DMR-данные (SMS) на TYT MD-UV380 под OpenGD77

Восстановлено из транскрипта сессии после потери рабочего дерева на сервере сборки.

Что внутри:
- четыре новых модуля прошивки (приём, сборка исходящего, меню Messages);
- правки существующих файлов в виде скриптов, с порядком применения;
- инструменты стенда: снятие экрана рации по USB, дамп кадров, эталонная сборка пакета;
- NOTES.md: что проверено в эфире, что не доделано, где грабли.

Работало в эфире в августе 2026: данные ходят в обе стороны в режиме хотспота,
рация принимает SMS и показывает на экране. Отправка собрана и сверена с эталоном
сервера побайтово, но в эфире не проверялась.

Claude-Session: https://claude.ai/code/session_01GJtvG5WwqYNe1b7s9LLtgy

patches/patch_txfix.py

@@ -0,0 +1,74 @@
+#!/usr/bin/env python3
+"""Fix: a stale "this frame is packet data" marker could suppress the Voice LC Header
+of an ordinary voice transmission.
+
+hrc.txDataType is only ever set on the hotspot path, and was only cleared after a data
+burst was actually sent. If a data transmission was cut short before that, the marker
+stayed set - and the next PTT from the radio itself started without a Voice LC Header,
+so receivers never recognised the call.
+
+Two guards:
+  - the start-of-transmission checks now only apply while acting as a hotspot modem;
+  - the marker is cleared whenever a normal (non-hotspot) transmission begins.
+"""
+W = "/home/viktor/ogd77rel/uv_src/OpenGD77_MDUV380_DM1701_20260130/MDUV380_firmware/application"
+p = W + "/source/hardware/HR-C6000.c"
+s = open(p).read()
+
+if "isHotspotDataTx" in s:
+    print("already patched")
+    raise SystemExit(0)
+
+# 1. helper right after the struct instance, so all three call sites can use it
+anchor = "\t\t.txDataType = HOTSPOT_TX_TYPE_VOICE,\n"
+assert anchor in s, "hrc initialiser not found"
+
+# find the end of the initialiser block to place the helper after it
+marker = "static void hrc6000SendPcOrTgLCHeader(void);"
+assert marker in s, "forward declaration anchor not found"
+helper = """// True only when the frame about to be sent is a packet data burst queued by the host.
+// Outside hotspot mode there is no such thing, so ordinary voice transmissions are never
+// affected by a leftover marker.
+static inline bool isHotspotDataTx(void)
+{
+	return ((settingsUsbMode == USB_MODE_HOTSPOT) && (hrc.txDataType != HOTSPOT_TX_TYPE_VOICE));
+}
+
+"""
+s = s.replace(marker, marker + "\n" + helper, 1)
+
+# 2. use it in the three start-of-transmission checks
+a1 = "\t\t\tif (hrc.txDataType != HOTSPOT_TX_TYPE_VOICE)\n\t\t\t{\n\t\t\t\t// A data burst is queued: don't announce a voice call in front of it."
+assert a1 in s, "TX_START_1 check not found"
+s = s.replace(a1, "\t\t\tif (isHotspotDataTx())\n\t\t\t{\n\t\t\t\t// A data burst is queued: don't announce a voice call in front of it.", 1)
+
+a2 = "((hrc.txDataType != HOTSPOT_TX_TYPE_VOICE) ? (DT_IDLE << 4) : 0x10)"
+n2 = s.count(a2)
+assert n2 == 2, "expected 2 shared checks, found %d" % n2
+s = s.replace(a2, "(isHotspotDataTx() ? (DT_IDLE << 4) : 0x10)")
+
+# 3. and in the ongoing-TX branch, so a stale marker cannot divert voice frames either
+a3 = "\t\t\t\t\tif (hrc.txDataType != HOTSPOT_TX_TYPE_VOICE)\n"
+assert a3 in s, "TX_2 data branch not found"
+s = s.replace(a3, "\t\t\t\t\tif (isHotspotDataTx())\n", 1)
+
+# 4. belt and braces: clear the marker when a non-hotspot transmission starts
+a4 = """				if (settingsUsbMode != USB_MODE_HOTSPOT)
+				{
+					codecInit(false);
+				}
+				else
+				{
+					hrc.hotspotPostponedFrameHandling = (HS_NUM_OF_SILENCE_SEQ_ON_STARTUP * 6);"""
+assert a4 in s, "TX start branch not found"
+s = s.replace(a4, """				if (settingsUsbMode != USB_MODE_HOTSPOT)
+				{
+					hrc.txDataType = HOTSPOT_TX_TYPE_VOICE;// plain voice call: never a data burst
+					codecInit(false);
+				}
+				else
+				{
+					hrc.hotspotPostponedFrameHandling = (HS_NUM_OF_SILENCE_SEQ_ON_STARTUP * 6);""", 1)
+
+open(p, "w").write(s)
+print("HR-C6000.c: voice transmissions can no longer be suppressed by a stale data marker")