vms-dfa: RMS fills XABFHC/XABALL from the real ODS-2 FAT (facade-risk -> real) - #1049
Merged
Conversation
baron-3dl
force-pushed
the
work/xab-fhc-all
branch
from
September 2, 2026 19:30
926c659 to
fe39b04
Compare
… -> real) Compat register facade-risk row rms$xab_silent_unsupported (rms-api, tier-1): a program chaining an XABFHC (file-header characteristics) or XABALL (allocation) onto its FAB and calling $OPEN/$DISPLAY got SILENCE -- rms_impl_display walked XABKEY/XABDAT/XABPRO and left any other cod untouched (no fill, no error), so XAB$C_FHC (29) / XAB$C_ALL (20) requests returned the caller's zero-init struct while RMS reported RMS$_NORMAL. The structs were not even defined (XAB$C_FHC/ALL were bare #defines). Fix (INV-6, from the real header over the executive ACP): - Define struct XABFHC and struct XABALL (+ cc$rms_xabfhc / cc$rms_xaball) in rms/xab.h, OVMX-native layout like XABKEY/XABDAT/XABPRO. - Retain the file-header characteristics decoded from the ODS-2 FAT (ATR$C_RECATTR) on the open handle at $OPEN: rms_acp_seed_handle now stores rfm/rat/lrl/ebk/ffb/mrz/dxq/gbc/verlimit/bkz alongside eof/hiblk, decoding at->recattr via struct ods2_recattr exactly as rms_file_attr already does. - $DISPLAY (rms_impl_display) fills a chained XABFHC (rfm, atr, lrl, hbk=hiblk, ebk, ffb, bkz, mrz, dxq, gbc, verlimit) and XABALL (alq=hiblk, bkz, deq) from the retained handle fields -- driven by the open handle, not a POSIX stat, so it is correct on the ACP path. Create-time XABALL input controls (aop/aln/loc) are left 0 (honest -- no on-disk source). Ground-source (tests/qemu/test_syssvc_rms_acp.c, real /dev/vms): open the FIX file created earlier, chain XABFHC+XABALL, $OPEN + $DISPLAY, and assert xab$b_rfm == FAB$C_FIX (read from the FAT, not the zero-init silent-skip), xab$l_ebk >= 1, xab$l_hbk >= xab$l_ebk, XABALL xab$l_alq == hiblk. A regression to the silent skip leaves xab$b_rfm == 0 and reddens the first assertion. Register: xab$fhc / xab$all absent -> implemented/real; rms$xab_silent_unsupported facade-risk -> real; sys$display note updated. Surface regenerated. Only tests chaining XABFHC/XABALL are new (zero existed), so no golden baseline changes. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
baron-3dl
force-pushed
the
work/xab-fhc-all
branch
from
September 2, 2026 19:38
fe39b04 to
d7b018f
Compare
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.
What
Compat register facade-risk → real for
rms$xab_silent_unsupported(RMS, tier-1), plusxab$fhc/xab$allabsent → implemented/real. A program chaining an XABFHC (file-header characteristics) or XABALL (allocation) onto its FAB and calling$OPEN/$DISPLAYgot silence:rms_impl_displaywalkedXABKEY/XABDAT/XABPROand left any other cod untouched — soXAB$C_FHC(29) /XAB$C_ALL(20) returned the caller's zero-init struct while RMS reportedRMS$_NORMAL. The structs weren't even defined.Fix (INV-6 — from the real header over the executive ACP)
struct XABFHC/struct XABALL(+cc$rms_xabfhc/cc$rms_xaball) inrms/xab.h, OVMX-native layout like the existing XABs.$OPEN:rms_acp_seed_handlenow storesrfm/rat/lrl/ebk/ffb/mrz/dxq/gbc/verlimit/bkzalongsideeof/hiblk, decodingat->recattr(the 32-byteATR$C_RECATTRFAT) viastruct ods2_recattr— exactly asrms_file_attralready does for DIRECTORY/F$FILE_ATTRIBUTES.$DISPLAYfills the XABs from the retained handle fields — XABFHC (rfm, atr, lrl, hbk=hiblk, ebk, ffb, bkz, mrz, dxq, gbc, verlimit) and XABALL (alq=hiblk, bkz, deq). Driven by the open handle, not a POSIXstat, so it's correct on the ACP path. Create-time XABALL input controls (aop/aln/loc) are left 0 — honest, no on-disk source. The arm only matches FHC/ALL cod, so existing XABDAT/XABPRO handling is untouched.Ground-source (
tests/qemu/test_syssvc_rms_acp.c, real/dev/vms)Open the FIX file created earlier, chain XABFHC+XABALL,
$OPEN+$DISPLAY, assertxab$b_rfm == FAB$C_FIX(read from the FAT, not the zero-init silent skip),xab$l_ebk >= 1,xab$l_hbk >= xab$l_ebk,XABALL xab$l_alq == hiblk. A regression to the silent skip leavesxab$b_rfm == 0and reddens the first assertion — teeth. Runs in the Kernel Executive shard.Register
xab$fhc/xab$all: absent → implemented/real;rms$xab_silent_unsupported: facade-risk → real;sys$displaynote updated. Surface regenerated. No test chained XABFHC/XABALL before (zero hits), so no golden baseline changes. Off-cluster, self-contained.🤖 Generated with Claude Code