CPU switching consists of the following steps:
1. Drain the system
2. Switch out old CPUs (cpu.switchOut())
3. Change the system timing mode to the mode the new CPUs require
4. Flush caches if switching to hardware virtualization
5. Inform new CPUs of the handover (cpu.takeOverFrom())
6. Resume the system
m5.switchCpus() previously only did step 2 & 5. Since information
about the new processors' memory system requirements is now exposed,
do all of the steps above.
This patch adds automatic memory system switching and flush (if
needed) to switchCpus(). Additionally, it adds optional draining to
switchCpus(). This has the following implications:
* changeToTiming and changeToAtomic are no longer needed, so they have
been removed.
* changeMemoryMode is only used internally, so it is has been renamed
to be private.
* switchCpus requires a reference to the system containing the CPUs as
its first parameter.
WARNING: This changeset breaks compatibility with existing
configuration scripts since it changes the signature of
m5.switchCpus().
This patch moves the default DRAM parameters from the SimpleDRAM class
to two different subclasses, one for DDR3 and one for LPDDR2. More can
be added as we go forward.
The regressions that previously used the SimpleDRAM are now using
SimpleDDR3 as this is the most similar configuration.
The actual statistical values are being updated for only two tests belonging
to sparc architecture and inorder cpu: 00.hello and 02.insttest. For others
the patch updates config.ini and name changes to statistical variables.
This changeset adds a set of tests that stress the CPU switching
code. It adds the following test configurations:
* tsunami-switcheroo-full -- Alpha system (atomic, timing, O3)
* realview-switcheroo-atomic -- ARM system (atomic<->atomic)
* realview-switcheroo-timing -- ARM system (timing<->timing)
* realview-switcheroo-o3 -- ARM system (O3<->O3)
* realview-switcheroo-full -- ARM system (atomic, timing, O3)
Reference data is provided for the 10.linux-boot test case. All of the
tests trigger a CPU switch once per millisecond during the boot
process.
The in-order CPU model was not included in any of the tests as it does
not support CPU handover.
This patch generalises the address range resolution for the I/O cache
and I/O bridge such that they do not assume a single memory. The patch
involves adding a parameter to the system which is then defined based
on the memories that are to be visible from the I/O subsystem, whether
behind a cache or a bridge.
The change is needed to allow interleaved memory controllers in the
system.
This patch adds support for reading input traces encoded using
protobuf according to what is done in the CommMonitor.
A follow-up patch adds a Python script that can be used to convert the
previously used ASCII traces to protobuf equivalents. The appropriate
regression input is updated as part of this patch.
The EIO tests depend on the EIO support from the "encumbered"
repository, which means that they are not normally built with
gem5. This causes all EIO related tests to fail, which is both
annoying and confusing. This patch addresses this by adding support
for skipping tests if certain conditions (e.g., the presence of a
SimObject) can not be met. It introduces the following Python
functions that can be called from within a test case:
* skip_test -- Skip a test and optionally print why the test was
skipped.
* has_sim_object -- Test if a SimObject exists.
* require_sim_object -- Test if a SimObject exists and skip, or
optionally fail, the test if not.
Additionally, this patch updates the EIO tests to check for the
presence of EioProcess.
This patch adds packet tracing to the communication monitor using a
protobuf as the mechanism for creating the trace.
If no file is specified, then the tracing is disabled. If a file is
specified, then for every packet that is successfully sent, a protobuf
message is serialized to the file.
This patch changes the traffic generator period such that it does not
completely saturate the DRAM controller and create an ever-growing
backlog in the queued port.
A separate patch updates the stats.
The ISA class on stores the contents of ID registers on many
architectures. In order to make reset values of such registers
configurable, we make the class inherit from SimObject, which allows
us to use the normal generated parameter headers.
This patch introduces a Python helper method, BaseCPU.createThreads(),
which creates a set of ISAs for each of the threads in an SMT
system. Although it is currently only needed when creating
multi-threaded CPUs, it should always be called before instantiating
the system as this is an obvious place to configure ID registers
identifying a thread/CPU.
Previous to this change we didn't always set the memory mode which worked as
long as we never attempted to switch CPUs or checked that a CPU was in a
memory system with the correct mode. Future changes will make CPUs verify
that they're operating in the correct mode and thus we need to always set it.
Most of the test cases currently contain a large amount of duplicated
boiler plate code. This changeset introduces a set of classes that
encapsulates most of the functionality when setting up a test
configuration.
The following base classes are introduced:
* BaseSystem - Basic system configuration that can be used for both
SE and FS simulation.
* BaseFSSystem - Basic FS configuration uni-processor and multi-processor
configurations.
* BaseFSSystemUniprocessor - Basic FS configuration for uni-processor
configurations. This is provided as a way
to make existing test cases backwards
compatible.
Architecture specific implementations are provided for ARM, Alpha, and
X86.
This patch unified the L1 and L2 caches used throughout the
regressions instead of declaring different, but very similar,
configurations in the different scripts.
The patch also changes the default L2 configuration to match what it
used to be for the fs and se scripts (until the last patch that
updated the regressions to also make use of the cache config). The
MSHRs and targets per MSHR are now set to a more realistic default of
20 and 12, respectively.
As a result of both the aforementioned changes, many of the regression
stats are changed. A follow-on patch will bump the stats.
This patch updates the stats to reflect the change in the default
system clock from 1 THz to 1GHz. The changes are due to the DMA
devices now injecting requests at a lower pace.
This patch bumps the stats to match the use of SimpleDRAM instead of
SimpleMemory in all inorder and O3 regressions, and also all
full-system regressions. A number of performance-related stats change,
and a whole bunch of stats are added for the memory controller.
This patch favours using SimpleDRAM with the default timing instead of
SimpleMemory for all regressions that involve the o3 or inorder CPU,
or are full system (in other words, where the actual performance of
the memory is important for the overall performance).
Moving forward, the solution for FSConfig and the users of fs.py and
se.py is probably something similar to what we use to choose the CPU
type. I envision a few pre-set configurations SimpleLPDDR2,
SimpleDDR3, etc that can be choosen by a dram_type option. Feedback on
this part is welcome.
This patch changes plenty stats and adds all the DRAM controller
related stats. A follow-on patch updates the relevant statistics. The
total run-time for the entire regression goes up with ~5% with this
patch due to the added complexity of the SimpleDRAM model. This is a
concious trade-off to ensure that the model is properly tested.
This patch uses the common L1, L2 and IOCache configuration for the
regressions that all share the same cache parameters. There are a few
regressions that use a slightly different configuration (memtest,
o3-timing=mp, simple-atomic-mp and simple-timing-mp), and the latter
are not changed in this patch. They will be updated in a future patch.
The common cache configurations are changed to match the ones used in
the regressions, and are slightly changed with respect to what they
were. Hopefully this means we can converge on a common base
configuration, used both in the normal user configurations and
regressions.
As only regressions that shared the same cache configuration are
updated, no regressions are affected.
This patch updates the stats to reflect the change in how cache
latencies are expressed. In addition, the latencies are now rounded to
multiples of the clock period, thus also affecting other stats.
This patch changes the cache-related latencies from an absolute time
expressed in Ticks, to a number of cycles that can be scaled with the
clock period of the caches. Ultimately this patch serves to enable
future work that involves dynamic frequency scaling. As an immediate
benefit it also makes it more convenient to specify cache performance
without implicitly assuming a specific CPU core operating frequency.
The stat blocked_cycles that actually counter in ticks is now updated
to count in cycles.
As the timing is now rounded to the clock edges of the cache, there
are some regressions that change. Plenty of them have very minor
changes, whereas some regressions with a short run-time are perturbed
quite significantly. A follow-on patch updates all the statistics for
the regressions.
This patch changes the memtest clock from 1THz (the default) to 2GHz,
similar to the CPUs in the other regressions. This is useful as the
caches will adopt the same clock as the CPU. The bus clock rate is
scaled accordingly, and the L1-L2 bus is kept at the CPU clock while
the memory bus is at half that frequency.
A separate patch updates the affected stats.
This patch unifies the full-system regression config scripts and uses
the BaseCPU convenience method addTwoLevelCacheHierarchy to connect up
the L1s and L2, and create the bus inbetween.
The patch is a step on the way to use the clock period to express the
cache latencies, as the CPU is now the parent of the L1, L2 and L1-L2
bus, and these modules thus use the CPU clock.
The patch does not change the value of any stats, but plenty names,
and a follow-up patch contains the update to the stats, chaning
system.l2c to system.cpu.l2cache.
In the current caches the hit latency is paid twice on a miss. This patch lets
a configurable response latency be set of the cache for the backward path.
This patch merely adds a clock other than the default 1 Tick for the
CPUs of both the test system and drive system for the twosys-tsunami
regression.
The CPU frequency of the driver system is choosed to be twice that of
the test system to ensure it is not the bottleneck (although in this
case it mostly serves as a demonstration of a two-system setup),
This patch adds a basic regression for the traffic generator. The
regression also serves as an example of the file formats used. More
complex regressions that make use of a DRAM controller model will
follow shortly.
This patch simply bumps the stats to avoid having failing
regressions. Someone with more insight in the changes should verify
that these differences all make sense.
This patch removes the NACKing in the bridge, as the split
request/response busses now ensure that protocol deadlocks do not
occur, i.e. the message-dependency chain is broken by always allowing
responses to make progress without being stalled by requests. The
NACKs had limited support in the system with most components ignoring
their use (with a suitable call to panic), and as the NACKs are no
longer needed to avoid protocol deadlocks, the cleanest way is to
simply remove them.
The bridge is the starting point as this is the only place where the
NACKs are created. A follow-up patch will remove the code that deals
with NACKs in the endpoints, e.g. the X86 table walker and DMA
port. Ultimately the type of packet can be complete removed (until
someone sees a need for modelling more complex protocols, which can
now be done in parts of the system since the port and interface is
split).
As a consequence of the NACK removal, the bridge now has to send a
retry to a master if the request or response queue was full on the
first attempt. This change also makes the bridge ports very similar to
QueuedPorts, and a later patch will change the bridge to use these. A
first step in this direction is taken by aligning the name of the
member functions, as done by this patch.
A bit of tidying up has also been done as part of the simplifications.
Surprisingly, this patch has no impact on any of the
regressions. Hence, there was never any NACKs issued. In a follow-up
patch I would suggest changing the size of the bridge buffers set in
FSConfig.py to also test the situation where the bridge fills up.
This patch changes the simple memory to have a single slave port
rather than a vector port. The simple memory makes no attempts at
modelling the contention between multiple ports, and any such
multiplexing and demultiplexing could be done in a bus (or crossbar)
outside the memory controller. This scenario also matches with the
ongoing work on a SimpleDRAM model, which will be a single-ported
single-channel controller that can be used in conjunction with a bus
(or crossbar) to create a multi-port multi-channel controller.
There are only very few regressions that make use of the vector port,
and these are all for functional accesses only. To facilitate these
cases, memtest and memtest-ruby have been updated to also have a
"functional" bus to perform the (de)multiplexing of the functional
memory accesses.
This patch bumps all the stats to reflect the bus changes, i.e. the
introduction of the state variable, the division into a request and
response layer, and the new default bus width of 8 bytes.
This patch introduces a class hierarchy of buses, a non-coherent one,
and a coherent one, splitting the existing bus functionality. By doing
so it also enables further specialisation of the two types of buses.
A non-coherent bus connects a number of non-snooping masters and
slaves, and routes the request and response packets based on the
address. The request packets issued by the master connected to a
non-coherent bus could still snoop in caches attached to a coherent
bus, as is the case with the I/O bus and memory bus in most system
configurations. No snoops will, however, reach any master on the
non-coherent bus itself. The non-coherent bus can be used as a
template for modelling PCI, PCIe, and non-coherent AMBA and OCP buses,
and is typically used for the I/O buses.
A coherent bus connects a number of (potentially) snooping masters and
slaves, and routes the request and response packets based on the
address, and also forwards all requests to the snoopers and deals with
the snoop responses. The coherent bus can be used as a template for
modelling QPI, HyperTransport, ACE and coherent OCP buses, and is
typically used for the L1-to-L2 buses and as the main system
interconnect.
The configuration scripts are updated to use a NoncoherentBus for all
peripheral and I/O buses.
A bit of minor tidying up has also been done.
--HG--
rename : src/mem/bus.cc => src/mem/coherent_bus.cc
rename : src/mem/bus.hh => src/mem/coherent_bus.hh
rename : src/mem/bus.cc => src/mem/noncoherent_bus.cc
rename : src/mem/bus.hh => src/mem/noncoherent_bus.hh
This patch updates the stats for parser to be aligned with the most
up-to-date behaviour. Somehow the wrong results got committed as part
of 8800b05e1cb3 (see details below) when fixing the no_value -> nan
stats.
changeset: 8983:8800b05e1cb3
user: Nathan Binkert <nate@binkert.org>
summary: stats: update stats for no_value -> nan
The kernel originally used to generate the stats is different from the one
at use on zizzer. This patch updates the stats with the correct kernel in
use.
