This change speeds up booting, especially in MP cases, by not executing
udelay() on the core but instead skipping ahead tha amount of time that is being
delayed.
There may not be a formally correct spelling for the past tense of mmap, but
mmapped is the spelling Google doesn't try to autocorrect. This makes sense
because it mirrors the past tense of map->mapped and not the past tense of
cape->caped.
--HG--
rename : src/arch/alpha/mmaped_ipr.hh => src/arch/alpha/mmapped_ipr.hh
rename : src/arch/arm/mmaped_ipr.hh => src/arch/arm/mmapped_ipr.hh
rename : src/arch/mips/mmaped_ipr.hh => src/arch/mips/mmapped_ipr.hh
rename : src/arch/power/mmaped_ipr.hh => src/arch/power/mmapped_ipr.hh
rename : src/arch/sparc/mmaped_ipr.hh => src/arch/sparc/mmapped_ipr.hh
rename : src/arch/x86/mmaped_ipr.hh => src/arch/x86/mmapped_ipr.hh
This change fixes an issue where a DTLB fault occurs and redirects fetch to
handle the fault and the ITLB requires a walk which delays translation. In this
case the status of the cpu isn't updated appropriately, and an additional
instruction fetch occurs. Eventually this hits an assert as multiple instruction
fetches are occuring in the system and when the second one returns the
processor is in the wrong state.
Some asserts below are removed because it was always true (typo) and the state
after the initiateAcc() the processor could be in any valid state when a
d-side fault occurs.
In sendSplitData, keep a pointer to the senderState that may be updated after
the call to handle*Packet. This way, if the receiver updates the packet
senderState, it can still be accessed in sendSplitData.
In the case of a split transaction and a cache that is faster than a CPU we
could get two responses before next_tick expires. Add an event that is
scheduled in this case and return false rather than asserting.
This change modifies the way prefetches work. They are now like normal loads
that don't writeback a register. Previously prefetches were supposed to call
prefetch() on the exection context, so they executed with execute() methods
instead of initiateAcc() completeAcc(). The prefetch() methods for all the CPUs
are blank, meaning that they get executed, but don't actually do anything.
On Alpha dead cache copy code was removed and prefetches are now normal ops.
They count as executed operations, but still don't do anything and IsMemRef is
not longer set on them.
On ARM IsDataPrefetch or IsInstructionPreftech is now set on all prefetch
instructions. The timing simple CPU doesn't try to do anything special for
prefetches now and they execute with the normal memory code path.
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.
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.
When a request is NO_ACCESS (x86 CDA microinstruction), the memory op
doesn't go to the cache, so TimingSimpleCPU::completeDataAccess needs
to handle the case where the current status of the CPU is Running
and not DcacheWaitResponse or DTBWaitResponse
These recordEvent() calls could cause crashes since they
access the req pointer after it's potentially been
deleted during a failed translation call. (Similar
problem to the traceData bug fixed in the previous cset.)
Moving them above the translation call (as was done
recentlyi in cset 8b2b8e5e7d35) avoids the crash
but doesn't work, since at that point we don't know if
the access is uncached or not.
It's not clear why these calls are there, and no one
seems to use them, so we'll just delete them. If they
are needed, they should be moved to somewhere that's
guaranteed to be after the translation completes but
before the request is possibly deleted, e.g., in
finishTranslation().
Accessing traceData (to call setAddress() and/or setData())
after initiating a timing translation was causing crashes,
since a failed translation could delete the traceData
object before returning.
It turns out that there was never a need to access traceData
after initiating the translation, as the traced data was
always available earlier; this ordering was merely
historical. Furthermore, traceData->setAddress() and
traceData->setData() were being called both from the CPU
model and the ISA definition, often redundantly.
This patch standardizes all setAddress and setData calls
for memory instructions to be in the CPU models and not
in the ISA definition. It also moves those calls above
the translation calls to eliminate the crashes.
This initiates a timing translation and passes the read or write on to the
processor before waiting for it to finish. Once the translation is finished,
the instruction's state is updated via the 'finish' function. A new
DataTranslation class is created to handle this.
The idea is taken from the implementation of timing translations in
TimingSimpleCPU by Gabe Black. This patch also separates out the timing
translations from this CPU and uses the new DataTranslation class.
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.
across the subclasses. generally make it so that member data is _cpuId and
accessor functions are cpuId(). The ID val comes from the python (default -1 if
none provided), and if it is -1, the index of cpuList will be given. this has
passed util/regress quick and se.py -n4 and fs.py -n4 as well as standard
switch.
The constructor no-longer schedules an event at construction and the implict conversion between int and bool was allowing the old code to compile without warning.
Signed-off By: Ali Saidi
A whole bunch of stuff has been converted to use the new params stuff, but
the CPU wasn't one of them. While we're at it, make some things a bit
more stylish. Most of the work was done by Gabe, I just cleaned stuff up
a bit more at the end.
The notIdleFraction statistic isn't updated when the statistics reset, probably because the cpu Status information
was pulled into the atomic and timing cpus. This changeset pulls Status back into the BaseSimpleCPU object. Anyone
care to comment on the odd naming of the Status instance? It shouldn't just be status because that is confusing
with Port::Status, but _status seems a bit strage too.
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
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
time it was responded to is curTick. Doesn't change the results, but it does make implementation of nextCycle() more difficult
--HG--
extra : convert_revision : 67ed6261a5451d17d96d5df45992590acc353afc
fix the timing cpu to handle receiving a nacked packet
src/cpu/simple/timing.cc:
make the timing cpu handle receiving a nacked packet
src/mem/bridge.cc:
src/mem/bridge.hh:
the bridge never returns false when recvTiming() is called on its ports now, it always returns true and nacks the packet if there isn't sufficient buffer space
--HG--
extra : convert_revision : 5e12d0cf6ce985a5f72bcb7ce26c83a76c34c50a
1. Make sure connectMemPorts() only gets called when the CPU's peer gets changed. This is done by making setPeer() virtual, and overriding it in the CPU's ports. When it gets called on a CPU's port (dcache specifically), it calls the normal setPeer() function, and also connectMemPorts().
2. Consolidate redundant code that handles switching in a CPU.
src/cpu/base.cc:
Move common code of switching over peers to base CPU.
src/cpu/base.hh:
Move common code of switching over peers to BaseCPU.
src/cpu/o3/cpu.cc:
Add in function that updates thread context's ports.
Also use updated function to takeOverFrom() in BaseCPU. This gets rid of some repeated code.
src/cpu/o3/cpu.hh:
Include function to update thread context's memory ports.
src/cpu/o3/lsq.hh:
Add function to dcache port that will update the memory ports upon getting a new peer.
Also include a function that will tell the CPU to update those memory ports.
src/cpu/o3/lsq_impl.hh:
Add function that will update the memory ports upon getting a new peer.
src/cpu/simple/atomic.cc:
src/cpu/simple/timing.cc:
Add function that will update thread context's memory ports upon getting a new peer.
Also use the new BaseCPU's take over from function.
src/cpu/simple/atomic.hh:
Add in function (and dcache port) that will allow the dcache to update memory ports when it gets assigned a new peer.
src/cpu/simple/timing.hh:
Add function that will update thread context's memory ports upon getting a new peer.
src/mem/port.hh:
Make setPeer virtual so that other classes can override it.
--HG--
extra : convert_revision : 2050f1241dd2e83875d281cfc5ad5c6c8705fdaf
Add support for a twin 64 bit int load
Add Memory barrier and write barrier flags as appropriate
Make atomic memory ops atomic
src/arch/alpha/isa/mem.isa:
src/arch/alpha/locked_mem.hh:
src/cpu/base_dyn_inst.hh:
src/mem/cache/cache_blk.hh:
src/mem/cache/cache_impl.hh:
rename store conditional stuff as extra data so it can be used for conditional swaps as well
src/arch/alpha/types.hh:
src/arch/mips/types.hh:
src/arch/sparc/types.hh:
add a largest read data type for statically allocating read buffers in atomic simple cpu
src/arch/isa_parser.py:
Add support for a twin 64 bit int load
src/arch/sparc/isa/decoder.isa:
Make atomic memory ops atomic
Add Memory barrier and write barrier flags as appropriate
src/arch/sparc/isa/formats/mem/basicmem.isa:
add post access code block and define a twinload format for twin loads
src/arch/sparc/isa/formats/mem/blockmem.isa:
remove old microcoded twin load coad
src/arch/sparc/isa/formats/mem/mem.isa:
swap.isa replaces the code in loadstore.isa
src/arch/sparc/isa/formats/mem/util.isa:
add a post access code block
src/arch/sparc/isa/includes.isa:
need bigint.hh for Twin64_t
src/arch/sparc/isa/operands.isa:
add a twin 64 int type
src/cpu/simple/atomic.cc:
src/cpu/simple/atomic.hh:
src/cpu/simple/base.hh:
src/cpu/simple/timing.cc:
add support for twinloads
add support for swap and conditional swap instructions
rename store conditional stuff as extra data so it can be used for conditional swaps as well
src/mem/packet.cc:
src/mem/packet.hh:
Add support for atomic swap memory commands
src/mem/packet_access.hh:
Add endian conversion function for Twin64_t type
src/mem/physical.cc:
src/mem/physical.hh:
src/mem/request.hh:
Add support for atomic swap memory commands
Rename sc code to extradata
--HG--
extra : convert_revision : 69d908512fb34a4e28b29a6e58b807fb1a6b1656
Created MemCmd class to wrap enum and provide handy methods to
check attributes, convert to string/int, etc.
--HG--
extra : convert_revision : 57f147ad893443e3a2040c6d5b4cdb1a8033930b
Right now this introduces a minor memory leak as old physPorts and virtPorts are not deleted when new ones are created. A flyspray task has been created for this issue. It can not be resolved until we determine how the bus will handle giving out ID's to functional ports that may be deleted.
src/cpu/o3/cpu.cc:
src/cpu/simple/atomic.cc:
src/cpu/simple/timing.cc:
Change the setup of the physPort and virtPort to instead happen every time the CPU has a context activated. This is a little high overhead, but keeps it working correctly when the CPU does not have a physical memory attached to it until it switches in (like the case of switch CPUs).
src/cpu/o3/thread_context.hh:
Change function from being called at init() to just being called whenever the memory ports need to be connected.
src/cpu/o3/thread_context_impl.hh:
Update this to not delete the port if it's the same as the virtPort.
src/cpu/thread_context.hh:
Change function from being called at init() to whenever the memory ports need to be connected.
src/cpu/thread_state.cc:
Instead of initializing the ports, simply connect them, deleting any old ports that might exist. This allows these functions to be called multiple times.
src/cpu/thread_state.hh:
Ports are no longer initialized, but rather connected at context activation time.
--HG--
extra : convert_revision : e399ce5dfbd6ad658c953a7c9c7b69b89a70219e
src/cpu/simple/timing.cc:
Various updates for deleting requests more properly.
The major change is moving the deletion of the fetch request/packet to after the instruction has executed and completed. This should fix a few bugs because Ron's memory system didn't expect a call for a functional access while a timing access was being processed.
--HG--
extra : convert_revision : c7cf114bb1ff3cdaa7b0a40ed4c5302dc9d3a522
src/cpu/base.cc:
Move clock phase drift code to the base CPU so that any CPU model can use it.
src/cpu/base.hh:
Added two functions to help get the next cycle the CPU should be scheduled.
src/cpu/simple/atomic.cc:
src/cpu/simple/timing.cc:
Use the function now in BaseCPU.
--HG--
extra : convert_revision : 444494b66ffc85fc473c23f57683c5f9458ad80c
configs/example/fs.py:
configs/example/se.py:
src/cpu/simple/base.cc:
src/cpu/simple/base.hh:
src/cpu/simple/timing.cc:
src/cpu/simple_thread.cc:
src/cpu/simple_thread.hh:
src/cpu/thread_state.cc:
src/cpu/thread_state.hh:
tests/configs/o3-timing-mp.py:
tests/configs/o3-timing.py:
tests/configs/simple-atomic-mp.py:
tests/configs/simple-atomic.py:
tests/configs/simple-timing-mp.py:
tests/configs/simple-timing.py:
tests/configs/tsunami-simple-atomic-dual.py:
tests/configs/tsunami-simple-atomic.py:
tests/configs/tsunami-simple-timing-dual.py:
tests/configs/tsunami-simple-timing.py:
No need for mem parameter any more.
src/cpu/checker/cpu.cc:
Use new constructor for simple thread (no more MemObject parameter).
src/cpu/checker/cpu.hh:
Remove MemObject parameter.
src/cpu/memtest/memtest.hh:
Ports now take in their MemObject owner.
src/cpu/o3/alpha/cpu_builder.cc:
Remove mem parameter.
src/cpu/o3/alpha/cpu_impl.hh:
Remove memory parameter and clean up handling of TranslatingPort.
src/cpu/o3/cpu.cc:
src/cpu/o3/cpu.hh:
src/cpu/o3/fetch.hh:
src/cpu/o3/fetch_impl.hh:
src/cpu/o3/mips/cpu_builder.cc:
src/cpu/o3/mips/cpu_impl.hh:
src/cpu/o3/params.hh:
src/cpu/o3/thread_state.hh:
src/cpu/ozone/cpu.hh:
src/cpu/ozone/cpu_builder.cc:
src/cpu/ozone/cpu_impl.hh:
src/cpu/ozone/front_end.hh:
src/cpu/ozone/front_end_impl.hh:
src/cpu/ozone/lw_lsq.hh:
src/cpu/ozone/lw_lsq_impl.hh:
src/cpu/ozone/simple_params.hh:
src/cpu/ozone/thread_state.hh:
src/cpu/simple/atomic.cc:
Remove memory parameter.
--HG--
extra : convert_revision : 43cb44a33b31320d44b69679dcf646c0380d07d3
Still a bug in atomic uni-coherence in FS.
src/cpu/o3/fetch_impl.hh:
src/cpu/o3/lsq_impl.hh:
src/cpu/simple/atomic.cc:
src/cpu/simple/timing.cc:
Make CPU models handle coherence requests
src/mem/cache/base_cache.cc:
Properly signal coherence CSHRs
src/mem/cache/coherence/uni_coherence.cc:
Only deallocate once
--HG--
extra : convert_revision : c4533de421c371c5532ee505e3ecd451511f5c99
into zed.eecs.umich.edu:/z/hsul/work/m5/newmem
src/cpu/simple/timing.cc:
hand merge
--HG--
extra : convert_revision : 083bf102249ad9bc63c447dbf85d3863f935f647
src/cpu/simple/atomic.cc:
memory mode assertion change so that it only goes off if it's supposed to.
src/cpu/simple/timing.cc:
some drain changes (kevin's) and some changes to memoryMode assertions so that they don't go off when they're not supposed to.
--HG--
extra : convert_revision : 007d8610f097e08f01367b905ada49f93cf37ca3
into zamp.eecs.umich.edu:/z/ktlim2/clean/o3-merge/newmem
src/cpu/memtest/memtest.cc:
src/cpu/memtest/memtest.hh:
src/cpu/simple/timing.hh:
tests/configs/o3-timing-mp.py:
Hand merge.
--HG--
extra : convert_revision : a58cc439eb5e8f900d175ed8b5a85b6c8723e558
If the cpu needs to update any state when it gets a functional write (LSQ??)
then that code needs to be written.
src/cpu/o3/fetch_impl.hh:
src/cpu/o3/lsq_impl.hh:
src/cpu/ozone/front_end_impl.hh:
src/cpu/ozone/lw_lsq_impl.hh:
src/cpu/simple/atomic.cc:
src/cpu/simple/timing.cc:
CPU's can recieve functional accesses, they need to determine if they need to do anything with them.
src/mem/bus.cc:
src/mem/bus.hh:
Make the fuctional path do the correct tye of snoop
--HG--
extra : convert_revision : 70d09f954b907a8aa9b8137579cd2b06e02ae2ff