revert(ata): rip the exFAT diagnostics back out of rolling -- they hang the page#532
Merged
Merged
Conversation
…ng the page Maintainer, on ROLLING: "Entering exFAT HDD page: Horrific lag in the 40%'s ... I enter the page and its stuck loading. This was supposed to be the one thing we DID NOT TAKE from experimental." He is right on both counts. #518 graduated EXP11's diagnostics into rolling and they are the regression: the crash marker writes to the MEMORY CARD on every integer step change -- EnsurePopstarterDir (folder probe + a 3-asset existence sweep) plus a full WriteAtomic tmp/rename/backup dance, ~25 times per page entry across the 4-attempt retry loop. That is ~200+ memory-card operations wedged inside the ATA probe, on every exFAT page entry, for every user, to diagnose a bug ONE tester has. The EXP11 adversarial review predicted this precisely ("New memory-card IO on every ata probe that did not exist before EXP11... ~24 writes") and I filed it as severity: note and shipped it anyway. That was the mistake; the tester is not the place to discover a warning I already had in hand. Removed from rolling: the staged 43-50 percents, the crash marker + its next-boot toast, the bdm_query pre-check, the iop128k readout, and the BuildMassRootIdentity attempt param. Verified byte-identical to the pre-#518 build he confirmed working: EnsureMassBackendsReady, BuildMassRootIdentity, BuildATAIdentityDeferred and GetATAMassRootNow all diff clean against d24168a. Two ui.lua deltas kept deliberately: PLDR.L on the "Locating..." string (translation, no I/O) and a pcall around InitATAPopsRoot (cannot hang, only stops a Lua error from stranding the busy overlay). Diagnostics belong in EXPERIMENTAL, where sAGA tests. That is the channel's entire purpose and I put them in the wrong place. lua54 parse-clean; host harness 21/21. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
Warning You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again! |
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.
Maintainer, on rolling: "Entering exFAT HDD page: Horrific lag in the 40%'s... I enter the page and its stuck loading. This was supposed to be the one thing we DID NOT TAKE from experimental."
Right on both counts. #518 graduated EXP11's diagnostics into rolling, and they are the regression. The crash marker writes to the memory card on every integer step change —
EnsurePopstarterDir(folder probe + 3-asset existence sweep) plus a fullWriteAtomictmp/rename/backup dance — roughly 25 times per page entry across the 4-attempt retry loop. That's 200+ memory-card operations wedged inside the ATA probe, on every exFAT page entry, for every user, to diagnose a bug one tester has.The EXP11 adversarial review called this exactly ("New memory-card IO on every ata probe that did not exist before EXP11... ~24 writes"). I filed it
severity: noteand shipped it. The tester is not where I should be discovering a warning I already had.Removed from rolling: staged 43–50 percents, the crash marker and its next-boot toast, the bdm_query pre-check, the iop128k readout, the
attemptparam.Verified byte-identical to the pre-#518 build he confirmed working (
d24168a):EnsureMassBackendsReady,BuildMassRootIdentity,BuildATAIdentityDeferred,GetATAMassRootNowall diff clean. Two ui.lua deltas kept on purpose:PLDR.Lon the "Locating..." string (translation, no I/O) and apcallaroundInitATAPopsRoot(can't hang; only stops a Lua error from stranding the overlay).The diagnostics belong in experimental, where sAGA tests. That's the channel's whole purpose and I put them in the wrong place.
lua54 parse-clean; host harness 21/21.
🤖 Generated with Claude Code