Artemis bus-agnostic discovery: signature format + idle-loop generalization (FABRIC-3.md §XXVI follow-on)

Step 1: new artemis_sig_t header format (magic 'ARTM', sibling to
homeblocks_sig_t, distinct so a generic scan can tell Artemis's own disk
apart from an identity thumbdrive by content alone) -- artemis_sig.h/.c,
wired into Makefile.starkernel.

Step 2: sk_repl_idle()'s existing per-USB-MSC-slot attach handling (the
pattern WIREBIND already uses for identity thumbdrives) now also checks
for the ARTM signature whenever a device's home-blocks check comes back
BLANK. On a match, once Artemis's own storage-attach round-trip
(HERA-BLK-ATTACH-REQ/BLK-ATTACH-ACK) confirms success,
capsule_zuse_boot_load_root_pubkey() runs -- the same call kernel_main.c's
synchronous QEMU-only virtio-blk path already makes, now reachable
without a hardcoded PCI vendor/device scan. That function is already
idempotent (no-op once zuse_root_pubkey_known is set), so no boot
restructuring was needed despite the initial concern that deferring
Artemis discovery to the idle loop would require one.

Verified: clean build + QEMU boot to [zuse@Hera] ok> on all three
architectures, zero regression to the existing virtio-blk/Zuse-thumbdrive
attach path.

Steps 3 (genesis-stamping onto disk/artemis.img) and 4 (growable
production log-persistence region) not yet started.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EXieurDfDSsDFdnSyusuWo

src/starkernel/repl.c

@@ -39,6 +39,7 @@
 #include "starkernel/blkio_usb.h"
 #include "starkernel/kmalloc.h"
 #include "starkernel/homeblocks_sig.h"
+#include "starkernel/artemis_sig.h"
 #include "starkernel/capsule_birth.h"
 #include "starkernel/capsule_zuse_boot.h"
 #include "starkernel/capsule_wirebind.h"
@@ -158,6 +159,22 @@ blkio_dev_t *sk_repl_get_attached_blk_dev(void) {
     return g_attached_blk_dev;
 }
 
+/* FABRIC-3.md §XXVI follow-on (2026-09-13): Artemis's own disk, once found
+ * generically via USB-MSC content signature (artemis_sig_t, 'ARTM') rather
+ * than the QEMU-only PCI virtio-blk vendor/device scan kernel_main.c still
+ * does synchronously at boot. Mirrors g_homeblocks_dev's own accessor
+ * shape. NULL on QEMU (virtio-blk finds Artemis before this file's idle
+ * loop ever runs) and on any boot where no USB-MSC device presents the
+ * 'ARTM' signature -- real bare-metal hardware is the case this exists
+ * for. Set once sk_word_blk_attach_ack() below confirms the storage-attach
+ * succeeded for a device this loop already recognized as Artemis's own;
+ * cleared on detach alongside g_homeblocks_dev/g_attached_blk_dev. */
+static blkio_dev_t *g_artemis_usb_dev = (void *)0;
+
+blkio_dev_t *sk_repl_get_artemis_usb_dev(void) {
+    return g_artemis_usb_dev;
+}
+
 /* Storage-attach messaging migration (Bob, 2026-09-07): Hera keeps
  * polling/sig-checking, but no longer registers a newly-attached drive
  * into the block subsystem herself -- that's Artemis's own domain now,
@@ -190,6 +207,14 @@ typedef struct {
     int                     pending;
     homeblocks_sig_result_t sig_rc;
     homeblocks_sig_t        sig;
+    /* FABRIC-3.md §XXVI follow-on: set when the attach loop below already
+     * recognized this device as Artemis's own disk (artemis_sig_t 'ARTM'
+     * check, only attempted when sig_rc is HOMEBLOCKS_SIG_BLANK -- a
+     * device can't be both an identity thumbdrive and Artemis's disk).
+     * The ack handler uses this to run capsule_zuse_boot_load_root_pubkey()
+     * once storage-attach is confirmed, same as kernel_main.c's own
+     * virtio-blk path already does synchronously. */
+    int                     is_artemis;
 } sk_blk_attach_pending_t;
 static sk_blk_attach_pending_t *g_blk_attach_pending = (void *)0;
 
@@ -227,8 +252,25 @@ static void sk_word_blk_attach_ack(VM *vm) {
     if (ms) ms->bot_msc_attached = 1;
     g_attached_blk_dev = dev;
 
-    homeblocks_sig_result_t sig_rc = g_blk_attach_pending[found_slot].sig_rc;
-    homeblocks_sig_t        sig    = g_blk_attach_pending[found_slot].sig;
+    homeblocks_sig_result_t sig_rc     = g_blk_attach_pending[found_slot].sig_rc;
+    homeblocks_sig_t        sig        = g_blk_attach_pending[found_slot].sig;
+    int                     is_artemis = g_blk_attach_pending[found_slot].is_artemis;
+
+    if (is_artemis) {
+        /* FABRIC-3.md §XXVI follow-on: bus-agnostic Artemis discovery.
+         * Mirrors kernel_main.c's own virtio-blk-found branch exactly
+         * (blk_subsys_attach_device() there is this device's equivalent,
+         * already done for us above via HERA-BLK-ATTACH-REQ/BLK-ATTACH --
+         * ok_flag being true is that confirmation). Safe to call even if
+         * virtio-blk already found Artemis first (the common QEMU case,
+         * since that path runs synchronously before this idle loop ever
+         * gets a beat): capsule_zuse_boot_load_root_pubkey() is a no-op
+         * once mama_vm->zuse_root_pubkey_known is already set. */
+        g_artemis_usb_dev = dev;
+        log_message(LOG_INFO, "xhci: Artemis's own disk attached via USB-MSC");
+        capsule_zuse_boot_load_root_pubkey((VM *)sk_get_mama_vm());
+    }
+
     capsule_zuse_boot_try_attach(dev, sig_rc, &sig, (VM *)sk_get_mama_vm());
     if (sig_rc == HOMEBLOCKS_SIG_OK) {
         capsule_wirebind_try_attach(dev, &sig, (VM *)sk_get_mama_vm());
@@ -441,6 +483,27 @@ static void sk_repl_idle(VM *active_vm)
                     break;
             }
 
+            /* FABRIC-3.md §XXVI follow-on: a device isn't an identity
+             * thumbdrive (sig_rc above came back BLANK, i.e. no 'LAHB'
+             * magic) -- check whether it's Artemis's own disk instead
+             * (distinct 'ARTM' magic, same devblock-1 convention). Only
+             * attempted on BLANK, not on every device: a drive that
+             * already checked out as home-blocks (or failed a home-blocks
+             * version/CRC check) can't also be Artemis's disk, and
+             * skipping the second read keeps the common identity-drive
+             * case down to one signature check per attach, same as
+             * before this feature existed. */
+            int is_artemis_disk = 0;
+            if (sig_rc == HOMEBLOCKS_SIG_BLANK) {
+                artemis_sig_t asig;
+                artemis_sig_result_t art_rc =
+                    artemis_sig_check(usb_blk_dev, ARTEMIS_SIG_START_FBLOCK, &asig);
+                if (art_rc == ARTEMIS_SIG_OK) {
+                    log_message(LOG_DEBUG, "xhci: USB drive recognized as Artemis's own disk");
+                    is_artemis_disk = 1;
+                }
+            }
+
             /* FABRIC-2.md §F.20/§F.21 / §F.5/§F.23 (WIREBIND): Zuse
              * genesis-mint/attach-authenticate and regular-identity
              * verify-then-birth-then-pair both used to run synchronously,
@@ -451,9 +514,10 @@ static void sk_repl_idle(VM *active_vm)
              * no longer happens on top of storage that might not have
              * registered. Stash what that deferred call needs. */
             if (g_blk_attach_pending) {
-                g_blk_attach_pending[slot_id].pending = 1;
-                g_blk_attach_pending[slot_id].sig_rc  = sig_rc;
-                g_blk_attach_pending[slot_id].sig     = sig;
+                g_blk_attach_pending[slot_id].pending    = 1;
+                g_blk_attach_pending[slot_id].sig_rc     = sig_rc;
+                g_blk_attach_pending[slot_id].sig        = sig;
+                g_blk_attach_pending[slot_id].is_artemis = is_artemis_disk;
             }
 
             /* Storage-attach registration (blk_subsys_attach_device(),