This adds a vector register type. The type is defined as a std::array of a
fixed number of uint64_ts. The isa_parser.py has been modified to parse vector
register operands and generate the required code. Different cpus have vector
register files now.
This is another step in the process of removing global variables
from Ruby to enable multiple RubySystem instances in a single simulation.
The list of abstract controllers is per-RubySystem and should be
represented that way, rather than as a global.
Since this is the last remaining Ruby global variable, the
src/mem/ruby/Common/Global.* files are also removed.
The drain() call currently passes around a DrainManager pointer, which
is now completely pointless since there is only ever one global
DrainManager in the system. It also contains vestiges from the time
when SimObjects had to keep track of their child objects that needed
draining.
This changeset moves all of the DrainState handling to the Drainable
base class and changes the drain() and drainResume() calls to reflect
this. Particularly, the drain() call has been updated to take no
parameters (the DrainManager argument isn't needed) and return a
DrainState instead of an unsigned integer (there is no point returning
anything other than 0 or 1 any more). Drainable objects should return
either DrainState::Draining (equivalent to returning 1 in the old
system) if they need more time to drain or DrainState::Drained
(equivalent to returning 0 in the old system) if they are already in a
consistent state. Returning DrainState::Running is considered an
error.
Drain done signalling is now done through the signalDrainDone() method
in the Drainable class instead of using the DrainManager directly. The
new call checks if the state of the object is DrainState::Draining
before notifying the drain manager. This means that it is safe to call
signalDrainDone() without first checking if the simulator has
requested draining. The intention here is to reduce the code needed to
implement draining in simple objects.
The drain state enum is currently a part of the Drainable
interface. The same state machine will be used by the DrainManager to
identify the global state of the simulator. Make the drain state a
global typed enum to better cater for this usage scenario.
Objects that are can be serialized are supposed to inherit from the
Serializable class. This class is meant to provide a unified API for
such objects. However, so far it has mainly been used by SimObjects
due to some fundamental design limitations. This changeset redesigns
to the serialization interface to make it more generic and hide the
underlying checkpoint storage. Specifically:
* Add a set of APIs to serialize into a subsection of the current
object. Previously, objects that needed this functionality would
use ad-hoc solutions using nameOut() and section name
generation. In the new world, an object that implements the
interface has the methods serializeSection() and
unserializeSection() that serialize into a named /subsection/ of
the current object. Calling serialize() serializes an object into
the current section.
* Move the name() method from Serializable to SimObject as it is no
longer needed for serialization. The fully qualified section name
is generated by the main serialization code on the fly as objects
serialize sub-objects.
* Add a scoped ScopedCheckpointSection helper class. Some objects
need to serialize data structures, that are not deriving from
Serializable, into subsections. Previously, this was done using
nameOut() and manual section name generation. To simplify this,
this changeset introduces a ScopedCheckpointSection() helper
class. When this class is instantiated, it adds a new /subsection/
and subsequent serialization calls during the lifetime of this
helper class happen inside this section (or a subsection in case
of nested sections).
* The serialize() call is now const which prevents accidental state
manipulation during serialization. Objects that rely on modifying
state can use the serializeOld() call instead. The default
implementation simply calls serialize(). Note: The old-style calls
need to be explicitly called using the
serializeOld()/serializeSectionOld() style APIs. These are used by
default when serializing SimObjects.
* Both the input and output checkpoints now use their own named
types. This hides underlying checkpoint implementation from
objects that need checkpointing and makes it easier to change the
underlying checkpoint storage code.
This changeset adds support for aarch64 in kvm. The CPU module
supports both checkpointing and online CPU model switching as long as
no devices are simulated by the host kernel. It currently has the
following limitations:
* The system register based generic timer can only be simulated by
the host kernel. Workaround: Use a memory mapped timer instead to
simulate the timer in gem5.
* Simulating devices (e.g., the generic timer) in the host kernel
requires that the host kernel also simulates the GIC.
* ID registers in the host and in gem5 must match for switching
between simulated CPUs and KVM. This is particularly important
for ID registers describing memory system capabilities (e.g.,
ASID size, physical address size).
* Switching between a virtualized CPU and a simulated CPU is
currently not supported if in-kernel device emulation is
used. This could be worked around by adding support for switching
to the gem5 (e.g., the KvmGic) side of the device models. A
simpler workaround is to avoid in-kernel device models
altogether.
This changeset adds a GIC implementation that uses the kernel's
built-in support for simulating the interrupt controller. Since there
is currently no support for state transfer between gem5 and the
kernel, the device model does not support serialization and CPU
switching (which would require switching to a gem5-simulated GIC).
There are cases (particularly when attaching GDB) when instruction
events are scheduled at the current instruction tick. This used to
trigger an assertion error in kvm. This changeset adds a check for
this condition and forces KVM to do a quick entry that completes any
pending IO operations, but does not execute any new instructions,
before servicing the event. We could check if we need to enter KVM at
all, but forcing a quick entry is makes the code slightly cleaner and
does not hurt correctness (performance is hardly an issue in these
cases).
This changeset moves the ARM-specific KVM CPU implementation to
arch/arm/kvm/. This change is expected to keep the source tree
somewhat cleaner as we start adding support for ARMv8 and KVM
in-kernel interrupt controller simulation.
--HG--
rename : src/cpu/kvm/ArmKvmCPU.py => src/arch/arm/kvm/ArmKvmCPU.py
rename : src/cpu/kvm/arm_cpu.cc => src/arch/arm/kvm/arm_cpu.cc
rename : src/cpu/kvm/arm_cpu.hh => src/arch/arm/kvm/arm_cpu.hh
The MinorCPU would count bubbles in Execute::issue as part of
the num_insts_issued and so sometimes reach the instruction
issue limit incorrectly.
Fixed by checking for a bubble in one new place.
The register dumping code in kvm tries to print the bytes in large
registers (128 bits and larger) instead of printing them as hex. This
changeset fixes that.
Three minor issues are resolved:
1. Apparently gcc 5.1 does not like negation of booleans followed by
bitwise AND.
2. Somehow the compiler also gets confused and warns about
NoopMachInst being unused (removing it causes compilation errors
though). Most likely a compiler bug.
3. There seems to be a number of instances where loop unrolling causes
false positives for the array-bounds check. For now, switch to
std::array. Potentially we could disable the warning for newer gcc
versions, but switching to std::array is probably a good move in
any case.
The Request::UNCACHEABLE flag currently has two different
functions. The first, and obvious, function is to prevent the memory
system from caching data in the request. The second function is to
prevent reordering and speculation in CPU models.
This changeset gives the order/speculation requirement a separate flag
(Request::STRICT_ORDER). This flag prevents CPU models from doing the
following optimizations:
* Speculation: CPU models are not allowed to issue speculative
loads.
* Write combining: CPU models and caches are not allowed to merge
writes to the same cache line.
Note: The memory system may still reorder accesses unless the
UNCACHEABLE flag is set. It is therefore expected that the
STRICT_ORDER flag is combined with the UNCACHEABLE flag to prevent
this behavior.
This patch takes a last step in fixing issues related to uncacheable
accesses. We do not separate uncacheable memory from uncacheable
devices, and in cases where it is really memory, there are valid
scenarios where we need to snoop since we do not support cache
maintenance instructions (yet). On snooping an uncacheable access we
thus provide data if possible. In essence this makes uncacheable
accesses IO coherent.
The snoop filter is also queried to steer the snoops, but not updated
since the uncacheable accesses do not allocate a block.
This patch fixes a recent issue with gcc 4.9 (and possibly more) being
convinced that indices outside the array bounds are used when
initialising the FUPool members.
Currently, each op class has a parameter issueLat that denotes the cycles after
which another op of the same class can be issued. As of now, this latency can
either be one cycle (fully pipelined) or same as execution latency of the op
(not at all pipelined). The fact that issueLat is a parameter of type Cycles
makes one believe that it can be set to any value. To avoid the confusion, the
parameter is being renamed as 'pipelined' with type boolean. If set to true,
the op would execute in a fully pipelined fashion. Otherwise, it would execute
in an unpipelined fashion.
This patch sets the default latency of the division microop to a single cycle
on x86. This is because the division instructions DIV and IDIV have been
implemented as loops of div microops, where each microop computes a single bit
of the quotient.
The o3 cpu instruction queue model uses the count variable to track the number
of unissued instructions in the queue. Previously, the squash method used
this variable to avoid executing the doSquash method when there were no
unissued instructions in the pipeline. A corner case problem exists when
only issued instructions exist in the pipeline and a squash occurs; the
doSquash code is not invoked and subsequently does not clean up state properly.
This patch takes the final step in removing the InOrderCPU from the
tree. Rest in peace.
The MinorCPU is now used to model an in-order microarchitecture, and
long term the MinorCPU will eventually be renamed InOrderCPU.
This patch ensures that the CPU progress Event is triggered for the new set of
switched_cpus that get scheduled (e.g. during fast-forwarding). it also avoids
printing the interval state if the cpu is currently switched out.
Committed by: Nilay Vaish <nilay@cs.wisc.edu>
The totalInstructions counter is only incremented when the whole instruction is
commited and not on every microop. It was incorrectly reset in atomic and
timing cpus.
Committed by: Nilay Vaish <nilay@cs.wisc.edu>"
Fix erroneous message format for fatal error.
Previously, code did not have type indicator (% instead of %d).
Also removed redundant fatal check.
Ran modified sweep.py with in range and out of range values to test.
Refactor the way that specific MemCmd values are generated for packets.
The new approach is a little more elegant in that we assign the right
value up front, and it's also more amenable to non-heap-allocated
Packet objects.
Also replaced the code in the Minor model that was still doing it the
ad-hoc way.
This is basically a refinement of http://repo.gem5.org/gem5/rev/711eb0e64249.
The variable is used in only one place and a whole new function setNextStatus()
has been defined just to compute the value of the variable. Instead of calling
the function, the value is now computed in the loop that preceded the function
call.
This patch introduces a few subclasses to the CoherentXBar and
NoncoherentXBar to distinguish the different uses in the system. We
use the crossbar in a wide range of places: interfacing cores to the
L2, as a system interconnect, connecting I/O and peripherals,
etc. Needless to say, these crossbars have very different performance,
and the clock frequency alone is not enough to distinguish these
scenarios.
Instead of trying to capture every possible case, this patch
introduces dedicated subclasses for the three primary use-cases:
L2XBar, SystemXBar and IOXbar. More can be added if needed, and the
defaults can be overridden.
This patch changes how the MMU and table walkers are created such that
a single port is used to connect the MMU and the TLBs to the memory
system. Previously two ports were needed as there are two table walker
objects (stage one and stage two), and they both had a port. Now the
port itself is moved to the Stage2MMU, and each TableWalker is simply
using the port from the parent.
By using the same port we also remove the need for having an
additional crossbar joining the two ports before the walker cache or
the L2. This simplifies the creation of the CPU cache topology in
BaseCPU.py considerably. Moreover, for naming and symmetry reasons,
the TLB walker port is connected through the stage-one table walker
thus making the naming identical to x86. Along the same line, we use
the stage-one table walker to generate the master id that is used by
all TLB-related requests.
Now, prior to the renaming, the instruction requests the exact amount of
registers it will need, and the rename_map decides whether the instruction is
allowed to proceed or not.
This patch fixes a long-standing isue with the port flow
control. Before this patch the retry mechanism was shared between all
different packet classes. As a result, a snoop response could get
stuck behind a request waiting for a retry, even if the send/recv
functions were split. This caused message-dependent deadlocks in
stress-test scenarios.
The patch splits the retry into one per packet (message) class. Thus,
sendTimingReq has a corresponding recvReqRetry, sendTimingResp has
recvRespRetry etc. Most of the changes to the code involve simply
clarifying what type of request a specific object was accepting.
The biggest change in functionality is in the cache downstream packet
queue, facing the memory. This queue was shared by requests and snoop
responses, and it is now split into two queues, each with their own
flow control, but the same physical MasterPort. These changes fixes
the previously seen deadlocks.
Have the traffic generator add its masterID as the PC address to the
requests. That way, prefetchers (and other components) that use a PC
for request classification will see per-tester streams of requests.
This enables us to test strided prefetchers with the memchecker, too.
To be able to use the TrafficGen in a system with caches we need to
allow it to sink incoming snoop requests. By default the master port
panics, so silently ignore any snoops.
The MemTest class really only tests false sharing, and as such there
was a lot of old cruft that could be removed. This patch cleans up the
tester, and also makes it more clear what the assumptions are. As part
of this simplification the reference functional memory is also
removed.
The regression configs using MemTest are updated to reflect the
changes, and the stats will be bumped in a separate patch. The example
config will be updated in a separate patch due to more extensive
re-work.
In a follow-on patch a new tester will be introduced that uses the
MemChecker to implement true sharing.
The TLB-related code is generally architecture dependent and should
live in the arch directory to signify that.
--HG--
rename : src/sim/BaseTLB.py => src/arch/generic/BaseTLB.py
rename : src/sim/tlb.cc => src/arch/generic/tlb.cc
rename : src/sim/tlb.hh => src/arch/generic/tlb.hh
This patch changes how the timing CPU deals with processing responses,
always scheduling an event, even if it is for the current tick. This
helps to avoid situations where a new request shows up before a
response is finished in the crossbar, and also is more in line with
any realistic behaviour.
While the IsFirstMicroop flag exists it was only occasionally used in the ARM
instructions that gem5 microOps and therefore couldn't be relied on to be correct.
This patch tidies up how we create and set the fields of a Request. In
essence it tries to use the constructor where possible (as opposed to
setPhys and setVirt), thus avoiding spreading the information across a
number of locations. In fact, setPhys is made private as part of this
patch, and a number of places where we callede setVirt instead uses
the appropriate constructor.
The ppCommit should notify the attached listener every time the cpu commits
a microop or non microcoded insturction. The listener can then decide
whether it will process only the last microop (eg. SimPoint probe).
Committed by: Nilay Vaish <nilay@cs.wisc.edu>
Minor was reporting the data cache access as ".inst" accesses.
This just switches the MasterPortID to dataMasterPortId.
Committed by: Nilay Vaish <nilay@cs.wisc.edu>
Only the instruction address is actually checked, so there's no need to check
repeatedly while we're working through the microops of a macroop and that's
not changing.
This patch fixes a case where a store in Minor's store buffer never
leaves the store buffer as it is pre-maturely counted as having been
issued, leading to the store buffer idling.
LSQ::StoreBuffer::numUnissuedAccesses should count the number of accesses
either in memory, or still in the store buffer after being completed.
For stores which are also barriers, the store will stay in the store
buffer for a cycle after it is completed and will be cleaned up by the
barrier clearing code (to ensure that barriers are completed in-order).
To acheive this, numUnissuedAccesses is not decremented when a store-barrier
is issued to memory, but when its barrier effect is cleared.
Without this patch, the correct behaviour happens when a memory transaction
is immediately accepted, but not if it needs a retry.
In case the memory subsystem sends a combined response with invalidate
(e.g. ReadRespWithInvalidate), we cannot ignore the invalidate part
of the response.
If we were to ignore the invalidate part, under certain circumstances
this effectively leads to reordering of loads to the same address
which is not permitted under any memory consistency model implemented
in gem5.
Consider the case where a later load's address is computed before an
earlier load in program order, and is therefore sent to the memory
subsystem first. At some point the earlier load's address is computed
and in doing so correctly marks the later load as a
possibleLoadViolation. In the meantime some other node writes and
sends invalidations to all other nodes. The invalidation races with
the later load's ReadResp, and arrives before ReadResp and is
deferred. Upon receipt of the ReadResp, the response is changed to
ReadRespWithInvalidate, and sent to the CPU. If we ignore the
invalidate part of the packet, we let the later load read the old
value of the address. Eventually the earlier load's ReadResp arrives,
but with new data. As there was no invalidate snoop (sunk into the
ReadRespWithInvalidate), and if we did not process the invalidate of
the ReadRespWithInvalidate, we obtain a load reordering.
A similar scenario can be constructed where the earlier load's address
is computed after ReadRespWithInvalidate arrives for the younger
load. In this case hitExternalSnoop needs to be set to true on the
ReadRespWithInvalidate, so that upon knowing the address of the
earlier load, checkViolations will cause the later load to be
squashed.
Finally we must account for the case where both loads are sent to the
memory subsystem (reordered), a snoop invalidate arrives and correctly
sets the later loads fault to ReExec. However, before the CPU
processes the fault, the later load's ReadResp arrives and the
writeback discards the outstanding fault. We must add a check to
ensure that we do not skip any unprocessed faults.
Move the packet deallocations in the O3 CPU so that the completeDataAccess
deals only with the LSQ specific parts and the generic recvTimingResp frees the
packet in all other cases.
This patch simplifies how we deal with dynamically allocated data in
the packet, always assuming that it is array allocated, and hence
should be array deallocated (delete[] as opposed to delete). The only
uses of dataDynamic was in the Ruby testers.
The ARRAY_DATA flag in the packet is removed accordingly. No
defragmentation of the flags is done at this point, leaving a gap in
the bit masks.
As the last part the patch, it renames dataDynamicArray to dataDynamic.
This patch takes a first step in tightening up how we use the data
pointer in write packets. A const getter is added for the pointer
itself (getConstPtr), and a number of member functions are also made
const accordingly. In a range of places throughout the memory system
the new member is used.
The patch also removes the unused isReadWrite function.
This patch removes the parameter that enables bypassing the null check
in the Packet::getPtr method. A number of call sites assume the value
to be non-null.
The one odd case is the RubyTester, which issues zero-sized
prefetches(!), and despite being reads they had no valid data
pointer. This is now fixed, but the size oddity remains (unless anyone
object or has any good suggestions).
Finally, in the Ruby Sequencer, appropriate checks are made for flush
packets as they have no valid data pointer.
This patch adds methods in KvmCPU model to handle KVM exits caused by syscall
instructions and page faults. These types of exits will be encountered if
KvmCPU is run in SE mode.
Another churn to clean up undefined behaviour, mostly ARM, but some
parts also touching the generic part of the code base.
Most of the fixes are simply ensuring that proper intialisation. One
of the more subtle changes is the return type of the sign-extension,
which is changed to uint64_t. This is to avoid shifting negative
values (undefined behaviour) in the ISA code.
Mwait works as follows:
1. A cpu monitors an address of interest (monitor instruction)
2. A cpu calls mwait - this loads the cache line into that cpu's cache.
3. The cpu goes to sleep.
4. When another processor requests write permission for the line, it is
evicted from the sleeping cpu's cache. This eviction is forwarded to the
sleeping cpu, which then wakes up.
Committed by: Nilay Vaish <nilay@cs.wisc.edu>
Fixes a bug where Minor drains in the midst of committing a
conditional store.
While committing a conditional store, lastCommitWasEndOfMacroop is true
(from the previous instruction) as we still haven't finished the conditional
store. If a drain occurs before the cache response, Minor would check just
lastCommitWasEndOfMacroop, which was true, and set drainState=DrainHaltFetch,
which increases the streamSeqNum. This caused the conditional store to be
squashed when the memory responded and it completed. However, to the memory
the store succeeded, while to the instruction sequence it never occurred.
In the case of an LLSC, the instruction sequence will replay the squashed
STREX, which will fail as the cache is no longer in LLSC. Then the
instruction sequence will loop back to a LDREX, which receives the updated
(incorrect) value.
Committed by: Nilay Vaish <nilay@cs.wisc.edu>
It is possible for the O3 CPU to consider itself drained and
later have a squashed instruction perform a writeback. This
patch re-adds tracking of in-flight instructions to prevent
falsely signaling a drained event.
IEW did not check the instQueue and memDepUnit to ensure
they were drained. This caused issues when drainSanityCheck()
did check those structures after asserting IEW was drained.
This patch fixes a bug where a completing load or store which is also a
barrier can push a barrier into the store buffer without first checking
that there is a free slot.
The bug was not fatal but would print a warning that the store buffer
was full when inserting.
This patch takes quite a large step in transitioning from the ad-hoc
RefCountingPtr to the c++11 shared_ptr by adopting its use for all
Faults. There are no changes in behaviour, and the code modifications
are mostly just replacing "new" with "make_shared".
This patch transitions the o3 MemDepEntry from the ad-hoc
RefCountingPtr to the c++11 shared_ptr. There are no changes in
behaviour, and the code modifications are mainly replacing "new" with
"make_shared".
This changeset adds probe points that can be used to implement PMU
counters for CPU stats. The following probes are supported:
* BaseCPU::ppCycles / Cycles
* BaseCPU::ppRetiredInsts / RetiredInsts
* BaseCPU::ppRetiredLoads / RetiredLoads
* BaseCPU::ppRetiredStores / RetiredStores
* BaseCPU::ppRetiredBranches RetiredBranches
This changeset adds probe points that can be used to implement PMU
counters for branch predictor stats. The following probes are
supported:
* BPRedUnit::ppBranches / Branches
* BPRedUnit::ppMisses / Misses
The Ozone CPU is now very much out of date and completely
non-functional, with no one actively working on restoring it. It is a
source of confusion for new users who attempt to use it before
realizing its current state. RIP
This patch optimises the passing of StaticInstPtr by avoiding copying
the reference-counting pointer. This avoids first incrementing and
then decrementing the reference-counting pointer.
Fix a number few minor issues to please gcc 4.9.1. Removing the
'-fuse-linker-plugin' flag means no libraries are part of the LTO
process, but hopefully this is an acceptable loss, as the flag causes
issues on a lot of systems (only certain combinations of gcc, ld and
ar work).
The call paths for de-scheduling a thread are halt() and suspend(), from
the thread context. There is no call to deallocateContext() in general,
though some CPUs chose to define it. This patch removes the function
from BaseCPU and the cores which do not require it.
activate(), suspend(), and halt() used on thread contexts had an optional
delay parameter. However this parameter was often ignored. Also, when used,
the delay was seemily arbitrarily set to 0 or 1 cycle (no other delays were
ever specified). This patch removes the delay parameter and 'Events'
associated with them across all ISAs and cores. Unused activate logic
is also removed.
This patch changes the name of the Bus classes to XBar to better
reflect the actual timing behaviour. The actual instances in the
config scripts are not renamed, and remain as e.g. iobus or membus.
As part of this renaming, the code has also been clean up slightly,
making use of range-based for loops and tidying up some comments. The
only changes outside the bus/crossbar code is due to the delay
variables in the packet.
--HG--
rename : src/mem/Bus.py => src/mem/XBar.py
rename : src/mem/coherent_bus.cc => src/mem/coherent_xbar.cc
rename : src/mem/coherent_bus.hh => src/mem/coherent_xbar.hh
rename : src/mem/noncoherent_bus.cc => src/mem/noncoherent_xbar.cc
rename : src/mem/noncoherent_bus.hh => src/mem/noncoherent_xbar.hh
rename : src/mem/bus.cc => src/mem/xbar.cc
rename : src/mem/bus.hh => src/mem/xbar.hh
Add new DRAM_ROTATE mode to traffic generator.
This mode will generate DRAM traffic that rotates across
banks per rank, command types, and ranks per channel
The looping order is illustrated below:
for (ranks per channel)
for (command types)
for (banks per rank)
// Generate DRAM Command Series
This patch also adds the read percentage as an input argument to the
DRAM sweep script. If the simulated read percentage is 0 or 100, the
middle for loop does not generate additional commands. This loop is
used only when the read percentage is set to 50, in which case the
middle loop will toggle between read and write commands.
Modified sweep.py script, which generates DRAM traffic.
Added input arguments and support for new DRAM_ROTATE mode.
The script now has input arguments for:
1) Read percentage
2) Number of ranks
3) Address mapping
4) Traffic generator mode (DRAM or DRAM_ROTATE)
The default values are:
100% reads, 1 rank, RoRaBaCoCh address mapping, and DRAM traffic gen mode
For the DRAM traffic mode, added multi-rank support.
This patch does a bit of housekeeping on the string helper functions
and relies on the C++11 standard library where possible. It also does
away with our custom string hash as an implementation is already part
of the standard library.
This patch changes how faults are passed between methods in an attempt
to copy as few reference-counting pointer instances as possible. This
should avoid unecessary copies being created, contributing to the
increment/decrement of the reference counters.
This patch fixes cases where uncacheable/memory type flags are not set
correctly on a memory op which is split in the LSQ. Without this
patch, request->request if freely used to check flags where the flags
should actually come from the accumulation of request fragment flags.
This patch also fixes a bug where an uncacheable access which passes
through tryToSendRequest more than once can increment
LSQ::numAccessesInMemorySystem more than once.
This patch closes a number of space gaps in debug messages caused by
the incorrect use of line continuation within strings. (There's also
one consistency change to a similar, but correct, use of line
continuation)
Static analysis unearther a bunch of uninitialised variables and
members, and this patch addresses the problem. In all cases these
omissions seem benign in the end, but at least fixing them means less
false positives next time round.
This patch tidies up random number generation to ensure that it is
done consistently throughout the code base. In essence this involves a
clean-up of Ruby, and some code simplifications in the traffic
generator.
As part of this patch a bunch of skewed distributions (off-by-one etc)
have been fixed.
Note that a single global random number generator is used, and that
the object instantiation order will impact the behaviour (the sequence
of numbers will be unaffected, but if module A calles random before
module B then they would obviously see a different outcome). The
dependency on the instantiation order is true in any case due to the
execution-model of gem5, so we leave it as is. Also note that the
global ranom generator is not thread safe at this point.
Regressions using the memtest, TrafficGen or any Ruby tester are
affected and will be updated accordingly.
For X86, the o3 CPU would get stuck with the commit stage not being
drained if an interrupt arrived while drain was pending. isDrained()
makes sure that pcState.microPC() == 0, thus ensuring that we are at
an instruction boundary. However, when we take an interrupt we
execute:
pcState.upc(romMicroPC(entry));
pcState.nupc(romMicroPC(entry) + 1);
tc->pcState(pcState);
As a result, the MicroPC is no longer zero. This patch ensures the drain is
delayed until no interrupts are present. Once draining, non-synchronous
interrupts are deffered until after the switch.
Analogous to ee049bf (for x86). Requires a bump of the checkpoint version
and corresponding upgrader code to move the condition code register values
to the new register file.
A small bug in the bimodal predictor caused significant degradation in
performance on some benchmarks. This was caused by using the wrong
globalHistoryReg during the update phase. This patches fixes the bug
and brings the performance to normal level.
This patch fixes the load blocked/replay mechanism in the o3 cpu. Rather than
flushing the entire pipeline, this patch replays loads once the cache becomes
unblocked.
Additionally, deferred memory instructions (loads which had conflicting stores),
when replayed would not respect the number of functional units (only respected
issue width). This patch also corrects that.
Improvements over 20% have been observed on a microbenchmark designed to
exercise this behavior.
O3 is supposed to stop fetching instructions once a quiesce is encountered.
However due to a bug, it would continue fetching instructions from the current
fetch buffer. This is because of a break statment that only broke out of the
first of 2 nested loops. It should have broken out of both.
The o3 cpu could attempt to schedule inactive threads under round-robin SMT
mode.
This is because it maintained an independent priority list of threads from the
active thread list. This priority list could be come stale once threads were
inactive, leading to the cpu trying to fetch/commit from inactive threads.
Additionally the fetch queue is now forcibly flushed of instrctuctions
from the de-scheduled thread.
Relevant output:
24557000: system.cpu: [tid:1]: Calling deactivate thread.
24557000: system.cpu: [tid:1]: Removing from active threads list
24557500: system.cpu:
FullO3CPU: Ticking main, FullO3CPU.
24557500: system.cpu.fetch: Running stage.
24557500: system.cpu.fetch: Attempting to fetch from [tid:1]
When a branch mispredicted gem5 would squash all history after and including
the mispredicted branch. However, the mispredicted branch is still speculative
and its history is required to rollback state if another, older, branch
mispredicts. This leads to things like RAS corruption.
This patch adds a fetch queue that sits between fetch and decode to the
o3 cpu. This effectively decouples fetch from decode stalls allowing it
to be more aggressive, running futher ahead in the instruction stream.
The o3 pipeline interlock/stall logic is incorrect. o3 unnecessicarily stalled
fetch and decode due to later stages in the pipeline. In general, a stage
should usually only consider if it is stalled by the adjacent, downstream stage.
Forcing stalls due to later stages creates and results in bubbles in the
pipeline. Additionally, o3 stalled the entire frontend (fetch, decode, rename)
on a branch mispredict while the ROB is being serially walked to update the
RAT (robSquashing). Only should have stalled at rename.
As highlighed on the mailing list gem5's writeback modeling can impact
performance. This patch removes the limitation on maximum outstanding issued
instructions, however the number that can writeback in a single cycle is still
respected in instToCommit().
We currently generate and compile one version of the ISA code per CPU
model. This is obviously wasting a lot of resources at compile
time. This changeset factors out the interface into a separate
ExecContext class, which also serves as documentation for the
interface between CPUs and the ISA code. While doing so, this
changeset also fixes up interface inconsistencies between the
different CPU models.
The main argument for using one set of ISA code per CPU model has
always been performance as this avoid indirect branches in the
generated code. However, this argument does not hold water. Booting
Linux on a simulated ARM system running in atomic mode
(opt/10.linux-boot/realview-simple-atomic) is actually 2% faster
(compiled using clang 3.4) after applying this patch. Additionally,
compilation time is decreased by 35%.
The namespace Message conflicts with the Message data type used extensively
in Ruby. Since Ruby is being moved to the same Master/Slave ports based
configuration style as the rest of gem5, this conflict needs to be resolved.
Hence, the namespace is being renamed to ProtoMessage.
The branch predictor is normally only built when a CPU that uses a
branch predictor is built. The list of CPUs is currently incomplete as
the simple CPUs support branch predictors (for warming, branch stats,
etc). In practice, all CPU models now use branch predictors, so this
changeset removes the CPU model check and replaces it with a check for
the NULL ISA.
RefCountingPtr is sometimes forward declared to avoid having to
include refcnt.hh. This does not work since we typically return
instances of RefCountingPtr rather than references to instances. The
only reason this currently works is that we include refcnt.hh in
cprintf.hh, which "leaks" the header to most other source files. This
changeset replaces such forward declarations with an include of
refcnt.hh.
This patch does some minor house keeping of the branch predictor by
adopting STL containers, and shifting some iterator to use range-based
for loops.
The predictor history is also changed from a list to a deque as we
never to insertion/deletion other than at the front and back.
This patch adds a check to ensure that packets which are not going to
a memory range are suppressed in the traffic generator. Thus, if a
trace is collected in full-system, the packets destined for devices
are not played back.
This patch contains a new CPU model named `Minor'. Minor models a four
stage in-order execution pipeline (fetch lines, decompose into
macroops, decompose macroops into microops, execute).
The model was developed to support the ARM ISA but should be fixable
to support all the remaining gem5 ISAs. It currently also works for
Alpha, and regressions are included for ARM and Alpha (including Linux
boot).
Documentation for the model can be found in src/doc/inside-minor.doxygen and
its internal operations can be visualised using the Minorview tool
utils/minorview.py.
Minor was designed to be fairly simple and not to engage in a lot of
instruction annotation. As such, it currently has very few gathered
stats and may lack other gem5 features.
Minor is faster than the o3 model. Sample results:
Benchmark | Stat host_seconds (s)
---------------+--------v--------v--------
(on ARM, opt) | simple | o3 | minor
| timing | timing | timing
---------------+--------+--------+--------
10.linux-boot | 169 | 1883 | 1075
10.mcf | 117 | 967 | 491
20.parser | 668 | 6315 | 3146
30.eon | 542 | 3413 | 2414
40.perlbmk | 2339 | 20905 | 11532
50.vortex | 122 | 1094 | 588
60.bzip2 | 2045 | 18061 | 9662
70.twolf | 207 | 2736 | 1036
Check for free entries in Load Queue and Store Queue separately to
avoid cases when load cannot be renamed due to full Store Queue and
vice versa.
This work was done while Binh was an intern at AMD Research.
Using '== true' in a boolean expression is totally redundant,
and using '== false' is pretty verbose (and arguably less
readable in most cases) compared to '!'.
It's somewhat of a pet peeve, perhaps, but I had some time
waiting for some tests to run and decided to clean these up.
Unfortunately, SLICC appears not to have the '!' operator,
so I had to leave the '== false' tests in the SLICC code.
This patch removes the stat totalCommittedInsts. This variable was used for
recording the total number of instructions committed across all the threads
of a core. The instructions committed by each thread are recorded invidually.
The total would now be generated by summing these individual counts.
This patch adds a the member function StaticInst::printFlags to allow all
of an instruction's flags to be printed without using the individual
is... member functions or resorting to exposing the 'flags' vector
It also replaces the enum definition StaticInst::Flags with a
Python-generated enumeration and adds to the enum generation mechanism
in src/python/m5/params.py to allow Enums to be placed in namespaces
other than Enums or, alternatively, in wrapper structs allowing them to
be inherited by other classes (so populating that class's name-space
with the enumeration element names).
The ARM TLBs have a bootUncacheability flag used to make some loads
and stores become uncacheable when booting in FS mode. Later the
flag is cleared to let those loads and stores operate as normal. When
doing a takeOverFrom(), this flag's state is not preserved and is
momentarily reset until the CPSR is touched. On single core runs this
is a non-issue. On multi-core runs this can lead to crashes on the O3
CPU model from the following series of events:
1) takeOverFrom executed to switch from Atomic -> O3
2) All bootUncacheability flags are reset to true
3) Core2 tries to execute a load covered by bootUncacheability, it
is flagged as uncacheable
4) Core2's load needs to replay due to a pipeline flush
3) Core1 core does an action on CPSR
4) The handling code for CPSR then checks all other cores
to determine if bootUncacheability can be set to false
5) Asynchronously set bootUncacheability on all cores to false
6) Core2 replays load previously set as uncacheable and notices
it is now flagged as cacheable, leads to a panic.
This patch implements takeOverFrom() functionality for the ARM TLBs
to preserve flag values when switching from atomic -> detailed.
For the o3, add instruction mix (OpClass) histogram at commit (stats
also already collected at issue). For the simple CPUs we add a
histogram of executed instructions
Allow the specification of a socket ID for every core that is reflected in the
MPIDR field in ARM systems. This allows studying multi-socket / cluster
systems with ARM CPUs.
In the O3 LSQ, data read/written is printed out in DPRINTFs. However,
the data field is treated as a character string with a null terminated.
However the data field is not encoded this way. This patch removes
that possibility by removing the data part of the print.
O3CPU has a compile-time maximum width set in o3/impl.hh, but checking
the configuration against this limit was not implemented anywhere
except for fetch. Configuring a wider pipe than the limit can silently
cause various issues during the simulation. This patch adds the proper
checking in the constructor of the various pipeline stages.
A number of calls to isEmpty() and numFreeEntries()
should be thread-specific.
In cpu.cc, the fact that tid is /*commented*/ out is a bug. Say the rob
has instructions from thread 0 (isEmpty() returns false), and none from
thread 1. If we are trying to squash all of thread 1, then
readTailInst(thread 1) will be called because rob->isEmpty() returns
false. The result is end_it is not in the list and the while
statement loops indefinitely back over the cpu's instList.
In iew_impl.hh, all threads are told they have the entire remaining IQ, when
each thread actually has a certain allocation. The result is extra stalls at
the iew dispatch stage which the rename stage usually takes care of.
In commit_impl.hh, rob->readHeadInst(thread 1) can be called if the rob only
contains instructions from thread 0. This returns a dummyInst (which may work
since we are trying to squash all instructions, but hardly seems like the right
way to do it).
In rob_impl.hh this fix skips the rest of the function more frequently and is
more efficient.
Committed by: Nilay Vaish <nilay@cs.wisc.edu>
Simulating a SMP or multicore requires devices to be shared between
multiple KVM vCPUs. This means that locking is required when accessing
devices. This changeset adds the necessary locking to allow devices to
execute correctly. It is implemented by temporarily migrating the KVM
CPU to the VM's (and devices) event queue when handling
MMIO. Similarly, the VM migrates to the interrupt controller's event
queue when delivering an interrupt.
The support for fast-forwarding of multicore simulations added by this
changeset assumes that all devices in a system are simulated in the
same thread and each vCPU has its own thread. Special care must be
taken to ensure that devices living under the CPU in the object
hierarchy (e.g., the interrupt controller) do not inherit the parent
CPUs thread and are assigned to device thread. The KvmVM object is
assumed to live in the same thread as the other devices in the system.
This patch fixes violation of TSO in the O3CPU, as all loads must be
ordered with all other loads. In the LQ, if a snoop is observed, all
subsequent loads need to be squashed if the system is TSO.
Prior to this patch, the following case could be violated:
P0 | P1 ;
MOV [x],mail=/usr/spool/mail/nilay | MOV EAX,[y] ;
MOV [y],mail=/usr/spool/mail/nilay | MOV EBX,[x] ;
exists (1:EAX=1 /\ 1:EBX=0) [is a violation]
The problem was found using litmus [http://diy.inria.fr].
Committed by: Nilay Vaish <nilay@cs.wisc.edu
This patch enables a new 'DRAM' mode to the existing traffic
generator, catered to generate specific requests to DRAM based on
required hit length (stride size) and bank utilization. It is an add on
to the Random mode.
The basic idea is to control how many successive packets target the
same page, and how many banks are being used in parallel. This gives a
two-dimensional space that stresses different aspects of the DRAM
timing.
The configuration file needed to use this patch has to be changed as
follow: (reference to Random Mode, LPDDR3 memory type)
'STATE 0 10000000000 RANDOM 50 0 134217728 64 3004 5002 0'
-> 'STATE 0 10000000000 DRAM 50 0 134217728 32 3004 5002 0 96 1024 8 6 1'
The last 4 parameters to be added are:
<stride size (bytes), page size(bytes), number of banks available in DRAM,
number of banks to be utilized, address mapping scheme>
The address mapping information is used to get the stride address
stream of the specified size and to know where to find the bank
bits. The configuration file has a parameter where '0'-> RoCoRaBaCh,
'1'-> RoRaBaCoCh/RoRaBaChCo address-mapping schemes. Note that the
generator currently assumes a single channel and a single rank. This
is to avoid overwhelming the traffic generator with information about
the memory organisation.
Prevent incomplete configuration of TrafficGen class from causing
segmentation faults. If an 'INIT' line is not present in the
configuration file then the currState variable will remain
uninitialized which may result in a crash.
KVM used to use two signals, one for instruction count exits and one
for timer exits. There is really no need to distinguish between the
two since they only trigger exits from KVM. This changeset unifies and
renames the signals and adds a method, kick(), that can be used to
raise the control signal in the vCPU thread. It also removes the early
timer warning since we do not normally see if the signal was
delivered.
--HG--
extra : rebase_source : cd0e45ca90894c3d6f6aa115b9b06a1d8f0fda4d
gem5 seems to store the PC as RIP+CS_BASE. This is not what KVM
expects, so we need to subtract CS_BASE prior to transferring the PC
into KVM. This changeset adds the necessary PC manipulation and
refactors thread context updates slightly to avoid reading registers
multiple times from KVM.
--HG--
extra : rebase_source : 3f0569dca06a1fcd8694925f75c8918d954ada44
This changeset adds support for INIT and STARTUP IPI handling. We
currently handle both of these interrupts in gem5 and transfer the
state to KVM. Since we do not have a BIOS loaded, we pretend that the
INIT interrupt suspends the CPU after reset.
--HG--
extra : rebase_source : 7f3b25f3801d68f668b6cd91eaf50d6f48ee2a6a
When transferring segment registers into kvm, we need to find the
value of the unusable bit. We used to assume that this could be
inferred from the selector since segments are generally unusable if
their selector is 0. This assumption breaks in some weird corner
cases. Instead, we just assume that segments are always usable. This
is what qemu does so it should work.
Signal handlers in KVM are controlled per thread and should be
initialized from the thread that is going to execute the CPU. This
changeset moves the initialization call from startup() to
startupThread().
A copyRegs() function is added to MIPS utilities
to copy architectural state from the old CPU to
the new CPU during fast-forwarding. This
addition alone enables fast-forwarding for the
o3 cpu model running MIPS.
The patch also adds takeOverFrom() and
drainResume() functions to the InOrderCPU to
enable it to take over from another CPU. This
change enables fast-forwarding for the inorder
cpu model running MIPS, but not for Alpha.
Committed by: Nilay Vaish <nilay@cs.wisc.edu>