Bump OpenROAD for rsz default max fanout backstop - #4462
Open
minjukim55 wants to merge 7 commits into
Open
Conversation
Signed-off-by: Minju Kim <mkim@precisioninno.com>
…max-fanout-50 Signed-off-by: Minju Kim <mkim@precisioninno.com> # Conflicts: # tools/OpenROAD
Points tools/OpenROAD at 257db71c3e, the rsz default max fanout backstop rebased onto the OpenROAD commit that ORFS master already pins, so the submodule carries master's engine plus this one change. Signed-off-by: Minju Kim <mkim@precisioninno.com>
…max-fanout-50 Signed-off-by: Minju Kim <mkim@precisioninno.com> # Conflicts: # tools/OpenROAD
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
…max-fanout-50 Signed-off-by: Minju Kim <mkim@precisioninno.com> # Conflicts: # tools/OpenROAD
Points tools/OpenROAD at f86f32d8b5, the rsz default max fanout backstop cherry-picked onto 6cb3f2b704, the OpenROAD commit ORFS master already pins, so the submodule carries master's engine plus this one change. Signed-off-by: Minju Kim <mkim@precisioninno.com>
Member
🔍 QoR checkMetrics reflect the PR merge build — i.e. what will land on the target branch. Advisory — results are log-only and do not affect build status. The authoritative QoR gate remains the local rules-file check. Commit 62 design(s) checked — 17 with regression(s), 0 without a comparable baseline.
❌ asap7/aes-block base — 3 failing metric(s)
❌ asap7/mock-alu base — 2 failing metric(s)
❌ asap7/riscv32i base — 1 failing metric(s)
❌ asap7/riscv32i-mock-sram base — 1 failing metric(s)
❌ gf180/aes base — 3 failing metric(s)
❌ gf180/aes-hybrid base — 2 failing metric(s)
❌ gf180/jpeg base — 1 failing metric(s)
❌ gt2n/aes base — 2 failing metric(s)
❌ nangate45/aes base — 1 failing metric(s)
❌ nangate45/ariane133 base — 3 failing metric(s)
❌ nangate45/bp_fe_top base — 1 failing metric(s)
❌ nangate45/swerv base — 1 failing metric(s)
❌ sky130hd/aes base — 3 failing metric(s)
❌ sky130hd/chameleon base — 5 failing metric(s)
❌ sky130hd/riscv32i base — 1 failing metric(s)
❌ sky130hs/aes base — 1 failing metric(s)
❌ sky130hs/riscv32i base — 4 failing metric(s)
|
Six designs fail their metadata checks with the backstop enabled. It buffers high fanout nets that used to stay on a single driver, so setup TNS, routed wirelength and antenna counts move against the current limits. Regenerated with genRuleFile.py --failing from the CI metadata.json of PR-4462-head build 2; no passing rule is tightened. | Design | Metric | Old | New | Type | | ------ | ------ | --- | --- | ---- | | asap7/riscv32i | globalroute__timing__setup__tns | -1890.0 | -3390.0 | Failing | | gf180/aes-hybrid | cts__timing__setup__tns | -132.0 | -137.0 | Failing | | gf180/aes-hybrid | globalroute__timing__setup__tns | -144.0 | -146.0 | Failing | | gf180/jpeg | cts__timing__setup__tns | -1.5 | -23.3 | Failing | | nangate45/bp_fe_top | detailedroute__route__wirelength | 1673870 | 1951663 | Failing | | nangate45/bp_fe_top | finish__timing__setup__tns | -5.17 | -5.53 | Failing | | nangate45/ibex | finish__timing__setup__tns | -0.44 | -0.957 | Failing | | sky130hd/chameleon | cts__timing__setup__tns | -136.0 | -198.0 | Failing | | sky130hd/chameleon | globalroute__timing__setup__tns | -48.8 | -186.0 | Failing | | sky130hd/chameleon | detailedroute__antenna__violating__nets | 0 | 1 | Failing | | sky130hd/chameleon | detailedroute__antenna_diodes_count | 106 | 118 | Failing | | sky130hd/chameleon | finish__timing__setup__tns | -47.4 | -156.0 | Failing | chameleon's two synth netlist hash rules also drift in this run, but they are warning level and pre-date this branch: the submodule pins master's engine plus the one rsz change, so synthesis is untouched. Left them alone. Signed-off-by: Minju Kim <mkim@precisioninno.com>
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
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.
Summary
Pins OpenROAD at the rsz default max fanout backstop, The-OpenROAD-Project-private/OpenROAD#3683. Submodule bump only, no flow script change.
repair_designnow builds a fanout tree on designs whose libraries and SDC carry no max fanout limit, using a default of 50 load pins. This PR is here for the QoR metrics.Type of change
Impact
Expect more buffers and a different area/slew tradeoff on designs with unconstrained fanout. Measured on the OpenROAD tests:
upf_aesgpl timing-driven repair_design area recovery -24.70% -> -22.92%; a 250 load reset net closes with a fanout tree instead of a wire split, area +2.8% against +2.2%, one -0.04 max slew violation left.Verification
OpenROAD ctest 8310/8311 on the pinned commit, see #3683. ORFS metrics pending on this PR.
Related Issues
The-OpenROAD-Project-private/OpenROAD#3683