temp #32
Open
temp #32
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
mem: small fixes to compile gem5 after rebase Change-Id: I6fc7c0b161b945f338dc60856864235d838951cb mem: Add Octopi cache test Change-Id: Ie7464e72e9bbcded0a58d23e56a4460759a08834 mem: small fixes Change-Id: I1513e2a39e8823a57db377b8b3a03954bdb18ab3 mem: dcache with ports -- working version Change-Id: I7aaa0468977742ad27196dfd9f8b40aa66ea048e stdlib: fix cache line size error Change-Id: I98fb1403a8da9dc204c9f8ea5fb2d9d2f11c981c stdlib: temp change Change-Id: I04921d596811dfad5df32c1fbbd9532d5da6a59f configs: Use DiskImageWorkload and fix path Change-Id: I017f3d7b7fd18fbe39311599fbfffbf6d9260d5f Signed-off-by: Jason Lowe-Power <[email protected]> stdlib: Sync changes Change-Id: Ie18db282c97fcc2cc2c1131de8b9cab09c0ece07 configs: HACK\! enable checkpointing in octotpi Signed-off-by: Jason Lowe-Power <[email protected]> mem: cache warmup exit handling misc: Add checkpoint script Change-Id: I8e00452aed65fd49e80e6a7816cb6f6dad94455e mem: some minor updates Change-Id: I56108f2d8b02b777c0c35146d0e7aeee5f6cbf17 mem: just the GAPBS and NPB scripts are updated mem: latest version, modified for write age be ignored for cases not needed mem: updated stats mem: just updated some stats and params mem: more updates on stats stdlib: checkpoint script Change-Id: I8792cd616fad975973e5de96addc0d912bfb1c4e mem: syncing changes Change-Id: Ic48c244896f5560886a1fe0d0b9e84049a923e41 mem: update scripts Change-Id: I7142400634f362fab80c2553a82cb3d79a9ff1b8 misc: documentation Change-Id: I24f00455adc2ef6697cc0c70ed76a292f76df774 stdlib: add missing local change Change-Id: I3a4981e8412c466c130165b4f68343039ac0aaab misc: doc update Change-Id: Ia8b68057f5ff1a25b5b935f43f0158b16cad8fb4 mem-ruby: Fix warmup with backing store When warming up with Ruby backing store enabled, you cannot access both the backing store and the mem_interface with the packet. This change now accesses one or the other. The downside to this approach is that with this change you cannot restore a checkpoint or switch CPUs and disable backing store. Change-Id: I2d9f542a759d47fea571abd0d3141ce5996a81c3 Signed-off-by: Jason Lowe-Power <[email protected]> arch-riscv: Copy miscregs on takeOverFrom Previously, the miscregs were not copied when switching CPUs. This caused failures after the switch because of an incorrect SATP. This change copies all of the miscregs from the old thread context to the new thread context. Change-Id: I4d1e9534b909e55a078fdfa579f4956d1c3a37db Signed-off-by: Jason Lowe-Power <[email protected]> ckpt restore script for gapbs o3: Add looppoint stuff in this branch Update restore script small changes syncing changes small changes mem: removed old files mem: created TDRAM_32 mem: modified scripts traff gen mem: added plot data mem: added some test cases mem: latest plots mem:latest version mem: updated scripts and merged with latest version of TDRAM mem: updated old scripts mem: updated old scripts python: Allow SimObjectVector to match proxy There are proxies (like System.memories) which want to find *all* of the SimObjects that are of that type. However, the code currently cannot deal with SimObjectVectors. This change now checks for SimObjectVectors and updates the proxy list with each SimObject in the vector. It's been tested with System.memories and a VectorParam.SimpleMemory. This is needed to allow multiple discontinuous phys_mem parameters for Ruby's backing store. Change-Id: I937fbfc41a8a611b1b3a491caec41e8a9dcb0947 Signed-off-by: Jason Lowe-Power <[email protected]> mem-ruby: Allow multiple backing store ranges This change enables Ruby to have multiple backing store ranges. The phys_mem parameter is now a list of simple memory objects, one for each discontiguous range. This is needed to have a second memory range above 4GiB for x86 simulation. Note that this is currently not used in Ruby.py, but it should be straightforward to set up for anyone's specific use case. Change-Id: I591a2b82a06bec69f583ba21f7267b6f535f48b1 Signed-off-by: Jason Lowe-Power <[email protected]> configs: Updates to system to work with more memory Signed-off-by: Jason Lowe-Power <[email protected]> mem: updated scripts mem: updated scripts mem: fixed the exitSimLoop error cpu-o3: Update looppoint analysis Improve the output for the recent PCs and update the types. Change-Id: Id86f7f0c2329cc9104912cf2655a0202541a450e Signed-off-by: Jason Lowe-Power <[email protected]> misc: Add region info for workloads For use with the loop point tracker Change-Id: Id51cc7907cbc064dfd75cb26b887d1a0f5752428 Signed-off-by: Jason Lowe-Power <[email protected]> just some updates on the scripts updated scripts, used for chkpts and restores configs: Add restore both script and update info Change-Id: Ib17f94cd59a8b152e5b8bf3d489e20236800a6e8 Signed-off-by: Jason Lowe-Power <[email protected]> cpu-o3: Update looppoint probe to keep order Make sure that the recent PCs are presented in order instead of randomly. Change-Id: I9b2efae991507ff49dcc9525c5aea87c78e12e2e Signed-off-by: Jason Lowe-Power <[email protected]> cpu-o3: Fix off by one in looppoint analysis We should insert PCs into the PC list with "1" access instead of 0 Change-Id: I9a773675a99e22d9b7603f00f28e4cc2dac766c2 Signed-off-by: Jason Lowe-Power <[email protected]> Revert "cpu-o3: Fix off by one in looppoint analysis" This should be reverted since the loop point counting also has this off-by-one "error" This reverts commit 27ba5eb. cpu: Update looppoint pc-count hash function TBH, I'm not sure this is needed or not Change-Id: I69f14c201c375fac76823b7bf8948aba0d426f9d Signed-off-by: Jason Lowe-Power <[email protected]> configs: add workload info Change-Id: Ie21cc50223797e6c6b056a0b2aeacea209c80db1 Signed-off-by: Jason Lowe-Power <[email protected]> configs: update info with 24 hour point Change-Id: Icec1036ce8cc839a6a459f0625d3e0cf33f1e580 Signed-off-by: Jason Lowe-Power <[email protected]> fixed some status for Oracle configs: Update info with D/25 classes Change-Id: I1764a50e3d6da7c47ebd2c6d1eaa864fde0e2bef Signed-off-by: Jason Lowe-Power <[email protected]> configs: Update restore both script Change-Id: Ieffb8924b8e2d9f6ddb19895733ab5a829fd7e20 Signed-off-by: Jason Lowe-Power <[email protected]> revised version for combined dirmap and set-asso, only tested for dirmap first version tested with traff gen and tc in GAP mem: added serialize and unserialize for chkpt and rstr latest version of set-associative mem: added data plot for set-assoc mem: added data plot for set-assoc fixed unserialization new set-assoc data mem: latest version fixing tag issue in unserialization data plots data plots data plots updated plots misc: Google traces files misc: gtraces config data plots for comparing running for longer hours data plots for comparing running for longer hours small changes in scripts for chkpt and pc analysis mem: updated google scripts mem: updated google scripts updated scripts with new pc updated scripts with new pc updated scripts backup parent 41af1e744aaf2a2dd291d32ad1ec1b672b8822b9 author mbabaie <[email protected]> 1689618387 -0700 committer mbabaie <[email protected]> 1693268224 -0700 plots mem: updated plots first version, incomplete latest version all 3 cases Rd Hit/MissCl/MissDirty are implemented and tested on traffgen mem: latest version few minor changes first version tested bfs tc bt sp latest version tested on calss C and 22 data plots, 1st data plots, 1st data plots, 1st fixed the plot for breakdown of probing fixed the plot for breakdown of probing fixed the plot for breakdown of probing latest version with new stats runs bash file latest files in the directory just some comments removed tiny changes in the script tested archs
Change-Id: I2f94d6f2aababb678c8f8d11b307bd91bc217d6b Signed-off-by: Jason Lowe-Power <[email protected]>
…ded LRU to serialize, added atomicNonCache
…ecific to each app in the scripts
…ck size, only for atomic
…kpoint iterations
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.