gem5.git
9 years agokvm: Fix dumping code for large registers
Andreas Sandberg [Sat, 23 May 2015 12:37:22 +0000 (13:37 +0100)]
kvm: Fix dumping code for large registers

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.

9 years agokvm, x86: Guard x86-specific APIs in KvmVM
Andreas Sandberg [Sat, 23 May 2015 12:37:20 +0000 (13:37 +0100)]
kvm, x86: Guard x86-specific APIs in KvmVM

Protect x86-specific APIs in KvmVM with compile-time guards to avoid
breaking ARM builds.

9 years agobuild: Don't test for KVM xsave support on ARM
Andreas Sandberg [Sat, 23 May 2015 12:37:18 +0000 (13:37 +0100)]
build: Don't test for KVM xsave support on ARM

The current build tests for KVM unconditionally check for xsave
support. This obviously never works on ARM since xsave is
x86-specific. This changeset refactors the build tests probing for KVM
support and moves the xsave test to an x86-specific section of
is_isa_kvm_compatible().

9 years agoarm: Workaround incorrect HDLCD register order in kernel
Andreas Sandberg [Sat, 23 May 2015 12:37:04 +0000 (13:37 +0100)]
arm: Workaround incorrect HDLCD register order in kernel

Some versions of the kernel incorrectly swap the red and blue color
select registers. This changeset adds a workaround for that by
swapping them when instantiating a PixelConverter.

9 years agobase: Redesign internal frame buffer handling
Andreas Sandberg [Sat, 23 May 2015 12:37:03 +0000 (13:37 +0100)]
base: Redesign internal frame buffer handling

Currently, frame buffer handling in gem5 is quite ad hoc. In practice,
we pass around naked pointers to raw pixel data and expect consumers
to convert frame buffers using the (broken) VideoConverter.

This changeset completely redesigns the way we handle frame buffers
internally. In summary, it fixes several color conversion bugs, adds
support for more color formats (e.g., big endian), and makes the code
base easier to follow.

In the new world, gem5 always represents pixel data using the Pixel
struct when pixels need to be passed between different classes (e.g.,
a display controller and the VNC server). Producers of entire frames
(e.g., display controllers) should use the FrameBuffer class to
represent a frame.

Frame producers are expected to create one instance of the FrameBuffer
class in their constructors and register it with its consumers
once. Consumers are expected to check the dimensions of the frame
buffer when they consume it.

Conversion between the external representation and the internal
representation is supported for all common "true color" RGB formats of
up to 32-bit color depth. The external pixel representation is
expected to be between 1 and 4 bytes in either big endian or little
endian. Color channels are assumed to be contiguous ranges of bits
within each pixel word. The external pixel value is scaled to an 8-bit
internal representation using a floating multiplication to map it to
the entire 8-bit range.

9 years agobase: Clean up bitmap generation code
Andreas Sandberg [Sat, 23 May 2015 12:37:01 +0000 (13:37 +0100)]
base: Clean up bitmap generation code

The bitmap generation code is hard to follow and incorrectly uses the
size of an enum member to calculate the size of a pixel. This
changeset cleans up the code and adds some documentation.

9 years agoruby: Fix RubySystem warm-up and cool-down scope
Joel Hestness [Tue, 19 May 2015 15:56:51 +0000 (10:56 -0500)]
ruby: Fix RubySystem warm-up and cool-down scope

The processes of warming up and cooling down Ruby caches are simulation-wide
processes, not just RubySystem instance-specific processes. Thus, the warm-up
and cool-down variables should be globally visible to any Ruby components
participating in either process. Make these variables static members and track
the warm-up and cool-down processes as appropriate.

This patch also has two side benefits:
1) It removes references to the RubySystem g_system_ptr, which are problematic
for allowing multiple RubySystem instances in a single simulation. Warmup and
cooldown variables being static (global) reduces the need for instance-specific
dereferences through the RubySystem.
2) From the AbstractController, it removes local RubySystem pointers, which are
used inconsistently with other uses of the RubySystem: 11 other uses reference
the RubySystem with the g_system_ptr. Only sequencers have local pointers.

9 years agoarm: Identify table-walker requests
Andreas Hansson [Fri, 15 May 2015 17:40:01 +0000 (13:40 -0400)]
arm: Identify table-walker requests

This patch ensures all page-table walks are flagged as such.

9 years agomisc: Appease gcc 5.1
Andreas Hansson [Fri, 15 May 2015 17:39:53 +0000 (13:39 -0400)]
misc: Appease gcc 5.1

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.

9 years agosim: Don't clear the active CPU vector in System::initState
Andreas Sandberg [Fri, 15 May 2015 17:39:44 +0000 (13:39 -0400)]
sim: Don't clear the active CPU vector in System::initState

The system class currently clears the vector of active CPUs in
initState(). CPUs are added to the list by registerThreadContext()
which is called from BaseCPU::init(). This obviously breaks when the
System object is initialized after the CPUs. This changeset removes
the offending clear() call since the list will be empty after it has
been instantiated anyway.

9 years agoconfig: Use null memory for DRAM sweep script
Andreas Hansson [Fri, 15 May 2015 17:38:46 +0000 (13:38 -0400)]
config: Use null memory for DRAM sweep script

Do not waste time when we do not care about the data.

9 years agoconfig: Add new MemConfig options to DRAM sweep script
Wendy Elsasser [Fri, 15 May 2015 17:38:45 +0000 (13:38 -0400)]
config: Add new MemConfig options to DRAM sweep script

Update script to match current MemConfig options with
external_memory_system option set to 0.

9 years agosyscall_emul: fix warn_once behavior
Steve Reinhardt [Tue, 5 May 2015 16:25:59 +0000 (09:25 -0700)]
syscall_emul: fix warn_once behavior

The current ignoreWarnOnceFunc doesn't really work as expected,
since it will only generate one warning total, for whichever
"warn-once" syscall is invoked first.  This patch fixes that
behavior by keeping a "warned" flag in the SyscallDesc object,
allowing suitably flagged syscalls to warn exactly once per
syscall.

9 years agostats, arm: Update stats for missing FPEXC.EN check
Andreas Hansson [Tue, 5 May 2015 07:22:48 +0000 (03:22 -0400)]
stats, arm: Update stats for missing FPEXC.EN check

Only one regression is affected.

9 years agoarm: Add missing FPEXC.EN check
Andreas Hansson [Tue, 5 May 2015 07:22:45 +0000 (03:22 -0400)]
arm: Add missing FPEXC.EN check

Add a missing check to ensure that exceptions are generated properly.

9 years agoarm: enable DCZVA by default in SE mode
Giacomo Gabrielli [Tue, 5 May 2015 07:22:42 +0000 (03:22 -0400)]
arm: enable DCZVA by default in SE mode

9 years agostats: Update stats to reflect cache changes
Andreas Hansson [Tue, 5 May 2015 07:22:39 +0000 (03:22 -0400)]
stats: Update stats to reflect cache changes

9 years agomem: Create a request copy for deferred snoops
Stephan Diestelhorst [Tue, 17 Mar 2015 11:50:55 +0000 (11:50 +0000)]
mem: Create a request copy for deferred snoops

Sometimes, we need to defer an express snoop in an MSHR, but the original
request might complete and deallocate the original pkt->req.  In those cases,
create a copy of the request so that someone who is inspecting the delayed
snoop can also inspect the request still.  All of this is rather hacky, but the
allocation / linking and general life-time management of Packet and Request is
rather tricky.  Deleting the copy is another tricky area, testing so far has
shown that the right copy is deleted at the right time.

9 years agoarm: Relax ordering for some uncacheable accesses
Andreas Sandberg [Tue, 5 May 2015 07:22:34 +0000 (03:22 -0400)]
arm: Relax ordering for some uncacheable accesses

We currently assume that all uncacheable memory accesses are strictly
ordered. Instead of always enforcing strict ordering, we now only
enforce it if the required memory type is device memory or strongly
ordered memory.

9 years agomem, cpu: Add a separate flag for strictly ordered memory
Andreas Sandberg [Tue, 5 May 2015 07:22:33 +0000 (03:22 -0400)]
mem, cpu: Add a separate flag for strictly ordered memory

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.

9 years agomem, alpha: Move Alpha-specific request flags
Andreas Sandberg [Tue, 5 May 2015 07:22:31 +0000 (03:22 -0400)]
mem, alpha: Move Alpha-specific request flags

Move Alpha-specific memory request flags to an architecture-specific
header and map them to the architecture specific flag bit range.

9 years agoarm: Remove unnecessary boot uncachability
Andreas Hansson [Tue, 5 May 2015 07:22:30 +0000 (03:22 -0400)]
arm: Remove unnecessary boot uncachability

With the recent patches addressing how we deal with uncacheable
accesses there is no longer need for the work arounds put in place to
enforce certain sections of memory to be uncacheable during boot.

9 years agomem: Snoop into caches on uncacheable accesses
Andreas Hansson [Tue, 5 May 2015 07:22:29 +0000 (03:22 -0400)]
mem: Snoop into caches on uncacheable accesses

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.

9 years agoarch, cpu: Do not forward snoops to table walker
Andreas Hansson [Tue, 5 May 2015 07:22:27 +0000 (03:22 -0400)]
arch, cpu: Do not forward snoops to table walker

This patch simplifies the overall CPU by changing the TLB caches such
that they do not forward snoops to the table walker port(s). Note that
only ARM and X86 are affected.

There is no reason for the ports to snoop as they do not actually take
any action, and from a performance point of view we are better of not
snooping more than we have to.

Should it at a later point be required to snoop for a particular TLB
design it is easy enough to add it back.

9 years agomem: Pass shared downstream through caches
Andreas Hansson [Tue, 5 May 2015 07:22:26 +0000 (03:22 -0400)]
mem: Pass shared downstream through caches

This patch ensures that we pass on information about a packet being
shared (rather than exclusive), when forwarding a packet downstream.

Without this patch there is a risk that a downstream cache considers
the line exclusive when it really isn't.

9 years agomem: Add forward snoop check for HardPFReqs
Ali Jafri [Tue, 5 May 2015 07:22:25 +0000 (03:22 -0400)]
mem: Add forward snoop check for HardPFReqs

We should always check whether the cache is supposed to be forwarding snoops
before generating snoops.

9 years agomem: Add missing stats update for uncacheable MSHRs
Andreas Hansson [Tue, 5 May 2015 07:22:24 +0000 (03:22 -0400)]
mem: Add missing stats update for uncacheable MSHRs

This patch adds a missing counter update for the uncacheable
accesses. By updating this counter we also get a meaningful average
latency for uncacheable accesses (previously inf).

9 years agomem: Tidy up BaseCache parameters
Andreas Hansson [Tue, 5 May 2015 07:22:22 +0000 (03:22 -0400)]
mem: Tidy up BaseCache parameters

This patch simply tidies up the BaseCache parameters and removes the
unused "two_queue" parameter.

9 years agomem: Remove templates in cache model
David Guillen [Tue, 5 May 2015 07:22:21 +0000 (03:22 -0400)]
mem: Remove templates in cache model

This patch changes the cache implementation to rely on virtual methods
rather than using the replacement policy as a template argument.

There is no impact on the simulation performance, and overall the
changes make it easier to modify (and subclass) the cache and/or
replacement policy.

9 years agocpu: Work around gcc 4.9 issues with Num_OpClasses
Andreas Hansson [Tue, 5 May 2015 07:22:19 +0000 (03:22 -0400)]
cpu: Work around gcc 4.9 issues with Num_OpClasses

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.

9 years agostats: Bring regression stats in line with actual behaviour
Andreas Hansson [Tue, 5 May 2015 07:22:17 +0000 (03:22 -0400)]
stats: Bring regression stats in line with actual behaviour

9 years agostats: arm: updates
Nilay Vaish [Thu, 30 Apr 2015 19:17:43 +0000 (14:17 -0500)]
stats: arm: updates

9 years agostats: x86: updates due to change in div latency
Nilay Vaish [Thu, 30 Apr 2015 03:35:23 +0000 (22:35 -0500)]
stats: x86: updates due to change in div latency

9 years agoarch, base, dev, kern, sym: FreeBSD support
Ruslan Bukin [Thu, 30 Apr 2015 03:35:23 +0000 (22:35 -0500)]
arch, base, dev, kern, sym: FreeBSD support

This adds support for FreeBSD/aarch64 FS and SE mode (basic set of syscalls only)

Committed by: Nilay Vaish <nilay@cs.wisc.edu>

9 years agomem: Simplify page close checks for adaptive policies
Rizwana Begum [Thu, 30 Apr 2015 03:35:22 +0000 (22:35 -0500)]
mem: Simplify page close checks for adaptive policies

Both open_adaptive and close_adaptive page polices keep the page
open if a row hit is found. If a row hit is not found, close_adaptive
page policy precharges the row, and open_adaptive policy precharges
the row only if there is a bank conflict request waiting in the queue.

This patch makes the checks for above conditions simpler.

Committed by: Nilay Vaish <nilay@cs.wisc.edu>

9 years agoruby: set: replace long by unsigned long
Nilay Vaish [Thu, 30 Apr 2015 03:35:22 +0000 (22:35 -0500)]
ruby: set: replace long by unsigned long
UBSan complains about negative value being shifted

9 years agocpu: o3: replace issueLatency with bool pipelined
Nilay Vaish [Thu, 30 Apr 2015 03:35:22 +0000 (22:35 -0500)]
cpu: o3: replace issueLatency with bool pipelined

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.

9 years agocpu: o3: single cycle default div microop latency on x86
Nilay Vaish [Thu, 30 Apr 2015 03:35:22 +0000 (22:35 -0500)]
cpu: o3: single cycle default div microop latency on x86

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.

9 years agox86: change divide-by-zero fault to divide-error
Nilay Vaish [Thu, 30 Apr 2015 03:35:22 +0000 (22:35 -0500)]
x86: change divide-by-zero fault to divide-error
Same exception is raised whether division with zero is performed or the
quotient is greater than the maximum value that the provided space can hold.
Divide-by-Zero is the AMD terminology, while Divide-Error is Intel's.

9 years agomisc: Appease gcc 5.1 without moving GDB_REG_BYTES
Andreas Hansson [Fri, 24 Apr 2015 07:30:08 +0000 (03:30 -0400)]
misc: Appease gcc 5.1 without moving GDB_REG_BYTES

This patch rolls back the move of the GDB_REG_BYTES constant, and
instead adds M5_VAR_USED.

9 years agoconfig: enable setting SE-mode environment variables from file
bpotter [Thu, 23 Apr 2015 20:40:18 +0000 (13:40 -0700)]
config: enable setting SE-mode environment variables from file

9 years agoarm, dev: Add a UFS device
Rene de Jong [Thu, 23 Apr 2015 17:37:50 +0000 (13:37 -0400)]
arm, dev: Add a UFS device

This patch introduces a UFS host controller and a UFS device. More
information about the UFS standard can be found at the JEDEC site:
http://www.jedec.org/standards-documents/results/jesd220

Note that the model does not implement the complete standard, and as
such is not an actual implementation of UFS. The following SCSI
commands are implemented: inquiry, read, read capacity, report LUNs,
start/stop, test unit ready, verify, write, format unit, send
diagnostic, synchronize cache, mode select, mode sense, request sense,
unmap, write buffer and read buffer. This is sufficient for usage with
Linux and Android.

To interact with this model a kernel version 3.9 or above is
needed.

9 years agoarm, dev: Add a NAND flash timing model
Rene de Jong [Thu, 23 Apr 2015 17:37:49 +0000 (13:37 -0400)]
arm, dev: Add a NAND flash timing model

This adds a NAND flash timing model. This model takes the number of
planes into account and is ultimately intended to be used as a
high-level performance model for any device using flash. To access the
memory, use either readMemory or writeMemory.

To make use of the model you will need an interface model
such as UFSHostDevice, which is part of a separate patch.

At the moment the flash device is part of the ARM device tree since
the only use if the UFSHostDevice, and that in turn relies on the ARM
GIC.

9 years agodev: Add support for i2c devices
Peter Enns [Thu, 23 Apr 2015 17:37:48 +0000 (13:37 -0400)]
dev: Add support for i2c devices

This patch adds an I2C bus and base device. I2C is used to connect a
variety of sensors, and this patch serves as a starting point to
enable a range of I2C devices.

9 years agomisc: Appease gcc 5.1
Andreas Hansson [Thu, 23 Apr 2015 17:37:46 +0000 (13:37 -0400)]
misc: Appease gcc 5.1

This patch fixes a few small issues to ensure gem5 compiles when using
gcc 5.1.

First, the GDB_REG_BYTES in the RemoteGDB header are, rather
surprisingly, flagged as unused for both ARM and X86. Removing them,
however, causes compilation errors as they are actually used in the
source file. Moving the constant into the class definition fixes the
issue. Possibly a gcc bug.

Second, we have an unused EthPktData constructor using auto_ptr, and
the latter is deprecated. Since the code is never used it is simply
removed.

9 years agostats: update for previous changeset
Steve Reinhardt [Thu, 23 Apr 2015 03:22:29 +0000 (20:22 -0700)]
stats: update for previous changeset

Very small differences in IQ-specific O3 stats.

9 years agocpu: remove conditional check (count > 0) on o3 IQ squashes
Brandon Potter [Wed, 22 Apr 2015 14:52:03 +0000 (07:52 -0700)]
cpu: remove conditional check (count > 0) on o3 IQ squashes

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.

9 years agosyscall_emul: implement clock_gettime system call
Brandon Potter [Wed, 22 Apr 2015 14:51:27 +0000 (07:51 -0700)]
syscall_emul: implement clock_gettime system call

9 years agosyscall_emul: update x86 syscall table
Monir Mozumder [Wed, 22 Apr 2015 14:51:27 +0000 (07:51 -0700)]
syscall_emul: update x86 syscall table
Update table with additional definitions through Linux 3.13.

9 years agosyscall_emul: update getrlimit to use warn
Brandon Potter [Wed, 22 Apr 2015 14:51:27 +0000 (07:51 -0700)]
syscall_emul: update getrlimit to use warn
Don't use std::cerr directly, and just return EINVAL instead of aborting.

9 years agosyscall_emul: fix warning with wrong syscall name
Brandon Potter [Wed, 22 Apr 2015 14:51:27 +0000 (07:51 -0700)]
syscall_emul: fix warning with wrong syscall name
Also nix extra whitespace.

9 years agobase: add new ChunkGenerator method to identify last chunk
Brandon Potter [Wed, 22 Apr 2015 14:51:27 +0000 (07:51 -0700)]
base: add new ChunkGenerator method to identify last chunk

9 years agostats: update a few stats from long O3 runs
Steve Reinhardt [Mon, 20 Apr 2015 22:09:43 +0000 (15:09 -0700)]
stats: update a few stats from long O3 runs

Very small changes to iew.predictedNotTakenIncorrect
and iew.branchMispredicts.  Looks like similar updates
were committed on April 3 (changeset 235ff1c046df), but
only for the quick tests.

9 years agocpu: Remove the InOrderCPU from the tree
Andreas Hansson [Mon, 20 Apr 2015 16:46:35 +0000 (12:46 -0400)]
cpu: Remove the InOrderCPU from the tree

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.

9 years agoconfig: Remove memory aliases and rely on class name
Andreas Hansson [Mon, 20 Apr 2015 16:46:29 +0000 (12:46 -0400)]
config: Remove memory aliases and rely on class name

Instead of maintaining two lists, rely entirely on the class
name. There is really no point in causing unecessary confusion.

9 years agoAdded tag stable_2015_04_15 for changeset e17949745150
Nilay Vaish [Wed, 15 Apr 2015 21:04:37 +0000 (16:04 -0500)]
Added tag stable_2015_04_15 for changeset e17949745150

9 years agostats: x86: changes due to recent patches
Nilay Vaish [Tue, 14 Apr 2015 16:01:11 +0000 (11:01 -0500)]
stats: x86: changes due to recent patches

The change in 20.parser is from new x87 instructions.  The change to
pc-o3-timing is not clear to me.  It seems that this test might be invoking
some undefined behavior.

9 years agoconfig, cpu: fix progress interval for switched CPUs
Malek Musleh [Tue, 14 Apr 2015 16:01:10 +0000 (11:01 -0500)]
config, cpu: fix progress interval for switched CPUs
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>

9 years agocpu: re-organizes the branch predictor structure.
Dibakar Gope [Mon, 13 Apr 2015 22:33:57 +0000 (17:33 -0500)]
cpu: re-organizes the branch predictor structure.

Committed by: Nilay Vaish <nilay@cs.wisc.edu>

9 years agox86: implements x87 mult/div instructions
Nilay Vaish [Mon, 13 Apr 2015 22:33:57 +0000 (17:33 -0500)]
x86: implements x87 mult/div instructions

9 years agoruby: allow restoring from checkpoint when using DRAMCtrl
Lena Olson [Mon, 13 Apr 2015 22:33:57 +0000 (17:33 -0500)]
ruby: allow restoring from checkpoint when using DRAMCtrl

Restoring from a checkpoint with ruby + the DRAMCtrl memory model was not
working, because ruby and DRAMCtrl disagreed on the current tick during warmup.
Since there is no reason to do timing requests during warmup, use functional
requests instead.

Committed by: Nilay Vaish <nilay@cs.wisc.edu>

9 years agosim: Use NULL instead of None for testing filenames.
Nilay Vaish [Mon, 13 Apr 2015 22:33:57 +0000 (17:33 -0500)]
sim: Use NULL instead of None for testing filenames.
The filenames are initialized with NULL.  So the test should be
checking for them to be == NULL instead == None.

9 years agosim: fix function for emulating dup()
Nilay Vaish [Mon, 13 Apr 2015 22:33:57 +0000 (17:33 -0500)]
sim: fix function for emulating dup()
The function was using the host fd to obtain the fd object from the simulated
process.

9 years agoconfig: Support full-system with SST's memory system
Curtis Dunham [Wed, 8 Apr 2015 20:56:06 +0000 (15:56 -0500)]
config: Support full-system with SST's memory system

This patch adds an example configuration in ext/sst/tests/ that allows
an SST/gem5 instance to simulate a 4-core AArch64 system with SST's
memHierarchy components providing all the caches and memories.

9 years agoext: Add SST connector
Curtis Dunham [Wed, 8 Apr 2015 20:56:06 +0000 (15:56 -0500)]
ext: Add SST connector

This patch adds a connector that allows gem5 to be used as a component
in SST (Structural Simulation Toolkit, sst-simulator.org). At a high
level, this allows memory traffic to pass between the two simulators.
SST Links are roughly analogous to gem5 Ports, although Links do not
have a notion of master and slave. This distinction is important to
gem5, so when connecting a gem5 CPU to an SST cache, an ExternalSlave
must be used, and similarly when connecting the memory side of SST cache
to a gem5 port (for memory <-> I/O), an ExternalMaster must be used.

These connectors handle the administrative aspects of gem5
(initialization, simulation, shutdown) as well as translating SST's
MemEvents into gem5 Packets and vice-versa.

9 years agostats: updates due to recent changesets.
Nilay Vaish [Fri, 3 Apr 2015 16:42:11 +0000 (11:42 -0500)]
stats: updates due to recent changesets.

9 years agodev: (un)serialize fix for the RTC and RTC Timer Interrupt events
Nikos Nikoleris [Fri, 3 Apr 2015 16:42:10 +0000 (11:42 -0500)]
dev: (un)serialize fix for the RTC and RTC Timer Interrupt events

Restoring from a checkpoint fails if either the RTC or the RTC Timer
Interrrupt event is disabled. The restored machine tried incorrectly
to schedule the next event with negative offset.

Committed by: Nilay Vaish <nilay@cs.wisc.edu>

9 years agosim: correct check for endianess
Ruslan Bukin [Fri, 3 Apr 2015 16:42:10 +0000 (11:42 -0500)]
sim: correct check for endianess

Committed by: Nilay Vaish <nilay@cs.wisc.edu>

9 years agodev: Extend access width for IDE control registers
Ruslan Bukin [Fri, 3 Apr 2015 16:42:10 +0000 (11:42 -0500)]
dev: Extend access width for IDE control registers

Add 32-bit access width for PrimaryTiming register and 16bit for UDMAControl
register as FreeBSD required.

Committed by: Nilay Vaish <nilay@cs.wisc.edu>

9 years agocpu: fix system total instructions accounting
Nikos Nikoleris [Fri, 3 Apr 2015 16:42:10 +0000 (11:42 -0500)]
cpu: fix system total instructions accounting

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>"

9 years agox86: fix debug trace output for mwait
Lena Olson [Fri, 3 Apr 2015 16:42:10 +0000 (11:42 -0500)]
x86: fix debug trace output for mwait

When running with the Exec flag, the mwait instruction attempted
to print out its source registers, which were never actually
initialized. This led to sporadic assertion failures when the
value stored there was invalid.

Committed by: Nilay Vaish <nilay@cs.wisc.edu>

9 years agoarm, configs: Do not forward snoops from I cache
Andreas Hansson [Fri, 27 Mar 2015 08:56:10 +0000 (04:56 -0400)]
arm, configs: Do not forward snoops from I cache

This fix simply tells the I cache to not forward snoops to the fetch
unit (since there is really no reason to do so).

9 years agomem: Support any number of master-IDs in stride prefetcher
Stephan Diestelhorst [Fri, 27 Mar 2015 08:56:03 +0000 (04:56 -0400)]
mem: Support any number of master-IDs in stride prefetcher

The stride prefetcher had a hardcoded number of contexts (i.e. master-IDs)
that it could handle. Since master IDs need to be unique per system, and
every core, cache etc. requires a separate master port, a static limit on
these does not make much sense.

Instead, this patch adds a small hash map that will map all master IDs to
the right prefetch state and dynamically allocates new state for new master
IDs.

9 years agomem: Allocate cache writebacks before new MSHRs
Andreas Hansson [Fri, 27 Mar 2015 08:56:02 +0000 (04:56 -0400)]
mem: Allocate cache writebacks before new MSHRs

This patch changes the order of writeback allocation such that any
writebacks resulting from a tag lookup (e.g. for an uncacheable
access), are added to the writebuffer before any new MSHR entries are
allocated. This ensures that the writebacks logically precedes the new
allocations.

The patch also changes the uncacheable flush to use proper timed (or
atomic) writebacks, as opposed to functional writes.

9 years agomem: Cleanup flow for uncacheable accesses
Andreas Hansson [Fri, 27 Mar 2015 08:56:01 +0000 (04:56 -0400)]
mem: Cleanup flow for uncacheable accesses

This patch simplifies the code dealing with uncacheable timing
accesses, aiming to align it with the existing miss handling. Similar
to what we do in atomic, a timing request now goes through
Cache::access (where the block is also flushed), and then proceeds to
ignore any existing MSHR for the block in question. This unifies the
flow for cacheable and uncacheable accesses, and for atomic and timing.

9 years agomem: Ignore uncacheable MSHRs when finding matches
Andreas Hansson [Fri, 27 Mar 2015 08:56:00 +0000 (04:56 -0400)]
mem: Ignore uncacheable MSHRs when finding matches

This patch changes how we search for matching MSHRs, ignoring any MSHR
that is allocated for an uncacheable access. By doing so, this patch
fixes a corner case in the MSHRs where incorrect data ended up being
copied into a (cacheable) read packet due to a first uncacheable MSHR
target of size 4, followed by a cacheable target to the same MSHR of
size 64. The latter target was filled with nonsense data.

9 years agomem: Remove redundant allocateUncachedReadBuffer in cache
Andreas Hansson [Fri, 27 Mar 2015 08:55:59 +0000 (04:55 -0400)]
mem: Remove redundant allocateUncachedReadBuffer in cache

This patch removes the no-longer-needed
allocateUncachedReadBuffer. Besides the checks it is exactly the same
as allocateMissBuffer and thus provides no value.

9 years agomem: Modernise MSHR iterators to C++11
Andreas Hansson [Fri, 27 Mar 2015 08:55:57 +0000 (04:55 -0400)]
mem: Modernise MSHR iterators to C++11

This patch updates the iterators in the MSHR and MSHR queues to use
C++11 range-based for loops. It also does a bit of additional house
keeping.

9 years agotests: Update stats for cache block alignment
Andreas Hansson [Fri, 27 Mar 2015 08:55:57 +0000 (04:55 -0400)]
tests: Update stats for cache block alignment

9 years agomem: Align all MSHR entries to block boundaries
Andreas Hansson [Fri, 27 Mar 2015 08:55:55 +0000 (04:55 -0400)]
mem: Align all MSHR entries to block boundaries

This patch aligns all MSHR queue entries to block boundaries to
simplify checks for matches. Previously there were corner cases that
could lead to existing entries not being identified as matches.

There are, rather alarmingly, a few regressions that change with this
patch.

9 years agomem: Rename PREFETCH_SNOOP_SQUASH flag to BLOCK_CACHED
Ali Jafri [Fri, 27 Mar 2015 08:55:54 +0000 (04:55 -0400)]
mem: Rename PREFETCH_SNOOP_SQUASH flag to BLOCK_CACHED

This patch subsumes the PREFETCH_SNOOP_SQUASH flag with the more
generic BLOCK_CACHED flag. Future patches implementing cache eviction
messages can use the BLOCK_CACHED flag in almost the same manner as
hardware prefetches use the PREFETCH_SNOOP_SQUASH flag. The
PREFTECH_SNOOP_FLAG is set if the prefetch target is found in the tags
or the MSHRs in any state, so we are simply replacing calls to
setPrefetchSquashed() with setBlockCached(). The case of where the
prefetch target is found in the writeback MSHRs of upper level caches
continues to be covered by the MEM_INHIBIT flag.

9 years agosim: Update limit_event reuse to final version
Curtis Dunham [Thu, 26 Mar 2015 15:16:44 +0000 (11:16 -0400)]
sim: Update limit_event reuse to final version

Matching final version on reviewboard.

9 years agocpu: Fix InstPBTrace inheritance
Andreas Hansson [Thu, 26 Mar 2015 15:16:43 +0000 (11:16 -0400)]
cpu: Fix InstPBTrace inheritance

This patch fixes an issue that prevented gem5 to be built with C++
config and without Python.

9 years agomem: rename Locked/LOCKED to LockedRMW/LOCKED_RMW
Steve Reinhardt [Mon, 23 Mar 2015 23:14:20 +0000 (16:14 -0700)]
mem: rename Locked/LOCKED to LockedRMW/LOCKED_RMW

Makes x86-style locked operations even more distinct from
LLSC operations.  Using "locked" by itself should be
obviously ambiguous now.

9 years agoconfig: expand '~' and '~user' in paths
Steve Reinhardt [Mon, 23 Mar 2015 23:14:19 +0000 (16:14 -0700)]
config: expand '~' and '~user' in paths

9 years agomisc: quote args in echoed command line
Steve Reinhardt [Mon, 23 Mar 2015 23:14:18 +0000 (16:14 -0700)]
misc: quote args in echoed command line

Currently if there are shell special characters in a
command-line argument, you can't copy and paste the
echoed command line onto a shell prompt because the
characters aren't quoted properly.  This patch fixes
that problem.

9 years agoconfig: Add ability to exit simulation after initialization
Curtis Dunham [Mon, 23 Mar 2015 10:57:38 +0000 (06:57 -0400)]
config: Add ability to exit simulation after initialization

When using gem5 as a slave simulator, it will not advance the
clock on its own and depends on the master simulator calling
simulate(). This new option lets us use the Python scripts
to do all the configuration while stopping short of actually
simulating anything.

9 years agosim: Reuse the same limit_event in simulate()
Curtis Dunham [Mon, 23 Mar 2015 10:57:36 +0000 (06:57 -0400)]
sim: Reuse the same limit_event in simulate()

This patch accomplishes two things:
1. Makes simulate()'s GlobalSimLoopExitEvent a singleton reused
   across calls. This is slightly more efficient than recreating
   it every time.
2. Gives callers to simulate() (especially other simulators) a
   foolproof way of knowing that the simulation period ended
   successfully by hitting the limit event. They can call
   getLimitEvent() and compare it to the return
   value of simulate().

This change was motivated by an ongoing effort to integrate gem5
and SST, with SST as the master sim and gem5 as the slave sim.

9 years agomem: Tidy up Request
Andreas Hansson [Mon, 23 Mar 2015 10:57:34 +0000 (06:57 -0400)]
mem: Tidy up Request

This patch does a bit of house keeping, fixing up typos, removing dead
code etc.

9 years agotests: Final reclassification of quick regressions
Andreas Hansson [Mon, 23 Mar 2015 10:57:31 +0000 (06:57 -0400)]
tests: Final reclassification of quick regressions

A few regressions were still considered long, but finished well within
the 180 seconds. They are only a handful (mostly mcf in atomic).

--HG--
rename : tests/long/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/config.ini => tests/quick/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/config.ini
rename : tests/long/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/simerr => tests/quick/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/simerr
rename : tests/long/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/simout => tests/quick/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/simout
rename : tests/long/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/stats.txt => tests/quick/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/stats.txt
rename : tests/long/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/system.terminal => tests/quick/fs/10.linux-boot/ref/arm/linux/realview-switcheroo-timing/system.terminal
rename : tests/long/se/10.mcf/ref/arm/linux/simple-atomic/chair.cook.ppm => tests/quick/se/10.mcf/ref/arm/linux/simple-atomic/chair.cook.ppm
rename : tests/long/se/10.mcf/ref/arm/linux/simple-atomic/config.ini => tests/quick/se/10.mcf/ref/arm/linux/simple-atomic/config.ini
rename : tests/long/se/10.mcf/ref/arm/linux/simple-atomic/mcf.out => tests/quick/se/10.mcf/ref/arm/linux/simple-atomic/mcf.out
rename : tests/long/se/10.mcf/ref/arm/linux/simple-atomic/simerr => tests/quick/se/10.mcf/ref/arm/linux/simple-atomic/simerr
rename : tests/long/se/10.mcf/ref/arm/linux/simple-atomic/simout => tests/quick/se/10.mcf/ref/arm/linux/simple-atomic/simout
rename : tests/long/se/10.mcf/ref/arm/linux/simple-atomic/stats.txt => tests/quick/se/10.mcf/ref/arm/linux/simple-atomic/stats.txt
rename : tests/long/se/10.mcf/ref/arm/linux/simple-timing/chair.cook.ppm => tests/quick/se/10.mcf/ref/arm/linux/simple-timing/chair.cook.ppm
rename : tests/long/se/10.mcf/ref/arm/linux/simple-timing/config.ini => tests/quick/se/10.mcf/ref/arm/linux/simple-timing/config.ini
rename : tests/long/se/10.mcf/ref/arm/linux/simple-timing/mcf.out => tests/quick/se/10.mcf/ref/arm/linux/simple-timing/mcf.out
rename : tests/long/se/10.mcf/ref/arm/linux/simple-timing/simerr => tests/quick/se/10.mcf/ref/arm/linux/simple-timing/simerr
rename : tests/long/se/10.mcf/ref/arm/linux/simple-timing/simout => tests/quick/se/10.mcf/ref/arm/linux/simple-timing/simout
rename : tests/long/se/10.mcf/ref/arm/linux/simple-timing/stats.txt => tests/quick/se/10.mcf/ref/arm/linux/simple-timing/stats.txt
rename : tests/long/se/10.mcf/ref/sparc/linux/simple-atomic/config.ini => tests/quick/se/10.mcf/ref/sparc/linux/simple-atomic/config.ini
rename : tests/long/se/10.mcf/ref/sparc/linux/simple-atomic/mcf.out => tests/quick/se/10.mcf/ref/sparc/linux/simple-atomic/mcf.out
rename : tests/long/se/10.mcf/ref/sparc/linux/simple-atomic/simerr => tests/quick/se/10.mcf/ref/sparc/linux/simple-atomic/simerr
rename : tests/long/se/10.mcf/ref/sparc/linux/simple-atomic/simout => tests/quick/se/10.mcf/ref/sparc/linux/simple-atomic/simout
rename : tests/long/se/10.mcf/ref/sparc/linux/simple-atomic/stats.txt => tests/quick/se/10.mcf/ref/sparc/linux/simple-atomic/stats.txt
rename : tests/long/se/10.mcf/ref/x86/linux/simple-atomic/config.ini => tests/quick/se/10.mcf/ref/x86/linux/simple-atomic/config.ini
rename : tests/long/se/10.mcf/ref/x86/linux/simple-atomic/mcf.out => tests/quick/se/10.mcf/ref/x86/linux/simple-atomic/mcf.out
rename : tests/long/se/10.mcf/ref/x86/linux/simple-atomic/simerr => tests/quick/se/10.mcf/ref/x86/linux/simple-atomic/simerr
rename : tests/long/se/10.mcf/ref/x86/linux/simple-atomic/simout => tests/quick/se/10.mcf/ref/x86/linux/simple-atomic/simout
rename : tests/long/se/10.mcf/ref/x86/linux/simple-atomic/stats.txt => tests/quick/se/10.mcf/ref/x86/linux/simple-atomic/stats.txt
rename : tests/long/se/10.mcf/test.py => tests/quick/se/10.mcf/test.py
rename : tests/long/se/30.eon/ref/alpha/tru64/simple-atomic/config.ini => tests/quick/se/30.eon/ref/alpha/tru64/simple-atomic/config.ini
rename : tests/long/se/30.eon/ref/alpha/tru64/simple-atomic/simerr => tests/quick/se/30.eon/ref/alpha/tru64/simple-atomic/simerr
rename : tests/long/se/30.eon/ref/alpha/tru64/simple-atomic/simout => tests/quick/se/30.eon/ref/alpha/tru64/simple-atomic/simout
rename : tests/long/se/30.eon/ref/alpha/tru64/simple-atomic/stats.txt => tests/quick/se/30.eon/ref/alpha/tru64/simple-atomic/stats.txt
rename : tests/long/se/30.eon/test.py => tests/quick/se/30.eon/test.py

9 years agostats: update Minor stats due to PF bug fix
Steve Reinhardt [Thu, 19 Mar 2015 12:41:32 +0000 (08:41 -0400)]
stats: update Minor stats due to PF bug fix

A recent changeset of mine (http://repo.gem5.org/gem5/rev/4cfe55719da5)
inadvertently fixed a bug in the Minor CPU model which caused it to treat
software prefetches as regular loads.  Prior to this changeset, Minor
did an ad-hoc generation of memory commands that left out the PF check;
because it now uses the common code that the other CPU models use,
it generates prefetches properly.  These stat changes reflect the fact
that the Minor model now issues SoftPFReqs.

9 years agotests: Recategorise regressions based on run time
Andreas Hansson [Thu, 19 Mar 2015 08:06:21 +0000 (04:06 -0400)]
tests: Recategorise regressions based on run time

This patch takes a first stab at recategorising the regression tests
based on actual run times. The simple-atomic and simple-timing runs of
vortex and twolf all finish in less than 180 s, and they are
consequently moved from long to quick. All realview64 linux-boot
regressions take more than 700 s, and they are therefore moved to
long.

Later patches will rename quick to short, and further divide the
regressions into short, medium and long.

--HG--
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/config.ini => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/config.ini
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/simerr => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/simerr
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/simout => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/simout
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/stats.txt => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/stats.txt
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/system.terminal => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic-dual/system.terminal
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/config.ini => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/config.ini
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/simerr => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/simerr
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/simout => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/simout
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/stats.txt => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/stats.txt
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/system.terminal => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-atomic/system.terminal
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/config.ini => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/config.ini
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/simerr => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/simerr
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/simout => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/simout
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/stats.txt => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/stats.txt
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/system.terminal => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing-dual/system.terminal
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/config.ini => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/config.ini
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/simerr => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/simerr
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/simout => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/simout
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/stats.txt => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/stats.txt
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/system.terminal => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-simple-timing/system.terminal
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/config.ini => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/config.ini
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/simerr => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/simerr
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/simout => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/simout
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/stats.txt => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/stats.txt
rename : tests/quick/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/system.terminal => tests/long/fs/10.linux-boot/ref/arm/linux/realview64-switcheroo-atomic/system.terminal
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-atomic/config.ini => tests/quick/se/50.vortex/ref/alpha/tru64/simple-atomic/config.ini
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-atomic/simerr => tests/quick/se/50.vortex/ref/alpha/tru64/simple-atomic/simerr
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-atomic/simout => tests/quick/se/50.vortex/ref/alpha/tru64/simple-atomic/simout
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-atomic/smred.msg => tests/quick/se/50.vortex/ref/alpha/tru64/simple-atomic/smred.msg
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-atomic/smred.out => tests/quick/se/50.vortex/ref/alpha/tru64/simple-atomic/smred.out
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-atomic/stats.txt => tests/quick/se/50.vortex/ref/alpha/tru64/simple-atomic/stats.txt
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-timing/config.ini => tests/quick/se/50.vortex/ref/alpha/tru64/simple-timing/config.ini
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-timing/simerr => tests/quick/se/50.vortex/ref/alpha/tru64/simple-timing/simerr
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-timing/simout => tests/quick/se/50.vortex/ref/alpha/tru64/simple-timing/simout
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-timing/smred.msg => tests/quick/se/50.vortex/ref/alpha/tru64/simple-timing/smred.msg
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-timing/smred.out => tests/quick/se/50.vortex/ref/alpha/tru64/simple-timing/smred.out
rename : tests/long/se/50.vortex/ref/alpha/tru64/simple-timing/stats.txt => tests/quick/se/50.vortex/ref/alpha/tru64/simple-timing/stats.txt
rename : tests/long/se/50.vortex/ref/arm/linux/simple-atomic/config.ini => tests/quick/se/50.vortex/ref/arm/linux/simple-atomic/config.ini
rename : tests/long/se/50.vortex/ref/arm/linux/simple-atomic/simerr => tests/quick/se/50.vortex/ref/arm/linux/simple-atomic/simerr
rename : tests/long/se/50.vortex/ref/arm/linux/simple-atomic/simout => tests/quick/se/50.vortex/ref/arm/linux/simple-atomic/simout
rename : tests/long/se/50.vortex/ref/arm/linux/simple-atomic/smred.out => tests/quick/se/50.vortex/ref/arm/linux/simple-atomic/smred.out
rename : tests/long/se/50.vortex/ref/arm/linux/simple-atomic/stats.txt => tests/quick/se/50.vortex/ref/arm/linux/simple-atomic/stats.txt
rename : tests/long/se/50.vortex/ref/arm/linux/simple-timing/config.ini => tests/quick/se/50.vortex/ref/arm/linux/simple-timing/config.ini
rename : tests/long/se/50.vortex/ref/arm/linux/simple-timing/simerr => tests/quick/se/50.vortex/ref/arm/linux/simple-timing/simerr
rename : tests/long/se/50.vortex/ref/arm/linux/simple-timing/simout => tests/quick/se/50.vortex/ref/arm/linux/simple-timing/simout
rename : tests/long/se/50.vortex/ref/arm/linux/simple-timing/smred.out => tests/quick/se/50.vortex/ref/arm/linux/simple-timing/smred.out
rename : tests/long/se/50.vortex/ref/arm/linux/simple-timing/stats.txt => tests/quick/se/50.vortex/ref/arm/linux/simple-timing/stats.txt
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-atomic/config.ini => tests/quick/se/50.vortex/ref/sparc/linux/simple-atomic/config.ini
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-atomic/simerr => tests/quick/se/50.vortex/ref/sparc/linux/simple-atomic/simerr
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-atomic/simout => tests/quick/se/50.vortex/ref/sparc/linux/simple-atomic/simout
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-atomic/smred.msg => tests/quick/se/50.vortex/ref/sparc/linux/simple-atomic/smred.msg
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-atomic/smred.out => tests/quick/se/50.vortex/ref/sparc/linux/simple-atomic/smred.out
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-atomic/stats.txt => tests/quick/se/50.vortex/ref/sparc/linux/simple-atomic/stats.txt
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-timing/config.ini => tests/quick/se/50.vortex/ref/sparc/linux/simple-timing/config.ini
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-timing/simerr => tests/quick/se/50.vortex/ref/sparc/linux/simple-timing/simerr
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-timing/simout => tests/quick/se/50.vortex/ref/sparc/linux/simple-timing/simout
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-timing/smred.msg => tests/quick/se/50.vortex/ref/sparc/linux/simple-timing/smred.msg
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-timing/smred.out => tests/quick/se/50.vortex/ref/sparc/linux/simple-timing/smred.out
rename : tests/long/se/50.vortex/ref/sparc/linux/simple-timing/stats.txt => tests/quick/se/50.vortex/ref/sparc/linux/simple-timing/stats.txt
rename : tests/long/se/50.vortex/test.py => tests/quick/se/50.vortex/test.py
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/config.ini => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/config.ini
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/simerr => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/simerr
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/simout => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/simout
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.out => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.out
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.pin => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.pin
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.pl1 => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.pl1
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.pl2 => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.pl2
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.sav => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.sav
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.sv2 => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.sv2
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.twf => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/smred.twf
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-atomic/stats.txt => tests/quick/se/70.twolf/ref/alpha/tru64/simple-atomic/stats.txt
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/config.ini => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/config.ini
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/simerr => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/simerr
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/simout => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/simout
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/smred.out => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/smred.out
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/smred.pin => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/smred.pin
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/smred.pl1 => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/smred.pl1
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/smred.pl2 => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/smred.pl2
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/smred.sav => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/smred.sav
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/smred.sv2 => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/smred.sv2
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/smred.twf => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/smred.twf
rename : tests/long/se/70.twolf/ref/alpha/tru64/simple-timing/stats.txt => tests/quick/se/70.twolf/ref/alpha/tru64/simple-timing/stats.txt
rename : tests/long/se/70.twolf/ref/arm/linux/simple-atomic/config.ini => tests/quick/se/70.twolf/ref/arm/linux/simple-atomic/config.ini
rename : tests/long/se/70.twolf/ref/arm/linux/simple-atomic/simerr => tests/quick/se/70.twolf/ref/arm/linux/simple-atomic/simerr
rename : tests/long/se/70.twolf/ref/arm/linux/simple-atomic/simout => tests/quick/se/70.twolf/ref/arm/linux/simple-atomic/simout
rename : tests/long/se/70.twolf/ref/arm/linux/simple-atomic/smred.out => tests/quick/se/70.twolf/ref/arm/linux/simple-atomic/smred.out
rename : tests/long/se/70.twolf/ref/arm/linux/simple-atomic/stats.txt => tests/quick/se/70.twolf/ref/arm/linux/simple-atomic/stats.txt
rename : tests/long/se/70.twolf/ref/arm/linux/simple-timing/config.ini => tests/quick/se/70.twolf/ref/arm/linux/simple-timing/config.ini
rename : tests/long/se/70.twolf/ref/arm/linux/simple-timing/simerr => tests/quick/se/70.twolf/ref/arm/linux/simple-timing/simerr
rename : tests/long/se/70.twolf/ref/arm/linux/simple-timing/simout => tests/quick/se/70.twolf/ref/arm/linux/simple-timing/simout
rename : tests/long/se/70.twolf/ref/arm/linux/simple-timing/smred.out => tests/quick/se/70.twolf/ref/arm/linux/simple-timing/smred.out
rename : tests/long/se/70.twolf/ref/arm/linux/simple-timing/stats.txt => tests/quick/se/70.twolf/ref/arm/linux/simple-timing/stats.txt
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/config.ini => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/config.ini
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/simerr => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/simerr
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/simout => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/simout
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/smred.out => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/smred.out
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/smred.pin => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/smred.pin
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/smred.pl1 => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/smred.pl1
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/smred.pl2 => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/smred.pl2
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/smred.sav => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/smred.sav
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/smred.sv2 => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/smred.sv2
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/smred.twf => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/smred.twf
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-atomic/stats.txt => tests/quick/se/70.twolf/ref/sparc/linux/simple-atomic/stats.txt
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/config.ini => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/config.ini
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/simerr => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/simerr
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/simout => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/simout
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/smred.out => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/smred.out
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/smred.pin => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/smred.pin
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/smred.pl1 => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/smred.pl1
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/smred.pl2 => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/smred.pl2
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/smred.sav => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/smred.sav
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/smred.sv2 => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/smred.sv2
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/smred.twf => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/smred.twf
rename : tests/long/se/70.twolf/ref/sparc/linux/simple-timing/stats.txt => tests/quick/se/70.twolf/ref/sparc/linux/simple-timing/stats.txt
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/config.ini => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/config.ini
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/simerr => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/simerr
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/simout => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/simout
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/smred.out => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/smred.out
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/smred.pin => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/smred.pin
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/smred.pl1 => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/smred.pl1
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/smred.pl2 => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/smred.pl2
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/smred.sav => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/smred.sav
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/smred.sv2 => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/smred.sv2
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/smred.twf => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/smred.twf
rename : tests/long/se/70.twolf/ref/x86/linux/simple-atomic/stats.txt => tests/quick/se/70.twolf/ref/x86/linux/simple-atomic/stats.txt
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/config.ini => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/config.ini
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/simerr => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/simerr
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/simout => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/simout
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/smred.out => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/smred.out
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/smred.pin => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/smred.pin
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/smred.pl1 => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/smred.pl1
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/smred.pl2 => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/smred.pl2
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/smred.sav => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/smred.sav
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/smred.sv2 => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/smred.sv2
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/smred.twf => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/smred.twf
rename : tests/long/se/70.twolf/ref/x86/linux/simple-timing/stats.txt => tests/quick/se/70.twolf/ref/x86/linux/simple-timing/stats.txt
rename : tests/long/se/70.twolf/test.py => tests/quick/se/70.twolf/test.py

9 years agotest, arm: Add scripts to test checkpoints
Andreas Sandberg [Thu, 19 Mar 2015 08:06:20 +0000 (04:06 -0400)]
test, arm: Add scripts to test checkpoints

Add a set of scripts to automatically test checkpointing in the
regression framework. The checkpointing tests are similar to the
switcheroo tests, but instead of switching between CPUs, they
checkpoint the system and restore from the checkpoint again. This is
done at regular intervals, typically while booting Linux.

The implementation is fairly straight forward, with the exception that
we have to work around gem5's inability to restore from a checkpoint
after a system has been instantiated. We work around this by forking
off child processes that does the actual simulation and never
instantiate a system in the parent process unless a maximum checkpoint
count is reached (in which case we just simulate the system to
completion in the parent).

Checkpoint testing is currently only enabled 32- and 64-bit ARM
systems using atomic CPUs.

Note: An unfortunate side-effect of forking is that every new process
will overwrite the stats and terminal output from the previous
process. This means that the output directory only contains data from
the last checkpoint.

9 years agoconfig: Add soak test for memtest.py
Andreas Hansson [Thu, 19 Mar 2015 08:06:18 +0000 (04:06 -0400)]
config: Add soak test for memtest.py

This patch adds a random option to memtest.py which allows the user to
easily test valid random tree topologies. The patch also adds a
wrapper script to run soak tests using the newly introduced option.

We also adjust the progress interval and progress limit check to make
the output less noisy, and avoid false positives.

Bring on the pain.

9 years agoarm: Add a GICv2m device
Matt Evans [Thu, 19 Mar 2015 08:06:17 +0000 (04:06 -0400)]
arm: Add a GICv2m device

This patch adds a new PIO-accessible GICv2m shim. This shim has a PIO
slave port on one side, and SPI 'wires' on the other. It accepts MSIs
from the system and triggers SPIs on the GIC. It is configurable with
a number of frames, each of which has a number of SPIs and a base SPI
offset.

A Linux driver for GICv2m is available upstream.

9 years agoarm: Remove the 'magic MSI register' in the GIC (PL390)
Matt Evans [Thu, 19 Mar 2015 08:06:16 +0000 (04:06 -0400)]
arm: Remove the 'magic MSI register' in the GIC (PL390)

This patch removes the code that added this magic register. A
follow-up patch provides a GICv2m MSI shim that gives the same
functionality in a standard ARM system architecture way.

9 years agoconfig: Specify OS type and release on command line
Chris Emmons [Thu, 19 Mar 2015 08:06:14 +0000 (04:06 -0400)]
config: Specify OS type and release on command line

This patch enables users to speficy --os-type on the command
line. This option is used to take specific actions for an OS type,
such as changing the kernel command line. This patch is part of the
Android KitKat enablement.

9 years agocpu: Fix TrafficGen message format
Wendy Elsasser [Thu, 19 Mar 2015 08:06:12 +0000 (04:06 -0400)]
cpu: Fix TrafficGen message format

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.

9 years agomem: Use emplace front/back for deferred packets
Andreas Hansson [Thu, 19 Mar 2015 08:06:11 +0000 (04:06 -0400)]
mem: Use emplace front/back for deferred packets

Embrace C++11 for the deferred packets as we actually store the
objects in the data structure, and not just pointers.

9 years agomem: Enable CommMonitor to output traces in atomic mode
Geoffrey Blake [Thu, 19 Mar 2015 08:06:10 +0000 (04:06 -0400)]
mem: Enable CommMonitor to output traces in atomic mode

The CommMonitor by default only allows memory traces to be gathered in
timing mode. This patch allows memory traces to be gathered in atomic
mode if all one needs is a functional trace of memory addresses used
and timing information is of a secondary concern.