summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAge
* docs: defer experimental host acceleration to post-1.0; record rationaleArthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | | | | The host-VF design is reverted. Independent adversarial, soundness and VM-support reviews falsified its load-bearing mechanisms: RAM-backed BARs cannot be ioremapped or claimed by the stock driver on x86_64 or aarch64; trap-free windows cannot observe edge-triggered writes or read side effects; multi-client GSP-RM is unproven; the MSI and bridge bring-up plans were incorrect; and the isolation and performance claims were overstated for the mechanism. Replace it with one self-contained post-1.0 document recording the findings, corrected candidate approaches and trade-offs, a minimal-risk future architecture, the compile-time + module-load experimental gating policy, a phased plan with abort gates, and the revisit criteria. No baseline design or code path is changed; all earlier doc edits are reverted.
* STATE-CURRENT: 23:12 check-back - guest-wiring P1 lifecycle landed and ↵Arthur Hanson Rasmusson11 days
| | | | exercised (fail-open); async boot FWSEC failure and a release-path freeze recurrence noted
* gsp-phys: mdev-owned backend lifecycle (P1)Arthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | | | Add the gsp_phys_mdev_owned ownership selector (default 0: unchanged load-time module-owned claim/boot). With 1, module load claims nothing; the first mdev create starts a refcounted backend session whose claim+boot runs asynchronously from a work item, so mdev create and init_work never block. The last mdev remove releases after the emulation thread is stopped. Failures are fail-open: the backend stays OFF and the mdev keeps serving the emulation path. nv_vgpu_exit now gates the session STOPPING under the session lock, drains the emulation threads via mdev_unregister_driver(), and only then releases the claim; the module-owned path keeps its previous release position. A start while STOPPING is refused and a concurrent second stopper cannot release twice. No forwarding is added.
* docs: record host VF acceleration as a deferred future goalArthur Hanson Rasmusson11 days
|
* STATE-CURRENT: 22:12 check-back - guest-wiring rung opened (inert session ↵Arthur Hanson Rasmusson11 days
| | | | API skeleton); PLAN A baseline re-verified 64/64
* gsp-phys: backend session API skeleton (P0, inert)Arthur Hanson Rasmusson11 days
|
* STATE-CURRENT + docs: first real copy executed by the physical CE; add the ↵Arthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | physical-backend guest-wiring design - STATE-CURRENT snapshot 2026-10-01 21:05: the host RM-client ladder now allocates VAS+PDB+TSG+channel and executes a real 4 KiB VA-mode CE copy (semaphore released, dst==src) on the physical engine; next is wiring the live session into the guest-facing mdev/vfio path. - docs/03-architecture/11-physical-backend-guest-wiring.md: phased design (P0-P6) for bridging the physical backend into the guest-facing emulation, with the seam table, lifecycle changes and static-config flow. - docs/README.md: reading order updated for the new document; execution status refreshed.
* gsp-phys: encode the GPFIFO GET field correctly so the pushbuffer fetch uses ↵Arthur Hanson Rasmusson11 days
| | | | the mapped VA
* gsp-phys: drain and decode the status queue on a failed CE submissionArthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | | | After a doorbell the GSP-RM status ring accumulated unconsumed records that nothing drained: the host readPtr stayed behind the writePtr and the engine's error text was never decoded. Add a bounded drain (max 64 records) using the existing validated ring walk: each record is logged with its function/status/seqNum, function 0x1006 (OS_ERROR_LOG) is decoded into exceptType/runlistId/chid/prevXid/errString, and function 0x1020 (GSP_POST_NOCAT_RECORD) into recType/errorCode/subsystem/ bugcheck/faultingEngine plus a bounded printable view of the diagnostic buffer. The host read pointer advances per consumed record (the normal consumer side), so the ring keeps free space. Stage 4's poll-timeout path drains before returning -ETIMEDOUT, and a new read-only debugfs file gsp-phys-status-drain triggers one bounded drain per read. No locks are taken and no device register is written.
* gsp-phys: submit the CE copy in virtual mode through the mapped VA windowArthur Hanson Rasmusson11 days
| | | | | | | | | | SET_SEMAPHORE_A/B always take a GPU VA and the channel is USER, so the release cannot target a raw DMA address and physical-mode src/dst is not permitted. Map the src/dst/semem pages at VA 0x102000/0x103000/0x104000 (small-PT slots 258/259/260) in the existing channel PDB, push the semaphore VA, drop the SET_SRC/DST_PHYS_MODE pairs and clear LAUNCH_DMA bits 12/13 (0x18d: Pipelined|Flush|Release-one-word|Pitch). The block is now 13 method pairs / 26 dwords; GPFIFO/PUT/doorbell/poll/memcmp are unchanged.
* gsp-phys: keep the channel USERD in host-writable sysmem for the ladderArthur Hanson Rasmusson11 days
| | | | | | | | | | | | Revert the USERD selection to the client sysmem chunk unconditionally: the ladder advances GP_PUT by writing that chunk directly, and there is no BAR1/PRAMIN aperture here to reach an FBMEM page, so an FBMEM USERD would leave the host write invisible to the hardware USERD the engine polls. Remove the now-dead FBMEM USERD allocator, handle and state; the FBMEM error-notifier allocation path is unchanged. The PDB hierarchy, VA-based gpFifoOffset and ring-entry GET, TSG schedule insurance, logs and asserts from the PDB change are kept. Logs now say userd=sysmem (host-writable) so the path in use is unambiguous.
* gsp-phys: build the channel page directory and submit the CE copy through ↵Arthur Hanson Rasmusson11 days
| | | | | | | | | | | | GPU VAs Fill the external VAS root's GP10X v2 hierarchy (root/PD2/PD1/PD0-small/PT) around a 2 MB VA window so the PDBMA/CE can fetch the GPFIFO and pushbuffer by GPU VA; gpFifoOffset and the GPFIFO ring entry GET now carry those VAs. Allocate the USERD page in FBMEM (class 0x40, matching the reference default) with the sysmem chunk as fallback, and add the advisory TSG-level GPFIFO_SCHEDULE 0xa06c0101 after the channel BIND/schedule. Log the window, table DMAs and entries, and expose them in the debugfs state snapshot.
* STATE-CURRENT: 20:12 check-back - first real GPFIFO work submitted to the ↵Arthur Hanson Rasmusson11 days
| | | | physical engine; channel now allocates; engine did not execute the block
* gsp-phys: allocate the FBMEM notifier with the RM client heap owner; make ↵Arthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | | | | the notif index advisory The class-0x40 NV_MEMORY_ALLOCATION_PARAMS reached the server with owner=0, which the class constructor rejects with NV_ERR_INVALID_OWNER (0x39). Set owner=HEAP_OWNER_RM_CLIENT_GENERIC (0xDEAF000C) and internalflags= NVOS32_ALLOC_INTERNAL_FLAGS_SKIP_SCRUB (0x4), matching the reference notifier allocation. Add static_asserts pinning owner/attr/size/offset/hVASpace/ internalflags to the reference 128-byte ABI so a layout drift breaks the build. SET_WORK_SUBMIT_TOKEN_NOTIF_INDEX (0xc36f010a, index 3) now logs an advisory warning on failure (no error-context object exists, so the server returns NV_ERR_OBJECT_NOT_FOUND 0x57) and the stage continues to GET_WORK_SUBMIT_TOKEN: the server writes the work-submit token before the notifier update, so the token is independent of the notifier object.
* gsp-phys: FBMEM error notifier and the bind/notif-index/token order before ↵Arthur Hanson Rasmusson11 days
| | | | submission
* gsp-phys: use the 4-entry GP10X root format for the VAS page directoryArthur Hanson Rasmusson11 days
| | | | | | | | | The TU104 GMMU root level has 2 index bits (4 PDEs of 2^47 bytes each), so the server's vaLimitNew = (numEntries << 47) - 1 must stay within its vaLimitMax = 2^49-1; numEntries=256 computed 2^55-1 and drew NV_ERR_INVALID_ARGUMENT (0x1f). Send 4 entries (32 bytes of the 4 KB page-aligned VASPDB chunk) and log the value on failure. The aperture/flags word is unchanged (0x0a, SYSMEM_NONCOH + ALL_CHANNELS).
* STATE-CURRENT: 19:12 check-back - host RM client allocates a VAS on the live ↵Arthur Hanson Rasmusson11 days
| | | | GSP-RM; wall moved to the address-space page-directory setup; clean release
* gsp-phys: allocate an externally-owned VAS with a page directory before the ↵Arthur Hanson Rasmusson11 days
| | | | channel ladder (V5)
* STATE-CURRENT: 18:47 check-back - first RM objects on the live GSP-RM; ↵Arthur Hanson Rasmusson11 days
| | | | channel allocation blocked; unsafe variant removed after fail-safe reboot
* gsp-phys: remove the GSP-owned channel variant that wedges the RM taskArthur Hanson Rasmusson11 days
|
* gsp-phys: channel-alloc variant matrix (device parent, skip-scrubber, ↵Arthur Hanson Rasmusson11 days
| | | | gsp-owned, server-allocated memdescs)
* gsp-phys: allocate the channel group under the device handle (valid parent) ↵Arthur Hanson Rasmusson11 days
| | | | and decode rm-call statuses
* gsp-phys: rm-call diagnostics (failed-reply payload, RM log dump) and the ↵Arthur Hanson Rasmusson11 days
| | | | static-config-handle ladder path
* gsp-phys: add the host RM-client ladder (gsp-phys-rm-call) for real object ↵Arthur Hanson Rasmusson11 days
| | | | allocation and a first real copy
* STATE-CURRENT: 16:47 check-back - ladder row updated for the GSP-RM init ↵Arthur Hanson Rasmusson11 days
| | | | milestone (log-ring and identity fixes)
* STATE-CURRENT: physical GSP-RM completes initialization and serves the ↵Arthur Hanson Rasmusson11 days
| | | | static-config RPC
* gsp-phys: register the RM log ring correctly, send raw config-space ↵Arthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | | | | | | identity, and request the full static-config payload The RM log-ring init argument carried a malformed libos id (LOGM), so the server never mapped the ring and its early failure path was invisible. The pre-boot system-info payload carried unpacked 16-bit identity fields; the server expects the raw config-space dwords (device<<16|vendor), so its GPU device-id read back as zero and the static-config device-name lookup returned NOT_SUPPORTED. The static-config RPC must be sized to the full reply payload, and the guest-system-info RPC is sent first, matching the reference order. Live result: GSP-RM completes initialization (init-done rpc_result=0), serves the 1656-byte static config (name, FB size, handles, PDE bases), and shuts down cleanly through the unload RPC.
* guest-diag: treat the benign TX-header status line as non-fatal in the claim ↵Arthur Hanson Rasmusson11 days
| | | | harness
* STATE-CURRENT: 15:47 check-back - source campaign closes host-side inputs; ↵Arthur Hanson Rasmusson11 days
| | | | PMU-provisioned chip identity is the one hard dependency
* STATE-CURRENT: cycles 4-10 - pre-boot RPCs/registry/timer/chipset landed; RM ↵Arthur Hanson Rasmusson11 days
| | | | init wall narrowed to static-config collection
* gsp-phys: populate the chipset/CL block of the pre-boot system infoArthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | | | | | | | | | Fill the GspSystemInfo chipset fields the production client syncs before the GSP boot instead of leaving them zero: - clPdbProperties (bitmask in the client's clSyncWithGsp order): bit 0 (config space accessible, from the OS MCFG/ECAM availability), bit 7 and bit 21 (Intel Skylake-family upstream-port setup: ASPM POR list, root-port nosnoop WAR), bit 13 (upstream port ASPM L1 capable), bit 27 (LTR supported+enabled up to the root port) and bit 32 (IO coherent). PDB_PROP_CL_UNSUPPORTED_CHIPSET is never assigned by the client, so the faithful unmatched-chipset value is 0. - Chipset = 0 (CS_UNKNOWN) for a host bridge with no chipsetInfo[] entry, the reference's no-match fallback. - FHBBusInfo/chipsetIDInfo: 10-byte BUSINFO blobs of the first host bridge of the GPU's domain (device, vendor, subdevice, subvendor, revision), with the subsystem-capability fallback on chipsetIDInfo. acpiMethodData is intentionally sent all-zero (bValid=0): no ACPI/AML calls are made and every consumer tolerates the no-data encoding. All device lookups use refcounted pci_get_device()/pci_dev_put().
* gsp-phys: drop RmDisableDisplay from the registry profile (display-enabled ↵Arthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | | | experiment) The working distribution driver runs with display enabled; the RmDisableDisplay=1 deviation changes board/display state and may be implicated in the devid/board validation failure blocking GSP-RM init. Regenerate the profile without it: the 35 host default parameters plus RMDevidCheckIgnore=1, 36 entries total, display brought up as a stock boot would. Same canonical packing: size 1344, numEntries 36, rpc.length 1376, record 1392 (8-byte aligned), one 4 KB element. The module is unchanged apart from comments and the prepared log line, which now reports 'host default profile + RMDevidCheckIgnore=1, display enabled'.
* STATE-CURRENT: physical GSP-RM init root-caused to devid validation; cycle 8 ↵Arthur Hanson Rasmusson11 days
| | | | under test
* gsp-phys: add RMDevidCheckIgnore to the registry profileArthur Hanson Rasmusson11 days
| | | | | | | | | | | | | Append RMDevidCheckIgnore=1 (dword) to the pre-boot SET_REGISTRY profile, after the 35 host default parameters and RmDisableDisplay=1. The key is the registry escape for the static-config devid/board check that currently fails GSP-RM init; every existing entry is unchanged. Regenerated kernel/gsp_phys_registry_img.h with the same canonical packing (values inline, names-only blob): size 1377, numEntries 37, rpc.length 1409, one 4 KB element. The module code is unchanged apart from comments: the submission and the prepared log line read the size and entry count from the generated blob/macros.
* STATE-CURRENT: 14:07 check-back - 0x59 root-caused (missing pre-boot init ↵Arthur Hanson Rasmusson11 days
| | | | handshake); error journal collapsed 851->4; cycles 4-7 release clean
* gsp-phys: program the GPU timer before the GSP kickArthur Hanson Rasmusson11 days
|
* gsp-phys: send the host default registry profile with RmDisableDisplay for ↵Arthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | | | | | | | | | | | | GSP-RM init The pre-boot SET_REGISTRY (function 73) submitted an empty PACKED_REGISTRY_TABLE. Replace it with the default registry profile the host driver sends before boot: the 35 default dword parameters from its parameter table plus RmDisableDisplay=1. The display key makes the GSP-RM display-presence check treat display as absent, so the display engine is removed during pre-init instead of being brought up on a headless boot. The payload is generated into kernel/gsp_phys_registry_img.h as a PACKED_REGISTRY_TABLE blob (8-byte header, 36 16-byte dword entries, NUL-terminated names after the array; values inline; no value bytes in the name blob, matching the reference packer): size 1342, numEntries 36, rpc.length 1374, one 4 KB element. fn72 is unchanged. The payload length is not 8-byte aligned (record 1390 -> 1392); the TX producer already zero-pads the kvzalloc'd staging record and folds the XOR checksum over the padded length, exactly like the reference producer zero-pads before _checkSum32, so no producer change is needed. The blob is static const data: no heap, no new stack frame, and a submit failure still aborts the boot. The prepared log line now prints the payload size and entry count.
* gsp-phys: queue the pre-boot GSP_SET_SYSTEM_INFO and SET_REGISTRY async RPCsArthur Hanson Rasmusson11 days
| | | | | | | | | | | | | | | | | | | The real-GSP boot path staged the message queue but never queued the two fire-and-forget init RPCs the production client places in it before the booter starts. GSP-RM therefore created its GPU object without any system information and aborted RM init (live GSP_INIT_DONE rpc_result=0x59), with the resource-manager's object-database lookups asserting from the first journal record. Submit function 72 (GSP_SET_SYSTEM_INFO, 928-byte v17_00 payload carrying the BAR physical bases, domain/bus/slot, PCI IDs, the BAR0 config-mirror offset, link/FLR capabilities and upstream-port flags read from the pci_dev) and function 73 (SET_REGISTRY, header-only empty table, which the server explicitly accepts) through the existing submit-only queue producer after all stage-3d preconditions and before the GSP falcon reset/Booter Load kick. The wire layouts are re-declared in the module with per-field static asserts against the reference ABI; payloads are heap-allocated, so no new stack frames are introduced.
* STATE-CURRENT: 12:12 check-back - cycles 2-3 ran the real GSP-RM sequencer; ↵Arthur Hanson Rasmusson11 days
| | | | RM init fails 0x59; clean release without reboot; NOCAT capture next
* gsp-phys: key the NOCAT summary on the truncated stringsArthur Hanson Rasmusson11 days
|
* guest-diag: add --nocat for the next diagnostic bootArthur Hanson Rasmusson11 days
|
* gsp-phys: capture GSP-RM NOCAT journal records from the status queueArthur Hanson Rasmusson11 days
|
* STATE-CURRENT: cycles 2-3 results - clean release without reboot; RM init ↵Arthur Hanson Rasmusson11 days
| | | | failure 0x59 under diagnosis
* guest-diag: allow selecting the staged GSP firmware with --fwArthur Hanson Rasmusson11 days
|
* gsp-phys: use the reference default for a zero sequencer poll timeoutArthur Hanson Rasmusson11 days
| | | | | | | | | | | | A zero REG_POLL timeout dword is the reference's GPU_TIMEOUT_DEFAULT sentinel: the reference substitutes the platform default (microseconds; 4 s in graphics mode, the default platform) before waiting. The executor clamped zero to the 1 us floor, i.e. one jiffy, so a poll that was not already satisfied failed the boot in 1 ms. The live 0x1002 sequencer element contains exactly one REG_POLL with timeout=0 (addr 0x110040); substitute the 4 s graphics-mode default and then keep the existing 1 us .. 10 s clamp.
* STATE-CURRENT: stage 3e parser fix after the live GSP-RM sequencer request; ↵Arthur Hanson Rasmusson11 days
| | | | next cycle prepared
* gsp-phys: accept the real GSP-RM sequencer buffer size in stage 3eArthur Hanson Rasmusson11 days
| | | | | | | | | The live 0x1002 element carries bufferSizeDWord=0x3FE2 (16354) - the size of GSP-RM's own command buffer, not the message. The executor's 0x1000 sanity cap refused it, so the sequencer was never run and GSP_INIT_DONE was never observed. The reference imposes no upper bound (only nonzero and greater than cmdIndex); raise the cap to a sanity ceiling and document that the real execution bound is cmdIndex against the element payload.
* docs: record next-cycle readiness for the physical GSP bootArthur Hanson Rasmusson11 days
| | | | STATE-CURRENT and the integration page reflect the implemented sequencer, clean-shutdown RPC and static-info probe awaiting the next supervised live cycle.
* gsp-phys: harden the RM transport and add the static-info probeArthur Hanson Rasmusson11 days
| | | | | | Applies the adversarial review fixes for the command-queue TX and unload RPC: the unload RPC is gated on GSP-RM init-done, debugfs fops pin the module, REG_POLL timeouts use microseconds, the host status readPtr is published, TX free space is checked, swap flags and barriers are corrected, and the legacy payload fallback is opt-in. Generalizes the transport to multi-element records and a reply-matching status-queue reader, and adds a debugfs probe that round-trips GET_GSP_STATIC_INFO (with a SET_GUEST_SYSTEM_INFO fallback), decodes the server configuration, and a read-only state snapshot file. All gated by gsp_phys_boot; the normal load path is unchanged.
* gsp-phys: command-queue TX and the GSP-RM unload RPC for clean teardownArthur Hanson Rasmusson11 days
| | | | Adds a minimal msgq producer on the command queue (element layout, checksum, sequence, dma_wmb before writePtr, doorbell) and submits the UNLOADING_GUEST_DRIVER RPC so GSP-RM shuts down and sets the scratch14 handoff bit the Booter Unload gate needs. Release attempts the RPC once before quiescing when a boot ran; on-demand testing is available through the debugfs file gsp-phys-unload-rpc while the module owns the GPU. GSP-RM processor suspend is polled on the GSP mailbox.
generated by cgit v1.2.3 (git 2.39.1) at 2026-10-12 23:41:46 +0000