Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

D$: fix two noisy-drop performance bugs #2739

Merged
merged 2 commits into from
Nov 21, 2020
Merged

Conversation

aswaterman
Copy link
Member

Related issue:

Type of change: bug report | feature request | other enhancement

Impact: no functional change | API addition (no impact on existing code) | API modification

Development Phase: proposal | implementation

Release Notes

If the C-channel TSHR is occupied and we wish to evict a clean line,
just silently drop the line rather than waiting for the TSHR to become
available.  Otherwise, we'd stretch the miss latency.
d_address_inc is only valid when d.valid is true (or, more precisely, when
d.bits.opcode is valid).  So, we weren't unblocking the core on the most
optimal cycle in some cases.  This only seems to happen in practice with
noisy drops, since there d.bits.opcode is sometimes ReleaseAck, whereas
otherwise d.bits.opcode is almost always GrantData.

Not a correctness issue.
@aswaterman aswaterman requested a review from hcook November 20, 2020 07:26
@aswaterman aswaterman merged commit ee9cdcd into master Nov 21, 2020
@aswaterman aswaterman deleted the noisy-drop-perf-fixes-2 branch November 21, 2020 01:36
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants