diff options
| author | Mitch Hayenga <mitch.hayenga@arm.com> | 2014-09-03 07:42:34 -0400 |
|---|---|---|
| committer | Mitch Hayenga <mitch.hayenga@arm.com> | 2014-09-03 07:42:34 -0400 |
| commit | 1716749c8cec6f9c9f10a0aeaff981be759bb4e5 (patch) | |
| tree | 0e789e02e642227ae170a18782daf05666f7316a /src/cpu/o3/decode.hh | |
| parent | 976f27487b57e968a326752fcf74747427733df6 (diff) | |
| download | gem5-1716749c8cec6f9c9f10a0aeaff981be759bb4e5.tar.xz | |
cpu: Fix o3 front-end pipeline interlock behavior
The o3 pipeline interlock/stall logic is incorrect. o3 unnecessicarily stalled
fetch and decode due to later stages in the pipeline. In general, a stage
should usually only consider if it is stalled by the adjacent, downstream stage.
Forcing stalls due to later stages creates and results in bubbles in the
pipeline. Additionally, o3 stalled the entire frontend (fetch, decode, rename)
on a branch mispredict while the ROB is being serially walked to update the
RAT (robSquashing). Only should have stalled at rename.
Diffstat (limited to 'src/cpu/o3/decode.hh')
| -rw-r--r-- | src/cpu/o3/decode.hh | 4 |
1 files changed, 1 insertions, 3 deletions
diff --git a/src/cpu/o3/decode.hh b/src/cpu/o3/decode.hh index 3424b1d07..006219a50 100644 --- a/src/cpu/o3/decode.hh +++ b/src/cpu/o3/decode.hh @@ -126,7 +126,7 @@ class DefaultDecode void drainSanityCheck() const; /** Has the stage drained? */ - bool isDrained() const { return true; } + bool isDrained() const; /** Takes over from another CPU's thread. */ void takeOverFrom() { resetStage(); } @@ -249,8 +249,6 @@ class DefaultDecode /** Source of possible stalls. */ struct Stalls { bool rename; - bool iew; - bool commit; }; /** Tracks which stages are telling decode to stall. */ |