target/ast10x0: Reword test comments to stand on their own

Name the test image "test image" and explain stubbed crypto by what it
waits for (the service) rather than a schedule.

Assisted-by: Claude (Anthropic)
diff --git a/target/ast10x0/tests/integration/full_update/system.json5 b/target/ast10x0/tests/integration/full_update/system.json5
index 9078a71..c1e2fc1 100644
--- a/target/ast10x0/tests/integration/full_update/system.json5
+++ b/target/ast10x0/tests/integration/full_update/system.json5
@@ -1,7 +1,7 @@
 // Licensed under the Apache-2.0 license
 // SPDX-License-Identifier: Apache-2.0
 
-// AST10x0 full update QEMU test: one image, the whole demo script.
+// AST10x0 full update QEMU test: boot, update, reboot in one image.
 //
 // Five processes:
 //
diff --git a/target/ast10x0/tests/integration/pldm_update/orchestrator_main.rs b/target/ast10x0/tests/integration/pldm_update/orchestrator_main.rs
index cf2d840..2afcd48 100644
--- a/target/ast10x0/tests/integration/pldm_update/orchestrator_main.rs
+++ b/target/ast10x0/tests/integration/pldm_update/orchestrator_main.rs
@@ -279,7 +279,7 @@
 /// The RoT's own check of the staged candidate.
 ///
 /// Signature checking belongs to the crypto service and is stubbed until
-/// after the demo, so this reads the staging region and compares it with the
+/// the service exists, so this reads the staging region and compares it with the
 /// pattern the update agent is expected to have sent. That makes the RoT's
 /// verdict depend on the bytes rather than on nothing, which is what keeps
 /// the authenticated path from passing vacuously.
diff --git a/target/ast10x0/tests/integration/pldm_update/pldm_fd_main.rs b/target/ast10x0/tests/integration/pldm_update/pldm_fd_main.rs
index dbfdc28..5527592 100644
--- a/target/ast10x0/tests/integration/pldm_update/pldm_fd_main.rs
+++ b/target/ast10x0/tests/integration/pldm_update/pldm_fd_main.rs
@@ -3,8 +3,8 @@
 
 //! The PLDM firmware device, running the real DSP0267 state machine.
 //!
-//! The same `FirmwareDevice` the demo ships, reached over the same
-//! `IpcMctpClient` call path. What differs from the hardware card is one
+//! The same `FirmwareDevice` the production image ships, reached over the
+//! same `IpcMctpClient` call path. What differs from the hardware card is one
 //! layer at the bottom and one at the side: the wire is the loopback bus
 //! rather than I2C, and the image is staged into RAM rather than SPI NOR.
 //!
@@ -579,7 +579,7 @@
         // running tally. A write that silently did not land, or landed
         // somewhere else, fails here instead of passing. Signature
         // checking belongs to the crypto service and is stubbed until
-        // after the demo, so this is a content check.
+        // until the crypto service exists, so this is a content check.
         let mut flash = self.flash.borrow_mut();
         let mut chunk = [0u8; READBACK_CHUNK];
         for base in (0..IMAGE_SIZE).step_by(READBACK_CHUNK) {
diff --git a/target/ast10x0/tests/integration/pldm_update/pldm_ua_main.rs b/target/ast10x0/tests/integration/pldm_update/pldm_ua_main.rs
index 24970d2..9fe539c 100644
--- a/target/ast10x0/tests/integration/pldm_update/pldm_ua_main.rs
+++ b/target/ast10x0/tests/integration/pldm_update/pldm_ua_main.rs
@@ -62,7 +62,7 @@
 /// The firmware device's EID, matching the bus's other side.
 const FD_EID: u8 = 8;
 
-/// Size of the demo image, in bytes. Must match the firmware device's.
+/// Size of the test image, in bytes. Must match the firmware device's.
 const IMAGE_SIZE: u32 = 1024;
 
 /// The UUID this agent expects the firmware device to report. It only updates
@@ -114,7 +114,7 @@
 
 const UA_BUF_SIZE: usize = 1024;
 
-/// The byte the demo image carries at `offset`. The firmware device generates
+/// The byte the test image carries at `offset`. The firmware device generates
 /// the same sequence and rejects anything that does not match.
 fn expected_byte(offset: usize) -> u8 {
     (offset % 251) as u8