Merge pull request 'feat(P4): media_manager vrai playback WAV + /game/media/play' (#9) from feat/p4-media-wav-playback into main

idf_zacus/components/game_endpoint/CMakeLists.txt

@@ -15,6 +15,8 @@ idf_component_register(
         log
         joltwallet__littlefs
         puzzle_state
+        media_manager
+        sd_storage
     PRIV_REQUIRES
         local_puzzles
 )

idf_zacus/components/game_endpoint/game_endpoint.c

@@ -18,6 +18,8 @@
 #include "esp_http_server.h"
 #include "esp_littlefs.h"
 #include "esp_log.h"
+#include "media_manager.h"
+#include "sd_storage.h"
 #include "esp_system.h"
 #include "freertos/FreeRTOS.h"
 #include "freertos/task.h"
@@ -866,6 +868,55 @@ static esp_err_t handle_step_post(httpd_req_t *req) {
     }
 }
 
+// ─── POST /game/media/play — trigger a media_manager WAV cue ─────────────────
+// Body: { "path": "apps/<id>/cue.wav" | "/littlefs/music/foo.wav" }. Streams
+// the file to the MAX98357A via media_manager (P4 real playback). Lets the
+// gateway / atelier fire audio cues without an NPC decision.
+static esp_err_t handle_media_play_post(httpd_req_t *req) {
+    char body[513] = {0};
+    if (req->content_len <= 0 || req->content_len > 512) {
+        return send_error(req, "400 Bad Request", "body must be 1..512 bytes");
+    }
+    int total = 0;
+    while (total < (int) req->content_len) {
+        int got = httpd_req_recv(req, body + total, req->content_len - total);
+        if (got <= 0) {
+            if (got == HTTPD_SOCK_ERR_TIMEOUT) continue;
+            return send_error(req, "400 Bad Request", "recv failed");
+        }
+        total += got;
+    }
+    body[total] = '\0';
+
+    cJSON *root = cJSON_Parse(body);
+    if (!root) {
+        return send_error(req, "400 Bad Request", "malformed json");
+    }
+    const cJSON *path_j = cJSON_GetObjectItemCaseSensitive(root, "path");
+    if (!cJSON_IsString(path_j) || !path_j->valuestring || path_j->valuestring[0] == '\0') {
+        cJSON_Delete(root);
+        return send_error(req, "400 Bad Request", "missing or empty path");
+    }
+    char path[160];
+    strncpy(path, path_j->valuestring, sizeof(path) - 1);
+    path[sizeof(path) - 1] = '\0';
+    cJSON_Delete(root);
+
+    esp_err_t err = media_manager_play(path);
+    if (err == ESP_OK) {
+        char buf[200];
+        snprintf(buf, sizeof(buf), "{\"status\":\"playing\",\"path\":\"%s\"}", path);
+        return send_json(req, "200 OK", buf);
+    }
+    if (err == ESP_ERR_NOT_FOUND) {
+        return send_error(req, "404 Not Found", "file_not_found");
+    }
+    if (err == ESP_ERR_INVALID_ARG) {
+        return send_error(req, "400 Bad Request", "invalid_path");
+    }
+    return send_error(req, "500 Internal Server Error", esp_err_to_name(err));
+}
+
 // ─── GET /game/puzzle_state ──────────────────────────────────────────────────
 //
 // Returns {"step_id":"STEP_X"|null, "solved":[1,3], "code":"125"}.
@@ -951,22 +1002,36 @@ static esp_err_t handle_file_post(httpd_req_t *req) {
                               sizeof(path_param)) != ESP_OK) {
         return send_error(req, "400 Bad Request", "missing path param");
     }
-    if (strncmp(path_param, "apps/", 5) != 0 || strstr(path_param, "..") ||
+    // Route: apps/<…> → /littlefs/apps/<…> (display shell tiles, ≤256 KiB);
+    //        sd/<…>   → /sdcard/<…>        (P4 large assets: audio, images).
+    const bool to_sd   = (strncmp(path_param, "sd/", 3) == 0);
+    const bool to_apps = (strncmp(path_param, "apps/", 5) == 0);
+    if ((!to_sd && !to_apps) || strstr(path_param, "..") ||
         path_param[strlen(path_param) - 1] == '/') {
-        return send_error(req, "403 Forbidden", "path must be under apps/");
+        return send_error(req, "403 Forbidden", "path must be under apps/ or sd/");
     }
-    if (req->content_len <= 0 ||
+    const long max_bytes = to_sd ? (8L * 1024 * 1024) : GAME_ENDPOINT_MAX_FILE_BYTES;
-        req->content_len > GAME_ENDPOINT_MAX_FILE_BYTES) {
+    if (req->content_len <= 0 || (long) req->content_len > max_bytes) {
-        return send_error(req, "400 Bad Request", "size 1..262144 bytes");
+        return send_error(req, "400 Bad Request", "file too large");
-    }
-    if (mount_storage_lazy() != ESP_OK) {
-        return send_error(req, "503 Service Unavailable", "storage_unavailable");
     }
 
-    char full[160];
+    char full[192];
-    snprintf(full, sizeof(full), "/littlefs/%s", path_param);
+    size_t root_len;
+    if (to_sd) {
+        if (!sd_storage_ready()) {
+            return send_error(req, "503 Service Unavailable", "sd_not_mounted");
+        }
+        snprintf(full, sizeof(full), "/sdcard/%s", path_param + 3);  // strip "sd/"
+        root_len = strlen("/sdcard/");
+    } else {
+        if (mount_storage_lazy() != ESP_OK) {
+            return send_error(req, "503 Service Unavailable", "storage_unavailable");
+        }
+        snprintf(full, sizeof(full), "/littlefs/%s", path_param);
+        root_len = strlen("/littlefs/");
+    }
     // mkdir -p for intermediate directories.
-    for (char *p = full + strlen("/littlefs/"); *p; p++) {
+    for (char *p = full + root_len; *p; p++) {
         if (*p == '/') {
             *p = '\0';
             mkdir(full, 0775);   // EEXIST is fine
@@ -1084,6 +1149,12 @@ esp_err_t game_endpoint_init(httpd_handle_t server) {
         .handler  = handle_espnow_cmd_post,
         .user_ctx = NULL,
     };
+    static const httpd_uri_t uri_media_play_post = {
+        .uri      = "/game/media/play",
+        .method   = HTTP_POST,
+        .handler  = handle_media_play_post,
+        .user_ctx = NULL,
+    };
 
     // Bring up the ESP-NOW mesh transport. The master is primarily a sender
     // (the relay handler) but we also register the apply adapter so a peer
@@ -1156,6 +1227,10 @@ esp_err_t game_endpoint_init(httpd_handle_t server) {
     if (err != ESP_OK) {
         ESP_LOGW(TAG, "register GET /game/puzzle_state: %s", esp_err_to_name(err));
     }
+    err = httpd_register_uri_handler(server, &uri_media_play_post);
+    if (err != ESP_OK) {
+        ESP_LOGW(TAG, "register POST /game/media/play: %s", esp_err_to_name(err));
+    }
 
     ESP_LOGI(TAG, "game endpoint registered "
                   "(GET+POST /game/group_profile, POST /game/scenario%s, "

idf_zacus/components/media_manager/CMakeLists.txt

@@ -17,5 +17,7 @@ idf_component_register(
         driver
         esp_timer
         esp_system
+        freertos
+        voice_pipeline
         joltwallet__littlefs
 )

idf_zacus/components/media_manager/idf_component.yml

@@ -0,0 +1,4 @@
+## Managed dependencies for media_manager.
+## Slice P4: helix MP3 decoder for the hotline_tts/*.mp3 cue pool.
+dependencies:
+  chmorgan/esp-libhelix-mp3: "^1.0.3"

idf_zacus/components/media_manager/media_manager.c

@@ -27,15 +27,25 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
+#include <strings.h>
 #include <sys/stat.h>
 #include <sys/types.h>
 #include <unistd.h>
 
+#include "mp3dec.h"   // P4: helix MP3 decoder (chmorgan/esp-libhelix-mp3)
+
+#include "freertos/FreeRTOS.h"
+#include "freertos/task.h"
+
 #include "esp_check.h"
 #include "esp_err.h"
 #include "esp_log.h"
 #include "esp_timer.h"
 
+// Slice P4: real WAV playback streams 16-bit PCM to the MAX98357A over the
+// I2S TX channel already owned by voice_pipeline (shared DAC, no codec init).
+#include "voice_pipeline.h"
+
 static const char *TAG = "media_manager";
 
 #define MEDIA_DEFAULT_MUSIC_DIR    "/littlefs/music"
@@ -59,6 +69,8 @@ static struct {
     uint32_t                   playback_ends_ms;
     FILE                      *recording_file;
     uint32_t                   recording_data_bytes;
+    TaskHandle_t               playback_task;   // P4: real WAV streamer, NULL when idle
+    volatile bool              playback_stop;   // request the streamer to abort
 } s_media;
 
 // ─── Helpers ─────────────────────────────────────────────────────────────────
@@ -149,10 +161,20 @@ static void resolve_play_path(char *out, size_t out_len, const char *path) {
         return;
     }
     if (path[0] == '/') {
-        copy_text(out, out_len, path);
+        copy_text(out, out_len, path);   // absolute: /sdcard/... or /littlefs/...
-    } else {
+        return;
-        snprintf(out, out_len, "%s/%s", s_media.config.music_dir, path);
     }
+    // Relative cue: prefer the microSD copy when present (P4 — large asset
+    // store), else fall back to the LittleFS music_dir. file_exists() returns
+    // false when no card is mounted, so this is transparent without a hard
+    // dependency on sd_storage.
+    char sd_try[MEDIA_PATH_MAX];
+    snprintf(sd_try, sizeof(sd_try), "/sdcard/music/%s", path);
+    if (file_exists(sd_try)) {
+        copy_text(out, out_len, sd_try);
+        return;
+    }
+    snprintf(out, out_len, "%s/%s", s_media.config.music_dir, path);
 }
 
 static void sanitize_filename(char *out, size_t out_len,
@@ -332,6 +354,211 @@ void media_manager_note_step_change(void) {
     }
 }
 
+// ─── P4: real WAV playback over the shared MAX98357A I2S TX ──────────────────
+//
+// Streams 16-bit PCM WAV from LittleFS to the DAC via voice_pipeline's TX
+// channel. WAV (not MP3) keeps this decoder-free; MP3 stays a later slice.
+// NOTE: shares the I2S TX with TTS — callers must not overlap a `speak_*`
+// exchange with music playback (the NPC coordination layer already serialises
+// cues). A TX mutex is the natural follow-up.
+
+#define MEDIA_PLAY_CHUNK_BYTES   1024U
+
+typedef struct {
+    uint32_t sample_rate;
+    uint16_t channels;
+    uint16_t bits;
+    uint32_t data_size;   // PCM bytes in the data chunk
+} wav_info_t;
+
+static uint32_t rd_u32(const uint8_t *p) {
+    return (uint32_t) p[0] | ((uint32_t) p[1] << 8) |
+           ((uint32_t) p[2] << 16) | ((uint32_t) p[3] << 24);
+}
+
+// Parse a RIFF/WAVE header, leaving `f` positioned at the PCM data. Accepts
+// 16-bit PCM mono/stereo only; returns false otherwise.
+static bool parse_wav_header(FILE *f, wav_info_t *out) {
+    uint8_t riff[12];
+    if (fread(riff, 1, 12, f) != 12) return false;
+    if (memcmp(riff, "RIFF", 4) != 0 || memcmp(riff + 8, "WAVE", 4) != 0) return false;
+
+    bool have_fmt = false;
+    uint8_t ck[8];
+    while (fread(ck, 1, 8, f) == 8) {
+        uint32_t csz = rd_u32(ck + 4);
+        if (memcmp(ck, "fmt ", 4) == 0) {
+            uint8_t fmt[16];
+            uint32_t want = csz < sizeof(fmt) ? csz : (uint32_t) sizeof(fmt);
+            if (fread(fmt, 1, want, f) != want) return false;
+            if (((uint16_t) fmt[0] | ((uint16_t) fmt[1] << 8)) != 1U) return false; // PCM
+            out->channels    = (uint16_t) fmt[2] | ((uint16_t) fmt[3] << 8);
+            out->sample_rate = rd_u32(fmt + 4);
+            out->bits        = (uint16_t) fmt[14] | ((uint16_t) fmt[15] << 8);
+            if (csz > want) fseek(f, (long) (csz - want), SEEK_CUR);
+            have_fmt = true;
+        } else if (memcmp(ck, "data", 4) == 0) {
+            out->data_size = csz;
+            return have_fmt && out->bits == 16U &&
+                   (out->channels == 1U || out->channels == 2U);
+        } else {
+            fseek(f, (long) (csz + (csz & 1U)), SEEK_CUR); // skip + word-align pad
+        }
+    }
+    return false;
+}
+
+static void apply_volume(int16_t *s, size_t count, uint8_t volume) {
+    if (volume >= 100U) return;
+    for (size_t i = 0; i < count; ++i) {
+        s[i] = (int16_t) (((int32_t) s[i] * (int32_t) volume) / 100);
+    }
+}
+
+static void clear_playback_state(void) {
+    s_media.snapshot.playing = false;
+    s_media.snapshot.playing_path[0] = '\0';
+    s_media.playback_ends_ms = 0U;
+    s_media.playback_task = NULL;
+}
+
+// Stream a 16-bit PCM WAV to the DAC. play_start/play_end bracket the session.
+static esp_err_t stream_wav(FILE *f, const char *path) {
+    wav_info_t wav = {0};
+    if (!parse_wav_header(f, &wav)) {
+        ESP_LOGE(TAG, "playback: %s is not a 16-bit PCM WAV", path);
+        return ESP_ERR_INVALID_ARG;
+    }
+    ESP_LOGI(TAG, "playback start: %s (wav %lu Hz, %uch, %lu B) vol=%u",
+             path, (unsigned long) wav.sample_rate, wav.channels,
+             (unsigned long) wav.data_size, s_media.volume);
+
+    esp_err_t err = voice_pipeline_play_start(wav.sample_rate, "wav");
+    if (err != ESP_OK) {
+        ESP_LOGW(TAG, "play_start failed: %s", esp_err_to_name(err));
+        return err;
+    }
+
+    uint8_t buf[MEDIA_PLAY_CHUNK_BYTES];
+    uint32_t remaining = wav.data_size, written = 0;
+    while (remaining > 0U && !s_media.playback_stop) {
+        size_t want = remaining < sizeof(buf) ? remaining : sizeof(buf);
+        size_t n = fread(buf, 1, want, f);
+        if (n == 0U) break;
+        remaining -= (uint32_t) n;
+        size_t pcm = n;
+        if (wav.channels == 2U) {              // downmix interleaved stereo16 → mono16
+            int16_t *s = (int16_t *) buf;
+            size_t frames = n / 4U;
+            for (size_t i = 0; i < frames; ++i)
+                s[i] = (int16_t) (((int32_t) s[2 * i] + s[2 * i + 1]) / 2);
+            pcm = frames * 2U;
+        }
+        apply_volume((int16_t *) buf, pcm / 2U, s_media.volume);
+        if (voice_pipeline_play_chunk(buf, pcm) != ESP_OK) break;
+        written += (uint32_t) pcm;
+    }
+    voice_pipeline_play_end();
+    ESP_LOGI(TAG, "playback done: %s (%lu B to DAC%s)", path,
+             (unsigned long) written, s_media.playback_stop ? ", stopped" : "");
+    return ESP_OK;
+}
+
+// Decode and stream an MP3 to the DAC via the helix decoder. Big work buffers
+// live on the heap (the decoder also uses a few KB of stack — task is sized
+// accordingly). Mono output is downmixed; volume is applied per frame.
+static esp_err_t stream_mp3(FILE *f, const char *path) {
+    HMP3Decoder dec = MP3InitDecoder();
+    const size_t IN_SZ = 2048U;
+    uint8_t *inbuf = malloc(IN_SZ);
+    int16_t *pcm = malloc(2304 * sizeof(int16_t)); // MAX_NCHAN*MAX_NGRAN*MAX_NSAMP
+    if (dec == NULL || inbuf == NULL || pcm == NULL) {
+        if (dec) MP3FreeDecoder(dec);
+        free(inbuf); free(pcm);
+        ESP_LOGE(TAG, "mp3: alloc failed");
+        return ESP_ERR_NO_MEM;
+    }
+
+    int bytes_left = 0;
+    uint8_t *rp = inbuf;
+    bool started = false, eof = false;
+    uint32_t written = 0;
+    esp_err_t result = ESP_OK;
+
+    while (!s_media.playback_stop) {
+        if (bytes_left < 1024 && !eof) {           // refill: keep tail, top up
+            memmove(inbuf, rp, (size_t) bytes_left);
+            size_t n = fread(inbuf + bytes_left, 1, IN_SZ - (size_t) bytes_left, f);
+            if (n == 0U) eof = true;
+            bytes_left += (int) n;
+            rp = inbuf;
+        }
+        if (bytes_left <= 0) break;
+        int off = MP3FindSyncWord(rp, bytes_left);
+        if (off < 0) { if (eof) break; bytes_left = 0; continue; }
+        rp += off; bytes_left -= off;
+        int derr = MP3Decode(dec, &rp, &bytes_left, pcm, 0);
+        if (derr == ERR_MP3_INDATA_UNDERFLOW) { if (eof) break; continue; }
+        if (derr != ERR_MP3_NONE) continue;        // skip a bad frame
+        MP3FrameInfo info;
+        MP3GetLastFrameInfo(dec, &info);
+        if (info.outputSamps <= 0) continue;
+        if (!started) {
+            ESP_LOGI(TAG, "playback start: %s (mp3 %d Hz, %dch) vol=%u",
+                     path, info.samprate, info.nChans, s_media.volume);
+            esp_err_t e = voice_pipeline_play_start((uint32_t) info.samprate, "mp3");
+            if (e != ESP_OK) { result = e; break; }
+            started = true;
+        }
+        size_t samps = (size_t) info.outputSamps, bytes;
+        if (info.nChans == 2) {                    // downmix stereo16 → mono16
+            size_t frames = samps / 2U;
+            for (size_t i = 0; i < frames; ++i)
+                pcm[i] = (int16_t) (((int32_t) pcm[2 * i] + pcm[2 * i + 1]) / 2);
+            bytes = frames * 2U;
+        } else {
+            bytes = samps * 2U;
+        }
+        apply_volume(pcm, bytes / 2U, s_media.volume);
+        if (voice_pipeline_play_chunk((uint8_t *) pcm, bytes) != ESP_OK) break;
+        written += (uint32_t) bytes;
+    }
+
+    if (started) voice_pipeline_play_end();
+    MP3FreeDecoder(dec);
+    free(inbuf); free(pcm);
+    ESP_LOGI(TAG, "playback done: %s (mp3, %lu B to DAC%s)", path,
+             (unsigned long) written, s_media.playback_stop ? ", stopped" : "");
+    if (!started && result == ESP_OK) result = ESP_ERR_NOT_SUPPORTED;
+    return result;
+}
+
+static void playback_task(void *arg) {
+    (void) arg;
+    char path[MEDIA_PATH_MAX];
+    copy_text(path, sizeof(path), s_media.snapshot.playing_path);
+
+    FILE *f = fopen(path, "rb");
+    if (f == NULL) {
+        ESP_LOGE(TAG, "playback: cannot open %s", path);
+        set_last_error("media_play_open");
+        clear_playback_state();
+        vTaskDelete(NULL);
+        return;
+    }
+
+    const char *ext = strrchr(path, '.');
+    esp_err_t err = (ext && strcasecmp(ext, ".mp3") == 0)
+                        ? stream_mp3(f, path)
+                        : stream_wav(f, path);
+    fclose(f);
+    if (err != ESP_OK) {
+        set_last_error("media_play_failed");
+    }
+    clear_playback_state();
+    vTaskDelete(NULL);
+}
+
 esp_err_t media_manager_play(const char *path) {
     if (!s_media.initialized) {
         return ESP_ERR_INVALID_STATE;
@@ -354,15 +581,27 @@ esp_err_t media_manager_play(const char *path) {
         return ESP_ERR_NOT_FOUND;
     }
 
-    // TODO(slice-4+): hand off to the real I2S MP3 decoder here.
+    // Stop any in-flight playback before starting the next cue.
-    const uint32_t now_ms = (uint32_t) (esp_timer_get_time() / 1000LL);
+    if (s_media.playback_task != NULL) {
+        (void) media_manager_stop();
+    }
+
+    s_media.playback_stop = false;
     s_media.snapshot.playing = true;
     copy_text(s_media.snapshot.playing_path,
               sizeof(s_media.snapshot.playing_path), resolved);
-    s_media.playback_ends_ms = now_ms + MEDIA_STUB_PLAYBACK_MS;
+    s_media.playback_ends_ms = 0U;   // real playback: tick must not auto-clear
     clear_last_error();
-    ESP_LOGI(TAG, "playing %s (simulated %ums) vol=%u",
+
-             resolved, MEDIA_STUB_PLAYBACK_MS, s_media.volume);
+    BaseType_t ok = xTaskCreate(playback_task, "media_play", 8192, NULL, 4,
+                                &s_media.playback_task);
+    if (ok != pdPASS) {
+        s_media.playback_task = NULL;
+        s_media.snapshot.playing = false;
+        s_media.snapshot.playing_path[0] = '\0';
+        set_last_error("media_play_task");
+        return ESP_ERR_NO_MEM;
+    }
     return ESP_OK;
 }
 
@@ -370,6 +609,12 @@ esp_err_t media_manager_stop(void) {
     if (!s_media.initialized) {
         return ESP_ERR_INVALID_STATE;
     }
+    if (s_media.playback_task != NULL) {
+        s_media.playback_stop = true;
+        for (int i = 0; i < 100 && s_media.playback_task != NULL; ++i) {
+            vTaskDelay(pdMS_TO_TICKS(10));   // let the streamer drain + self-delete
+        }
+    }
     if (s_media.snapshot.playing) {
         ESP_LOGI(TAG, "stop %s", s_media.snapshot.playing_path);
     }

idf_zacus/components/sd_storage/CMakeLists.txt

@@ -0,0 +1,11 @@
+## Zacus sd_storage — microSD (SDMMC 1-bit) FAT mount at /sdcard (P4).
+idf_component_register(
+    SRCS
+        "sd_storage.c"
+    INCLUDE_DIRS
+        "."
+    REQUIRES
+        fatfs
+        sdmmc
+        esp_driver_sdmmc
+)

idf_zacus/components/sd_storage/sd_storage.c

@@ -0,0 +1,74 @@
+#include "sd_storage.h"
+
+#include "esp_log.h"
+#include "esp_vfs_fat.h"
+#include "sdmmc_cmd.h"
+#include "driver/sdmmc_host.h"
+
+static const char *TAG = "sd_storage";
+
+// Freenove ESP32-S3 WROOM N16R8 microSD (SDMMC 1-bit). Mirrors
+// ui_freenove_allinone/include/ui_freenove_config.h FREENOVE_SDMMC_*.
+#define SD_PIN_CMD 38
+#define SD_PIN_CLK 39
+#define SD_PIN_D0  40
+
+static sdmmc_card_t *s_card = NULL;
+
+esp_err_t sd_storage_mount(void) {
+    if (s_card != NULL) {
+        return ESP_OK;
+    }
+
+    const esp_vfs_fat_sdmmc_mount_config_t mount_cfg = {
+        .format_if_mount_failed = false,   // never wipe a user's card
+        .max_files = 5,
+        .allocation_unit_size = 16 * 1024,
+    };
+
+    sdmmc_host_t host = SDMMC_HOST_DEFAULT();
+    host.flags = SDMMC_HOST_FLAG_1BIT;     // Freenove wires D0 only
+    host.max_freq_khz = SDMMC_FREQ_DEFAULT;
+
+    sdmmc_slot_config_t slot = SDMMC_SLOT_CONFIG_DEFAULT();
+    slot.width = 1;
+    slot.clk = SD_PIN_CLK;
+    slot.cmd = SD_PIN_CMD;
+    slot.d0  = SD_PIN_D0;
+    slot.flags |= SDMMC_SLOT_FLAG_INTERNAL_PULLUP;
+
+    esp_err_t err = esp_vfs_fat_sdmmc_mount(SD_STORAGE_MOUNT_POINT, &host,
+                                            &slot, &mount_cfg, &s_card);
+    if (err != ESP_OK) {
+        ESP_LOGW(TAG, "mount failed: %s — no card inserted or wiring issue",
+                 esp_err_to_name(err));
+        s_card = NULL;
+        return err;
+    }
+
+    ESP_LOGI(TAG, "microSD mounted at %s — %lu MiB (%s)",
+             SD_STORAGE_MOUNT_POINT, (unsigned long) sd_storage_capacity_mb(),
+             s_card->cid.name);
+    return ESP_OK;
+}
+
+esp_err_t sd_storage_unmount(void) {
+    if (s_card == NULL) {
+        return ESP_OK;
+    }
+    esp_err_t err = esp_vfs_fat_sdcard_unmount(SD_STORAGE_MOUNT_POINT, s_card);
+    s_card = NULL;
+    return err;
+}
+
+bool sd_storage_ready(void) {
+    return s_card != NULL;
+}
+
+uint32_t sd_storage_capacity_mb(void) {
+    if (s_card == NULL) {
+        return 0U;
+    }
+    uint64_t bytes = (uint64_t) s_card->csd.capacity * s_card->csd.sector_size;
+    return (uint32_t) (bytes / (1024ULL * 1024ULL));
+}

idf_zacus/components/sd_storage/sd_storage.h

@@ -0,0 +1,35 @@
+// Zacus sd_storage — microSD mount for the Freenove ESP32-S3 master.
+//
+// The Freenove WROOM N16R8 exposes the card on the SDMMC peripheral in
+// 1-bit mode (CMD=GPIO38, CLK=GPIO39, D0=GPIO40 — mirrors the Arduino
+// `ui_freenove_allinone` FREENOVE_SDMMC_* pinout). Mounts FAT at /sdcard so
+// media_manager and the asset endpoints can stage large assets off the 5 MB
+// LittleFS partition (P4). Mount is best-effort: a missing card is not fatal.
+#pragma once
+
+#include <stdbool.h>
+#include <stdint.h>
+#include "esp_err.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#define SD_STORAGE_MOUNT_POINT "/sdcard"
+
+// Mount the microSD card (SDMMC 1-bit) at /sdcard. Returns ESP_OK on success;
+// ESP_ERR_NOT_FOUND / ESP_FAIL when no card is present or wiring is wrong.
+esp_err_t sd_storage_mount(void);
+
+// Unmount and release the card.
+esp_err_t sd_storage_unmount(void);
+
+// True once a card is mounted.
+bool sd_storage_ready(void);
+
+// Card capacity in MiB (0 when not mounted).
+uint32_t sd_storage_capacity_mb(void);
+
+#ifdef __cplusplus
+}
+#endif

idf_zacus/components/voice_pipeline/voice_pipeline.c

@@ -21,6 +21,7 @@
 #include "esp_mac.h"
 #include "freertos/FreeRTOS.h"
 #include "freertos/task.h"
+#include "freertos/semphr.h"
 
 #include "esp_afe_config.h"
 #include "esp_afe_sr_iface.h"
@@ -64,6 +65,7 @@ static struct {
     i2s_chan_handle_t        tx_chan;
     bool                     tx_enabled;        // channel currently enabled
     uint32_t                 tx_sample_rate;    // current configured rate
+    SemaphoreHandle_t        play_lock;         // serialises TTS vs media playback (one session at a time)
     voice_state_t            state;
     // capture_task / capture_run removed: mic_broker's task delivers frames
     // via on_npc_frame(); broker mode (MIC_NPC_LISTEN/MIC_IDLE) gates it.
@@ -241,6 +243,9 @@ static esp_err_t i2s_tx_setup(void) {
         return err;
     }
     s_pipe.tx_enabled = false;
+    if (s_pipe.play_lock == NULL) {
+        s_pipe.play_lock = xSemaphoreCreateMutex();
+    }
     return ESP_OK;
 }
 
@@ -575,6 +580,13 @@ esp_err_t voice_pipeline_play_start(uint32_t sample_rate, const char *format) {
         ESP_LOGW(TAG, "play_start: no TX channel (enable_tts_playback=false?)");
         return ESP_ERR_INVALID_STATE;
     }
+    // Serialise the shared TX channel: only one playback session (TTS or
+    // media_manager) may own the DAC at a time. Held until play_end().
+    if (s_pipe.play_lock &&
+        xSemaphoreTake(s_pipe.play_lock, pdMS_TO_TICKS(2000)) != pdTRUE) {
+        ESP_LOGW(TAG, "play_start: busy — another playback owns the I2S TX");
+        return ESP_ERR_TIMEOUT;
+    }
     // Reconfigure the I2S clock if the bridge announces a different rate.
     // F5-TTS produces 24 kHz; the mic side runs at 16 kHz by default.
     if (sample_rate != 0 && sample_rate != s_pipe.tx_sample_rate) {
@@ -587,13 +599,17 @@ esp_err_t voice_pipeline_play_start(uint32_t sample_rate, const char *format) {
         if (err != ESP_OK) {
             ESP_LOGW(TAG, "play_start: clk reconfig %u Hz failed: %s",
                      (unsigned) sample_rate, esp_err_to_name(err));
+            if (s_pipe.play_lock) xSemaphoreGive(s_pipe.play_lock);
             return err;
         }
         s_pipe.tx_sample_rate = sample_rate;
     }
     if (!s_pipe.tx_enabled) {
         esp_err_t err = i2s_channel_enable(s_pipe.tx_chan);
-        if (err != ESP_OK) return err;
+        if (err != ESP_OK) {
+            if (s_pipe.play_lock) xSemaphoreGive(s_pipe.play_lock);
+            return err;
+        }
         s_pipe.tx_enabled = true;
     }
     voice_pipeline_set_state(VOICE_STATE_SPEAKING);
@@ -625,5 +641,6 @@ esp_err_t voice_pipeline_play_end(void) {
     }
     voice_pipeline_set_state(VOICE_STATE_IDLE);
     ESP_LOGI(TAG, "play_end");
+    if (s_pipe.play_lock) xSemaphoreGive(s_pipe.play_lock);
     return ESP_OK;
 }

idf_zacus/main/CMakeLists.txt

@@ -6,6 +6,7 @@ idf_component_register(
     REQUIRES
         joltwallet__littlefs
         ota_server
+        sd_storage
         media_manager
         npc_engine
         hints_client

idf_zacus/main/main.c

@@ -40,6 +40,7 @@
 
 #include "ota_server.h"
 #include "media_manager.h"
+#include "sd_storage.h"
 #include "npc_engine.h"
 #include "hints_client.h"
 #include "voice_pipeline.h"
@@ -484,6 +485,13 @@ void app_main(void) {
         }
     }
 
+    // P4: best-effort microSD mount (SDMMC 1-bit on /sdcard). A missing card
+    // is non-fatal — assets fall back to LittleFS.
+    if (sd_storage_mount() == ESP_OK) {
+        ESP_LOGI(TAG, "microSD ready (%lu MiB) at %s",
+                 (unsigned long) sd_storage_capacity_mb(), SD_STORAGE_MOUNT_POINT);
+    }
+
     if (mount_littlefs() == ESP_OK) {
         list_littlefs_root();
 

idf_zacus/sdkconfig.defaults

@@ -70,3 +70,8 @@ CONFIG_LV_USE_FS_STDIO=y
 CONFIG_LV_FS_STDIO_LETTER=76
 CONFIG_LV_FS_STDIO_PATH="/littlefs"
 CONFIG_LV_USE_PNG=y
+
+# P4: FATFS Long File Names (heap) — assets sur SD ont des noms > 8.3
+# (SCENE_WIN.mp3, sonar_hint.mp3, hotline_tts/…). Sans ça fopen échoue.
+CONFIG_FATFS_LFN_HEAP=y
+CONFIG_FATFS_MAX_LFN=255