martino/MARTe-Integrated-Components
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

added interface and added join to multicast

Browse Source
This commit is contained in:
Martino Ferrari
2026-07-25 12:26:51 +02:00
parent 2d5ca20ae4
commit 3e0a481c13
7 changed files with 1854 additions and 1844 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Docs/UDPStreamer.md
+28 -28
View File
@@ -66,34 +66,34 @@ thread.
### Top-level Parameters
| Parameter | Type | Default | Description |
|-----------|------|---------|-------------|
| `Port` | uint16 | 44500 | UDP server port |
| `MaxPayloadSize` | uint32 | 1400 | Max payload bytes per UDP datagram (min 18) |
| `CPUMask` | uint32 | 0 (any) | Background thread CPU affinity |
| `StackSize` | uint32 | 1 048 576 | Background thread stack size in bytes |
| Parameter | Type | Default | Description |
| ---------------- | ------ | --------- | ------------------------------------------- |
| `Port` | uint16 | 44500 | UDP server port |
| `MaxPayloadSize` | uint32 | 1400 | Max payload bytes per UDP datagram (min 18) |
| `CPUMask` | uint32 | 0 (any) | Background thread CPU affinity |
| `StackSize` | uint32 | 1 048 576 | Background thread stack size in bytes |
### Per-signal Parameters
| Parameter | Type | Default | Applies to |
|-----------|------|---------|------------|
| `Unit` | string | `""` | Any type — informational, forwarded to client in CONFIG |
| `RangeMin` | float64 | 0.0 | float32/float64 with `QuantizedType` |
| `RangeMax` | float64 | 1.0 | float32/float64 with `QuantizedType` |
| `QuantizedType` | string | `none` | float32/float64 only |
| `TimeMode` | string | `PacketTime` | Signals with `NumberOfElements > 1` |
| `TimeSignal` | string | — | Required when `TimeMode` ≠ `PacketTime` |
| `SamplingRate` | float64 | 0.0 | Required when `TimeMode` = `FirstSample` or `LastSample` |
| Parameter | Type | Default | Applies to |
| --------------- | ------- | ------------ | -------------------------------------------------------- |
| `Unit` | string | `""` | Any type — informational, forwarded to client in CONFIG |
| `RangeMin` | float64 | 0.0 | float32/float64 with `QuantizedType` |
| `RangeMax` | float64 | 1.0 | float32/float64 with `QuantizedType` |
| `QuantizedType` | string | `none` | float32/float64 only |
| `TimeMode` | string | `PacketTime` | Signals with `NumberOfElements > 1` |
| `TimeSignal` | string | — | Required when `TimeMode` ≠ `PacketTime` |
| `SamplingRate` | float64 | 0.0 | Required when `TimeMode` = `FirstSample` or `LastSample` |
### Quantization Types
| Value | Wire type | Bit depth | Notes |
|-------|-----------|-----------|-------|
| `none` | same as source | — | Raw copy, no quantization |
| `uint8` | uint8 | 8-bit | Maps `[RangeMin, RangeMax]` → `[0, 255]` |
| `int8` | int8 | 8-bit | Maps `[RangeMin, RangeMax]` → `[-127, 127]` |
| `uint16` | uint16 | 16-bit | Maps `[RangeMin, RangeMax]` → `[0, 65 535]` |
| `int16` | int16 | 16-bit | Maps `[RangeMin, RangeMax]` → `[-32 767, 32 767]` |
| Value | Wire type | Bit depth | Notes |
| -------- | -------------- | --------- | ------------------------------------------------- |
| `none` | same as source | — | Raw copy, no quantization |
| `uint8` | uint8 | 8-bit | Maps `[RangeMin, RangeMax]` → `[0, 255]` |
| `int8` | int8 | 8-bit | Maps `[RangeMin, RangeMax]` → `[-127, 127]` |
| `uint16` | uint16 | 16-bit | Maps `[RangeMin, RangeMax]` → `[0, 65 535]` |
| `int16` | int16 | 16-bit | Maps `[RangeMin, RangeMax]` → `[-32 767, 32 767]` |
Quantization formula (unsigned, e.g. uint16):
@@ -104,12 +104,12 @@ wire_value = (uint16)(normalized × 65535)
### Time Modes
| Value | Meaning | Requirements |
|-------|---------|--------------|
| `PacketTime` | The HRT counter captured at `Synchronise()` time is used as the packet timestamp. No per-signal time metadata. | — |
| `FullArray` | `TimeSignal` carries one timestamp per element (same `NumberOfElements`). | `TimeSignal` must have the same `NumberOfElements`. |
| `FirstSample` | `TimeSignal` is a scalar giving the timestamp of element `[0]`. Elements `[1..N-1]` are inferred at `1/SamplingRate` intervals. | Scalar `TimeSignal`; `SamplingRate > 0`. |
| `LastSample` | Same as `FirstSample` but `TimeSignal` is the timestamp of element `[N-1]`. | Scalar `TimeSignal`; `SamplingRate > 0`. |
| Value | Meaning | Requirements |
| ------------- | ------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------- |
| `PacketTime` | The HRT counter captured at `Synchronise()` time is used as the packet timestamp. No per-signal time metadata. | — |
| `FullArray` | `TimeSignal` carries one timestamp per element (same `NumberOfElements`). | `TimeSignal` must have the same `NumberOfElements`. |
| `FirstSample` | `TimeSignal` is a scalar giving the timestamp of element `[0]`. Elements `[1..N-1]` are inferred at `1/SamplingRate` intervals. | Scalar `TimeSignal`; `SamplingRate > 0`. |
| `LastSample` | Same as `FirstSample` but `TimeSignal` is the timestamp of element `[N-1]`. | Scalar `TimeSignal`; `SamplingRate > 0`. |
---
README.md
+30 -29
View File
@@ -9,15 +9,15 @@ for control applications built with [MARTe2](https://vcis.f4e.europa.eu/marte2-d
This repository integrates two complementary capabilities:
| Capability | Component | Purpose |
|---|---|---|
| **Signal streaming** | `UDPStreamer` DataSource | Continuously stream selected signals to a browser-based oscilloscope over UDP |
| **Signal debugging** | `DebugService` Interface | On-demand signal tracing, value forcing, and conditional breakpoints — zero application code changes required |
| **Sine generation** | `SineArrayGAM` | Generate continuous sine-wave arrays for testing and simulation |
| **Time stamping** | `TimeArrayGAM` | Provide time-reference arrays aligned to an RT cycle |
| **Log forwarding** | `TCPLogger` Interface | Forward `REPORT_ERROR` log events to TCP clients in real time |
| **Integrated client** | `Common/Client/go` | Go packages for UDPS protocol and WebSocket hub |
| **Debug web client** | `Client/debugger` | Browser-based debug UI communicating with `DebugService` |
| Capability | Component | Purpose |
| --------------------- | ------------------------ | ------------------------------------------------------------------------------------------------------------- |
| **Signal streaming** | `UDPStreamer` DataSource | Continuously stream selected signals to a browser-based oscilloscope over UDP |
| **Signal debugging** | `DebugService` Interface | On-demand signal tracing, value forcing, and conditional breakpoints — zero application code changes required |
| **Sine generation** | `SineArrayGAM` | Generate continuous sine-wave arrays for testing and simulation |
| **Time stamping** | `TimeArrayGAM` | Provide time-reference arrays aligned to an RT cycle |
| **Log forwarding** | `TCPLogger` Interface | Forward `REPORT_ERROR` log events to TCP clients in real time |
| **Integrated client** | `Common/Client/go` | Go packages for UDPS protocol and WebSocket hub |
| **Debug web client** | `Client/debugger` | Browser-based debug UI communicating with `DebugService` |
---
@@ -49,9 +49,9 @@ MARTe_Integrated_components/
### UDPStreamer DataSource
Streams MARTe2 signals over UDP using the UDPS binary protocol. Clients register by
Streams MARTe2 signals over UDP using the UDPS binary protocol. Clients register by
sending a `CONNECT` packet; the server then sends `CONFIG` (signal metadata) and continuous
`DATA` packets. Features:
`DATA` packets. Features:
- Optional 16-bit quantization (configurable per signal: `QuantizedType`)
- Packed high-frequency bursts (`NumberOfElements > 1` with `SamplingRate`)
@@ -61,8 +61,8 @@ See `Docs/UDPStreamer.md` and `Docs/Protocol.md`.
### SineArrayGAM
Generates a continuous float32 sine-wave array every RT cycle. Used as a signal
source for testing and demo applications. Configurable: `Frequency`, `Amplitude`,
Generates a continuous float32 sine-wave array every RT cycle. Used as a signal
source for testing and demo applications. Configurable: `Frequency`, `Amplitude`,
`Phase`, `SamplingRate`, `NumberOfElements`.
See `Docs/SineArrayGAM.md`.
@@ -75,12 +75,13 @@ configured `SamplingRate`.
### DebugService Interface
Instruments a running MARTe2 application **without modifying its source code**. On
Instruments a running MARTe2 application **without modifying its source code**. On
`Initialise()` it patches the `ClassRegistryDatabase` to wrap all standard
`MemoryMap*Broker` types. When `RealTimeApplication::ConfigureApplication()` runs
`MemoryMap*Broker` types. When `RealTimeApplication::ConfigureApplication()` runs
afterward the application transparently uses the wrapped brokers.
Capabilities accessible over TCP (port 8080 by default):
- `DISCOVER` — enumerate all signals with type and alias metadata
- `TRACE` — enable/disable high-speed UDP telemetry per signal (with decimation)
- `FORCE` / `UNFORCE` — inject persistent values into signals on the RT path
@@ -98,7 +99,7 @@ See `Docs/DebugService.md`.
### TCPLogger Interface
A `LoggerConsumerI` that forwards every MARTe2 `REPORT_ERROR` call to up to 8 TCP
clients on a configurable port. Works as a sidecar to `DebugService`.
clients on a configurable port. Works as a sidecar to `DebugService`.
### StreamHub Application
@@ -116,7 +117,7 @@ See `Docs/StreamHub-UserGuide.md`, `Docs/StreamHub-API.md` and
### UDPS Protocol
The `Common/UDP/UDPSProtocol.h` header defines the shared binary wire format used by
both `UDPStreamer` and `DebugService`. It is intentionally free of MARTe2-specific
both `UDPStreamer` and `DebugService`. It is intentionally free of MARTe2-specific
dependencies so it can also be used by Go clients (via `Common/Client/go/udpsprotocol`).
See `Docs/Protocol.md`.
@@ -224,18 +225,18 @@ Open `http://localhost:9090`, explore the object tree, trace signals, force valu
## Documentation
| Document | Contents |
|---|---|
| `Docs/Protocol.md` | UDPS binary wire protocol specification |
| `Docs/UDPStreamer.md` | UDPStreamer DataSource configuration reference |
| `Docs/SineArrayGAM.md` | SineArrayGAM configuration reference |
| `Docs/DebugService.md` | DebugService TCP API and architecture |
| `Docs/Tutorial.md` | Step-by-step tutorial covering both components |
| `Docs/WebUI.md` | Web client user guide |
| `Docs/StreamHub-UserGuide.md` | StreamHub oscilloscope user guide (web + ImGui clients) |
| `Docs/StreamHub-API.md` | StreamHub WebSocket protocol (commands, events, binary frames) |
| `Docs/StreamHub-Developer.md` | StreamHub internals, threading, time base, build & E2E tests |
| `ARCHITECTURE.md` | System architecture overview |
| Document | Contents |
| ----------------------------- | -------------------------------------------------------------- |
| `Docs/Protocol.md` | UDPS binary wire protocol specification |
| `Docs/UDPStreamer.md` | UDPStreamer DataSource configuration reference |
| `Docs/SineArrayGAM.md` | SineArrayGAM configuration reference |
| `Docs/DebugService.md` | DebugService TCP API and architecture |
| `Docs/Tutorial.md` | Step-by-step tutorial covering both components |
| `Docs/WebUI.md` | Web client user guide |
| `Docs/StreamHub-UserGuide.md` | StreamHub oscilloscope user guide (web + ImGui clients) |
| `Docs/StreamHub-API.md` | StreamHub WebSocket protocol (commands, events, binary frames) |
| `Docs/StreamHub-Developer.md` | StreamHub internals, threading, time base, build & E2E tests |
| `ARCHITECTURE.md` | System architecture overview |
---
Source/Components/DataSources/UDPStreamer/UDPStreamer.cpp
+1120 -1100
View File
File diff suppressed because it is too large Load Diff
Source/Components/DataSources/UDPStreamer/UDPStreamer.h
+11 -10
View File
@@ -140,16 +140,17 @@ struct UDPStreamerSignalInfo {
* fragmented into multiple datagrams if payload exceeds MaxPayloadSize).
*
* @par Top-level configuration parameters
* | Parameter | Type | Default | Description |
* |-----------------|---------|---------|-------------|
* | Port | uint16 | 44500 | TCP control port (multicast) or UDP server port (unicast). Values ≤ 1024 produce a warning. |
* | MulticastGroup | string | *(absent)* | **Enables multicast mode.** IPv4 multicast address, e.g. `"239.0.0.1"`. Must be in 224.0.0.0/4. Absent or empty = unicast. |
* | DataPort | uint16 | Port+1 | UDP port for multicast DATA datagrams. Ignored in unicast mode. Must be non-zero and differ from Port. |
* | MaxPayloadSize | uint32 | 1400 | Maximum bytes of signal payload per UDP datagram (excluding the 17-byte header). Larger signals are fragmented. |
* | PublishingMode | string | Strict | `Strict`: send one packet every Synchronise() call. `Auto`: rate-limited; flush only when MinRefreshRate interval has elapsed. |
* | MinRefreshRate | float64 | — | Required when PublishingMode = Auto. Flush frequency in Hz (e.g. 120.0). |
* | MaxBatchSize | uint32 | 1 | Optional when PublishingMode = Auto. Number of RT cycles to accumulate before flushing one packet. Scalar signals are expanded to arrays of MaxBatchSize elements; the first scalar with Unit="us" or "ns" is auto-promoted as the per-sample FullArray timestamp reference for all other scalars. When omitted or 1, the most-recent single value is sent at MinRefreshRate. |
* | CPUMask | uint32 | 0xFFFFFFFF | CPU affinity bitmask for the background thread. |
* | Parameter | Type | Default | Description |
* |-----------------|---------|------------------|-------------|
* | Port | uint16 | 44500 | TCP control port (multicast) or UDP server port (unicast). Values ≤ 1024 produce a warning. |
* | MulticastGroup | string | *(absent)* | **Enables multicast mode.** IPv4 multicast address, e.g. `"239.0.0.1"`. Must be in 224.0.0.0/4. Absent or empty = unicast. |
* | Interface | string | *(absent)* | Multicast binded interface **ONLY FOR MULTICAST** |
* | DataPort | uint16 | Port+1 | UDP port for multicast DATA datagrams. Ignored in unicast mode. Must be non-zero and differ from Port. |
* | MaxPayloadSize | uint32 | 1400 | Maximum bytes of signal payload per UDP datagram (excluding the 17-byte header). Larger signals are fragmented. |
* | PublishingMode | string | Strict | `Strict`: send one packet every Synchronise() call. `Auto`: rate-limited; flush only when MinRefreshRate interval has elapsed. |
* | MinRefreshRate | float64 | — | Required when PublishingMode = Auto. Flush frequency in Hz (e.g. 120.0). |
* | MaxBatchSize | uint32 | 1 | Optional when PublishingMode = Auto. Number of RT cycles to accumulate before flushing one packet. Scalar signals are expanded to arrays of MaxBatchSize elements; the first scalar with Unit="us" or "ns" is auto-promoted as the per-sample FullArray timestamp reference for all other scalars. When omitted or 1, the most-recent single value is sent at MinRefreshRate. |
* | CPUMask | uint32 | 0xFFFFFFFF | CPU affinity bitmask for the background thread. |
* | StackSize | uint32 | (MARTe2 default) | Stack size in bytes for the background thread. |
*
* @par Per-signal configuration parameters
Source/Components/Interfaces/UDPStream/UDPSServer.cpp
+664 -677
View File
File diff suppressed because it is too large Load Diff
Source/Components/Interfaces/UDPStream/UDPSServer.h
+1
View File
@@ -235,6 +235,7 @@ private:
uint16 port;
uint32 maxPayloadSize;
StreamString multicastGroup;
StreamString interface;
uint16 dataPort;
bool useMulticast;
uint64 clientTimeoutTicks; ///< 0 = disabled
vgcore.26968
BIN
View File
Binary file not shown.