Implementation of C-Pack, as described in "C-Pack: A High-
Performance Microprocessor Cache Compression Algorithm".
C-Pack uses pattern matching schemes to detect and compress
frequently appearing data patterns. As in the original paper,
it divides the input in 32-bit words, and uses 6 patterns to
match with its dictionary.
For the patterns, each letter represents a byte: Z is a null
byte, M is a dictionary match, X is a new value. The patterns
are ZZZZ, XXXX, MMMM, MMXX, ZZZX, MMMX.
Change-Id: I2efc9db2c862620dcc1155300e39be558f9017e0
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/11105
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This reverts commit bf0a722acd.
Reason for revert: This patch introduces a bug:
The problem here is that the insertion of block A may cause the
eviction of block B, which on the lower level may cause the
eviction of block A. Since A is not marked as present yet, A is
"safely" removed from the snoop filter
However, by reverting it, using atomic and a Tags sub-class that
can generate multiple evictions at once becomes broken when using
Atomic mode and shall be fixed in a future patch.
Change-Id: I5b27e54b54ae5b50255588835c1a2ebf3015f002
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/19088
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This patch adds a meta-prefetcher that enables gem5's cache models to
connect to multiple prefetchers. Sub-prefetchers still use the
probes-based interface and training can be controlled
independently. However, when the cache requests a prefetch packet, the
adaptor traverses the priority list of prefetchers and uses the first
prefetcher that is able to generate a prefetch.
Kudos to Mitch Hayenga for the original version of this patch.
Change-Id: I25569a834997e5404c7183ec995d212912c5dcdf
Signed-off-by: Andreas Sandberg <andreas.sandberg@arm.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/18868
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This changeset adds support for partial (or masked) loads/stores, i.e.
loads/stores that can disable accesses to individual bytes within the
target address range. In addition, this changeset extends the code to
crack memory accesses across most CPU models (TimingSimpleCPU still
TBD), so that arbitrarily wide memory accesses are supported. These
changes are required for supporting ISAs with wide vectors.
Additional authors:
- Gabor Dozsa <gabor.dozsa@arm.com>
- Tiago Muck <tiago.muck@arm.com>
Change-Id: Ibad33541c258ad72925c0b1d5abc3e5e8bf92d92
Signed-off-by: Giacomo Gabrielli <giacomo.gabrielli@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/13518
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Previously all atomic writebacks concerned a single block,
therefore, when a block was evicted, no other block would be
pending eviction. With sector tags (and compression),
however, a single replacement can generate many evictions.
This can cause problems, since a writeback that evicts a block
may evict blocks in the lower cache. If one of these conflict
with one of the blocks pending eviction in the higher level, the
snoop must inform it to the lower level. Since atomic mode does
not have a writebuffer, this kind of conflict wouldn't be noticed.
Therefore, instead of evicting multiple blocks at once, we
do it one by one.
Change-Id: I2fc2f9eb0f26248ddf91adbe987d158f5a2e592b
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/18209
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
When a block in compressed form is overwriten, it may change
its size. If the new compressed size is bigger, and the total
size becomes bigger than the block size, one or more blocks
will have to be evicted. This is called data expansion, or
fat writes.
This change assumes that a first level cache cannot have a
compressor, since otherwise data expansion should have been
handled for atomic operations and writes. As such, data
expansions should only be seen on writebacks. As writebacks
are forwarded to the next level when failed, there should
be no data expansions when servicing misses either.
This patch adds the functionality to handle data expansions
by evicting the co-allocated blocks to make room for an
expanded block.
Change-Id: I0bd77bf6446bfae336889940b2f75d6f0c87e533
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/12087
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Implement a co-allocation function in compressed tags, so
that compressed blocks can be co-allocated in a superblock.
Co-allocation is possible when compression ratio (CR) blocks
that share a superblock tag can be compressed to up to (100/CR)%
of their size.
Change-Id: I937cc1fcbb488e70309cb5478c12db65f1b4b23f
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/11411
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Add compression statistics to the compressors. It tracks
the number of blocks that can fit into a certain power
of two size, and the number of decompressions.
For example, if a block is compressed to 100 bits, it will
belong to the 128-bits compression size. Although it could
also fit bigger sizes, they are not taken into account for
the stats (i.e., the 100-bit compression will fit only the
128-bits size, not 256 or higher).
We save stats for compressions that fail (i.e., compressed
size is bigger than original cache line size).
Change-Id: Idab71a40a660e33259908ccd880e42a880b5ee06
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/11103
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Create a stub of a compression framework where we can have
multiple data blocks per tag entry. Only consecutive blocks
can share a tag as of now.
For each tag entry there can be multiple data blocks. We have
the same number of tags a conventional cache would have, but
we instantiate the maximum number of data blocks (according to
the compression ratio) per tag, to virtually implement
compression without increasing the complexity of the simulator.
Change-Id: I549940c7afb2f744ab293ff8bb283967e7551a11
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/10763
Tested-by: kokoro <noreply+kokoro@google.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
This is a workaround for a bug introduced from the change:
59e3585a8 arch-arm: We add PRFM PST instruction for arm
which can cause deadlocks in the memory system.
The design of the classic memory system in gem5 makes the folloing two
assumptions:
* A cache that fetches a block with an intention to modify it, becomes
the point of ordering and therefore commits to respond to any snoop
requests [1].
* A cache that fetches an exclusive copy of the block, does so with
the intention to modify it [2]. Immediately after it receives the
block, it will write to it and mark it as dirty. As the point of
ordering, it responds to any outstanding snoops.
The current implementation of prefetch exclusive request breaks the
second assumption. A cache can fetch an exclusive block without an
immediate intention to modify it. If the block is not modified, it
will not be marked as dirty. However, the cache has committed to
respond to outstanding snoops, and if the block is clean it
won't. This can result in deadlocks where a snoop gets stuck waiting
for responses.
One solution (implemented by this patch) is to unconditionally mark
the block dirty when filling due to a prefetch exclusive request.
This makes the PrefetchExReq behave like a WriteReq. However, as it
may mark as dirty a clean block, it creates the requirement for an
uncessary WritebackDirty in the future. In practice, this shouldn't be
a big problem unless the application is unnecessarily using prefetch
exclusive instructions.
Other solutions, would require deeper changes to the design of the
memory system to handle this properly.
[1]: When a cache commits to respond, it "informs" the xbar/PoC (point
of coherence) and the other caches of its intention to respond. As a
result the request will not be send to the main memory.
[2]: In fact the assumption is that in the needsWritable MSHR there is
at least one WriteReq before any snoops from other caches.
Change-Id: I378d3c0dadf25fc52e430b67102347b44d2f18ea
Signed-off-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/17729
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Tested-by: kokoro <noreply+kokoro@google.com>
MemObject doesn't provide anything beyond its base ClockedObject any
more, so this change removes it from most inheritance hierarchies.
Occasionally MemObject is replaced with SimObject when I was fairly
confident that the extra functionality of ClockedObject wasn't needed.
Change-Id: Ic014ab61e56402e62548e8c831eb16e26523fdce
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/18289
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Anthony Gutierrez <anthony.gutierrez@amd.com>
Maintainer: Gabe Black <gabeblack@google.com>
Commit 7976b561de tried fixing
replacement update when a single location can be associated to
multiple blocks.
Although the comment of the correct action was added, the proper
validation check was forgotten. This change adds that check and
moves doing the eviction to when there is a valid block.
Change-Id: I31d8bb914ccfd1849e9d97464d70a58a62f59533
Signed-off-by: Daniel <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/18210
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Maintainer: Jason Lowe-Power <jason@lowepower.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Replacements should be increased when there is any evicted
block, which does not necessarily have to be the victim.
For example, assume a superblock contains 4 blocks, and both
A and C are stored compressed (belonging to SB_1). Then F,
from SB2 needs to make room by replacing SB1. If F map to
location 2, the number of replacements should be increased,
even though 2 had no valid blocks:
Tag Data Tag Data
|SB_1|--|A|X|C|X| --> |SB_2| |X|F|X|X|
1 2 3 4 1 2 3 4
Change-Id: I7b3735d28a35faa8d8fa613a1555bb258da65859
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/18208
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Maintainer: Jason Lowe-Power <jason@lowepower.com>
Having the caller decide the matching logic is error-prone, and
frequently ends up with the secure bit being forgotten. This
change adds matching functions to the QueueEntry to avoid this
problem.
As a side effect the signature of findPending has been changed.
Change-Id: I6e494a821c1e6e841ab103ec69632c0e1b269a08
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/17530
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
The RRPV values for RRIP and NRU replacment policies.
Long re-rereference interval was used instead of
distant re-rereference interval and vice-versa.
The btp value permit to choose beetwen distant and
long insertion ratio. A btp value of 0 force the
policy to always insert at a distant re-reference
interval and a btp value of 100 force the policy to
always insert at a long (intermediate) re-rereference
interval.
Change-Id: I516098f73942b769dcc31fe0edfe07c3e9c3effd
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/17851
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Jason Lowe-Power <jason@lowepower.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Previously satisfied clean requests would not snoop in-service
MSHRs. This is a problem when a clean request is also invalidating, in
which case we have to post-invalidate or post-downgrade outstanding
requests. This changes fixes this bug.
Change-Id: I31e42aa94dd3637b2818e00fbaae68c810145eaf
Signed-off-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/17728
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Reference:
Stephen Somogyi, Thomas F. Wenisch, Anastasia Ailamaki, and
Babak Falsafi. 2009. Spatio-temporal memory streaming.
In Proceedings of the 36th annual international symposium on
Computer architecture (ISCA '09). ACM, New York, NY, USA, 69-80.
Change-Id: I58cea1a7faa9391f8aa4469eb4973feabd31097a
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/16423
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Error:
build/X86/mem/cache/prefetch/indirect_memory.cc:56:24:
error: result of comparison of constant -1 with expression
of type 'const ByteOrder' is always false
[-Werror,-Wtautological-constant-out-of-range-compare]
fatal_if(byteOrder == -1, "This prefetcher requires a defined ISA\n");
~~~~~~~~~ ^ ~~
build/X86/base/logging.hh:205:14: note: expanded from macro 'fatal_if'
if ((cond)) { \
^~~~
1 error generated.
Fix:
cast of constant (-1) used in comparison
Change-Id: I3deb154c2fe5b92c4ddf499176cb185c4ec7cf64
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/17388
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Added additional information to the PrefetchInfo data structure
- Whether the event is triggered by a cache miss
- Whether the event is a write or a read
- Size of the data accessed
- Data accessed by the request
Change-Id: I070f3ffe837ea960a357388e7f2b8a61d7b2196c
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/16583
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
Some accesses only need to search for a tag in the tag array, with
no need to touch the data array. This is the case for CleanEvicts,
evicts that don't find a corresponding block entry (since a write
cannot be done in parallel with tag lookup), and maintenance
operations.
Change-Id: I7365a915500b5d7ab636d49a9acc627072a7f58e
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/14878
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
When dealing with writebacks, as soon as the packet metadata arrives
there will be a tag lookup, done sequentially because a write can't
be done in parallel. While the tag lookup is being done, the payload
will arrive. When both the payload are present and the tag is correct
block entry is determined the fill happens.
Change-Id: If1a0085d742458b675bfc012b6d908d9d9a25e32
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/14877
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
