Commit graph

11861 commits

Author SHA1 Message Date
Steve Reinhardt
a2c875c746 x86: implement rcpps and rcpss SSE insts
These are packed single-precision approximate reciprocal operations,
vector and scalar versions, respectively.

This code was basically developed by copying the code for
sqrtps and sqrtss.  The mrcp micro-op was simplified relative to
msqrt since there are no double-precision versions of this operation.
2015-10-06 17:26:50 -07:00
Steve Reinhardt
57b9f53afa x86: implement fild, fucomi, and fucomip x87 insts
fild loads an integer value into the x87 top of stack register.
fucomi/fucomip compare two x87 register values (the latter
also doing a stack pop).
These instructions are used by some versions of GNU libstdc++.
2015-10-06 17:26:50 -07:00
Curtis Dunham
83e07c07f9 ext: fix SST connector
The renamings in changesets 8f5993cf (2015-03-23) "mem: rename
Locked/LOCKED to LockedRMW/LOCKED_RMW" and fdd4a895 (2015-07-03)
"mem: Split WriteInvalidateReq into write and invalidate" broke the
SST connector.  This commit repeats those renamings in ext/sst.
2015-10-06 18:08:28 -05:00
Dylan Johnson
71b1c6ce76 sim: Add ability to break at specific kernel function
Adds a GDB callable function that sets a breakpoint at
the beginning of a kernel function.
2015-09-02 13:34:19 -05:00
Andreas Sandberg
17dbb49294 tests: Update SMT tests to correctly configure CPUs
The 01.hello-2T-smt test case for the O3 CPU didn't correctly setup
the number of threads before creating interrupt controllers, which
confused the constructor in BaseCPU. This changeset adds SMT support
to the test configuration infrastructure.

--HG--
rename : tests/configs/o3-timing.py => tests/configs/o3-timing-mt.py
rename : tests/quick/se/01.hello-2T-smt/ref/alpha/linux/o3-timing/config.ini => tests/quick/se/01.hello-2T-smt/ref/alpha/linux/o3-timing-mt/config.ini
rename : tests/quick/se/01.hello-2T-smt/ref/alpha/linux/o3-timing/simerr => tests/quick/se/01.hello-2T-smt/ref/alpha/linux/o3-timing-mt/simerr
rename : tests/quick/se/01.hello-2T-smt/ref/alpha/linux/o3-timing/simout => tests/quick/se/01.hello-2T-smt/ref/alpha/linux/o3-timing-mt/simout
rename : tests/quick/se/01.hello-2T-smt/ref/alpha/linux/o3-timing/stats.txt => tests/quick/se/01.hello-2T-smt/ref/alpha/linux/o3-timing-mt/stats.txt
2015-10-05 13:13:23 -05:00
Steve Reinhardt
d1811cc6cf stats: update EIO stats for snoop filter changes 2015-10-02 01:04:12 -04:00
Andreas Hansson
7dd171ba96 config: Fix 'learning gem5' configs after SMT push
This patch updates the 'learning gem5' example scripts to match the
recent push of the SMT patches.
2015-10-01 04:07:15 -04:00
Curtis Dunham
02881a7bf3 base: remove Trace::enabled flag
The DTRACE() macro tests both Trace::enabled and the specific flag. This
change uses the same administrative interface for enabling/disabling
tracing, but masks the SimpleFlags settings directly. This eliminates a
load for every DTRACE() test, e.g. DPRINTF.
2015-09-30 15:21:55 -05:00
Mitch Hayenga
ccf4f6c3d7 arm: Change TLB Software Caching
In ARM, certain variables are only updated when a necessary change is
detected.  Having 2 SMT threads share a TLB resulted in these not being
updated as required.  This patch adds a thread context identifer to
assist in the invalidation of these variables.
2015-09-30 11:14:19 -05:00
Mitch Hayenga
9e07a7504c cpu,isa,mem: Add per-thread wakeup logic
Changes wakeup functionality so that only specific threads on SMT
capable cpus are woken.
2015-09-30 11:14:19 -05:00
Mitch Hayenga
a5c4eb3de9 isa,cpu: Add support for FS SMT Interrupts
Adds per-thread interrupt controllers and thread/context logic
so that interrupts properly get routed in SMT systems.
2015-09-30 11:14:19 -05:00
Mitch Hayenga
e255fa053f arm: SMT MPIDR Setting
Changes assignment of the MPIDR for multi-threaded systems only.
2015-09-30 11:14:19 -05:00
Mitch Hayenga
fafa83ed32 cpu: Add per-thread monitors
Adds per-thread address monitors to support FullSystem SMT.
2015-09-30 11:14:19 -05:00
Mitch Hayenga
582a0148b4 config,cpu: Add SMT support to Atomic and Timing CPUs
Adds SMT support to the "simple" CPU models so that they can be
used with other SMT-supported CPUs. Example usage: this enables
the TimingSimpleCPU to be used to warmup caches before swapping to
detailed mode with the in-order or out-of-order based CPU models.
2015-09-30 11:14:19 -05:00
Mitch Hayenga
52d521e433 cpu: Change thread assignments for heterogenous SMT
Trying to run an SE system with varying threads per core (SMT cores + Non-SMT
cores) caused failures due to the CPU id assignment logic.  The comment
about thread assignment (worrying about core 0 not having tid 0) seems
not to be valid given that our configuration scripts initialize them in
order.

This removes that constraint so a heterogenously threaded sytem can work.
2015-09-30 11:14:19 -05:00
Joel Hestness
c05d268cfa ruby: Fix CacheMemory allocate leak
If a cache entry permission was previously set to NotPresent, but the entry was
not deleted, a following cache allocation can cause the entry to be leaked by
setting the entry pointer to a newly allocated entry. To eliminate this
possibility, check if the new entry is different from the old one, and if so,
delete the old one.
2015-09-29 09:28:26 -05:00
Joel Hestness
0ecaab4ea8 arch, x86: Delete packet in IntDevice::recvResponse
IntDevice::recvResponse is called from two places in current mainline: (1) the
short circuit path of X86ISA::IntDevice::IntMasterPort::sendMessage for atomic
mode, and (2) the full request->response path to and from the x86 interrupts
device (finally called from MessageMasterPort::recvTimingResp). In the former
case, the packet was deleted correctly, but in the latter case, the packet and
request leak. To fix the leak, move request and packet deletion into IntDevice
inherited class implementations of recvResponse.
2015-09-29 09:28:26 -05:00
Joel Hestness
b80024ee7d ruby: RubyPort delete snoop requests
In RubyPort::ruby_eviction_callback, prior changes fixed a memory leak caused
by instantiating separate packets for each port that the eviction was forwarded
to. That change, however, left the instantiated request to also leak. Allocate
it on the stack to avoid the leak.
2015-09-29 09:28:25 -05:00
Joel Hestness
7b70fa02ae ruby: Fix memory leak in AbstractController
Recent changes to memory access queuing allocate requests for packets sent to
memory controllers, but did not free the requests. Delete them to avoid leaks.
2015-09-29 09:28:25 -05:00
Joel Hestness
501705eaf0 ruby: RubyMemoryControl delete requests
Changes to the RubyMemoryControl removed the dequeue function, which deleted
MemoryNode instances. This results in leaked MemoryNode instances. Correctly
delete these instances.
2015-09-29 09:25:29 -05:00
Joel Hestness
395b31f518 syscall_emul: Bandage readlink /proc/self/exe
The recent changeset to readlink() to handle reading the /proc/self/exe link
introduces a number of problems. This patch fixes two:

1) Because readlink() called on /proc/self/exe now uses LiveProcess::progName()
to find the binary path, it will only get the zeroth parameter of the simulated
system command line. However, if a config script also specifies the process'
executable, the executable parameter is used to create the LiveProcess rather
than the zeroth command line parameter. Thus, the zeroth command line parameter
is not necessarily the correct path to the binary executing in the simulated
system. To fix this, add a LiveProcess data member, 'executable', which is
correctly set during instantiation and returned from progName().

2) If a config script allows a user to pass a relative path as the zeroth
simulated system command line parameter or process executable, readlink() will
incorrecly return a relative path when called on '/proc/self/exe'.
/proc/self/exe is always set to a full path, so running benchmarks can fail if
a relative path is returned. To fix this, clean up the handling of
LiveProcess::progName() within readlink() to get the full binary path.

NOTE: This patch still leaves the potential problem that host full path to the
binary bleeds into the simulated system, potentially causing the appearance of
non-deterministic simulated system execution.
2015-09-29 09:25:20 -05:00
Andreas Hansson
9a0129dcbf mem: Add PacketInfo to be used for packet probe points
This patch fixes a use-after-delete issue in the packet probe points
by adding a PacketInfo struct to retain the key fields before passing
the packet onwards. We want to probe the packet after it is
successfully sent, but by that time the fields may be modified, and
the packet may even be deleted.

Amazingly enough the issue has gone undetected for months, and only
recently popped up in our regressions.
2015-09-25 13:25:34 -04:00
Andreas Hansson
806e1fbf0f stats: Update stats to reflect snoop-filter changes 2015-09-25 07:27:03 -04:00
Andreas Hansson
a9a7002a3b mem: Add check for block status on WriteLineReq fill
More checks to help with understanding of functionality.
2015-09-25 07:26:58 -04:00
Andreas Hansson
012dd52dc2 mem: Fix WriteLineReq fill behaviour
This patch fixes issues in the interactions between deferred snoops
and WriteLineReq. More specifically, the patch addresses an issue
where deferred snoops caused assertion failures when being serviced on
the arrival of an InvalidateResp. The response packet was perceived to
be invalidating, when actually it is not for the cache that sent out
the original invalidation request.
2015-09-25 07:26:58 -04:00
Andreas Hansson
5570aa9e9a mem: Comment clean-up for the snoop filter
Merely fixing up some style issues and adding more comments.
2015-09-25 07:26:57 -04:00
Andreas Hansson
7d4e89d4e0 mem: Avoid adding and then removing empty snoop-filter items
This patch tidies up how we access the snoop filter for snoops, and
avoids adding items only to later remove them.
2015-09-25 07:26:57 -04:00
Andreas Hansson
ca163a80e2 mem: Only track snooping ports in the snoop filter
This patch changes the tracking of ports in the snoop filter to use
local dense port IDs so that we can have 64 snooping ports (rather
than crossbar slave ports). This is achieved by adding a simple
remapping vector that translates the actal port IDs into the local
slave IDs used in the SnoopMask.

Ultimately this patch allows us to scale to much larger systems
without introducing a hierarchy of crossbars.
2015-09-25 07:26:57 -04:00
Ali Jafri
3aa87251d7 mem: Add snoop filters to L2 crossbars, and check size
This patch adds a snoop filter to the L2XBar. For now we refrain from
globally adding a snoop filter to the SystemXBar, since the latter is
also used in systems without caches. In scenarios without caches the
snoop filter will not see any writeback/clean evicts from the CPU
ports, despite the fact that they are snooping. To avoid inadvertent
use of the snoop filter in these cases we leave it out for now.

A size check is added to the snoop filter, merely to ensure it does
not grow beyond the total capacity of the caches above it. The size
has to be set manually, and a value of 8 MByte is choosen as suitably
high default.
2015-09-25 07:26:57 -04:00
Andreas Hansson
0c5a98f9d1 mem: Store snoop filter lookup result to avoid second lookup
This patch introduces a private member storing the iterator from the
lookupRequest call, such that it can be re-used when the request
eventually finishes. The method previously called updateRequest is
renamed finishRequest to make it more clear that the two functions
must be called together.
2015-09-25 07:26:57 -04:00
Ali Jafri
ceec2bb02c mem: Add snoops for CleanEvicts and Writebacks in atomic mode
This patch mirrors the logic in timing mode which sends up snoops to
check for cached copies before sending CleanEvicts and Writebacks down
the memory hierarchy. In case there is a copy in a cache above,
discard CleanEvicts and set the BLOCK_CACHED flag in Writebacks so
that writebacks do not reset the cache residency bit in the snoop
filter below.
2015-09-25 07:26:57 -04:00
Ali Jafri
6ac356f93b mem: Add CleanEvict and Writeback support to snoop filters
This patch adds the functionality to properly track CleanEvicts and
Writebacks in the snoop filter. Previously there were no CleanEvicts, and
Writebacks did not send up snoops to ensure there were no copies in
caches above. Hence a writeback could never erase an entry from the
snoop filter.

When a CleanEvict message reaches a snoop filter, it confirms that the
BLOCK_CACHED flag is not set and resets the bits corresponding to the
CleanEvict address and port it arrived on. If none of the other peer
caches have (or have requested) the block, the snoop filter forwards
the CleanEvict to lower levels of memory. In case of a Writeback
message, the snoop filter checks if the BLOCK_CACHED flag is not set
and only then resets the bits corresponding to the Writeback
address. If any of the other peer caches have (or has requested) the
same block, the snoop filter sets the BLOCK_CACHED flag in the
Writeback before forwarding it to lower levels of memory heirarachy.
2015-09-25 07:26:57 -04:00
Ali Jafri
79d3dbcea8 mem: Add check for snooping ports in the snoop filter
This patch prevents the snoop filter from creating items for requests
originating from non-snooping ports. The allocation decision is thus
based both on the cacheability of the line, and the snooping status of
the source port. Ultimately we should check if the source of the
packet is caching, since also the CPU ports are snooping (but not
allocating). Thus, at the moment we rely on the snoop filter being
used together with caches.

The patch also transitions to use the Packet::getBlockAddr in
determining the line address.
2015-09-25 07:26:57 -04:00
Andreas Hansson
462c288a75 mem: Make the coherent crossbar account for timing snoops
This patch introduces the concept of a snoop latency. Given the
requirement to snoop and forward packets in zero time (due to the
coherency mechanism), the latency is accounted for later.

On a snoop, we establish the latency, and later add it to the header
delay of the packet. To allow multiple caches to contribute to the
snoop latency, we use a separate variable in the packet, and then take
the maximum before adding it to the header delay.
2015-09-25 07:13:54 -04:00
Andreas Hansson
3bd78a141e mem: Do not include snoop-filter latency in crossbar occupancy
This patch ensures that the snoop-filter latency only contributes to
the packet latency, and not to the crossbar throughput/occupancy. In
essence we treat the snoop-filter lookup as pipelined.
2015-09-25 06:45:52 -04:00
Andreas Hansson
40fcb59c54 util: Fix minor issues in DRAM sweep scripts
This patch fixes a few issues in the sweep scripts, bringing them
up-to-date with the latest memory configs and options.
2015-09-25 06:45:50 -04:00
Nilay Vaish
4647e4e961 ruby: simple network: refactor code
Drops an unused variable and marks three variables as const.
2015-09-24 08:41:24 -05:00
Nilay Vaish
b3a3b0b6cf ruby: garnet: refactor code in network links 2015-09-23 11:23:11 -05:00
Nilay Vaish
6bd7aa1f20 ruby: bloom filters: refactor code 2015-09-23 11:23:10 -05:00
Nilay Vaish
c2376918a5 ruby: abstract controller: mark some variables as const 2015-09-23 11:23:10 -05:00
Wendy Elsasser
61c38524ce mem: Add initial HBM configurations
Created the following HBM configurations:
1) HBM gen1 (x128/CH), 2Gb die, 4H stack, 1Gbps, 8 channels
2) HBM gen2 (x64/PC), 8Gb die, 4H stack, 1Gbps, 16 pseudo-channels

The configuration values are based on:
- The HBM gen1 public JEDEC spec
- Publically released data from MemCon presentations
- Timing extrapolated from existing LPDDR configurations

Will adjust once specs become available.
2015-09-22 13:17:53 -05:00
Nilay Vaish
8975053864 ruby: garnet: mark some variables as const 2015-09-18 13:27:48 -05:00
Nilay Vaish
96c999fe88 ruby: print addresses in hex
Changeset 4872dbdea907 replaced Address by Addr, but did not make changes to
print statements.  So the addresses which were being printed in hex earlier
along with their line address, were now being printed in decimals.  This patch
adds a function printAddress(Addr) that can be used to print the address in hex
along with the lines address.  This function has been put to use in some of the
places.  At other places, change has been made to print just the address in
hex.
2015-09-18 13:27:47 -05:00
Nilay Vaish
216529bf18 ruby: slicc: derive DataMember class from Var instead of PairContainer
The DataMember class in Type.py was being derived from PairContainer.  A
separate Var object was also created for the DataMember.  This meant some
duplication of across the members of these two classes (Var and DataMember).
This patch changes DataMember from Var instead.  There is no obvious reason to
derive from PairContainer which can only hold pairs, something that Var class
already supports.  The only thing that DataMember has over Var is init_code,
which is being retained.  This change would later on help in having pointers
in DataMembers.
2015-09-18 13:27:47 -05:00
Tony Gutierrez
b3eb0d1423 ruby: update WireBuffer API to match that of MessageBuffer
this patch updates the WireBuffer API to mirror the changes in revision 11111
2015-09-17 14:00:33 -04:00
Nilay Vaish
928d4b4ba8 stats: updates due to changes to MOESI_hammer 2015-09-16 22:17:54 -05:00
Lena Olson
3225379cc0 ruby: Add missing block deallocations in MOESI_hammer
Some blocks in MOESI hammer were not getting deallocated when they were set to
an idle state (e.g. by invalidate or other_getx/s messages).  While
functionally correct, this caused some bad effects on performance, such as
blocks in I in the L1s getting sent to the L2 upon eviction, in turn evicting
valid blocks.  Also, if a valid block was in LRU, that block could be evicted
rather than a block in I.  This patch adds in the missing deallocations.

Committed by: Nilay Vaish<nilay@cs.wisc.edu>
2015-09-16 20:18:40 -05:00
Joe Gross
950e431d87 ruby: fix message buffer init order
The recent changes to make MessageBuffers SimObjects required them to be
initialized in a particular order, which could break some protocols. Fix this
by calling initNetQueues on the external nodes of each external link in the
constructor of Network.

This patch also refactors the duplicated code for checking network allocation
and setting net queues (which are called by initNetQueues) from the simple and
garnet networks to be in Network.
2015-09-16 13:10:42 -04:00
Nilay Vaish
c5058c0c00 stats: slight changes to MOESI_CMP_token.
Due slight change to latency for the reissue table.
2015-09-16 11:59:57 -05:00
Nilay Vaish
cd9e445813 ruby: message buffer, timer table: significant changes
This patch changes MessageBuffer and TimerTable, two structures used for
buffering messages by components in ruby.  These structures would no longer
maintain pointers to clock objects.  Functions in these structures have been
changed to take as input current time in Tick.  Similarly, these structures
will not operate on Cycle valued latencies for different operations.  The
corresponding functions would need to be provided with these latencies by
components invoking the relevant functions.  These latencies should also be
in Ticks.

I felt the need for these changes while trying to speed up ruby.  The ultimate
aim is to eliminate Consumer class and replace it with an EventManager object in
the MessageBuffer and TimerTable classes.  This object would be used for
scheduling events.  The event itself would contain information on the object and
function to be invoked.

In hindsight, it seems I should have done this while I was moving away from use
of a single global clock in the memory system.  That change led to introduction
of clock objects that replaced the global clock object.  It never crossed my
mind that having clock object pointers is not a good design.  And now I really
don't like the fact that we have separate consumer, receiver and sender
pointers in message buffers.
2015-09-16 11:59:56 -05:00