Resource-optimization RTL + legacy-Python removal [stacks on #87] - #88
Draft
bengineerd wants to merge 22 commits into
Draft
Resource-optimization RTL + legacy-Python removal [stacks on #87]#88bengineerd wants to merge 22 commits into
bengineerd wants to merge 22 commits into
Conversation
Delete Legacy board support (ColumnModule, RowModule, WarmTdmCore v1, RowDacDriver v1, Ad9106, RowSelect) and remove dead numRowSelects/ numChipSelects parameters from the entire stack. Simplify GroupConfig to hold only columnBoards, rowBoards, maxRows, host. Extract variable helper classes to _GroupVariables.py. Fix broken tuning diagnostic functions (sq1Ramp, tesRamp) to use ActivateRowIndex/DeactivateRowIndex directly. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
It was a stub that did nothing — the linked set/get were already removed and the variable just held False permanently. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add Vivado power-analysis TCL scripts for the ColumnFpgaBoard325 Coordinator target (SAIF generation, post-route runs, vectorless activity) and a UCLI/FSDB dump script for the GroupTb simulation to feed switching activity into Vivado power estimation. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Bring in the MAP RAM address width fix for RowDacDriver2. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…keep cleanup FW; remove legacy) Brings the operations refactor (PR #78) and the target-cleanup firmware reorg onto the cleanup branch, while preserving cleanup's unique firmware work and re-applying cleanup's "remove legacy Python" on top. Merge, not rebase. 27 conflicts resolved, batched: Software / structure (wtj-refactor canonical): - operations package, _GroupConfig/_GroupVariables (rowAddrBits model, GroupBroadcastVariable, G3/G4/G6 graduations), _Tuning, warmTdmServer: took wtj/fp-pid versions. - useFloatPid threading preserved via fp-pid's reconciled versions. - submodules -> wtj (surf 0ca723d, ruckus 66fd4a9). Target reorg: - kept cleanup's resource generics (GEN_PID_DEBUG_G, RSSI_WINDOW_ADDR_SIZE_G, ROW_ADDR_BITS_G=6, Power_DefaultOpt) on 160Coord/325Coord. - moved cleanup's 5 vivado power-analysis scripts into the renamed 325Coord10G/vivado/; deleted the reorg-obsoleted old FP target. - releases.yaml RogueScripts -> wtj (correct ../ paths + reorg script list). Firmware (preserve cleanup's RTL work): - AdcDsp.vhd keeps cleanup's GEN_PID_DEBUG_G generate-blocks + wtj's pidStateRamAddr resize. firmware/python drivers took fp-pid's rows= naming (no cleanup-unique content there; the generics live in the RTL). Remove legacy Python (cleanup's f917982, applied on top of wtj): - deleted 10 legacy modules (_Ad9106, _ColumnModule, _RowModule, _RowModuleDacs, _RowDacDriver, _RowSelect, _SaBiasOffset, _TesBias, _WarmTdmCommon, _WarmTdmCore). - _Amplifiers.py: wtj's version (with the R_CABLE->CableR G3 rename) minus the two legacy amp classes (ColumnBoardC00SaAmp, FEAmplifier3). - _ArgParser.py: wtj operations body + rowAddrBits, with Legacy/LegacyCh0Feb board-type choices and dict entries trimmed. - _FrontEnds.py, firmware __init__.py: cleanup's post-removal versions. Verified: all tracked python compiles; ColumnModule/_Ad9106 confirmed gone; the emulate tree builds in both fixed-point and --floatPid modes (--floatPid selects AdcDspFp); cleanup's GEN_PID_DEBUG_G present in both RTL and target ruckus.tcl. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Merges the current fp-pid tip into cleanup so cleanup can open a clean, mergeable PR against fp-pid (the stacked-PR base). cleanup and fp-pid were siblings (both = wtj-refactor + FP PID, divergent), so this reconciles the ~10 overlapping files (_HardwareGroup, _RowDacDriver2, _RowFpgaBoard, _ArgParser, _Group, _GroupRoot, warmTdmGui, submodules, the FP target). Recursive auto-merge resolved consistently; verified the invariants hold: - legacy Python stays removed (fp-pid carries it; NOT resurrected here); - cleanup's resource-optimization RTL preserved (GEN_PID_DEBUG_G in AdcDsp + target ruckus.tcl, EthCore/PgpEthCore/WarmTdmCore2 generics); - legacy board types absent from _ArgParser; - FP target ColumnFpgaBoard325Coord10G keeps USE_FLOAT_PID_G=true; - submodules at surf 0ca723d / ruckus 66fd4a9. Verified: tracked python compiles; ColumnModule gone; emulate tree builds in both fixed-point and --floatPid modes (--floatPid -> AdcDspFp). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
bengineerd
changed the title
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
1 participant
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.
Part of Issue #70 (Floating-point PID firmware). Stacks on PR #87 (
fp-pid → wtj-refactor) — review/merge that first.What this adds on top of #87
This is the
cleanup-branch remainder: everything the FP-PID base (#87) does, plus the resource-utilization firmware work, legacy-Python removal, and build/analysis tooling. Diff is ~42 files vsfp-pid.Resource-optimization RTL (the main content)
Generics to trim FPGA utilization, across the DSP and comms paths:
GEN_PID_DEBUG_G— conditionally generate the per-row PID-debug stream (AdcDsp/AdcDspFp), so builds that don't need debug reclaim the resources.RSSI_WINDOW_ADDR_SIZE_G,MEMORY_TYPE_G/SYNTH_MODE_Gtuning acrossEthCore,PgpEthCore,WarmTdmCore2,PgpCore,FastDacDriver,EventBuilder,BiquadFilter.FastDacDriver.160Coord/325CoordcarryGEN_PID_DEBUG_G,RSSI_WINDOW_ADDR_SIZE_G,ROW_ADDR_BITS_G=6,Power_DefaultOptstrategy.Legacy Python removal
Removes the legacy module layer (
_Ad9106,_ColumnModule,_RowModule,_RowModuleDacs,_RowDacDriver,_RowSelect,_SaBiasOffset,_TesBias,_WarmTdmCommon,_WarmTdmCore) and theLegacy/LegacyCh0Febboard types, trims_Amplifiers/_FrontEnds. The operations-eraR_CABLE→CableRrename (from #78) is preserved.Tooling / docs
Vivado power-analysis + VCD-dump scripts (in
ColumnFpgaBoard325Coord10G/vivado/), PGP-ring-network doc, plan docs, extra logging.Base branch / stacking
Targets
fp-pid(PR #87). Once #87 merges (→wtj-refactor, thenpre-release), retarget this to follow. Reviewing this against #87 isolates the resource/legacy/tooling delta from the FP-PID base and the operations refactor beneath it.Status — draft, needs hardware validation
Verified in emulate: tree builds in both fixed-point and
--floatPidmodes; legacy classes confirmed gone;GEN_PID_DEBUG_Gpresent in both RTL and target ruckus.tcl; both Python packages compile clean. Remaining gates (Issue #70, firmware/bench):🤖 Generated with Claude Code