to do it:
in .ini file set binned_fns=foo null bar foo
what this says is, track foo when it is called by null (i.e. anything). bin bar only when it is called by foo.
essentially, if you have a path of functions to track, the 0th, 2nd, 4th fn listed are the fns to track, and the 1st, 3rd, 5th are their respective callers.
base/statistics.hh:
turn it back to FS_MEASURE since we already have a build in place
kern/tru64/tru64_events.cc:
remove FS_MEASURE #defs. add more DPRINTF's. manage an anomaly that happens when tracking idle_thread.
kern/tru64/tru64_events.hh:
remove FS_MEASURE #defs
kern/tru64/tru64_system.cc:
make DumpState print all the time, but only with DPRINTF. add a new parameter to tru64System a vector<string> binned_fns, to read in from .ini file. now all this binning stuff is dynamically generated.
kern/tru64/tru64_system.hh:
remove all static binning stuff, add support for dynamic
sim/system.cc:
change nonPath bin name to Kernel, remove FS_MEASURE
sim/system.hh:
change nonPath to Kernel
--HG--
extra : convert_revision : 9ee813c0a64273bab4125815b7bc8145c5897ec1
code.
base/statistics.hh:
We're getting rid of FS_MEASURE, but for now, we're going
to still use a compile time flag to turn on and off binning
of statistics. (The flag is STATS_BINNING)
cpu/exec_context.cc:
cpu/exec_context.hh:
kern/tru64/tru64_system.cc:
get rid of FS_MEASURE
cpu/simple_cpu/simple_cpu.cc:
yank the function call tracking code out of the cpu and move
it into the software context class itself.
kern/tru64/tru64_system.hh:
get rid of FS_MEASURE
move all of the tacking stuff to the same place.
sim/system.hh:
cleanup
--HG--
extra : convert_revision : 73d3843afe1b3ba0d5445421c39c1148d3f4e7c0
sim/sim_object.cc:
serialize objects in the reverse order of their creation.
This causes a SimObject that is dependent on another
to be serialized first. For example, the ethernet device
writes back some data to physical memory while serializing,
so this will cause the physical memory to be serialized
after that, preserving the data written back.
--HG--
extra : convert_revision : f9a37b63ce777df1cfecefa80f94f8fc69e42448
benchmarks for alpha-linux.
arch/alpha/alpha_linux_process.cc:
Added some more ioctl commands to ignore.
Set unlink and rename to the new functions.
Ignore setrlimit, times and rt_sigaction.
Should eventually provide a function for times.
arch/alpha/alpha_tru64_process.cc:
Added some more ioctl commands to ignore.
Set unlink and rename to the new functions.
Ignore setrlimit.
sim/syscall_emul.cc:
Added implementations for unlink and rename.
sim/syscall_emul.hh:
Added unlink and rename functions.
Added a couple more ioctl requests to ignore.
Print out the PC of any ioctl commands that fail.
--HG--
extra : convert_revision : 8af21c7fa7d0645d3f9324c9ce70ad33590c3c8e
execution pipeline (Alpha trapb & excb).
Add support for write memory barriers (mostly impacts
store buffer).
Add StaticInst flag to indicate memory barriers, though
this is not modeled in the pipeline yet.
arch/alpha/isa_desc:
Implement trapb, excb, mb, and wmb as insts with
no execution effect (empty execute() function) but
with flags that indicate their side effects.
Also make sure every instruction that needs to go to
the execute stage has a real opClass value, since we
are now using No_OpClass to signal insts that can get
dropped at dispatch.
StaticInst::branchTarget() is now a const method.
cpu/static_inst.hh:
Add flags to indicate serializing insts (trapb, excb) and
memory and write barriers.
Also declare some StaticInst methods as const methods.
dev/etherlink.hh:
sim/eventq.hh:
sim/serialize.cc:
sim/serialize.hh:
sim/sim_object.hh:
Make name() return value const.
--HG--
extra : convert_revision : 39520e71469fa20e0a7446b2e06b494eec17a02c
The last change only caught the ones with types that
started with capitals. This pass catches the rest
(mostly STL and uint*_t types).
base/cprintf_formats.hh:
cpu/simple_cpu/simple_cpu.cc:
sim/serialize.cc:
sim/serialize.hh:
Change "foo_t& foo" to "foo_t &foo".
--HG--
extra : convert_revision : fc7f7425db2aef33e490f952b5ce74c8c36d0d41
This primarily to be internally consistent (sometimes we used one,
sometimes the other, even within the same line of code!).
I picked the latter to be symmetric with "Foo *foo".
base/cprintf_formats.hh:
base/range.hh:
base/refcnt.hh:
base/res_list.hh:
base/statistics.hh:
base/str.hh:
cpu/exec_context.hh:
cpu/simple_cpu/simple_cpu.cc:
cpu/simple_cpu/simple_cpu.hh:
sim/serialize.cc:
sim/serialize.hh:
sim/syscall_emul.hh:
Change "Foo& foo" declarations to "Foo &foo".
--HG--
extra : convert_revision : ca1b0e85a578b539214bda3b8d61ac23792f2e87
statStream catchall that we had before)
Also provide an optional output directory that multiple simulator output
files can be written to.
Make most output files use the output directory
base/misc.cc:
send warnings to the outputStream as well
base/trace.cc:
dprintf_stream defaults to cerr
dev/console.cc:
use the output directory for the console output if it exists
dev/etherdump.cc:
dump to the output directory if it exists
sim/builder.cc:
sim/builder.hh:
move constructor and destructor to .cc file
use a function to get the stream that the builder dumps its
output to, and create a separate file in the output directory
if able
sim/main.cc:
statStream -> outputStream
sim/serialize.cc:
dump checkpoints to the output directory if specified
sim/universe.cc:
provide an output stream for simulator output. (This takes place of the
statStream catchall that we had before)
Also provide an optional output directory that multiple simulator output
files can be written to.
--HG--
extra : convert_revision : 03abce20edbbf7ec19c9ddd8d69ec8485c383532
arch/alpha/isa_desc:
don't say warn: Warning:
base/misc.cc:
avoid printing two newlines in a row
sim/main.cc:
print out a message just before we enter the event queue
--HG--
extra : convert_revision : 2a824d4b67661903fc739a0fb0759aa91d72382c
Move global checkpoint-related functions and vars into Checkpoint class (as statics).
arch/alpha/pseudo_inst.cc:
dev/disk_image.cc:
Move global checkpoint-related functions and vars
into Checkpoint class (as statics).
sim/serialize.cc:
Move global checkpoint-related functions and vars
into Checkpoint class (as statics).
Checkpoint constructor now takes checkpoint directory name instead
of file name.
Make serialize:dir parameter actually set checkpoint directory name
instead of directory in which checkpoint directory is created. If
the value contains a '%', the curTick value is sprintf'd into the
format to create the directory name. The default is backwards compatible
with the old fixed name ("m5.%012d").
sim/serialize.hh:
Move global checkpoint-related functions and vars
into Checkpoint class (as statics).
Checkpoint constructor now takes checkpoint directory name instead
of file name.
--HG--
extra : convert_revision : d0aa87b62911f405a4f5811271b9e6351fdd9fe4
base/str.hh:
this should really be inline
base/time.cc:
base/time.hh:
clean up how the time class works. Export only one, and let
people calculate their own elapsed times, etc.
sim/main.cc:
sim/sim_time.X -> base/time.X
--HG--
rename : sim/sim_time.cc => base/time.cc
rename : sim/sim_time.hh => base/time.hh
extra : convert_revision : f3888fe3a1fdd1022084c282b58407c631a6d9a0
This lets us centralize priorities so we can see what's going on.
- Shift serialize & cpu-switch events to happen before CPU ticks (to be
consistent with starting new CPU on same cycle instead of next cycle).
- Get rid of unnecessary bus stats reset callback.
cpu/simple_cpu/simple_cpu.cc:
sim/debug.cc:
sim/eventq.hh:
sim/serialize.cc:
sim/sim_events.cc:
sim/sim_events.hh:
Switch events to use a priority enum instead of integers.
This lets us centralize priorities so we can see what's going on.
--HG--
extra : convert_revision : 510d79b43c0a1c97a10eb65916f7335b1de8b956
a checkpoint now gives identical results to running from scratch
and doing at switchover at the same cycle!
- CPUs start at cycle 0 again, not cycle 1.
- curTick is now serialized & unserialized.
- Stats get reset in main (before event loop). Since this is done
after curTick is unserialized, simTicks gets set correctly for
running from a checkpoint.
- Simplify serialization to happen in a single pass.
- s/Serializeable/Serializable/
arch/alpha/isa_traits.hh:
dev/etherlink.hh:
sim/eventq.cc:
sim/eventq.hh:
s/Serializeable/Serializable/
kern/tru64/tru64_system.cc:
sim/process.cc:
Make initial CPU activation on cycle 0 again (not 1).
sim/main.cc:
Reset stats before getting started.
Make error message on falling out of event loop
more meaningful.
sim/serialize.cc:
sim/serialize.hh:
Get rid of now-useless initial pass; serialization is
done in a single pass now.
Serialize & unserialize curTick.
Wrap curTick and mainEventQueue in a "globals" Serializable object.
s/Serializeable/Serializable/
sim/sim_object.cc:
Add static function to serialize all SimObjects.
sim/sim_object.hh:
Add static function to serialize all SimObjects.
s/Serializeable/Serializable/
--HG--
extra : convert_revision : 9dcc411d0009b54b8eb61c3a509680b81b9f6f68
and SimpleCPU::setStatus() into separate functions. For example,
setStatus(Active) is now activate().
--HG--
extra : convert_revision : 4392e07caf6c918db0b535f613175109681686fe
on test5 because of a wrong exit code; fixed that.
sim/main.cc:
No need to ignore SIGPIPE, as far as I know.
sim/sim_events.cc:
A CountedExitEvent is a normal termination, so
should have an exit code of 0.
--HG--
extra : convert_revision : 8b5072aca54f3ca08b2815b73ac01c00e4da49b6
I noticed that checkpoints dumped from m5.debug and m5.opt had
differences, which is no longer the case (other than addresses
encoded in names).
cpu/full_cpu/smt.hh:
Get rid of unused SMT_MAX_CPUS.
sim/debug.cc:
Use AutoDelete for DebugBreakEvent.
sim/serialize.cc:
sim/serialize.hh:
Move constructor & destructor to header so they can be inlined.
--HG--
extra : convert_revision : bc68c2c9b053b1de0d655ed555734419fafd0b83
interfaces, and specific support for Alpha Linux. Split syscall emulation
functions into several groups, based on whether they depend on the specific
OS and/or architecture (and all combinations of above), including the use of
template functions to support syscalls with slightly different constants
or interface structs.
arch/alpha/alpha_tru64_process.cc:
Incorporate full Tru64 object definition here, including structure and constant definitions.
This way we can wrap all of the functions inside the object, and not worry about namespace
conflicts because no one outside this file will ever see it.
base/loader/aout_object.cc:
base/loader/aout_object.hh:
base/loader/ecoff_object.cc:
base/loader/ecoff_object.hh:
base/loader/elf_object.cc:
base/loader/elf_object.hh:
base/loader/object_file.cc:
base/loader/object_file.hh:
Add enums to ObjectFile to indicate the object's architecture and operating system.
cpu/exec_context.cc:
prog.hh is now process.hh
cpu/exec_context.hh:
prog.hh is now process.hh
move architecture-specific syscall arg accessors into ExecContext
cpu/simple_cpu/simple_cpu.cc:
No need to include prog.hh (which has been renamed)
sim/process.cc:
sim/process.hh:
LiveProcess is now effectively an abstract base class.
New LiveProcess::create() function takes an object file and dynamically picks the
appropriate subclass of LiveProcess to handle the syscall interface that file expects
(currently Tru64 or Linux).
--HG--
rename : arch/alpha/fake_syscall.cc => arch/alpha/alpha_tru64_process.cc
rename : sim/prog.cc => sim/process.cc
rename : sim/prog.hh => sim/process.hh
extra : convert_revision : 4a03ca7d94a34177cb672931f8aae83a6bad179a
no need for multiple bin classes. If multiple bins are needed, we
can always do it with ini type config instead.
kern/tru64/tru64_events.hh:
sim/system.cc:
sim/system.hh:
No more GenBin always use MainBin
--HG--
extra : convert_revision : 8c466f302324c33b59d47d0da04583b2517fc72c
there aren't so many files littering your directory
dev/disk_image.cc:
Checkpoints now in a directory
sim/serialize.hh:
Make it so that we create a directory for each checkpoint so that
there aren't so many files littering your directory.
Remove unused variable
--HG--
extra : convert_revision : 261824eee62f7b68f6ae6e3dbd49ad5128ced148
base/statistics.cc:
formatting
cpu/simple_cpu/simple_cpu.cc:
cpu/simple_cpu/simple_cpu.hh:
Make numInsts reset by adding a resetStats function
sim/sim_object.cc:
Register the reset callback in a slightly cleaner way to avoid
potential static member constructor ordering issues
--HG--
extra : convert_revision : 408073b4b0397fbf9dfd9c548a313f1c8c3fc031
minor gratuitous cleanup in printAllExtraOutput.
(only create one end iterator)
Fix average stats reset
base/statistics.hh:
Shouldn't reset average stats this way. The current stat value
should stay the same.
sim/sim_object.cc:
Ok, actually call resetStats on all stats
minor gratuitous cleanup in printAllExtraOutput.
(only create one end iterator)
--HG--
extra : convert_revision : 13090ebe490a93757b8eb7d7c6a9697983095e41
from that will get called any time stats are reset.
sim/sim_object.cc:
sim/sim_object.hh:
add a virtual function resetStats that any simObject can reset
when a statistics reset is initiated
--HG--
extra : convert_revision : fdad673142f6ff811f84c246d80e5d41e3c9d4d1
make SIGUSR2 dump and reset stats
Make resetting time work
base/statistics.cc:
Fix statistics reset so that it works again, and correctly
reset bins as well. (The old code wouldn't reset if you didn't
have any bins, and then would actually only reset the first
bin)
cpu/simple_cpu/simple_cpu.cc:
cpu/simple_cpu/simple_cpu.hh:
convert idleCycles/idleFraction into a single Average stat
to make reset work more simply
sim/main.cc:
handle SIGUSR2 to dump and reset stats
(SIGUSR1 only dumps them)
sim/sim_time.cc:
sim/sim_time.hh:
Add support for resetting the time
--HG--
extra : convert_revision : ea43e03c50c0a4bb826dc0842a8c4fa1a9289e0a
command line:
m5.* <config file> <args> --server.system:bin=true to track function calls in the server
m5.* <config file> <args> --client.system:bin=true to track function calls in the client
base/statistics.cc:
make an adjustment to the way stats are printed for FS_MEASURE
base/statistics.hh:
add a name() virtual function to GenBin. add a debug printf for activate().
add amake MainBin the default bin when FS_MEASURE.
cpu/exec_context.cc:
initialize swCtx to null upon creation of an xc
cpu/exec_context.hh:
add a SWContext pointer to every execution context.
cpu/simple_cpu/simple_cpu.cc:
process calls and returns for FS_MEASURE
cpu/simple_cpu/simple_cpu.hh:
add this so idleCycles will not be accessed before all stats are constructed
kern/tru64/tru64_events.cc:
add a FnEvent that fires whenever a function we're tracking is called. implement the process() virtual function for it.
kern/tru64/tru64_events.hh:
add FnEvent
kern/tru64/tru64_system.cc:
send bin parameter to System constructor. add bin parameter to Tru64System object. initialize all the FnEvent and MainBin members of Tru64system. also, populate the calling map that indicates whether a function call is on the path we're tracking.
kern/tru64/tru64_system.hh:
modify the Tru64System class to support FS_MEASURE
sim/system.cc:
add a bin parameter to System class. initialize a MainBin to hold the stats for nonPath.
sim/system.hh:
add a map of to match bins to function names. add a swCtx map to map pcb addresses to SWContext *s. Add some supporting functions.
--HG--
extra : convert_revision : af3eadd798cb2d2aed9b54e1059dcedf244dd526
object in the config that was not in the checkpointed config.
This code compiles, but I haven't tested it... I'm committing it so Ron
can have it. Should not effect anything that currently works.
base/inifile.cc:
base/inifile.hh:
Add sectionExists() method so you can query whether a section exists
without knowing any of the entry names that would be in it.
sim/serialize.cc:
sim/serialize.hh:
Add Checkpoint::sectionExists() (pass through to IniFile).
--HG--
extra : convert_revision : 905db122afdfe55946ab8493ccac0b1e715bc7c6
arch/alpha/isa_desc:
Move the pseudo instructions out of the isa_desc, into their own
file and call out to them when they're to be accessed
sim/sim_events.cc:
sim/sim_events.hh:
sim/sim_exit.hh:
move SimExit to sim_exit.cc
--HG--
extra : convert_revision : 1c393adb1c18bd0fef065057d7f4e9cf60ac4197
resetstats
dumpstats
dumpresetstats
m5checkpoint
Lots of cleanup of serialization and stats dumping/resetting to
work with these new instructions
arch/alpha/isa_desc:
Implement more m5 pseduo opcodes:
resetstats
dumpstats
dumpresetstats
m5checkpoint
All of these functions take two optional parameters, the first is a delay,
and the second is a period. The delay tells the simulator to wait the
specified number of nanoseconds before triggering the event, the period
tells the simulator to repeat the event with a specified frequency
base/statistics.cc:
base/statistics.hh:
regReset RegResetCallback
dev/disk_image.cc:
serializeFilename -> CheckpointFile()
sim/debug.cc:
Move this debugging statement to sim_stats.cc
sim/eventq.cc:
Don't AutoDelete an event if it is scheduled since the process()
function could potentially schedule the event again.
sim/main.cc:
DumpStatsEvent is now Statistics::SetupEvent(Dump, curTick)
sim/serialize.cc:
Change the serialize event so that it's possible to cause the
event to repeat. Also make the priority such that the event
happens just before the simulator would exit if both events
were scheduled for the same cycle.
get rid of the serializeFilename variable and provide a CheckpointFile()
function. This function takes a basename that is set in the
configuration, and appends the current cycle to the name so that
multiple checkpoints can be dumped from the same simulation.
Also, don't exit the simulation when a checkpoint file is dumped.
sim/serialize.hh:
serializeFilename -> CheckpointFile()
SetupCheckpoint function to tell the simulator to prepare
to checkpoint at a certain time with a certain period
sim/sim_events.cc:
DumpStatsEvent stuff gets move to sim_stats.(cc|hh)
The context stuff gets moved into the already existing
stats context in stat_context.cc
sim/sim_events.hh:
DumpStatsEvent stuff gets move to sim_stats.(cc|hh)
sim/universe.cc:
Provide some simple functions for converting times into
ticks. These use floating point math to get as close as
possible to the real values. Multipliers are set up ahead
of time
--HG--
extra : convert_revision : d06ef26a9237529a1e5060cb1ac2dcc04d4ec252
Implement a new m5exit instruction with an optional delay
arch/alpha/isa_desc:
move m5exit to m5exit old. The old version of the
instruction is now deprecated
Implement the new exit instruction with the optional delay
sim/sim_events.cc:
sim/sim_events.hh:
Make SimExit take a cycle
sim/universe.cc:
provide ticksPerMS, ticksPerUS, and ticksPerNS so we don't
have to do math during the cycle
--HG--
extra : convert_revision : e2ed47a2e5cfcd57c82086c6fcb4a28bf801c214
Cleaned up serialization
sim/eventq.hh:
sim/eventq.cc:
Cleaned up serialization
--HG--
extra : convert_revision : b75696d75f1aee16ebca2076fdd3cd4913593762
objName as a member, instead it has the pure virtual function name(). SimObject
now has a objName member, and all classes derived directly from Serializeable
have to implement a name() function (which now makes them unique by pointer value)
cpu/simple_cpu/simple_cpu.cc:
Change initialization of Event to get rid of Serializeable naming
dev/etherlink.cc:
dev/etherlink.hh:
Seralizeable derived naming changes
sim/eventq.cc:
Serializeable derived naming changes, also changed serialization process so it
doesn't need to use nameChildren
sim/eventq.hh:
Serializeable derived naming changes, remove constructor for specifying event name
sim/serialize.cc:
Serializeable derived naming changes, remove setName function and the child naming
pass for serialization
sim/serialize.hh:
Serializeable derived naming changes, removed nameChildren, setName
sim/sim_object.cc:
sim/sim_object.hh:
Serializeable derived naming changes
--HG--
extra : convert_revision : 67bcc275b6c210f7049f98a1ad0d22e8f5596a63
base/predictor.hh:
base/sat_counter.hh:
sim/param.hh:
Get rid of spurious references to stat_sdb_t.
--HG--
extra : convert_revision : 0018a940c69b6e31b85fd85354b9d5ce2fd0aa6f
Can now serialize & unserialize DmaRequestEvents and DmaTransferEvents.
Also support serialize/unserialize of pointers to SimObjects and
other Serializable objects.
arch/alpha/alpha_memory.cc:
arch/alpha/alpha_memory.hh:
arch/alpha/isa_traits.hh:
cpu/exec_context.cc:
cpu/exec_context.hh:
cpu/simple_cpu/simple_cpu.hh:
dev/alpha_access.h:
dev/alpha_console.cc:
dev/alpha_console.hh:
dev/console.cc:
dev/console.hh:
unserialize() now takes a Checkpoint* instead of an IniFile*.
cpu/simple_cpu/simple_cpu.cc:
unserialize() now takes a Checkpoint* instead of an IniFile*.
Put ExecContext in its own section so its _status fields doesn't conflict.
sim/eventq.cc:
sim/eventq.hh:
unserialize() now takes a Checkpoint* instead of an IniFile*.
Events get serialized by the event queue only if they're marked
as AutoSerialize... others are assumed to be serialized by something
else (e.g. an owning SimObject) or to not matter.
sim/param.cc:
Shift 'const' in case T is a ptr type.
sim/serialize.cc:
sim/serialize.hh:
Define Checkpoint object to encapsulate everything you need to know
about a checkpoint. Use it to allow lookups of named Serializable
objects (and SimObjects) during unserialization.
unserialize() now takes a Checkpoint* instead of an IniFile*.
--HG--
extra : convert_revision : 8e6baab32405f8f548bb67a097b2f713296537a5
arch/alpha/alpha_memory.cc:
arch/alpha/alpha_memory.hh:
Serialize TLB contents.
cpu/simple_cpu/simple_cpu.cc:
cpu/simple_cpu/simple_cpu.hh:
Complete serialization of SimpleCPU (including owned events).
sim/eventq.cc:
sim/eventq.hh:
Basic serialization for events.
Still need to handle dynamic events (not owned by a SimObject).
sim/serialize.cc:
sim/serialize.hh:
Export serialization filename so PhysicalMemory can
derive its filename from that.
--HG--
extra : convert_revision : 4db851c5880f73f576ca092d5e5ad4256048eb51
Add support for serializing enums.
arch/alpha/isa_traits.hh:
Add serialize/unserialize functions for RegFile
(defined in new isa_traits.cc).
cpu/exec_context.cc:
Flesh out serialize/unserialize.
sim/serialize.hh:
Add {UN}SERIALIZE_ENUM().
--HG--
extra : convert_revision : 9e30c7e7b3b290dc8ea0888ba3636fc93ee89052
