Fix/adjust diffusion: public_load_pending self-check for PR #5754

Round 35 P1: _raise_if_helper_advisor_busy published a new public
pending marker without first checking public_load_pending(). Two
public workloads (e.g. training + diffusion) could both pass
their idle helper-busy snapshot concurrently, then both run
through destructive owner teardown before either flipped its
own visibility flag (is_training_active, current_checkpoint,
loading_model_identifier, diffusion is_loading).

Add the missing self-check under _HELPER_ADVISOR_START_LOCK so
the second public workload sees the first's pending marker and
gets a 503 retry instead of racing for VRAM. Helper / advisor
already checked public_load_pending() on its side via
_gpu_workload_busy_for_helper; this closes the symmetric public
-> public window.

86 backend tests pass + smoke test confirms second public load
is refused with 503 while first is pending, and the next public
load is permitted once the first clears.

studio/backend/routes/inference.py

@@ -375,6 +375,7 @@ def _raise_if_helper_advisor_busy(workload: str) -> None:
             _HELPER_ADVISOR_START_LOCK,
             _publish_public_load_pending,
             helper_advisor_busy,
+            public_load_pending,
         )
     except Exception:
         return
@@ -402,6 +403,21 @@ def _raise_if_helper_advisor_busy(workload: str) -> None:
                     f"Wait for it to finish before starting {workload}."
                 ),
             )
+        # Round 35 P1: also refuse when another public workload is
+        # already mid-handoff (passed its own helper-busy snapshot
+        # but has not yet flipped is_training_active /
+        # current_checkpoint / loading_model_identifier /
+        # diffusion is_loading). Without this two public loads can
+        # both pass their idle snapshots concurrently and race
+        # destructive owner teardown.
+        if public_load_pending():
+            raise HTTPException(
+                status_code = 503,
+                detail = (
+                    f"Another GPU workload is mid-handoff. Wait for it to "
+                    f"finish before starting {workload}."
+                ),
+            )
         _publish_public_load_pending(workload)