig 2009-07-18 18:20:03 -05:00
Derek Hower
4bd7fe4c53 ruby: removed dead files 2009-07-18 18:18:37 -05:00
Derek Hower
f3d8d29342 ruby: removed dead files 2009-07-18 18:17:48 -05:00
Derek Hower
926ab6e6db merge 2009-07-18 17:40:20 -05:00
Derek Hower
4b7ea4cb51 ruby: fixed dma sequencer bug
The DMASequencer was still using a parameter from the old RubyConfig,
causing an offset error when the requested data wasn't block aligned.
This changeset also includes a fix to MI_example for a similar bug.
2009-07-18 17:03:51 -05:00
Derek Hower
340845b139 ruby: better debug print for DataBlock 2009-07-18 16:58:33 -05:00
Derek Hower
7433029cd5 slicc: made coherence profilers per-controller 2009-07-18 16:54:45 -05:00
Polina Dudnik
e557b4beb5 merge 2009-07-16 15:40:48 -05:00
Polina Dudnik
23a405f5d8 Tester update 2009-07-15 10:46:22 -05:00
Polina Dudnik
289cd00326 Changed the state machine to generate code such that multiple processors can make atomic requests at once 2009-07-13 18:39:32 -05:00
Polina Dudnik
5f551d9ca2 1. Got rid of unused functions in DirectoryMemory
2. Reintroduced RMW_Read and RMW_Write
3. Defined -2 in the Sequencer as well as made a note about mandatory queue

Did not address the issues in the slicc because remaking the atomics altogether to allow
multiple processors to issue atomic requests at once
2009-07-13 17:22:29 -05:00
Derek Hower
100da6b326 merge 2009-07-13 14:49:51 -05:00
Derek Hower
d51445490d regression: updated memtest-ruby stats
This also includes a change to the default Ruby random seed, which was
previously set using the wall clock.  It is now set to 1234 so that
the stat files don't change for the regression tester.
2009-07-13 14:45:15 -05:00
Polina Dudnik
9a675a0391 Changes to add tracing and replaying command-line options
Trace is automatically ended upon a manual checkpoint
2009-07-13 12:50:10 -05:00
Polina Dudnik
b28058917c Locked requests should actually be converted to ST rather than ATOMIC, because ATOMIC is for RMW. 2009-07-13 12:11:17 -05:00
Polina Dudnik
7a6bf67e47 Added atomics implementation which would work for MI_example 2009-07-13 12:06:23 -05:00
Polina Dudnik
c66af9f474 Minor fixes for compiling 2009-07-13 11:59:13 -05:00
Polina Dudnik
7606c71ea5 Replaced RMW with Locked. RMW will be used for the coherence-aided atomics other than LLSC 2009-07-13 11:37:56 -05:00
Polina Dudnik
faf823f947 Moved the lock check and clearing the lock into makeRequest 2009-07-13 11:34:38 -05:00
Polina Dudnik
86ce60e5cd Forgot to replace one of the RubyRequest_RMW 2009-07-13 11:25:23 -05:00
Polina Dudnik
226981b2a6 Reintegrated Derek's functional implementation of atomics with a minor change: don't clear lock on failure 2009-07-13 11:13:29 -05:00
Gabe Black
b398b8ff1b Registers: Add a registers.hh file as an ISA switched header.
This file is for register indices, Num* constants, and register types.
copyRegs and copyMiscRegs were moved to utility.hh and utility.cc.

--HG--
rename : src/arch/alpha/regfile.hh => src/arch/alpha/registers.hh
rename : src/arch/arm/regfile.hh => src/arch/arm/registers.hh
rename : src/arch/mips/regfile.hh => src/arch/mips/registers.hh
rename : src/arch/sparc/regfile.hh => src/arch/sparc/registers.hh
rename : src/arch/x86/regfile.hh => src/arch/x86/registers.hh
2009-07-08 23:02:21 -07:00
Derek Hower
15afc87f7c slicc: fixed MI_example bug. The directory wasn't deallocating the TBE, leading to a leak. Also increased the default max TBE size to 256 to allow memtest to pass the regression. 2009-07-08 08:40:32 -05:00
Derek Hower
6a83bd5a03 ruby: set the default values of the debug object so that nothing is printed 2009-07-08 00:34:40 -05:00
Derek Hower
2f9d8bff5b slicc: Fixed MI_example bug. The directory was not writing data to DRAM after a PUTX. 2009-07-08 00:31:33 -05:00
Derek Hower
96c36afea9 removed stray debug print 2009-07-07 23:01:35 -05:00
Nathan Binkert
da704f52e5 ruby: Fix RubyMemory to work with the newer ruby. 2009-07-06 15:49:47 -07:00
Nathan Binkert
a7904e2cf3 ruby: apply some fixes that were overwritten by the recent ruby import. 2009-07-06 15:49:47 -07:00
Nathan Binkert
5b080ae046 slicc: update parser.py for changes in slicc language. 2009-07-06 15:49:47 -07:00
Nathan Binkert
1f6933503d scons: update SCons files for changes in ruby. 2009-07-06 15:49:47 -07:00
Nathan Binkert
92de70b69a ruby: Import the latest ruby changes from gems.
This was done with an automated process, so there could be things that were
done in this tree in the past that didn't make it.  One known regression
is that atomic memory operations do not seem to work properly anymore.
2009-07-06 15:49:47 -07:00
Nathan Binkert
05f6a4a6b9 ruby: replace strings that were missed in original ruby import. 2009-07-06 15:49:47 -07:00
Nathan Binkert
d3d8a5a83b copyright: I missed some copyrights during ruby integration 2009-06-10 00:41:56 -07:00
Nathan Binkert
6faf377b53 types: clean up types, especially signed vs unsigned 2009-06-04 23:21:12 -07:00
Nathan Binkert
a0104b6ff6 request: add accessor and constructor for setting time other than curTick 2009-05-29 15:30:16 -07:00
Nathan Binkert
47877cf2db types: add a type for thread IDs and try to use it everywhere 2009-05-26 09:23:13 -07:00
Nathan Binkert
8d2e51c7f5 includes: sort includes again 2009-05-17 14:34:52 -07:00
Nathan Binkert
709d859530 includes: use base/types.hh not inttypes.h or stdint.h 2009-05-17 14:34:51 -07:00
Nathan Binkert
eef3a2e142 types: Move stuff for global types into src/base/types.hh
--HG--
rename : src/sim/host.hh => src/base/types.hh
2009-05-17 14:34:50 -07:00
Nathan Binkert
5207586b26 ruby: deal with printf warnings and convert some to cprintf 2009-05-12 22:33:05 -07:00
Nathan Binkert
016d472c46 ruby: remove random uint typedef and use unsigned 2009-05-12 22:33:05 -07:00
Nathan Binkert
7389dc63b2 ruby: Make ruby's Map use hashmap.hh to simplify things. 2009-05-12 22:33:05 -07:00
Nathan Binkert
0c2b9cf90d slicc: work around improper initialization of a global in slicc. 2009-05-12 22:33:05 -07:00
Nathan Binkert
d923ce0f8c slicc: clean up the slicc environment so things build properly on mac. 2009-05-12 22:33:04 -07:00
Nathan Binkert
f21e80ec72 ruby: assert(false) should be panic.
This also fixes some compiler warnings
2009-05-11 16:32:32 -07:00
Nathan Binkert
cf6b4ef734 ruby: add RUBY sticky option that must be set to add ruby to the build
Default is false
2009-05-11 10:38:46 -07:00
Daniel Sanchez
93f2f69657 ruby: Working M5 interface and updated Ruby interface.
This changeset also includes a lot of work from Derek Hower <drh5@cs.wisc.edu>

RubyMemory is now both a driver for Ruby and a port for M5.  Changed
makeRequest/hitCallback interface. Brought packets (superficially)
into the sequencer. Modified tester infrastructure to be packet based.
and Ruby can be used together through the example ruby_se.py
script. SPARC parallel applications work, and the timing *seems* right
from combined M5/Ruby debug traces. To run,
% build/ALPHA_SE/m5.debug configs/example/ruby_se.py -c
  tests/test-progs/hello/bin/alpha/linux/hello -n 4 -t
2009-05-11 10:38:46 -07:00
Steve Reinhardt
ebf2f5aadd ruby: Check stderr and not stdin before hanging on an assert. 2009-05-11 10:38:46 -07:00
Polina Dudnik
7769cc9092 ruby: decommission code
1. Set.* and BigSet.* are replaced with OptBigSet.* which was renamed Set.*
2. Decomissioned all bloom filters
3. Decomissioned ruby/simics directory
2009-05-11 10:38:46 -07:00
Derek Hower
0ccf8f35a5 ruby: removed dead functions from the sequencer 2009-05-11 10:38:46 -07:00
Polina Dudnik
29f82f265a ruby: Removed g_SIMULATING flag
1. removed checks from tester files
2. removed else clause in Sequencer and DirectoryMemory else clause is
needed by the tester, it is up to Derek to revive it elsewhere when he
gets to it

Also:
1. Changed m_entries in DirectoryMemory to a map
2. And replaced SIMICS_read_physical_memory with a call to now-dummy
Derek's-to-be readPhysMem function
2009-05-11 10:38:46 -07:00
Polina Dudnik
b271090923 ruby: Remove transactional access types (e.g. LD_XACT) from CacheRequestType
1. Modified enumeration
2. Also modified profiler
3. Remove transactions from Tester
4. Edited XACT_MEM out of Synthetic Driver
2009-05-11 10:38:46 -07:00
Polina Dudnik
9f34659c52 ruby: reordered Debug and RubyConfig::init to fix segfault
due to uninitialized output file pointer.
2009-05-11 10:38:46 -07:00
Dan Gibson
8cbf8df5b7 ruby: Disabled RubyEventQueue's deletion of its home-grown priority heap.
Temporarily to fix unusual memory problem.
2009-05-11 10:38:46 -07:00
Nathan Binkert
7311fd7182 ruby: Migrate all of ruby and slicc to SCons.
Add the PROTOCOL sticky option sets the coherence protocol that slicc
will parse and therefore ruby will use.  This whole process was made
difficult by the fact that the set of files that are output by slicc
are not easily known ahead of time.  The easiest thing wound up being
to write a parser for slicc that would tell me.  Incidentally this
means we now have a slicc grammar written in python.
2009-05-11 10:38:46 -07:00
Nathan Binkert
e40b8e34c8 ruby: clean up a few warnings 2009-05-11 10:38:45 -07:00
Dan Gibson
8b9f70b9e4 ruby: Fixed some unresolved references. 2009-05-11 10:38:45 -07:00
Nathan Binkert
24da30e317 ruby: Make ruby #includes use full paths to the files they're including.
This basically means changing all #include statements and changing
autogenerated code so that it generates the correct paths.  Because
slicc generates #includes, I had to hard code the include paths to
mem/protocol.
2009-05-11 10:38:45 -07:00
Dan Gibson
d8c592a05d ruby: remove unnecessary code.
1) Removing files from the ruby build left some unresovled
symbols. Those have been fixed.

2) Most of the dependencies on Simics data types and the simics
interface files have been removed.

3) Almost all mention of opal is gone.

4) Huge chunks of LogTM are now gone.

5) Handling 1-4 left ~hundreds of unresolved references, which were
fixed, yielding a snowball effect (and the massive size of this
delta).
2009-05-11 10:38:45 -07:00
Derek Hower
6ceaffd724 ruby: Cleaned up sequencer. Removed LogTM specific code. 2009-05-11 10:38:45 -07:00
Derek Hower
3d2acc547c ruby: added Packet interface to makeRequest and isReady.
Also pushed Packet usage into the Sequencer
2009-05-11 10:38:45 -07:00
Nathan Binkert
e1915f16d1 ruby: fold the debugging options into Debug.cc 2009-05-11 10:38:45 -07:00
Derek Hower
6e8373fad6 ruby: Renamed Ruby's EventQueue to RubyEventQueue
--HG--
rename : src/mem/ruby/eventqueue/EventQueue.cc => src/mem/ruby/eventqueue/RubyEventQueue.cc
rename : src/mem/ruby/eventqueue/EventQueue.hh => src/mem/ruby/eventqueue/RubyEventQueue.hh
rename : src/mem/ruby/eventqueue/EventQueueNode.cc => src/mem/ruby/eventqueue/RubyEventQueueNode.cc
rename : src/mem/ruby/eventqueue/EventQueueNode.hh => src/mem/ruby/eventqueue/RubyEventQueueNode.hh
2009-05-11 10:38:45 -07:00
Daniel Sanchez
ab5e4a22b3 ruby: Removed System name clash by renaming ruby's System to RubySystem 2009-05-11 10:38:44 -07:00
Nathan Binkert
84a18e7fdc ruby: rename config.include to config.hh and clean up the macro stuff.
I did the macro cleanup because I was worried that the SCons scanner
would get confused.  This code will hopefully go away soon anyway.

--HG--
rename : src/mem/ruby/config/config.include => src/mem/ruby/config/config.hh
2009-05-11 10:38:44 -07:00
Nathan Binkert
b05da09cd6 ruby: strip out some unused defines 2009-05-11 10:38:44 -07:00
Nathan Binkert
2f30950143 ruby: Import ruby and slicc from GEMS
We eventually plan to replace the m5 cache hierarchy with the GEMS
hierarchy, but for now we will make both live alongside eachother.
2009-05-11 10:38:43 -07:00
Steve Reinhardt
7c056e44e5 request: reorganize flags to group related flags together. 2009-04-23 06:44:32 -07:00
Steve Reinhardt
6629d9b2bc mem: use single BadAddr responder per system.
Previously there was one per bus, which caused some coherence problems
when more than one decided to respond.  Now there is just one on
the main memory bus.  The default bus responder on all other buses
is now the downstream cache's cpu_side port.  Caches no longer need
to do address range filtering; instead, we just have a simple flag
to prevent snoops from propagating to the I/O bus.
2008-07-16 11:10:33 -07:00
Steve Reinhardt
97b6947eb7 Minor tweaks for future Ruby compatibility. 2009-04-21 08:17:36 -07:00
Steve Reinhardt
eb3b6935d3 request: add PREFETCH flag. 2009-04-21 08:17:10 -07:00
Steve Reinhardt
3083268d60 request: rename INST_READ to INST_FETCH. 2009-04-20 18:54:02 -07:00
Steve Reinhardt
7f8ea68a30 request: split public and private flags into separate fields.
This frees up needed space for more public flags.  Also:
- remove unused Request accessor methods
- make Packet use public Request accessors, so it need not be a friend
2009-04-20 18:40:00 -07:00
Gabe Black
9e9a34fed1 Mem: Fill out the comment that describes the LOCKED request flag. 2009-04-19 22:00:24 -07:00
Gabe Black
bd6f2bb538 Mem: Change isLlsc to isLLSC. 2009-04-19 21:44:15 -07:00
Gabe Black
742c3f045e Memory: Add a LOCKED flag back in for x86 style locking. 2009-04-19 04:39:25 -07:00
Gabe Black
3e5f487663 Memory: Rename LOCKED for load locked store conditional to LLSC. 2009-04-19 04:25:01 -07:00
Gabe Black
bdda224d41 X86: Add a function which gets called when an interrupt message has been delivered. 2009-04-19 03:54:11 -07:00
Gabe Black
3031af21c7 X86: Fix the flags for interrupt response messages. 2009-04-19 03:53:29 -07:00
Steve Reinhardt
758bfe4eb5 cache: set dirty bit on swaps (oops!) 2009-03-11 23:05:26 -07:00
Steve Reinhardt
a94c68228a prefetch: don't panic on requests w/o contextID (e.g., writebacks). 2009-03-10 17:37:15 -07:00
Nathan Binkert
cc95b57390 stats: Fix all stats usages to deal with template fixes 2009-03-05 19:09:53 -08:00
Gabe Black
a1aba01a02 CPU: Get rid of translate... functions from various interface classes. 2009-02-25 10:15:34 -08:00
Lisa Hsu
5d029ff11e sycalls: implement mremap() and add DATA flag for getrlimit(). mremap has been tested on Alpha, compiles for the rest but not tested. I don't see why it wouldn't work though. 2009-02-16 17:47:39 -05:00
Steve Reinhardt
89a7fb0393 Fixes to get prefetching working again.
Apparently we broke it with the cache rewrite and never noticed.
Thanks to Bao Yungang <baoyungang@gmail.com> for a significant part
of these changes (and for inspiring me to work on the rest).
Some other overdue cleanup on the prefetch code too.
2009-02-16 08:56:40 -08:00
Nathan Binkert
8153790d00 SCons: centralize the Dir() workaround for newer versions of scons.
Scons bug id: 2006 M5 Bug id: 308
2009-01-13 14:17:50 -08:00
Nathan Binkert
e141cb7441 flags: Change naming of functions to be clearer 2008-12-06 14:18:18 -08:00
Steve Reinhardt
4514f565e3 syscalls: fix latent brk/obreak bug.
Bogus calls to ChunkGenerator with negative size were triggering
a new assertion that was added there.
Also did a little renaming and cleanup in the process.
2008-11-15 09:30:10 -08:00
Steve Reinhardt
640b415688 Cache: get rid of obsolete Tag methods.
I think readData() and writeData() were used for Erik's compression
work, but that code is gone, these aren't called anymore, and they
don't even really do what their names imply.
2008-11-14 14:14:35 -08:00
Nathan Binkert
5711282f87 Fix a bunch of bugs I introduced when I changed the flags stuff for packets.
I did some of the flags and assertions wrong. Thanks to Brad Beckmann
for pointing this out.  I should have run the opt regressions instead
of the fast. I also screwed up some of the logical functions in the Flags
class.
2008-11-14 04:55:30 -08:00
Gabe Black
7a4d75bae3 CPU: Refactor read/write in the simple timing CPU. 2008-11-13 23:30:37 -08:00
Nathan Binkert
c25d966b06 Clean up the SimpleTimingPort class a little bit.
Move the constructor into the .cc file and get rid of the typedef for
SendEvent.
2008-11-10 11:51:18 -08:00
Nathan Binkert
3535d746ab style: clean up the Packet stuff 2008-11-10 11:51:17 -08:00
Steve Reinhardt
63127cbf37 mem: Assert that requests have non-negative size.
Would have saved me much debugging time if these
had been in there previously.
2008-11-10 14:11:07 -08:00
Steve Reinhardt
42bd460d7f Cache: Refactor packet forwarding a bit.
Makes adding write-through operations easier.
2008-11-10 14:10:28 -08:00
Lisa Hsu
c68032ddcb decouple eviction from insertion in the cache. 2008-11-04 11:35:58 -05:00
Lisa Hsu
4ab52cb986 Change the findBlock(addr, lat) to accessBlock, which I think has better connotations for what is really happening and how it should be used. 2008-11-04 11:35:57 -05:00
Lisa Hsu
dd99ff23c6 get rid of all instances of readTid() and getThreadNum(). Unify and eliminate
redundancies with threadId() as their replacement.
2008-11-04 11:35:42 -05:00
Lisa Hsu
d857faf073 Add in Context IDs to the simulator. From now on, cpuId is almost never used,
the primary identifier for a hardware context should be contextId().  The
concept of threads within a CPU remains, in the form of threadId() because
sometimes you need to know which context within a cpu to manipulate.
2008-11-02 21:57:07 -05:00
Lisa Hsu
8788d703f8 s/cpu_id/cpuId in o3 (to be consistent and match style), also fix some typos in
comments.
2008-10-23 16:49:17 -04:00
Lisa Hsu
546a6c0c1b probe function no longer used anywhere. 2008-10-23 16:49:13 -04:00
Lisa Hsu
7a28ab2d18 remove the totally obsolete split cache 2008-10-23 16:11:28 -04:00
Lisa Hsu
90e40ca982 This function declaration isn't used anywhere.
HG: user: Lisa Hsu <hsul@eecs.umich.edu> HG: branch default HG: changed
src/mem/cache/cache.hh
2008-10-14 17:22:03 -04:00
Gabe Black
34ca72d16d Get rid of some commented out code. 2008-10-12 23:50:22 -07:00
Gabe Black
e459013182 Create a message port for sending messages as apposed to reading/writing a memory range. 2008-10-12 12:08:51 -07:00
Nathan Binkert
b556dc4119 mem: Add a method for setting the time on a packet. 2008-10-09 04:58:24 -07:00
Nathan Binkert
e06321091d eventq: convert all usage of events to use the new API.
For now, there is still a single global event queue, but this is
necessary for making the steps towards a parallelized m5.
2008-10-09 04:58:24 -07:00
Nathan Binkert
8291d9db0a eventq: Major API change for the Event and EventQueue structures.
Since the early days of M5, an event needed to know which event queue
it was on, and that data was required at the time of construction of
the event object.  In the future parallelized M5, this sort of
requirement does not work well since the proper event queue will not
always be known at the time of construction of an event.  Now, events
are created, and the EventQueue itself has the schedule function,
e.g. eventq->schedule(event, when).  To simplify the syntax, I created
a class called EventManager which holds a pointer to an EventQueue and
provides the schedule interface that is a proxy for the EventQueue.
The intent is that objects that frequently schedule events can be
derived from EventManager and then they have the schedule interface.
SimObject and Port are examples of objects that will become
EventManagers.  The end result is that any SimObject can just call
schedule(event, when) and it will just call that SimObject's
eventq->schedule function.  Of course, some objects may have more than
one EventQueue, so this interface might not be perfect for those, but
they should be relatively few.
2008-10-09 04:58:23 -07:00
Nathan Binkert
9838be2521 When nesting if statements, use braces to avoid ambiguous else clauses. 2008-09-26 08:18:57 -07:00
Ali Saidi
3a3e356f4e style: Remove non-leading tabs everywhere they shouldn't be. Developers should configure their editors to not insert tabs 2008-09-10 14:26:15 -04:00
Nathan Binkert
9cf8ad3a17 params: Get rid of the remnants of the old style parameter configuration stuff. 2008-08-11 12:22:17 -07:00
Steve Reinhardt
62c08a75ad Make default PhysicalMemory latency slightly more realistic.
Also update stats to reflect change.
2008-08-03 18:13:29 -04:00
Steve Reinhardt
8e7ddce284 Use ReadResp instead of LoadLockedResp for LoadLockedReq responses. 2008-07-15 14:38:51 -04:00
Steve Reinhardt
6262e0d909 Add missing newlines to Bus DPRINTFs. 2008-07-15 14:38:51 -04:00
Ali Saidi
a4a7a09e96 Remove delVirtPort() and make getVirtPort() only return cached version. 2008-07-01 10:25:07 -04:00
Ali Saidi
c5fbbf376a Change everything to use the cached virtPort rather than created their own each time.
This appears to work, but I don't want to commit it until it gets tested a lot more.
I haven't deleted the functionality in this patch that will come later, but one question
is how to enforce encourage objects that call getVirtPort() to not cache the virtual port
since if the CPU changes out from under them it will be worse than useless. Perhaps a null
function like delVirtPort() is still useful in that case.
2008-07-01 10:24:19 -04:00
Steve Reinhardt
96bbccc36b Automated merge after backout. 2008-06-28 13:20:00 -04:00
Steve Reinhardt
caaac16803 Backed out changeset 94a7bb476fca: caused memory leak. 2008-06-28 13:19:38 -04:00
Steve Reinhardt
1434b86943 Make bus address conflict error more informative 2008-06-21 01:06:27 -04:00
Steve Reinhardt
6b45238316 Generate more useful error messages for unconnected ports.
Force all non-default ports to provide a name and an
owner in the constructor.
2008-06-21 01:04:43 -04:00
Nathan Binkert
fa8f91fdc0 physmem: Add a null option to physical memory so it doesn't store data. 2008-06-15 21:39:29 -07:00
Nathan Binkert
e3c267a3db port: Clean up default port setup and port switchover code. 2008-06-15 21:34:32 -07:00
Nathan Binkert
fe325c7f43 MemReq: Add option to reset the time on a request. 2008-06-14 19:39:01 -07:00
Steve Reinhardt
caccbd1edc Get rid of bogus bus assertion.
It runs out that if a MemObject turns around and does a send in its
receive callback, and there are other sends already scheduled, then
it could observe a state where it's not at the head of the list but
the bus's sendEvent is not scheduled (because we're still in the
middle of processing the prior sendEvent).
2008-06-13 01:33:49 -04:00
Steve Reinhardt
024ec4c5c3 Get rid of bogus cache assertion.
I was asserting that the only reason you would defer targets is if
a write came in while you had an outstanding read miss, but there's
another case where you could get a read access after you've snooped
an invalidation and buffered it because it applies to a prior
outstanding miss.
2008-06-13 01:29:20 -04:00
Ali Saidi
e71a5270a2 Make sure that output files are always checked success before they're used.
Make OutputDirectory::resolve() private and change the functions using
resolve() to instead use create().

--HG--
extra : convert_revision : 36d4be629764d0c4c708cec8aa712cd15f966453
2008-05-15 19:10:26 -04:00
Ali Saidi
8af6dc118c SCons: add comments to SConscript documenting bug workaround
--HG--
extra : convert_revision : e6cdffe953d56b96c76c7ff14d2dcc3de3ccfcc3
2008-04-10 15:38:10 -04:00
Ali Saidi
fe12f38353 PhysicalMemory: Add parameter for variance in memory delay.
--HG--
extra : convert_revision : b931472e81dedb650b7accb9061cb426f1c32e66
2008-04-10 14:44:52 -04:00
Ali Saidi
ed27c4c521 SCons: Manually specifying header only directories with Dir() works around the problem
--HG--
extra : convert_revision : d9713228d934cf4a45114a972603b8bca2bd27d3
2008-04-08 11:08:26 -04:00
Steve Reinhardt
29be31ce31 Fix handling of writeback-induced writebacks in atomic mode.
--HG--
extra : convert_revision : 4fa64f8a929f1aa36a9d5a71b8d1816b497aca4c
2008-03-25 10:01:21 -04:00
Steve Reinhardt
623dd7ed3a Delete the Request for a no-response Packet
when the Packet is deleted, since the requester
can't possibly do it.

--HG--
extra : convert_revision : 8571b144ecb3c70efc06d09faa8b3161fb58352d
2008-03-24 01:08:02 -04:00
Steve Reinhardt
93ab43288a Don't FastAlloc MSHRs since we don't allocate them on the fly.
--HG--
extra : convert_revision : 02775cfb460afe6df0df0938c62cccd93a71e775
2008-03-24 01:08:02 -04:00
Steve Reinhardt
407710d387 Fix cache problem with writes to tempBlock
getting wrong writeback address.

--HG--
extra : convert_revision : 023dfb69c227c13a69bfe2744c6af75a45828b0b
2008-03-22 22:17:15 -04:00
Steve Reinhardt
b051ae6acc Fix a few Packet memory leaks.
--HG--
extra : convert_revision : 00db19f0698c0786f0dff561eea9217860a5a05a
2008-03-17 03:08:28 -04:00
Steve Reinhardt
131c65f429 Restructure bus timing calcs to cope with pkt being deleted by target.
--HG--
extra : convert_revision : db8497e73a44f2a06aab121e797e88b4c0c31330
2008-03-17 03:07:38 -04:00
Steve Reinhardt
19c367fa8f Fix subtle cache bug where read could return stale data
if a prior write miss arrived while an even earlier
read miss was still outstanding.

--HG--
extra : convert_revision : 4924e145829b2ecf4610b88d33f4773510c6801a
2008-03-15 05:03:55 -07:00
Steve Reinhardt
e6d6adc731 Revamp cache timing access mshr check to make stats sane again.
--HG--
extra : convert_revision : 37009b8ee536807073b5a5ca07ed1d097a496aea
2008-02-26 22:03:28 -08:00
Steve Reinhardt
bdf3323915 Cache: better comments particularly regarding writeback situation.
--HG--
extra : convert_revision : 59ff9ee63ee0fec5a7dfc27b485b737455ccf362
2008-02-26 20:17:26 -08:00
Gabe Black
ec1a4cbbc7 Bus: Fix the bus timing to be more realistic.
--HG--
extra : convert_revision : acd70dc98ab840e55b114706fbb6afb2a95e54bc
2008-02-26 02:20:08 -05:00
Steve Reinhardt
4597a71cef Make L2+ caches allocate new block for writeback misses
instead of forwarding down the line.

--HG--
extra : convert_revision : b0d6e7862c92ea7a2d21f817d30398735e7bb8ba
2008-02-16 14:58:03 -05:00
Nicolas Zea
4c7eb21119 Bus: Only update port cache when there is an item to update it with.
--HG--
extra : convert_revision : 84848fd48bb9e6693a0518c862364142b1969aa8
2008-02-10 19:41:03 -05:00
Steve Reinhardt
9d7a69c582 Fix #include lines for renamed cache files.
--HG--
extra : convert_revision : b5008115dc5b34958246608757e69a3fa43b85c5
2008-02-10 14:45:25 -08:00
Steve Reinhardt
d56e77c180 Rename cache files for brevity and consistency with rest of tree.
--HG--
rename : src/mem/cache/base_cache.cc => src/mem/cache/base.cc
rename : src/mem/cache/base_cache.hh => src/mem/cache/base.hh
rename : src/mem/cache/cache_blk.cc => src/mem/cache/blk.cc
rename : src/mem/cache/cache_blk.hh => src/mem/cache/blk.hh
rename : src/mem/cache/cache_builder.cc => src/mem/cache/builder.cc
rename : src/mem/cache/miss/mshr.cc => src/mem/cache/mshr.cc
rename : src/mem/cache/miss/mshr.hh => src/mem/cache/mshr.hh
rename : src/mem/cache/miss/mshr_queue.cc => src/mem/cache/mshr_queue.cc
rename : src/mem/cache/miss/mshr_queue.hh => src/mem/cache/mshr_queue.hh
rename : src/mem/cache/prefetch/base_prefetcher.cc => src/mem/cache/prefetch/base.cc
rename : src/mem/cache/prefetch/base_prefetcher.hh => src/mem/cache/prefetch/base.hh
rename : src/mem/cache/prefetch/ghb_prefetcher.cc => src/mem/cache/prefetch/ghb.cc
rename : src/mem/cache/prefetch/ghb_prefetcher.hh => src/mem/cache/prefetch/ghb.hh
rename : src/mem/cache/prefetch/stride_prefetcher.cc => src/mem/cache/prefetch/stride.cc
rename : src/mem/cache/prefetch/stride_prefetcher.hh => src/mem/cache/prefetch/stride.hh
rename : src/mem/cache/prefetch/tagged_prefetcher.cc => src/mem/cache/prefetch/tagged.cc
rename : src/mem/cache/prefetch/tagged_prefetcher.hh => src/mem/cache/prefetch/tagged.hh
rename : src/mem/cache/tags/base_tags.cc => src/mem/cache/tags/base.cc
rename : src/mem/cache/tags/base_tags.hh => src/mem/cache/tags/base.hh
rename : src/mem/cache/tags/Repl.py => src/mem/cache/tags/iic_repl/Repl.py
rename : src/mem/cache/tags/repl/gen.cc => src/mem/cache/tags/iic_repl/gen.cc
rename : src/mem/cache/tags/repl/gen.hh => src/mem/cache/tags/iic_repl/gen.hh
rename : src/mem/cache/tags/repl/repl.hh => src/mem/cache/tags/iic_repl/repl.hh
extra : convert_revision : ff7a35cc155a8d80317563c45cebe405984eac62
2008-02-10 14:15:42 -08:00
Stephen Hines
6cc1573923 Make the Event::description() a const function
--HG--
extra : convert_revision : c7768d54d3f78685e93920069f5485083ca989c0
2008-02-06 16:32:40 -05:00
Geoffrey Blake
f9c54d5a4b Temporary fix for ll/sc bug see flyspray task for more info:
http://www.m5sim.org/flyspray/task/197

Signed-off by: Ali Saidi <saidi@eecs.umich.edu>

--HG--
extra : convert_revision : cdeece7e3163de9abf2c6c7435f1bc93570fab81
2008-01-06 00:19:45 -05:00
Steve Reinhardt
6c5a3ab8b2 Add ReadRespWithInvalidate to handle multi-level coherence situation
where we defer a response to a read from a far-away cache A, then later
defer a ReadExcl from a cache B on the same bus as us.  We'll assert
MemInhibit in both cases, but in the latter case MemInhibit will keep
the invalidation from reaching cache A.  This special response tells
cache A that it gets the block to satisfy its read, but must immediately
invalidate it.

--HG--
extra : convert_revision : f85c8b47bb30232da37ac861b50a6539dc81161b
2008-01-02 15:22:38 -08:00
Steve Reinhardt
bf9b3821bd Mark cache-to-cache MSHRs as downstreamPending when necessary.
Don't mark upstream MSHR as pending if downstream MSHR is already in service.

--HG--
extra : convert_revision : e1c135ff00217291db58ce8a06ccde34c403d37f
2008-01-02 15:18:33 -08:00
Steve Reinhardt
538da9e24d Don't DPRINTF in the middle of a PrintReq.
--HG--
extra : convert_revision : 6358c014d14a19a34111c39827b05987507544bb
2008-01-02 14:42:42 -08:00
Steve Reinhardt
87e5fd1755 Bug fix: functional cache port now needs otherPort set.
--HG--
extra : convert_revision : fb007df73a77535a5dba19341f7b0b32e8c99548
2008-01-02 14:42:24 -08:00
Steve Reinhardt
cde5a79eab Additional comments and helper functions for PrintReq.
--HG--
extra : convert_revision : 7eadf9b7db8c0289480f771271b6efe2400006d4
2008-01-02 13:46:22 -08:00
Steve Reinhardt
3952e41ab1 Add functional PrintReq command for memory-system debugging.
--HG--
extra : convert_revision : 73b753e57c355b7e6873f047ddc8cb371c3136b7
2008-01-02 12:20:15 -08:00
Steve Reinhardt
659aef3eb8 Fix formatting and comments in cache_impl.hh
--HG--
extra : convert_revision : 26d71cca5420ad03e16bf174e15dabe7f902da41
2008-01-02 12:15:48 -08:00
Gabe Black
8a020d40d3 Make ports that aren't connected to anything fail more gracefully.
--HG--
extra : convert_revision : 3803b28fb2fdfd729f01f1a44df2ae02ef83a2fc
2007-11-28 14:39:19 -08:00
Ali Saidi
8026ecbb8e Memory: Cache the physical memory start and size so we don't need a dynamic cast on every access.
--HG--
extra : convert_revision : d6c3e93718991e7b68248242c80d8e6ac637ac51
2007-11-19 18:23:43 -05:00
Steve Reinhardt
7d83cf35e1 Tweak check for writable block fill.
--HG--
extra : convert_revision : c04281bcfc4cd23c7613aeccb21dc74452bcc951
2007-11-16 20:10:33 -08:00
Steve Reinhardt
f03a62008a Fix bug on exclusive response to ReadReq with pending WriteReq.
--HG--
extra : convert_revision : 5429cd7ca84cf6348813a4607fa16f76aa5df7e0
2007-11-16 20:10:32 -08:00
Korey Sewell
375ddf8d25 branch merge
--HG--
extra : convert_revision : 1c56f3c6f2c50d642d2de5ddde83a55234455cec
2007-11-15 00:14:20 -05:00
Ali Saidi
7c8e4ca3a3 Checkpointing: Name SE page table entries better so that there isn't a problem if multiple workloads are being run at once.
--HG--
extra : convert_revision : 3bac9bd7fd93fcadf764e2991c5b029f2c745c08
2007-11-14 23:42:08 -05:00
Korey Sewell
bfdd2f379b remove unnecessary debug messages I added
--HG--
extra : convert_revision : 5c23218fd1b899fa7fe42701f7cb2f6033f7a583
2007-11-14 06:18:58 -05:00
Korey Sewell
2692590049 Add in files from merge-bare-iron, get them compiling in FS and SE mode
--HG--
extra : convert_revision : d4e19afda897bc3797868b40469ce2ec7ec7d251
2007-11-13 16:58:16 -05:00
Ali Saidi
c8123cef1b Cache: Fix for OS X 10.5 compiling.
--HG--
extra : convert_revision : ba3c33ed524367280eefc096177d767168ac2cf6
2007-11-04 18:57:36 -05:00
Ali Saidi
333ac6cc32 DRAM: Make latency parameters be Param.Latency instead of ints.
--HG--
extra : convert_revision : 553b86cc4653da089d7aa0045a3f3bdcabf6c4d8
2007-11-01 17:30:50 -04:00
Steve Reinhardt
90f42bf3ad Merge in bus DPRINTF changes.
--HG--
extra : convert_revision : 3bbd7c0745b31bb2a628b604ab1627cd9c61643c
2007-10-31 00:39:16 -07:00
Ali Saidi
538fae951b Traceflags: Add SCons function to created a traceflag instead of having one file with them all.
--HG--
extra : convert_revision : 427f6bd8f050861ace3bc0d354a1afa5fc8319e6
2007-10-31 01:21:54 -04:00
Gabe Black
fddfa71658 TLB: Fix serialization issues with the tlb entries and make the page table store the process, not the system.
--HG--
extra : convert_revision : 2421af11f62f60fb48faeee6bddadac2987df0e8
2007-10-25 19:04:44 -07:00
Ali Saidi
0711f4f17a SE: Fix page table and system serialization, don't reinit process if this is a checkpoint restore.
--HG--
extra : convert_revision : 03dcf3c088e57b7abab60efe700d947117888306
2007-10-25 20:13:35 -04:00
Steve Reinhardt
76c4c5fabc mem: clean up bus/cache DPRINTFs a bit
Not so much noise on failed sends, and more complete
info when grepping a trace using an address.

--HG--
extra : convert_revision : 05a8261c9452072ca08b906200c6322b33e2b9f1
2007-09-16 16:46:38 -07:00
Ali Saidi
bf7c01d43d Bus: Fix drain code; old method could return 1 in atomic mode and never call de->process().
--HG--
extra : convert_revision : 35990e5eaf93f7a95a0ec72e9f92034a042def3e
2007-09-05 17:12:41 -04:00
Miles Kaufmann
54cc0053f0 params: Deprecate old-style constructors; update most SimObject constructors.
SimObjects not yet updated:
- Process and subclasses
- BaseCPU and subclasses

The SimObject(const std::string &name) constructor was removed.  Subclasses
that still rely on that behavior must call the parent initializer as
  : SimObject(makeParams(name))

--HG--
extra : convert_revision : d6faddde76e7c3361ebdbd0a7b372a40941c12ed
2007-08-30 15:16:59 -04:00
Gabe Black
7227ab5f22 Merge with head
--HG--
extra : convert_revision : cc73b9aaf73e9dacf52f3350fa591e67ca4ccee6
2007-08-26 21:45:40 -07:00
Gabe Black
9b49a78cfd Address translation: Make the page table more flexible.
The page table now stores actual page table entries. It is still a templated
class here, but this will be corrected in the near future.

--HG--
extra : convert_revision : 804dcc6320414c2b3ab76a74a15295bd24e1d13d
2007-08-26 20:33:57 -07:00
Ali Saidi
02353a60ee MemorySystem: Fix the use of ?: to produce correct results.
--HG--
extra : convert_revision : 31aad7170b35556a4c984f4ebc013137d55d85eb
2007-08-12 19:43:54 -04:00
Ali Saidi
06a9f58c68 DMA: Add IOCache and fix bus bridge to optionally only send requests one
way so a cache can handle partial block requests for i/o devices.

--HG--
extra : convert_revision : a68b5ae826731bc87ed93eb7ef326a2393053964
2007-08-10 16:14:01 -04:00
Ali Saidi
5c38668ed6 Bus: Only call end() on an stl object once in a loop
--HG--
extra : convert_revision : 238dcd6da7577b533e52ada2107591c4e9168ebd
2007-08-10 16:14:01 -04:00
Vincentius Robby
3d40cba8d4 Port, StaticInst: Revert unnecessary changes.
--HG--
extra : convert_revision : e6ef262bbbc5ad53498e55caac1897e6cc2a61e6
2007-08-08 14:54:02 -04:00
Vincentius Robby
13d10e844c alpha: Make the TLB cache to actually work.
Improve MRU checking for StaticInst, Bus, TLB

--HG--
extra : convert_revision : 9116b5655cd2986aeb4205438aad4a0f5a440006
2007-08-08 14:18:09 -04:00
Vincentius Robby
1db9e1fb8f port: Implement cache for port interfaces and ranges
--HG--
extra : convert_revision : d7cbec7c277fb8f4d8846203caae36ce629602d5
2007-08-04 16:05:55 -04:00
Steve Reinhardt
62aa1d7f55 cache: get rid of obsolete params from python.
--HG--
extra : convert_revision : cd40e0ef938ef6da1cccedf7be01c3ac5b4883fb
2007-08-03 03:51:13 -04:00
Steve Reinhardt
2f93db6f95 memory system: fix functional access bug.
Make sure not to keep processing functional accesses
after they've been responded to.
Also use checkFunctional() return value instead of checking
packet command field where possible, mostly just for consistency.

--HG--
extra : convert_revision : 29fc76bc18731bd93a4ed05a281297827028ef75
2007-07-29 20:17:03 -07:00
Steve Reinhardt
4a7d0c4b79 bus: take out response prioritization (timing was messed up).
Also make express snoops not occupy bus (since they're magic).

--HG--
extra : convert_revision : 75aa5211a59380026d1e3f122778425e48e2edcd
2007-07-29 13:24:48 -07:00
Steve Reinhardt
8705b0799b packet: get rid of unused intersect() function.
--HG--
extra : convert_revision : f0a2947ccc49e0d18bc17a59371fa396d9ebd6c0
2007-07-27 12:46:55 -07:00
Steve Reinhardt
0cbcb715e0 cache/memtest: fixes for functional accesses.
--HG--
extra : convert_revision : 688ba4d882cad2c96cf44c9e46999f74266e02ee
2007-07-27 12:46:45 -07:00
Steve Reinhardt
01c9d34a0b cache: Get rid of unused variable.
--HG--
extra : convert_revision : 394adc12fbd7ea10280a1b8d6bc3cb15ee019f27
2007-07-27 03:51:15 -04:00
Nathan Binkert
f0fef8f850 Merge python and x86 changes with cache branch
--HG--
extra : convert_revision : e06a950964286604274fba81dcca362d75847233
2007-07-26 23:15:49 -07:00
Steve Reinhardt
6b73ff43ff Have owner respond to UpgradeReq to avoid race.
--HG--
extra : convert_revision : 30916fca6978c73d8a14558f2d7288c1eab54ad4
2007-07-26 17:04:17 -07:00
Steve Reinhardt
c3bf59dcb9 Add downward express snoops for invalidations.
--HG--
extra : convert_revision : 4916fa9721d727d8416ad8c07df3a8171d02b2b4
2007-07-26 17:04:17 -07:00
Steve Reinhardt
f1b5c8fb57 Continue snooping after a writeback is encountered.
--HG--
extra : convert_revision : 8411338a6c0fdd7072dd32bdffacdace62d5de90
2007-07-26 17:04:16 -07:00
Steve Reinhardt
58250b8e5f bus: Fix default port handling.
--HG--
extra : convert_revision : 121b6e31cddff17c51fc4f3df20e7e2bde87d04f
2007-07-26 17:04:12 -07:00
Steve Reinhardt
c1097d06f7 Can't block on memInhibit packets
(now that bus no longer filters them for us).

--HG--
extra : convert_revision : 34e7eaf5ee1e739f5557a2d417e569ed2ceb14b3
2007-07-25 07:47:37 -07:00
Steve Reinhardt
de52eebd3b Integrate snoop loop functions into their respective call sites.
Also some additional cleanup of Bus::recvTiming().

--HG--
extra : convert_revision : 156814119f75d04c2e954aec2d7ed6fdc186c26f
2007-07-24 22:37:41 -07:00
Steve Reinhardt
d094f9cf27 Don't delete request at target... requester still needs it.
--HG--
extra : convert_revision : 76377ca2e4d7ea70d1d54d325a63ce710e260b93
2007-07-24 22:36:10 -07:00
Steve Reinhardt
1f9ea6e122 A couple more minor bug fixes for multilevel coherence.
--HG--
extra : convert_revision : 370f9e34911157765be6fd49e826fa1af589b466
2007-07-23 22:28:40 -07:00
Nathan Binkert
abc76f20cb Major changes to how SimObjects are created and initialized. Almost all
creation and initialization now happens in python.  Parameter objects
are generated and initialized by python.  The .ini file is now solely for
debugging purposes and is not used in construction of the objects in any
way.

--HG--
extra : convert_revision : 7e722873e417cb3d696f2e34c35ff488b7bff4ed
2007-07-23 21:51:38 -07:00
Steve Reinhardt
82e2a35576 Replace lowerMSHRPending flag with more robust scheme
based on following Packet senderState links.

--HG--
extra : convert_revision : 9027d59bd7242aa0e4275bf94d8b1fb27bd59d79
2007-07-22 21:43:38 -07:00
Steve Reinhardt
1c2d5f5e64 Replace DeferredSnoop flag with LowerMSHRPending flag.
Turns out DeferredSnoop isn't quite the right bit of info
we needed... see new comment in cache_impl.hh.

--HG--
extra : convert_revision : a38de8c1677a37acafb743b7074ef88b21d3b7be
2007-07-22 08:09:24 -07:00
Steve Reinhardt
1edd143b68 A few minor non-debug compilation issues.
--HG--
extra : convert_revision : d59a5cad6187a2229dddd1a48282ebd2923d1d8d
2007-07-22 03:07:26 -04:00
Steve Reinhardt
92ce2b5974 Deal with invalidations intersecting outstanding upgrades.
If the invalidation beats the upgrade at a lower level
then the upgrade must be converted to a read exclusive
"in the field".
Restructure target list & deferred target list to
factor out some common code.

--HG--
extra : convert_revision : 7bab4482dd6c48efdb619610f0d3778c60ff777a
2007-07-21 18:18:42 -07:00
Steve Reinhardt
9117860094 Several more fixes for multi-level timing coherence.
- Add "deferred snoop" flag to Packet so upper-level caches
  can distinguish whether lower-level cache request was
  in-service or not at the time of the original snoop.
- Revamp response handling to properly handle deferred snoops
  on non-cache-fill requests (i.e. upgrades).
- Make sure forwarded writebacks are kept in write buffer at
  lower-level caches so they get snooped properly.

--HG--
extra : convert_revision : 17f8a3772a1ae31a16991a53f8225ddf54d31fc9
2007-07-21 13:45:17 -07:00
Steve Reinhardt
a67a0025b3 Make sure responses never get blocked.
--HG--
extra : convert_revision : 29f359d743994a94dc403aa0621ba72cd137d1a1
2007-07-17 08:15:23 -07:00
Steve Reinhardt
a25f3ac67f Forward cache-to-cache responses through other caches.
--HG--
extra : convert_revision : 5b6a02255bccd98b00949703cf4ba4b221553cea
2007-07-17 06:33:28 -07:00
Steve Reinhardt
ff13827ccb Assert that an mshr has a target in getTarget().
--HG--
extra : convert_revision : 08091670fc319876012ed139fcd2584c364a980c
2007-07-17 06:23:11 -07:00
Steve Reinhardt
884807a68a Fix up a bunch of multilevel coherence issues.
Atomic mode seems to work.  Timing is closer but not there yet.

--HG--
extra : convert_revision : 0dea5c3d4b973d009e9d4a4c21b9cad15961d56f
2007-07-15 20:11:06 -07:00
Steve Reinhardt
f790f34fe3 Make Bus::findPort() a little more useful.
Move check for loops outside, since half the call sites
end up working around it anyway.  Return integer port ID
instead of port object pointer.

--HG--
extra : convert_revision : 4c31fe9930f4d1aa4919e764efb7c50d43792ea3
2007-07-15 20:09:03 -07:00
Steve Reinhardt
15a51d0cae Add CacheRepl trace flag and move a couple DPRINTFs to it.
--HG--
extra : convert_revision : 31724d19ebdf2cdc2a2bafff83d17845b3a0b183
2007-07-14 13:28:52 -07:00
Steve Reinhardt
abd194df5c Move a couple of DPRINTFs from Cache to CachePort.
--HG--
extra : convert_revision : 55a0d26660aeb8f63b41897d53e6b2d1f0a163be
2007-07-14 13:16:58 -07:00
Steve Reinhardt
3b4ff75939 Fix bug in copying packet with static data pointer.
--HG--
extra : convert_revision : 2fcf99f050d73e007433c1db2475f2893c5961a0
2007-07-14 13:14:53 -07:00
Steve Reinhardt
a51e16dc89 Merge of DPRINTF fixes from head.
--HG--
extra : convert_revision : f99a33b2df6a6c5592856d17d00e73ee83267442
2007-07-14 12:09:37 -07:00
Steve Reinhardt
7cd6c7ee05 Fix & tweak DPRINTFs for tracediff w/new cache code.
Note that we should *not* print pointer values in DPRINTFs as
these needlessly clutter tracediff output.

--HG--
extra : convert_revision : 25a448f1b3ac8d453a717a104ad6dc0112fb30bb
2007-07-14 11:48:30 -07:00
Steve Reinhardt
4738649e32 Delete packets when we're done with them.
--HG--
extra : convert_revision : b8894d26e1ca7a6c9b736500accdaa53bfb09558
2007-07-03 00:40:31 -04:00
Steve Reinhardt
4b68652c87 Couple more minor bug fixes for FS timing mode.
src/cpu/simple/timing.cc:
    Fix another SC problem.
src/mem/cache/cache_impl.hh:
    Forgot to call makeTimingResponse() on uncached timing responses.

--HG--
extra : convert_revision : 5a5a58ca2053e4e8de2133205bfd37de15eb4209
2007-07-02 13:57:45 -07:00
Steve Reinhardt
e9c04dad60 Fix a couple LL/SC bugs that only affected timing mode.
src/cpu/simple/timing.cc:
    Fix swap/stq_c command bug.
src/mem/packet.cc:
    Fix incorrect LoadLockedReq command response field.

--HG--
extra : convert_revision : 7a4523be900bc2c9b1bdf2d372ce55f89ae58ae5
2007-07-02 09:26:36 -07:00
Steve Reinhardt
ffd697e149 bus.cc:
Fix atomic timing issue.

src/mem/bus.cc:
    Fix atomic timing issue.

--HG--
extra : convert_revision : a22ff80cd75f83c785b0604c2a4fde2e2e9f71ef
2007-07-02 01:02:35 -07:00
Steve Reinhardt
5e59739416 Don't propagate snoops across bridges. Wouldn't work anyway.
--HG--
extra : convert_revision : af29fc7d0c134f5e89dd2e814c819151350fcb38
2007-06-30 18:03:17 -07:00
Steve Reinhardt
07f091d6ed Get rid of remaining traces of obsolete CoherenceProtocol object.
--HG--
extra : convert_revision : c5555b00bef1b304a84886188ad2c0dcb4d7c5b9
2007-06-30 17:59:45 -07:00
Steve Reinhardt
d10a843723 Get rid of obsolete fixPacket() functions.
Handled by Packet::checkFunctional() now.

--HG--
extra : convert_revision : 63642254e2789c80a369ac269f317ec054ffe3c0
2007-06-30 17:51:29 -07:00
Steve Reinhardt
ee54ad318a Event descriptions should not end in "event"
(they function as adjectives not nouns)

--HG--
extra : convert_revision : 6506474ff3356ae8c80ed276c3608d8a4680bfdb
2007-06-30 17:45:58 -07:00
Steve Reinhardt
f0c4dd7920 Factor out a little more common code.
--HG--
extra : convert_revision : 626255a91679d534030c91bcdb4fc1bed36ceb9b
2007-06-30 13:56:25 -07:00
Steve Reinhardt
6babda7123 Fix up a few statistics problems.
Stats pretty much line up with old code, except:
- bug in old code included L1 latency in L2 miss time, making it too high
- UniCoherence did cache-to-cache transfers even from non-owner caches,
so occasionally the icache would get a block from the dcache not the L2
- L2 can now receive ReadExReq from L1 since L1s have coherence

--HG--
extra : convert_revision : 5052c1a1767b5a662f30a88f16012165a73b791c
2007-06-30 13:34:16 -07:00
Steve Reinhardt
6ab53415ef Get rid of Packet result field. Error responses are
now encoded in cmd field.

--HG--
extra : convert_revision : d67819b7e3ee4b9a5bf08541104de0a89485e90b
2007-06-30 10:16:18 -07:00
Steve Reinhardt
9117c94f9c Get rid of coherence protocol object.
--HG--
extra : convert_revision : 4ff144342dca23af9a12a2169ca318a002654b42
2007-06-27 20:54:13 -07:00
Steve Reinhardt
c4903e0882 Revamp replacement-of-upgrade handling.
--HG--
extra : convert_revision : 9bc09d8ae6d50e6dfbb4ab21514612f9aa102a2e
2007-06-26 23:30:30 -07:00