| Age | Commit message (Collapse) | Author |
|---|
|
Tags do not need to be aware of caches.
Change-Id: Ib6a082b74dcd9b2f10852651634b59512732fb2a
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/14296
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Access latency was not being calculated properly, as it was
always assuming that for hits reads take as long as writes,
and that parallel accesses would produce the same latency
for read and write misses.
By moving the calculation to the Cache we can use the write/
read information, reduce latency variables duplication and
remove Cache dependency from Tags.
The tag lookup latency is still calculated by the Tags.
Change-Id: I71bc68fb5c3515b372c3bf002d61b6f048a45540
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/13697
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Since the tag classes are subclasses of SimObject, they inherit an
init function which does generic initialization at simulation startup
and which doesn't take any parameters. A new function was added which
does take a parameter, and which is just for doing tag specific
initialization as triggered by the base cache. These two names clashed,
and clang complained that the tag local name was hiding the SimObject
name (which it was).
Change-Id: I399775aceaf8f1a8e2646d434facef22e6d3e7d0
Reviewed-on: https://gem5-review.googlesource.com/c/13875
Reviewed-by: Gabe Black <gabeblack@google.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Gabe Black <gabeblack@google.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Change-Id: I3b902045433ca56b3e62c251158e784b5fa9e4d7
Signed-off-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/13600
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
|
|
ReplaceableEntry is referenced by many classes that do
not necessarily need access to the replacement policies.
Therefore, in order to allow better compilation units,
we factor it out to a new file.
Change-Id: I0823567bf1ca336ffcdf783682ef473e8878d7fd
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/13418
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Encapsulate and virtualize block print, so that relevant
information can be easily printed anywhere.
Change-Id: I91109c29c126755183a0fd2b4446f5335e64076b
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/13415
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Replaceable entries belong to table-like structures, and therefore
they should be indexable by combining a row and a column. These,
using conventional cache nomenclature translate to sets and ways.
Make these entries aware of their sets and ways. The idea is to
make indexing policies usable by other table-like structures. In
order to do so we move sets and ways to ReplaceableEntry, which
will be the common base among table entries.
Change-Id: If0e3dacf9ea2f523af9cface067469ccecf82648
Reviewed-on: https://gem5-review.googlesource.com/c/12764
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Maintainer: Jason Lowe-Power <jason@lowepower.com>
|
|
Having the blocks initialized in the constructor makes it harder
to apply inheritance in the tags classes. This patch decouples
the block initialization functionality from the constructor by
using an init() function. It also sets the parent cache.
Change-Id: I0da7fdaae492b1177c7cc3bda8639f79921fbbeb
Reviewed-on: https://gem5-review.googlesource.com/c/11509
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Jason Lowe-Power <jason@lowepower.com>
|
|
Decouple Tags from Packets, only extracting the necessary
functionality for block insertion. As a side effect, create
a new function to update common insertion statistics.
Change-Id: I5c58f7c17de3255beee531f72a3fd25a30d74c90
Reviewed-on: https://gem5-review.googlesource.com/c/11098
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Jason Lowe-Power <jason@lowepower.com>
|
|
The block was being invalidated before the hash could erase
its entry, therefore it was using invalid values (tag was
being assigned MaxAddr and the secure bit was reset).
This change reorders the calls, so that the appropriate hash
entry is erased.
Change-Id: I161463df0f8f5220179bc68d7be12051e5390d01
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/13210
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
The checking function should not be able to modify either
the head and tail pointers nor should it modify its class.
Change-Id: I2ad495f0c8c6b778d48512143e94b4c9a353f22e
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/13209
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Fully associative caches use an unordered map to enable efficient
lookups of existing blocks. Previously this map was indexed using the
tag of the block. Security extentions allow secure and non secure
versions of a block with the same tag to co-exist in the cache. This
patch amends the block map to allow correct lookups for FALRU caches.
Change-Id: Iccf07464deab56d1d270bae14bb3b154047e3556
Reviewed-on: https://gem5-review.googlesource.com/11309
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
The packet should not be modified within insertBlock.
Change-Id: If7d2b01fe131f9923194efd155c9e85eeab24d5a
Reviewed-on: https://gem5-review.googlesource.com/10811
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Sector caches must know if there was a sector hit in order
to decide whether a victim's sector must be fully evicted
to give place to a new sector or not.
In order to do so it needs the tag and secure information.
Change-Id: Ib554169e25fa131d6bf986561f7970b787c56874
Reviewed-on: https://gem5-review.googlesource.com/10722
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Move tagsInUse to children, as sector caches have different
tag invalidation and insertion, and thus they must handle
updating this variable.
Change-Id: I875c9b7364a909c76daf610d1e226c4e82063870
Reviewed-on: https://gem5-review.googlesource.com/10721
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
For both sector and compressed caches multiple blocks may need
to be evicted in order to make room for a new block.
For example, when replacing a sector, all the blocks in this
sector must be evicted. A replacement, however, does not always
need to evict multiple blocks, as it is in the case of an
insertion of a block whose sector is already present in the cache
(i.e., its corresponding entry in the sector had not been brought
in yet, so it was invalid).
This patch creates the cache framework for that to happen.
Change-Id: I77bedf69637cf899fef4d9432eb6da8529ea398b
Reviewed-on: https://gem5-review.googlesource.com/10142
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
With a sector cache you can't find a block using only its set
and way, as there is the sector offset to take into account. As
all of these blocks inherit from ReplaceableEntry, the return
type of this function has been updated.
This function has also been declared closer to findBlock() due
to their similar functionality.
Change-Id: I4730a2b4ebb5738f7fc118201e208a1b9c3ba8e8
Reviewed-on: https://gem5-review.googlesource.com/10141
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Change-Id: I111b0f662897c43974aadb08da1ed85c7542585c
Reviewed-on: https://gem5-review.googlesource.com/10433
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
An access must perform a block search, which is done by findBlock.
The tagHash is indexed by tags, so use extractTag instead of re-
implementing its functionality.
Change-Id: Ib5abacbc65cddf0f2d7e4440eb5355b56998a585
Reviewed-on: https://gem5-review.googlesource.com/10082
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Maintainer: Jason Lowe-Power <jason@lowepower.com>
|
|
FALRU's findBlock() must use the secure flag to assure proper
functionality.
Change-Id: I54e9fbd3c9093b3e8043c4c6c850b74a8f1f5ec0
Reviewed-on: https://gem5-review.googlesource.com/10081
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
This change fixes a few bugs and refactors the mechanism by which
caches that use the FALRU tags can output statistics for multiple
cache sizes ranging from the minimum cache of interest up to the
actual configured cache size.
Change-Id: Ibea029cf275a8c068c26eceeb06c761fc53aede2
Reviewed-on: https://gem5-review.googlesource.com/9826
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
FALRU was missing MoveToTail functionality within its invalidate
function, and MoveToHead was doing unnecessary passes when the
moved block was the head already.
Besides, added some comments to make the code understandable.
Change-Id: I2430d82b5d53c88b102a62610ea38b46d6e03a55
Reviewed-on: https://gem5-review.googlesource.com/9541
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Removal of unused/barely used 'using namespace' from C++ files.
Change-Id: I66dc548c04506db2e41180b9ea7ab5abd7d5375a
Reviewed-on: https://gem5-review.googlesource.com/9601
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Block insertion is being done in the getCandidates function, while the
insertBlock function does not do anything.
Besides, BaseTags' stats weren't being updated.
Change-Id: Iadab9c1ea61519214f66fa24c4b91c4fc95604c0
Reviewed-on: https://gem5-review.googlesource.com/8882
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
FALRU didn't initialize the blocks' data, causing seg faults.
This patch does not make FALRU functional yet.
Change-Id: I10cbcf5afc3f8bc357eeb8b7cb46789dec47ba8b
Reviewed-on: https://gem5-review.googlesource.com/9302
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Replacement policies (LRU, Random) are currently considered as array
indexing methods, but have completely different functionalities:
- Array indexers determine the possible locations for block allocation.
This information is used to generate replacement candidates when
conflicts happen.
- Replacement policies determine which of the replacement candidates
should be evicted to make room for new allocations.
For this reason, they were split into different classes. Advantages:
- Easier and more straightforward to implement other replacement
policies (RRIP, LFU, ARC, ...)
- Allow easier future implementation of cache organization schemes
As now we can't assure the use of sets, the previous way to create a
true LRU is not viable. Now a timestamp_bits parameter controls how
many bits are dedicated for the timestamp, and a true LRU can be
achieved through an infinite number of bits (although a few bits suffice
in practice).
Change-Id: I23750db121f1474d17831137e6ff618beb2b3eda
Reviewed-on: https://gem5-review.googlesource.com/8501
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
invalidate was defined as a separate function in the base associative
and fully-associative tags classes although both functions should
implement identical functionality. This patch moves the invalidate
function in the base tags class.
Change-Id: I206ee969b00ab9e05873c6d87531474fcd712907
Reviewed-by: Andreas Sandberg <andreas.sandberg@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/8286
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
Change-Id: Id3afec0a62446d6d0f44ccb655032343037637e0
Reviewed-by: Curtis Dunham <curtis.dunham@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/8281
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
The warmupPercentage is the percentage of different tags (based on the
cache size) that need to be touched in order to warm up the cache.
If Warmup failed (i.e., not enough tags were touched), warmup_cycle = 0.
The warmup is not being taken into account to calculate the stats (i.e.,
stats acquisition starts before cache is warmed up). Maybe in the future
this functionality should be added.
Change-Id: I2b93a99c19fddb99a4c60e6d4293fa355744d05e
Reviewed-on: https://gem5-review.googlesource.com/8061
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
|
|
These files aren't a collection of miscellaneous stuff, they're the
definition of the Logger interface, and a few utility macros for
calling into that interface (panic, warn, etc.).
Change-Id: I84267ac3f45896a83c0ef027f8f19c5e9a5667d1
Reviewed-on: https://gem5-review.googlesource.com/6226
Reviewed-by: Brandon Potter <Brandon.Potter@amd.com>
Maintainer: Gabe Black <gabeblack@google.com>
|
|
Change-Id: I79c2662fc81630ab321db8a75be6cd15fa07d372
Reviewed-by: Andreas Hansson <andreas.hansson@arm.com>
Signed-off-by: Andreas Sandberg <andreas.sandberg@arm.com>
|
|
If the cache access mode is parallel, i.e. "sequential_access" parameter
is set to "False", tags and data are accessed in parallel. Therefore,
the hit_latency is the maximum latency between tag_latency and
data_latency. On the other hand, if the cache access mode is
sequential, i.e. "sequential_access" parameter is set to "True",
tags and data are accessed sequentially. Therefore, the hit_latency
is the sum of tag_latency plus data_latency.
Signed-off-by: Jason Lowe-Power <jason@lowepower.com>
|
|
Change-Id: Ia57cc104978861ab342720654e408dbbfcbe4b69
Reviewed-by: Andreas Sandberg <andreas.sandberg@arm.com>
|
|
Change-Id: I5042410be54935650b7d05c84d8d9efbfcc06e70
Reviewed-by: Andreas Sandberg <andreas.sandberg@arm.com>
|
|
A few minor fixes to issues identified by the clang static analyzer.
|
|
Cleaning up dead code. The CLREX stores zero directly to
MISCREG_LOCKFLAG and so the request flag is no longer needed. The
corresponding functionality in the cache tags is also removed.
|
|
This patch makes cache sets aware of the way number. This enables
some nice features such as the ablity to restrict way allocation. The
implemented mechanism allows to set a maximum way number to be
allocated 'k' which must fulfill 0 < k <= N (where N is the number of
ways). In the future more sophisticated mechasims can be implemented.
|
|
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.
|
|
This patch adds some much-needed clarity in the specification of the
cache timing. For now, hit_latency and response_latency are kept as
top-level parameters, but the cache itself has a number of local
variables to better map the individual timing variables to different
behaviours (and sub-components).
The introduced variables are:
- lookupLatency: latency of tag lookup, occuring on any access
- forwardLatency: latency that occurs in case of outbound miss
- fillLatency: latency to fill a cache block
We keep the existing responseLatency
The forwardLatency is used by allocateInternalBuffer() for:
- MSHR allocateWriteBuffer (unchached write forwarded to WriteBuffer);
- MSHR allocateMissBuffer (cacheable miss in MSHR queue);
- MSHR allocateUncachedReadBuffer (unchached read allocated in MSHR
queue)
It is our assumption that the time for the above three buffers is the
same. Similarly, for snoop responses passing through the cache we use
forwardLatency.
|
|
Static analysis unearther a bunch of uninitialised variables and
members, and this patch addresses the problem. In all cases these
omissions seem benign in the end, but at least fixing them means less
false positives next time round.
|
|
The patch
(1) removes the redundant writeback argument from findVictim()
(2) fixes the description of access() function
Committed by: Nilay Vaish <nilay@cs.wisc.edu>
|
|
This patch adds the basic building blocks required to support e.g. ARM
TrustZone by discerning secure and non-secure memory accesses.
|
|
This patch reorganizes the cache tags to allow more flexibility to
implement new replacement policies. The base tags class is now a
clocked object so that derived classes can use a clock if they need
one. Also having deriving from SimObject allows specialized Tag
classes to be swapped in/out in .py files.
The cache set is now templatized to allow it to contain customized
cache blocks with additional informaiton. This involved moving code to
the .hh file and removing cacheset.cc.
The statistics belonging to the cache tags are now including ".tags"
in their name. Hence, the stats need an update to reflect the change
in naming.
|
|
This patch address the most important name shadowing warnings (as
produced when using gcc/clang with -Wshadow). There are many
locations where constructor parameters and function parameters shadow
local variables, but these are left unchanged.
|
|
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 seperates the functionality to clear the state in a block into
blk.hh and the functionality to udpate the tag information into the
tags. This gets rid of the case where calling invalidateBlk on an
already-invalid block does something different than calling it on a
valid block, which was confusing.
|
|
This patch is the result of static analysis identifying a number of
memory leaks. The leaks are all benign as they are a result of not
deallocating memory in the desctructor. The fix still has value as it
removes false positives in the static analysis.
|
|
This step makes it easy to replace the accessor functions
(which still access a global variable) with ones that access
per-thread curTick values.
|
|
when it in received
|
|
On the config end, if a shared L2 is created for the system, it is
parameterized to have n sharers as defined by option.num_cpus. In addition to
making the cache sharing aware so that discriminating tag policies can make use
of context_ids to make decisions, I added an occupancy AverageStat and an occ %
stat to each cache so that you could know which contexts are occupying how much
cache on average, both in terms of blocks and percentage. Note that since
devices have context_id -1, having an array of occ stats that correspond to
each context_id will break here, so in FS mode I add an extra bucket for device
blocks. This bucket is explicitly not added in SE mode in order to not only
avoid ugliness in the stats.txt file, but to avoid broken stats (some formulas
break when a bucket is 0).
|
