cpu-o3: Fix non-decoupled frontend #561
Open
+276
−17
Conversation
This file contains hidden or 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
- Fix non-decoupled fetch path in O3:
- Always call lookupAndUpdateNextPC() to update next_pc and run the
predictor in both modes. This replaces the pc.branching() shortcut that left next_pc stale with non-decoupled BPs.
- For non-decoupled frontend, issue a new I-cache request when the
current PC falls outside the fetch buffer window or the buffer is
invalid. Keep the FTQ-based path unchanged for decoupled frontends.
- Instantiate non-decoupled BP classes in configs:
- configs/example/xiangshan.py, configs/example/fs.py:
use bpClass = ObjectList.bp_list.get(...); branchPred = bpClass()
instead of assigning the class object directly.
Before this change:
- TournamentBP/LocalBP failed under Xiangshan: difftest PC mismatch at
start and eventual commit stall (>40k cycles). Fetch logs showed PC
repeatedly outside [startPC, startPC+fetchBufferSize) without issuing
new I-cache requests.
- Also hit TypeError when branchPred was set to the class object.
After this change:
- Non-decoupled BPs (TournamentBP, LocalBP, TAGE, etc.) run correctly
with Xiangshan configs; decoupled predictors remain unaffected.
Change-Id: I808f3ed6301a3104a46a32e4848f5d56791d7511
|
[Generated by GEM5 Performance Robot] Ideal BTB PerformanceOverall Score
|
|
[Generated by GEM5 Performance Robot] Standard PerformanceOverall Score
|
|
[Generated by GEM5 Performance Robot] Ideal BTB PerformanceOverall Score
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
Fix non-decoupled fetch path in O3:
Instantiate non-decoupled BP classes in configs:
TournamentBP/LocalBP failed under Xiangshan: difftest PC mismatch at start and eventual commit stall (>40k cycles). Fetch logs showed PC repeatedly outside [startPC, startPC+fetchBufferSize) without issuing new I-cache requests.
Also hit TypeError when branchPred was set to the class object. After this change:
Non-decoupled BPs (TournamentBP, LocalBP, TAGE, etc.) run correctly with Xiangshan configs; decoupled predictors remain unaffected.
Change-Id: I808f3ed6301a3104a46a32e4848f5d56791d7511