Repository navigation
ci(fpga): fpga-bitstream is an Arty A7-100T smoke build -- never offer it for the Wukong - #5960
Merged
Merged
Conversation
…ver offer it for the Wukong The fpga-bitstream job builds for xc7a100tcsg324-1. It generates a csg324 chipdb, runs `t27c fpga-build --minimal --device xc7a100tcsg324-1` (pinned in 1328787, #2226), and --minimal writes Arty pins into its XDC (bootstrap/src/main.rs:7752-7768). Even so, its step summary said "Board | QMTECH XC7A100T", and the FLASHING.md uploaded with the .bit offered `t27c fpga-flash --board qmtech-a100t --profile minimal`. The QMTech board this repository flashes is the Wukong V1, an XC7A200T, and fpga/HARDWARE_SSOT.md:247-249 says the Arty's csg324 package must not be mixed into Wukong flows. Neither guide command ran as written: qmtech-a100t is not a known board, and fpga-flash has no --profile. Relabelled rather than retargeted. Retargeting to xc7a200tfbg676-1 would also need a Wukong pin set in t27c's --minimal XDC, and nobody has run a 200T chipdb export on a CI runner. - fpga-build.yml: a job comment names the Arty and points at the SSOT as the place that decides the flash target. The summary Board row now reads "Arty A7-100T (xc7a100tcsg324-1), CI smoke; not for the Wukong". FLASHING.md offers the file only for the Arty, as `t27c fpga-flash --board arty-a7 --bitstream <the .bit found>`, and says not to flash it to the Wukong. - fpga/HARDWARE_SSOT.md section 6: a dated 2026-10-04 entry for this correction. No workflow file, job or run: step was added or removed; `tri census pin --gate` passes unchanged. The edited step was run locally against a stub .bit in a scratch directory. No synthesis was run. Refs #5959 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Contributor
|
📓 NotebookLM Notebook linked to this PR
This notebook contains session context, decisions, and artifacts for this work. |
Contributor
This was referenced Oct 4, 2026
Closed
Merged
Merged
This was referenced Oct 4, 2026
Merged
Merged
gHashTag
added a commit
that referenced
this pull request
Oct 4, 2026
…er's moves shell: run: steps 279 -> 284, jobs 81 -> 82, workflow files 62 -> 63. One step is this PR's (loop-tools-gate: The t27b steward decides in t27). The rest came in from master: #4498 added wasm-explorer.yml and #5960 reworked fpga-build.yml without re-blessing, so master's own census gate is red too (tri census pin --gate on 0011bb0: workflow files 62 -> 63). quiet: workflow files 62 -> 63, named-but-not-quiet 152 -> 153, same cause. Refs #6202 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.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
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.
Closes #5959
Refs #5906
The
fpga-bitstreamjob builds an Arty A7-100T (xc7a100tcsg324-1) bitstream, but it labelled it "QMTECH XC7A100T". Its uploadedFLASHING.mdoffered the file for a QMTech board, and the QMTech board this repository flashes is the Wukong V1 (XC7A200T).fpga/HARDWARE_SSOT.md:247-249forbids mixing the Arty's csg324 package into Wukong flows. The SSOT wins (:3-5), so this PR fixes the workflow. The evidence, with file:line, is in #5959.Decision: relabel, do not retarget
fpga-build.yml:567-570), an explicit--device xc7a100tcsg324-1(:585, pinned on purpose in 1328787 / ci(fpga): pin --device to the board the bitstream job actually stages #2226), and Arty pins.--minimalmakes t27c write its own XDC,bootstrap/src/main.rs:7752-7768, whose comment at:7756names the Arty.xc7a200tfbg676-1would take more than a CI change. prjxray-db does carry that part, but a 200T chipdb export has never been run on a runner, and--minimalwould still emit Arty pins. A Wukong CI bitstream needs a Wukong pin set in t27c first.Changed
.github/workflows/fpga-build.yml, jobfpga-bitstream:fpga/HARDWARE_SSOT.md's call. The comment points at the SSOT and does not restate it.Boardrow changes fromQMTECH XC7A100TtoArty A7-100T (xc7a100tcsg324-1), CI smoke; not for the Wukong.FLASHING.mdnow says the file is an Arty A7-100T CI smoke build, not a Wukong bitstream, and that the Wukong's part and flash path are in the SSOT. It offers one command:t27c fpga-flash --board arty-a7 --bitstream <the .bit this step found>. The heredoc is now unquoted so that it can fill in the name, and nothing else in it expands.qmtech-a100tis not a known board (main.rs:7121-7124), andfpga-flashhas no--profileflag (main.rs:1873-1898). TheVerifylink todocs/fpga/QMTECH_A100T_SMOKE.mdpointed at the Wukong smoke test, so it is dropped.fpga/HARDWARE_SSOT.md, section 6 "Known-stale docs corrected by this SSOT": one dated (2026-10-04) entry for this correction.No job, step,
needs:, artifact name or path changes.Checked
bash -eagainst a stubzerodsp_top.bitin a scratch directory, with GNUsed -iswapped for macOSsed -i ''. Exit 0. The summary row andFLASHING.mdcome out as intended, with--bitstream zerodsp_top.bitfilled in.scripts/ci/check_untrusted_shell_interp.pyandscripts/ci/check_pr_branch_filters.pypass. The added lines are ASCII.Seen, not changed (listed in #5959)
fpga-build.yml:383(fpga-synthesis, synth-only at the default device) still writesBoard | QMTECH XC7A100T, andfpga-reportat:969and:975says "QMTECH"..github/workflows/vivado-synth.yml:93,106(dispatch-only) passes--device xc7a100tcsg324-1, againstHARDWARE_SSOT.md:251-252.Not verified
fpga-buildrun.🤖 Generated with Claude Code