This patch adds a mechanism to collect run time samples for specific portions
of a benchmark, using work_begin and work_end pseudo instructions.It also enhances
the histogram stat to report geometric mean.
This patch adds a function for replacing the event at the head of the queue
with another event. This helps in running a different set of events. Events
already scheduled can processed by replacing the original head event back.
This function has been specifically added to support cache warmup and
cooldown required for creating and restoring checkpoints.
--HG--
extra : rebase_source : ed6e2905720b6bfdefd020fab76235ccf33d28d1
This parameter depends on a number of coincidences to work properly. First,
there must be an array assigned to system called "cpu" even though there's no
parameter called that. Second, the items in the "cpu" array have to have a
"clock" parameter which has a "frequency" member. This is true of the normal
CPUs, but isn't true of the memory tester CPUs. This happened to work before
because the memory tester CPUs were only used in SE mode where this parameter
was being excluded. Since everything is being pulled into a common binary,
this won't work any more. Since the boot_cpu_frequency parameter is only used
by Alpha's Linux System object (and Mips's through copy and paste), the
definition of that parameter is moved down to those objects specifically.
PageTable supported an allocate() call that called back
through the Process to allocate memory, but did not have
a method to map addresses without allocating new pages.
It makes more sense for Process to do the allocation, so
this method was renamed allocateMem() and moved to Process,
and uses a new map() call on PageTable.
The remaining uses of the process pointer in PageTable
were only to get the name and the PID, so by passing these
in directly in the constructor, we can make PageTable
completely independent of Process.
Replace the (broken as of previous changeset) swig_objdecl() method
that allowed/forced you to substitute a whole new C++ struct
definition for SWIG to wrap with a set of export_method* hooks
that let you just declare a set of C++ methods (or other declarations)
that get inserted in the auto-generated struct.
Restore the System get/setMemoryMode methods, and use this mechanism
to specialize SimObject as well, eliminating teh need for sim_object.i.
Needed bits of sim_object.i are moved to the new pyobject.i.
Also sucked a little SimObject specialization into cxx_param_decl()
allowing us to get rid of src/sim/sim_object_params.hh. Now the
generation and wrapping of the base SimObject param struct is more
in line with how derived objects are handled.
--HG--
rename : src/python/swig/sim_object.i => src/python/swig/pyobject.i
- Move the random bits of SWIG code generation out of src/SConscript
file and into methods on the objects being wrapped.
- Cleaned up some variable naming and added some comments to make
the process a little clearer.
- Did a little generated file/module renaming:
- vptype_Foo now Foo_vector
- init_Foo is now Foo_init
This makes it easier to see all the Foo-related files in a
sorted directory listing.
- Made cxx_predecls and swig_predecls normal SimObject classmethods.
- Got rid of swig_objdecls hook, even though this breaks the System
objects get/setMemoryMode method exports. Will be fixing this in
a future changeset.
In order for a system object to work in SE mode and FS mode, it has to either
always require a platform object even in SE mode, or get rid of the
requirement all together. Making SE mode carry around unnecessary/unused bits
of FS seems less than ideal, so I decided to go with the second option. The
platform pointer in the System class was used for exactly one purpose, a path
for the Alpha Linux system object to get to the real time clock and read its
frequency so that it could short cut the loops_per_jiffy calculation. There
was also a copy and pasted implementation in MIPS, but since it was only there
because it was there in Alpha I still count that as one use.
This change reverses the mechanism that communicates the RTC frequency so that
the Tsunami platform object pushes it up to the AlphaSystem object. This is
slightly less specific than it could be because really only the
AlphaLinuxSystem uses it. Because the intrFrequency function on the Platform
class was no longer necessary (and unimplemented on anything but Alpha) it was
eliminated.
After this change, a platform will need to have a system, but a system won't
have to have a platform.
All of the classes will now be available in both modes, and only
GenericPageTableFault will continue to check the mode for conditional
compilation. It uses a process object to handle the fault in SE mode, and
for now those aren't available in FS mode.
This constant will have the same value as FULL_SYSTEM but will not be usable
by the preprocessor. It can be substituted into places where FULL_SYSTEM is
used in a C++ context and will make it easier to find which parts of the
simulator still use FULL_SYSTEM with the preprocessor using grep.
Initialize flags via the Event constructor instead of calling
setFlags() in the body of the derived class's constructor. I
forget exactly why, but this made life easier when implementing
multi-queue support.
Also rename Event::getFlags() to isFlagSet() to better match
common usage, and get rid of some unused Event methods.
Use exitSimLoop() function instead of explicitly scheduling
on mainEventQueue (which won't work once we go to multiple
event queues). Also introduced a local params variable to
shorten a lot of expressions.
It was technically possible but clumsy to determine what endianness a guest
was configured with using the state in byteswap.hh. This change makes that
information available more directly.
Also get rid of unused (and mildly redundant) ByteOrderDiffers constant.
Only create a memory ordering violation when the value could have changed
between two subsequent loads, instead of just when loads go out-of-order
to the same address. While not very common in the case of Alpha, with
an architecture with a hardware table walker this can happen reasonably
frequently beacuse a translation will miss and start a table walk and
before the CPU re-schedules the faulting instruction another one will
pass it to the same address (or cache block depending on the dendency
checking).
This patch has been tested with a couple of self-checking hand crafted
programs to stress ordering between two cores.
The performance improvement on SPEC benchmarks can be substantial (2-10%).
Do some minor cleanup of some recently added comments, a warning, and change
other instances of stack extension to be like what's now being done for x86.
Nothing big here, but when you have an address that is not in the page table request to be allocated, if it falls outside of the maximum stack range all you get is a page fault and you don't know why. Add a little warn() to explain it a bit. Also add some comments and alter logic a little so that you don't totally ignore the return value of checkAndAllocNextPage().
We were getting a spurious warning in the regressions that turned
out to be due to having the wrong value for TGT_MAP_ANONYMOUS for
Power Linux, but in the process of tracking it down I ended up
doing some cleanup of the mmap handling in general.
This is similar to guards on mercurial queues and they're used for selecting
which files are compiled into some given object. We already do something
similar, but it's mostly hard coded for the m5 binary and the m5 library
and I'd like to make it more flexible to better support the unittests
At the same time, rename the trace flags to debug flags since they
have broader usage than simply tracing. This means that
--trace-flags is now --debug-flags and --trace-help is now --debug-help
This patch prevents not executed conditional instructions marked as
IsQuiesce from stalling the pipeline indefinitely. If the instruction
is not executed the quiesceSkip psuedoinst is called which schedules a
wakes up call to the fetch stage.
Some ISAs (like ARM) relies on hardware page table walkers. For those ISAs,
when a TLB miss occurs, initiateTranslation() can return with NoFault but with
the translation unfinished.
Instructions experiencing a delayed translation due to a hardware page table
walk are deferred until the translation completes and kept into the IQ. In
order to keep track of them, the IQ has been augmented with a queue of the
outstanding delayed memory instructions. When their translation completes,
instructions are re-executed (only their initiateAccess() was already
executed; their DTB translation is now skipped). The IEW stage has been
modified to support such a 2-pass execution.
Setup initial timesync event in initState or loadState so that curTick has
been updated to the new value, otherwise the event is scheduled in the past.
Moving the definition of NoFault into fault.hh doesn't bring any new
dependencies with it, and allows some files to include just fault.hh which has
less baggage. NoFault will still be available to everything that includes
faults.hh because it includes fault.hh.
M5 skips over any simulated time where it doesn't have any work to do. When
the simulation is active, the time skipped is short and the work done at any
point in time is relatively substantial. If the time between events is long
and/or the work to do at each event is small, it's possible for simulated time
to pass faster than real time. When running a benchmark that can be good
because it means the simulation will finish sooner in real time. When
interacting with the real world through, for instance, a serial terminal or
bridge to a real network, this can be a problem. Human or network response time
could be greatly exagerated from the perspective of the simulation and make
simulated events happen "too soon" from an external perspective.
This change adds the capability to force the simulation to run no faster than
real time. It does so by scheduling a periodic event that checks to see if
its simulated period is shorter than its real period. If it is, it stalls the
simulation until they're equal. This is called time syncing.
A future change could add pseudo instructions which turn time syncing on and
off from within the simulation. That would allow time syncing to be used for
the interactive parts of a session but then turned off when running a
benchmark using the m5 utility program inside a script. Time syncing would
probably not happen anyway while running a benchmark because there would be
plenty of work for M5 to do, but the event overhead could be avoided.
There's no reason for it to derive from SimLoopExitEvent.
This whole drain thing needs to be redone eventually,
but this is a stopgap to make later changes to
SimLoopExitEvent feasible.
Avoid direct references to mainEventQueue in pseudo-insts
by indirecting through associated CPU object.
Made exitSimLoop() more flexible to enable some of these.
Ran all the source files through 'perl -pi' with this script:
s|\s*(};?\s*)?/\*\s*(end\s*)?namespace\s*(\S+)\s*\*/(\s*})?|} // namespace $3|;
s|\s*};?\s*//\s*(end\s*)?namespace\s*(\S+)\s*|} // namespace $2\n|;
s|\s*};?\s*//\s*(\S+)\s*namespace\s*|} // namespace $1\n|;
Also did a little manual editing on some of the arch/*/isa_traits.hh files
and src/SConscript.
This change is a low level and pervasive reorganization of how PCs are managed
in M5. Back when Alpha was the only ISA, there were only 2 PCs to worry about,
the PC and the NPC, and the lsb of the PC signaled whether or not you were in
PAL mode. As other ISAs were added, we had to add an NNPC, micro PC and next
micropc, x86 and ARM introduced variable length instruction sets, and ARM
started to keep track of mode bits in the PC. Each CPU model handled PCs in
its own custom way that needed to be updated individually to handle the new
dimensions of variability, or, in the case of ARMs mode-bit-in-the-pc hack,
the complexity could be hidden in the ISA at the ISA implementation's expense.
Areas like the branch predictor hadn't been updated to handle branch delay
slots or micropcs, and it turns out that had introduced a significant (10s of
percent) performance bug in SPARC and to a lesser extend MIPS. Rather than
perpetuate the problem by reworking O3 again to handle the PC features needed
by x86, this change was introduced to rework PC handling in a more modular,
transparent, and hopefully efficient way.
PC type:
Rather than having the superset of all possible elements of PC state declared
in each of the CPU models, each ISA defines its own PCState type which has
exactly the elements it needs. A cross product of canned PCState classes are
defined in the new "generic" ISA directory for ISAs with/without delay slots
and microcode. These are either typedef-ed or subclassed by each ISA. To read
or write this structure through a *Context, you use the new pcState() accessor
which reads or writes depending on whether it has an argument. If you just
want the address of the current or next instruction or the current micro PC,
you can get those through read-only accessors on either the PCState type or
the *Contexts. These are instAddr(), nextInstAddr(), and microPC(). Note the
move away from readPC. That name is ambiguous since it's not clear whether or
not it should be the actual address to fetch from, or if it should have extra
bits in it like the PAL mode bit. Each class is free to define its own
functions to get at whatever values it needs however it needs to to be used in
ISA specific code. Eventually Alpha's PAL mode bit could be moved out of the
PC and into a separate field like ARM.
These types can be reset to a particular pc (where npc = pc +
sizeof(MachInst), nnpc = npc + sizeof(MachInst), upc = 0, nupc = 1 as
appropriate), printed, serialized, and compared. There is a branching()
function which encapsulates code in the CPU models that checked if an
instruction branched or not. Exactly what that means in the context of branch
delay slots which can skip an instruction when not taken is ambiguous, and
ideally this function and its uses can be eliminated. PCStates also generally
know how to advance themselves in various ways depending on if they point at
an instruction, a microop, or the last microop of a macroop. More on that
later.
Ideally, accessing all the PCs at once when setting them will improve
performance of M5 even though more data needs to be moved around. This is
because often all the PCs need to be manipulated together, and by getting them
all at once you avoid multiple function calls. Also, the PCs of a particular
thread will have spatial locality in the cache. Previously they were grouped
by element in arrays which spread out accesses.
Advancing the PC:
The PCs were previously managed entirely by the CPU which had to know about PC
semantics, try to figure out which dimension to increment the PC in, what to
set NPC/NNPC, etc. These decisions are best left to the ISA in conjunction
with the PC type itself. Because most of the information about how to
increment the PC (mainly what type of instruction it refers to) is contained
in the instruction object, a new advancePC virtual function was added to the
StaticInst class. Subclasses provide an implementation that moves around the
right element of the PC with a minimal amount of decision making. In ISAs like
Alpha, the instructions always simply assign NPC to PC without having to worry
about micropcs, nnpcs, etc. The added cost of a virtual function call should
be outweighed by not having to figure out as much about what to do with the
PCs and mucking around with the extra elements.
One drawback of making the StaticInsts advance the PC is that you have to
actually have one to advance the PC. This would, superficially, seem to
require decoding an instruction before fetch could advance. This is, as far as
I can tell, realistic. fetch would advance through memory addresses, not PCs,
perhaps predicting new memory addresses using existing ones. More
sophisticated decisions about control flow would be made later on, after the
instruction was decoded, and handed back to fetch. If branching needs to
happen, some amount of decoding needs to happen to see that it's a branch,
what the target is, etc. This could get a little more complicated if that gets
done by the predecoder, but I'm choosing to ignore that for now.
Variable length instructions:
To handle variable length instructions in x86 and ARM, the predecoder now
takes in the current PC by reference to the getExtMachInst function. It can
modify the PC however it needs to (by setting NPC to be the PC + instruction
length, for instance). This could be improved since the CPU doesn't know if
the PC was modified and always has to write it back.
ISA parser:
To support the new API, all PC related operand types were removed from the
parser and replaced with a PCState type. There are two warts on this
implementation. First, as with all the other operand types, the PCState still
has to have a valid operand type even though it doesn't use it. Second, using
syntax like PCS.npc(target) doesn't work for two reasons, this looks like the
syntax for operand type overriding, and the parser can't figure out if you're
reading or writing. Instructions that use the PCS operand (which I've
consistently called it) need to first read it into a local variable,
manipulate it, and then write it back out.
Return address stack:
The return address stack needed a little extra help because, in the presence
of branch delay slots, it has to merge together elements of the return PC and
the call PC. To handle that, a buildRetPC utility function was added. There
are basically only two versions in all the ISAs, but it didn't seem short
enough to put into the generic ISA directory. Also, the branch predictor code
in O3 and InOrder were adjusted so that they always store the PC of the actual
call instruction in the RAS, not the next PC. If the call instruction is a
microop, the next PC refers to the next microop in the same macroop which is
probably not desirable. The buildRetPC function advances the PC intelligently
to the next macroop (in an ISA specific way) so that that case works.
Change in stats:
There were no change in stats except in MIPS and SPARC in the O3 model. MIPS
runs in about 9% fewer ticks. SPARC runs with 30%-50% fewer ticks, which could
likely be improved further by setting call/return instruction flags and taking
advantage of the RAS.
TODO:
Add != operators to the PCState classes, defined trivially to be !(a==b).
Smooth out places where PCs are split apart, passed around, and put back
together later. I think this might happen in SPARC's fault code. Add ISA
specific constructors that allow setting PC elements without calling a bunch
of accessors. Try to eliminate the need for the branching() function. Factor
out Alpha's PAL mode pc bit into a separate flag field, and eliminate places
where it's blindly masked out or tested in the PC.
In the process make add skipFuction() to handle isa specific function skipping
instead of ifdefs and other ugliness. For almost all ABIs, 64 bit arguments can
only start in even registers. Size is now passed to getArgument() so that 32
bit systems can make decisions about register selection for 64 bit arguments.
The number argument is now passed by reference because getArgument() will need
to change it based on the size of the argument and the current argument number.
For ARM, if the argument number is odd and a 64-bit register is requested the
number must first be incremented to because all 64 bit arguments are passed
in an even argument register. Then the number will be incremented again to
access both halves of the argument.
This reduces the scope of those includes and makes it less likely for there to
be a dependency loop. This also moves the hashing functions associated with
ExtMachInst objects to be with the ExtMachInst definitions and out of
utility.hh.
Also move the "Fault" reference counted pointer type into a separate file,
sim/fault.hh. It would be better to name this less similarly to sim/faults.hh
to reduce confusion, but fault.hh matches the name of the type. We could change
Fault to FaultPtr to match other pointer types, and then changing the name of
the file would make more sense.
Instead of putting all object files into m5/object/__init__.py, interrogate
the importer to find out what should be imported.
Instead of creating a single file that lists all of the embedded python
modules, use static object construction to put those objects onto a list.
Do something similar for embedded swig (C++) code.
This allows one two different OS requirements for the same ISA to be handled.
Some OSes are compiled for a virtual address and need to be loaded into physical
memory that starts at address 0, while other bare metal tools generate
images that start at address 0.
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().
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.
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!
