audio: msm8909w caf release LW.BR.1.0-00410-8x09w.0
MSM8909w Audio HAL code copied from CAF release
LW.BR.1.0-00410-8x09w.0
dbcce50 hal: Port wcd9326 changes to 8909
410c530 hal: update error handling for pcm_prepare failures
ff79309 hal: fix compilation issues with audio FM extention
762d7eb policy_hal: add support for fm device loopback
7c418f9 audio_policy: modify few methods to appropriately override base
8b12163 audio: Add support to enable split A2DP
a0559fa Revert "Revert "policy_hal: Function prototype correction for custom policy"."
Fixed makefiles to be compatible with PDK without kernel source
Change-Id: I9c6f2139adee62426b877516deeb41d4ed8052b2
diff --git a/msm8909/visualizer/offload_visualizer.c b/msm8909/visualizer/offload_visualizer.c
new file mode 100644
index 0000000..94c44a5
--- /dev/null
+++ b/msm8909/visualizer/offload_visualizer.c
@@ -0,0 +1,1267 @@
+/*
+ * Copyright (C) 2013 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#define LOG_TAG "offload_visualizer"
+/*#define LOG_NDEBUG 0*/
+#include <assert.h>
+#include <math.h>
+#include <stdlib.h>
+#include <string.h>
+#include <time.h>
+#include <sys/prctl.h>
+
+#include <cutils/list.h>
+#include <cutils/log.h>
+#include <system/thread_defs.h>
+#include <tinyalsa/asoundlib.h>
+#include <audio_effects/effect_visualizer.h>
+
+
+enum {
+ EFFECT_STATE_UNINITIALIZED,
+ EFFECT_STATE_INITIALIZED,
+ EFFECT_STATE_ACTIVE,
+};
+
+typedef struct effect_context_s effect_context_t;
+typedef struct output_context_s output_context_t;
+
+/* effect specific operations. Only the init() and process() operations must be defined.
+ * Others are optional.
+ */
+typedef struct effect_ops_s {
+ int (*init)(effect_context_t *context);
+ int (*release)(effect_context_t *context);
+ int (*reset)(effect_context_t *context);
+ int (*enable)(effect_context_t *context);
+ int (*disable)(effect_context_t *context);
+ int (*start)(effect_context_t *context, output_context_t *output);
+ int (*stop)(effect_context_t *context, output_context_t *output);
+ int (*process)(effect_context_t *context, audio_buffer_t *in, audio_buffer_t *out);
+ int (*set_parameter)(effect_context_t *context, effect_param_t *param, uint32_t size);
+ int (*get_parameter)(effect_context_t *context, effect_param_t *param, uint32_t *size);
+ int (*command)(effect_context_t *context, uint32_t cmdCode, uint32_t cmdSize,
+ void *pCmdData, uint32_t *replySize, void *pReplyData);
+} effect_ops_t;
+
+struct effect_context_s {
+ const struct effect_interface_s *itfe;
+ struct listnode effects_list_node; /* node in created_effects_list */
+ struct listnode output_node; /* node in output_context_t.effects_list */
+ effect_config_t config;
+ const effect_descriptor_t *desc;
+ audio_io_handle_t out_handle; /* io handle of the output the effect is attached to */
+ uint32_t state;
+ bool offload_enabled; /* when offload is enabled we process VISUALIZER_CMD_CAPTURE command.
+ Otherwise non offloaded visualizer has already processed the command
+ and we must not overwrite the reply. */
+ effect_ops_t ops;
+};
+
+typedef struct output_context_s {
+ struct listnode outputs_list_node; /* node in active_outputs_list */
+ audio_io_handle_t handle; /* io handle */
+ struct listnode effects_list; /* list of effects attached to this output */
+} output_context_t;
+
+
+/* maximum time since last capture buffer update before resetting capture buffer. This means
+ that the framework has stopped playing audio and we must start returning silence */
+#define MAX_STALL_TIME_MS 1000
+
+#define CAPTURE_BUF_SIZE 65536 /* "64k should be enough for everyone" */
+
+#define DISCARD_MEASUREMENTS_TIME_MS 2000 /* discard measurements older than this number of ms */
+
+/* maximum number of buffers for which we keep track of the measurements */
+#define MEASUREMENT_WINDOW_MAX_SIZE_IN_BUFFERS 25 /* note: buffer index is stored in uint8_t */
+
+typedef struct buffer_stats_s {
+ bool is_valid;
+ uint16_t peak_u16; /* the positive peak of the absolute value of the samples in a buffer */
+ float rms_squared; /* the average square of the samples in a buffer */
+} buffer_stats_t;
+
+typedef struct visualizer_context_s {
+ effect_context_t common;
+
+ uint32_t capture_idx;
+ uint32_t capture_size;
+ uint32_t scaling_mode;
+ uint32_t last_capture_idx;
+ uint32_t latency;
+ struct timespec buffer_update_time;
+ uint8_t capture_buf[CAPTURE_BUF_SIZE];
+ /* for measurements */
+ uint8_t channel_count; /* to avoid recomputing it every time a buffer is processed */
+ uint32_t meas_mode;
+ uint8_t meas_wndw_size_in_buffers;
+ uint8_t meas_buffer_idx;
+ buffer_stats_t past_meas[MEASUREMENT_WINDOW_MAX_SIZE_IN_BUFFERS];
+} visualizer_context_t;
+
+
+extern const struct effect_interface_s effect_interface;
+
+/* Offload visualizer UUID: 7a8044a0-1a71-11e3-a184-0002a5d5c51b */
+const effect_descriptor_t visualizer_descriptor = {
+ {0xe46b26a0, 0xdddd, 0x11db, 0x8afd, {0x00, 0x02, 0xa5, 0xd5, 0xc5, 0x1b}},
+ {0x7a8044a0, 0x1a71, 0x11e3, 0xa184, {0x00, 0x02, 0xa5, 0xd5, 0xc5, 0x1b}},
+ EFFECT_CONTROL_API_VERSION,
+ (EFFECT_FLAG_TYPE_INSERT | EFFECT_FLAG_HW_ACC_TUNNEL ),
+ 0, /* TODO */
+ 1,
+ "QCOM MSM offload visualizer",
+ "The Android Open Source Project",
+};
+
+const effect_descriptor_t *descriptors[] = {
+ &visualizer_descriptor,
+ NULL,
+};
+
+
+pthread_once_t once = PTHREAD_ONCE_INIT;
+int init_status;
+
+/* list of created effects. Updated by visualizer_hal_start_output()
+ * and visualizer_hal_stop_output() */
+struct listnode created_effects_list;
+/* list of active output streams. Updated by visualizer_hal_start_output()
+ * and visualizer_hal_stop_output() */
+struct listnode active_outputs_list;
+
+/* thread capturing PCM from Proxy port and calling the process function on each enabled effect
+ * attached to an active output stream */
+pthread_t capture_thread;
+/* lock must be held when modifying or accessing created_effects_list or active_outputs_list */
+pthread_mutex_t lock;
+/* thread_lock must be held when starting or stopping the capture thread.
+ * Locking order: thread_lock -> lock */
+pthread_mutex_t thread_lock;
+/* cond is signaled when an output is started or stopped or an effect is enabled or disable: the
+ * capture thread will reevaluate the capture and effect rocess conditions. */
+pthread_cond_t cond;
+/* true when requesting the capture thread to exit */
+bool exit_thread;
+/* 0 if the capture thread was created successfully */
+int thread_status;
+
+
+#define DSP_OUTPUT_LATENCY_MS 0 /* Fudge factor for latency after capture point in audio DSP */
+
+/* Retry for delay for mixer open */
+#define RETRY_NUMBER 10
+#define RETRY_US 500000
+
+#define MIXER_CARD 0
+#define SOUND_CARD 0
+#define CAPTURE_DEVICE 8
+
+/* Proxy port supports only MMAP read and those fixed parameters*/
+#define AUDIO_CAPTURE_CHANNEL_COUNT 2
+#define AUDIO_CAPTURE_SMP_RATE 48000
+#define AUDIO_CAPTURE_PERIOD_SIZE (768)
+#define AUDIO_CAPTURE_PERIOD_COUNT 32
+
+struct pcm_config pcm_config_capture = {
+ .channels = AUDIO_CAPTURE_CHANNEL_COUNT,
+ .rate = AUDIO_CAPTURE_SMP_RATE,
+ .period_size = AUDIO_CAPTURE_PERIOD_SIZE,
+ .period_count = AUDIO_CAPTURE_PERIOD_COUNT,
+ .format = PCM_FORMAT_S16_LE,
+ .start_threshold = AUDIO_CAPTURE_PERIOD_SIZE / 4,
+ .stop_threshold = INT_MAX,
+ .avail_min = AUDIO_CAPTURE_PERIOD_SIZE / 4,
+};
+
+
+/*
+ * Local functions
+ */
+
+static void init_once() {
+ list_init(&created_effects_list);
+ list_init(&active_outputs_list);
+
+ pthread_mutex_init(&lock, NULL);
+ pthread_mutex_init(&thread_lock, NULL);
+ pthread_cond_init(&cond, NULL);
+ exit_thread = false;
+ thread_status = -1;
+
+ init_status = 0;
+}
+
+int lib_init() {
+ pthread_once(&once, init_once);
+ return init_status;
+}
+
+bool effect_exists(effect_context_t *context) {
+ struct listnode *node;
+
+ list_for_each(node, &created_effects_list) {
+ effect_context_t *fx_ctxt = node_to_item(node,
+ effect_context_t,
+ effects_list_node);
+ if (fx_ctxt == context) {
+ return true;
+ }
+ }
+ return false;
+}
+
+output_context_t *get_output(audio_io_handle_t output) {
+ struct listnode *node;
+
+ list_for_each(node, &active_outputs_list) {
+ output_context_t *out_ctxt = node_to_item(node,
+ output_context_t,
+ outputs_list_node);
+ if (out_ctxt->handle == output) {
+ return out_ctxt;
+ }
+ }
+ return NULL;
+}
+
+void add_effect_to_output(output_context_t * output, effect_context_t *context) {
+ struct listnode *fx_node;
+
+ list_for_each(fx_node, &output->effects_list) {
+ effect_context_t *fx_ctxt = node_to_item(fx_node,
+ effect_context_t,
+ output_node);
+ if (fx_ctxt == context)
+ return;
+ }
+ list_add_tail(&output->effects_list, &context->output_node);
+ if (context->ops.start)
+ context->ops.start(context, output);
+}
+
+void remove_effect_from_output(output_context_t * output, effect_context_t *context) {
+ struct listnode *fx_node;
+
+ list_for_each(fx_node, &output->effects_list) {
+ effect_context_t *fx_ctxt = node_to_item(fx_node,
+ effect_context_t,
+ output_node);
+ if (fx_ctxt == context) {
+ if (context->ops.stop)
+ context->ops.stop(context, output);
+ list_remove(&context->output_node);
+ return;
+ }
+ }
+}
+
+bool effects_enabled() {
+ struct listnode *out_node;
+
+ list_for_each(out_node, &active_outputs_list) {
+ struct listnode *fx_node;
+ output_context_t *out_ctxt = node_to_item(out_node,
+ output_context_t,
+ outputs_list_node);
+
+ list_for_each(fx_node, &out_ctxt->effects_list) {
+ effect_context_t *fx_ctxt = node_to_item(fx_node,
+ effect_context_t,
+ output_node);
+ if (fx_ctxt->state == EFFECT_STATE_ACTIVE && fx_ctxt->ops.process != NULL)
+ return true;
+ }
+ }
+ return false;
+}
+
+int configure_proxy_capture(struct mixer *mixer, int value) {
+ const char *proxy_ctl_name = "AFE_PCM_RX Audio Mixer MultiMedia4";
+ struct mixer_ctl *ctl;
+
+ ctl = mixer_get_ctl_by_name(mixer, proxy_ctl_name);
+ if (ctl == NULL) {
+ ALOGW("%s: could not get %s ctl", __func__, proxy_ctl_name);
+ return -EINVAL;
+ }
+ if (mixer_ctl_set_value(ctl, 0, value) != 0)
+ ALOGW("%s: error setting value %d on %s ", __func__, value, proxy_ctl_name);
+
+ return 0;
+}
+
+
+void *capture_thread_loop(void *arg)
+{
+ int16_t data[AUDIO_CAPTURE_PERIOD_SIZE * AUDIO_CAPTURE_CHANNEL_COUNT * sizeof(int16_t)];
+ audio_buffer_t buf;
+ buf.frameCount = AUDIO_CAPTURE_PERIOD_SIZE;
+ buf.s16 = data;
+ bool capture_enabled = false;
+ struct mixer *mixer;
+ struct pcm *pcm = NULL;
+ int ret;
+ int retry_num = 0;
+
+ ALOGD("thread enter");
+
+ prctl(PR_SET_NAME, (unsigned long)"visualizer capture", 0, 0, 0);
+
+ pthread_mutex_lock(&lock);
+
+ mixer = mixer_open(MIXER_CARD);
+ while (mixer == NULL && retry_num < RETRY_NUMBER) {
+ usleep(RETRY_US);
+ mixer = mixer_open(MIXER_CARD);
+ retry_num++;
+ }
+ if (mixer == NULL) {
+ pthread_mutex_unlock(&lock);
+ return NULL;
+ }
+
+ for (;;) {
+ if (exit_thread) {
+ break;
+ }
+ if (effects_enabled()) {
+ if (!capture_enabled) {
+ ret = configure_proxy_capture(mixer, 1);
+ if (ret == 0) {
+ pcm = pcm_open(SOUND_CARD, CAPTURE_DEVICE,
+ PCM_IN|PCM_MMAP|PCM_NOIRQ, &pcm_config_capture);
+ if (pcm && !pcm_is_ready(pcm)) {
+ ALOGW("%s: %s", __func__, pcm_get_error(pcm));
+ pcm_close(pcm);
+ pcm = NULL;
+ configure_proxy_capture(mixer, 0);
+ } else {
+ capture_enabled = true;
+ ALOGD("%s: capture ENABLED", __func__);
+ }
+ }
+ }
+ } else {
+ if (capture_enabled) {
+ if (pcm != NULL)
+ pcm_close(pcm);
+ configure_proxy_capture(mixer, 0);
+ ALOGD("%s: capture DISABLED", __func__);
+ capture_enabled = false;
+ }
+ pthread_cond_wait(&cond, &lock);
+ }
+ if (!capture_enabled)
+ continue;
+
+ pthread_mutex_unlock(&lock);
+ ret = pcm_mmap_read(pcm, data, sizeof(data));
+ pthread_mutex_lock(&lock);
+
+ if (ret == 0) {
+ struct listnode *out_node;
+
+ list_for_each(out_node, &active_outputs_list) {
+ output_context_t *out_ctxt = node_to_item(out_node,
+ output_context_t,
+ outputs_list_node);
+ struct listnode *fx_node;
+
+ list_for_each(fx_node, &out_ctxt->effects_list) {
+ effect_context_t *fx_ctxt = node_to_item(fx_node,
+ effect_context_t,
+ output_node);
+ if (fx_ctxt->ops.process != NULL)
+ fx_ctxt->ops.process(fx_ctxt, &buf, &buf);
+ }
+ }
+ } else {
+ ALOGW("%s: read status %d %s", __func__, ret, pcm_get_error(pcm));
+ }
+ }
+
+ if (capture_enabled) {
+ if (pcm != NULL)
+ pcm_close(pcm);
+ configure_proxy_capture(mixer, 0);
+ }
+ mixer_close(mixer);
+ pthread_mutex_unlock(&lock);
+
+ ALOGD("thread exit");
+
+ return NULL;
+}
+
+/*
+ * Interface from audio HAL
+ */
+
+__attribute__ ((visibility ("default")))
+int visualizer_hal_start_output(audio_io_handle_t output, int pcm_id) {
+ int ret;
+ struct listnode *node;
+
+ ALOGV("%s output %d pcm_id %d", __func__, output, pcm_id);
+
+ if (lib_init() != 0)
+ return init_status;
+
+ pthread_mutex_lock(&thread_lock);
+ pthread_mutex_lock(&lock);
+ if (get_output(output) != NULL) {
+ ALOGW("%s output already started", __func__);
+ ret = -ENOSYS;
+ goto exit;
+ }
+
+ output_context_t *out_ctxt = (output_context_t *)malloc(sizeof(output_context_t));
+ if (out_ctxt == NULL) {
+ ALOGE("%s fail to allocate memory", __func__);
+ ret = -ENOMEM;
+ goto exit;
+ }
+ out_ctxt->handle = output;
+ list_init(&out_ctxt->effects_list);
+
+ list_for_each(node, &created_effects_list) {
+ effect_context_t *fx_ctxt = node_to_item(node,
+ effect_context_t,
+ effects_list_node);
+ if (fx_ctxt->out_handle == output) {
+ if (fx_ctxt->ops.start)
+ fx_ctxt->ops.start(fx_ctxt, out_ctxt);
+ list_add_tail(&out_ctxt->effects_list, &fx_ctxt->output_node);
+ }
+ }
+ if (list_empty(&active_outputs_list)) {
+ exit_thread = false;
+ thread_status = pthread_create(&capture_thread, (const pthread_attr_t *) NULL,
+ capture_thread_loop, NULL);
+ }
+ list_add_tail(&active_outputs_list, &out_ctxt->outputs_list_node);
+ pthread_cond_signal(&cond);
+
+exit:
+ pthread_mutex_unlock(&lock);
+ pthread_mutex_unlock(&thread_lock);
+ return ret;
+}
+
+__attribute__ ((visibility ("default")))
+int visualizer_hal_stop_output(audio_io_handle_t output, int pcm_id) {
+ int ret;
+ struct listnode *node;
+ struct listnode *fx_node;
+ output_context_t *out_ctxt;
+
+ ALOGV("%s output %d pcm_id %d", __func__, output, pcm_id);
+
+ if (lib_init() != 0)
+ return init_status;
+
+ pthread_mutex_lock(&thread_lock);
+ pthread_mutex_lock(&lock);
+
+ out_ctxt = get_output(output);
+ if (out_ctxt == NULL) {
+ ALOGW("%s output not started", __func__);
+ ret = -ENOSYS;
+ goto exit;
+ }
+ list_for_each(fx_node, &out_ctxt->effects_list) {
+ effect_context_t *fx_ctxt = node_to_item(fx_node,
+ effect_context_t,
+ output_node);
+ if (fx_ctxt->ops.stop)
+ fx_ctxt->ops.stop(fx_ctxt, out_ctxt);
+ }
+ list_remove(&out_ctxt->outputs_list_node);
+ pthread_cond_signal(&cond);
+
+ if (list_empty(&active_outputs_list)) {
+ if (thread_status == 0) {
+ exit_thread = true;
+ pthread_cond_signal(&cond);
+ pthread_mutex_unlock(&lock);
+ pthread_join(capture_thread, (void **) NULL);
+ pthread_mutex_lock(&lock);
+ thread_status = -1;
+ }
+ }
+
+ free(out_ctxt);
+
+exit:
+ pthread_mutex_unlock(&lock);
+ pthread_mutex_unlock(&thread_lock);
+ return ret;
+}
+
+
+/*
+ * Effect operations
+ */
+
+int set_config(effect_context_t *context, effect_config_t *config)
+{
+ if (config->inputCfg.samplingRate != config->outputCfg.samplingRate) return -EINVAL;
+ if (config->inputCfg.channels != config->outputCfg.channels) return -EINVAL;
+ if (config->inputCfg.format != config->outputCfg.format) return -EINVAL;
+ if (config->inputCfg.channels != AUDIO_CHANNEL_OUT_STEREO) return -EINVAL;
+ if (config->outputCfg.accessMode != EFFECT_BUFFER_ACCESS_WRITE &&
+ config->outputCfg.accessMode != EFFECT_BUFFER_ACCESS_ACCUMULATE) return -EINVAL;
+ if (config->inputCfg.format != AUDIO_FORMAT_PCM_16_BIT) return -EINVAL;
+
+ context->config = *config;
+
+ if (context->ops.reset)
+ context->ops.reset(context);
+
+ return 0;
+}
+
+void get_config(effect_context_t *context, effect_config_t *config)
+{
+ *config = context->config;
+}
+
+
+/*
+ * Visualizer operations
+ */
+
+uint32_t visualizer_get_delta_time_ms_from_updated_time(visualizer_context_t* visu_ctxt) {
+ uint32_t delta_ms = 0;
+ if (visu_ctxt->buffer_update_time.tv_sec != 0) {
+ struct timespec ts;
+ if (clock_gettime(CLOCK_MONOTONIC, &ts) == 0) {
+ time_t secs = ts.tv_sec - visu_ctxt->buffer_update_time.tv_sec;
+ long nsec = ts.tv_nsec - visu_ctxt->buffer_update_time.tv_nsec;
+ if (nsec < 0) {
+ --secs;
+ nsec += 1000000000;
+ }
+ delta_ms = secs * 1000 + nsec / 1000000;
+ }
+ }
+ return delta_ms;
+}
+
+int visualizer_reset(effect_context_t *context)
+{
+ visualizer_context_t * visu_ctxt = (visualizer_context_t *)context;
+
+ visu_ctxt->capture_idx = 0;
+ visu_ctxt->last_capture_idx = 0;
+ visu_ctxt->buffer_update_time.tv_sec = 0;
+ visu_ctxt->latency = DSP_OUTPUT_LATENCY_MS;
+ memset(visu_ctxt->capture_buf, 0x80, CAPTURE_BUF_SIZE);
+ return 0;
+}
+
+int visualizer_init(effect_context_t *context)
+{
+ int32_t i;
+
+ visualizer_context_t * visu_ctxt = (visualizer_context_t *)context;
+
+ context->config.inputCfg.accessMode = EFFECT_BUFFER_ACCESS_READ;
+ context->config.inputCfg.channels = AUDIO_CHANNEL_OUT_STEREO;
+ context->config.inputCfg.format = AUDIO_FORMAT_PCM_16_BIT;
+ context->config.inputCfg.samplingRate = 44100;
+ context->config.inputCfg.bufferProvider.getBuffer = NULL;
+ context->config.inputCfg.bufferProvider.releaseBuffer = NULL;
+ context->config.inputCfg.bufferProvider.cookie = NULL;
+ context->config.inputCfg.mask = EFFECT_CONFIG_ALL;
+ context->config.outputCfg.accessMode = EFFECT_BUFFER_ACCESS_ACCUMULATE;
+ context->config.outputCfg.channels = AUDIO_CHANNEL_OUT_STEREO;
+ context->config.outputCfg.format = AUDIO_FORMAT_PCM_16_BIT;
+ context->config.outputCfg.samplingRate = 44100;
+ context->config.outputCfg.bufferProvider.getBuffer = NULL;
+ context->config.outputCfg.bufferProvider.releaseBuffer = NULL;
+ context->config.outputCfg.bufferProvider.cookie = NULL;
+ context->config.outputCfg.mask = EFFECT_CONFIG_ALL;
+
+ visu_ctxt->capture_size = VISUALIZER_CAPTURE_SIZE_MAX;
+ visu_ctxt->scaling_mode = VISUALIZER_SCALING_MODE_NORMALIZED;
+
+ // measurement initialization
+ visu_ctxt->channel_count = popcount(context->config.inputCfg.channels);
+ visu_ctxt->meas_mode = MEASUREMENT_MODE_NONE;
+ visu_ctxt->meas_wndw_size_in_buffers = MEASUREMENT_WINDOW_MAX_SIZE_IN_BUFFERS;
+ visu_ctxt->meas_buffer_idx = 0;
+ for (i=0 ; i<visu_ctxt->meas_wndw_size_in_buffers ; i++) {
+ visu_ctxt->past_meas[i].is_valid = false;
+ visu_ctxt->past_meas[i].peak_u16 = 0;
+ visu_ctxt->past_meas[i].rms_squared = 0;
+ }
+
+ set_config(context, &context->config);
+
+ return 0;
+}
+
+int visualizer_get_parameter(effect_context_t *context, effect_param_t *p, uint32_t *size)
+{
+ visualizer_context_t *visu_ctxt = (visualizer_context_t *)context;
+
+ p->status = 0;
+ *size = sizeof(effect_param_t) + sizeof(uint32_t);
+ if (p->psize != sizeof(uint32_t)) {
+ p->status = -EINVAL;
+ return 0;
+ }
+ switch (*(uint32_t *)p->data) {
+ case VISUALIZER_PARAM_CAPTURE_SIZE:
+ ALOGV("%s get capture_size = %d", __func__, visu_ctxt->capture_size);
+ *((uint32_t *)p->data + 1) = visu_ctxt->capture_size;
+ p->vsize = sizeof(uint32_t);
+ *size += sizeof(uint32_t);
+ break;
+ case VISUALIZER_PARAM_SCALING_MODE:
+ ALOGV("%s get scaling_mode = %d", __func__, visu_ctxt->scaling_mode);
+ *((uint32_t *)p->data + 1) = visu_ctxt->scaling_mode;
+ p->vsize = sizeof(uint32_t);
+ *size += sizeof(uint32_t);
+ break;
+ case VISUALIZER_PARAM_MEASUREMENT_MODE:
+ ALOGV("%s get meas_mode = %d", __func__, visu_ctxt->meas_mode);
+ *((uint32_t *)p->data + 1) = visu_ctxt->meas_mode;
+ p->vsize = sizeof(uint32_t);
+ *size += sizeof(uint32_t);
+ break;
+ default:
+ p->status = -EINVAL;
+ }
+ return 0;
+}
+
+int visualizer_set_parameter(effect_context_t *context, effect_param_t *p, uint32_t size)
+{
+ visualizer_context_t *visu_ctxt = (visualizer_context_t *)context;
+
+ if (p->psize != sizeof(uint32_t) || p->vsize != sizeof(uint32_t))
+ return -EINVAL;
+
+ switch (*(uint32_t *)p->data) {
+ case VISUALIZER_PARAM_CAPTURE_SIZE:
+ visu_ctxt->capture_size = *((uint32_t *)p->data + 1);
+ ALOGV("%s set capture_size = %d", __func__, visu_ctxt->capture_size);
+ break;
+ case VISUALIZER_PARAM_SCALING_MODE:
+ visu_ctxt->scaling_mode = *((uint32_t *)p->data + 1);
+ ALOGV("%s set scaling_mode = %d", __func__, visu_ctxt->scaling_mode);
+ break;
+ case VISUALIZER_PARAM_LATENCY:
+ /* Ignore latency as we capture at DSP output
+ * visu_ctxt->latency = *((uint32_t *)p->data + 1); */
+ ALOGV("%s set latency = %d", __func__, visu_ctxt->latency);
+ break;
+ case VISUALIZER_PARAM_MEASUREMENT_MODE:
+ visu_ctxt->meas_mode = *((uint32_t *)p->data + 1);
+ ALOGV("%s set meas_mode = %d", __func__, visu_ctxt->meas_mode);
+ break;
+ default:
+ return -EINVAL;
+ }
+ return 0;
+}
+
+/* Real process function called from capture thread. Called with lock held */
+int visualizer_process(effect_context_t *context,
+ audio_buffer_t *inBuffer,
+ audio_buffer_t *outBuffer)
+{
+ visualizer_context_t *visu_ctxt = (visualizer_context_t *)context;
+
+ if (!effect_exists(context))
+ return -EINVAL;
+
+ if (inBuffer == NULL || inBuffer->raw == NULL ||
+ outBuffer == NULL || outBuffer->raw == NULL ||
+ inBuffer->frameCount != outBuffer->frameCount ||
+ inBuffer->frameCount == 0) {
+ return -EINVAL;
+ }
+
+ // perform measurements if needed
+ if (visu_ctxt->meas_mode & MEASUREMENT_MODE_PEAK_RMS) {
+ // find the peak and RMS squared for the new buffer
+ uint32_t inIdx;
+ int16_t max_sample = 0;
+ float rms_squared_acc = 0;
+ for (inIdx = 0 ; inIdx < inBuffer->frameCount * visu_ctxt->channel_count ; inIdx++) {
+ if (inBuffer->s16[inIdx] > max_sample) {
+ max_sample = inBuffer->s16[inIdx];
+ } else if (-inBuffer->s16[inIdx] > max_sample) {
+ max_sample = -inBuffer->s16[inIdx];
+ }
+ rms_squared_acc += (inBuffer->s16[inIdx] * inBuffer->s16[inIdx]);
+ }
+ // store the measurement
+ visu_ctxt->past_meas[visu_ctxt->meas_buffer_idx].peak_u16 = (uint16_t)max_sample;
+ visu_ctxt->past_meas[visu_ctxt->meas_buffer_idx].rms_squared =
+ rms_squared_acc / (inBuffer->frameCount * visu_ctxt->channel_count);
+ visu_ctxt->past_meas[visu_ctxt->meas_buffer_idx].is_valid = true;
+ if (++visu_ctxt->meas_buffer_idx >= visu_ctxt->meas_wndw_size_in_buffers) {
+ visu_ctxt->meas_buffer_idx = 0;
+ }
+ }
+
+ /* all code below assumes stereo 16 bit PCM output and input */
+ int32_t shift;
+
+ if (visu_ctxt->scaling_mode == VISUALIZER_SCALING_MODE_NORMALIZED) {
+ /* derive capture scaling factor from peak value in current buffer
+ * this gives more interesting captures for display. */
+ shift = 32;
+ int len = inBuffer->frameCount * 2;
+ int i;
+ for (i = 0; i < len; i++) {
+ int32_t smp = inBuffer->s16[i];
+ if (smp < 0) smp = -smp - 1; /* take care to keep the max negative in range */
+ int32_t clz = __builtin_clz(smp);
+ if (shift > clz) shift = clz;
+ }
+ /* A maximum amplitude signal will have 17 leading zeros, which we want to
+ * translate to a shift of 8 (for converting 16 bit to 8 bit) */
+ shift = 25 - shift;
+ /* Never scale by less than 8 to avoid returning unaltered PCM signal. */
+ if (shift < 3) {
+ shift = 3;
+ }
+ /* add one to combine the division by 2 needed after summing
+ * left and right channels below */
+ shift++;
+ } else {
+ assert(visu_ctxt->scaling_mode == VISUALIZER_SCALING_MODE_AS_PLAYED);
+ shift = 9;
+ }
+
+ uint32_t capt_idx;
+ uint32_t in_idx;
+ uint8_t *buf = visu_ctxt->capture_buf;
+ for (in_idx = 0, capt_idx = visu_ctxt->capture_idx;
+ in_idx < inBuffer->frameCount;
+ in_idx++, capt_idx++) {
+ if (capt_idx >= CAPTURE_BUF_SIZE) {
+ /* wrap around */
+ capt_idx = 0;
+ }
+ int32_t smp = inBuffer->s16[2 * in_idx] + inBuffer->s16[2 * in_idx + 1];
+ smp = smp >> shift;
+ buf[capt_idx] = ((uint8_t)smp)^0x80;
+ }
+
+ /* XXX the following two should really be atomic, though it probably doesn't
+ * matter much for visualization purposes */
+ visu_ctxt->capture_idx = capt_idx;
+ /* update last buffer update time stamp */
+ if (clock_gettime(CLOCK_MONOTONIC, &visu_ctxt->buffer_update_time) < 0) {
+ visu_ctxt->buffer_update_time.tv_sec = 0;
+ }
+
+ if (context->state != EFFECT_STATE_ACTIVE) {
+ ALOGV("%s DONE inactive", __func__);
+ return -ENODATA;
+ }
+
+ return 0;
+}
+
+int visualizer_command(effect_context_t * context, uint32_t cmdCode, uint32_t cmdSize,
+ void *pCmdData, uint32_t *replySize, void *pReplyData)
+{
+ visualizer_context_t * visu_ctxt = (visualizer_context_t *)context;
+
+ switch (cmdCode) {
+ case VISUALIZER_CMD_CAPTURE:
+ if (pReplyData == NULL || *replySize != visu_ctxt->capture_size) {
+ ALOGV("%s VISUALIZER_CMD_CAPTURE error *replySize %d context->capture_size %d",
+ __func__, *replySize, visu_ctxt->capture_size);
+ return -EINVAL;
+ }
+
+ if (!context->offload_enabled)
+ break;
+
+ if (context->state == EFFECT_STATE_ACTIVE) {
+ int32_t latency_ms = visu_ctxt->latency;
+ const uint32_t delta_ms = visualizer_get_delta_time_ms_from_updated_time(visu_ctxt);
+ latency_ms -= delta_ms;
+ if (latency_ms < 0) {
+ latency_ms = 0;
+ }
+ const uint32_t delta_smp = context->config.inputCfg.samplingRate * latency_ms / 1000;
+
+ int32_t capture_point = visu_ctxt->capture_idx - visu_ctxt->capture_size - delta_smp;
+ int32_t capture_size = visu_ctxt->capture_size;
+ if (capture_point < 0) {
+ int32_t size = -capture_point;
+ if (size > capture_size)
+ size = capture_size;
+
+ memcpy(pReplyData,
+ visu_ctxt->capture_buf + CAPTURE_BUF_SIZE + capture_point,
+ size);
+ pReplyData = (void *)((size_t)pReplyData + size);
+ capture_size -= size;
+ capture_point = 0;
+ }
+ memcpy(pReplyData,
+ visu_ctxt->capture_buf + capture_point,
+ capture_size);
+
+
+ /* if audio framework has stopped playing audio although the effect is still
+ * active we must clear the capture buffer to return silence */
+ if ((visu_ctxt->last_capture_idx == visu_ctxt->capture_idx) &&
+ (visu_ctxt->buffer_update_time.tv_sec != 0)) {
+ if (delta_ms > MAX_STALL_TIME_MS) {
+ ALOGV("%s capture going to idle", __func__);
+ visu_ctxt->buffer_update_time.tv_sec = 0;
+ memset(pReplyData, 0x80, visu_ctxt->capture_size);
+ }
+ }
+ visu_ctxt->last_capture_idx = visu_ctxt->capture_idx;
+ } else {
+ memset(pReplyData, 0x80, visu_ctxt->capture_size);
+ }
+ break;
+
+ case VISUALIZER_CMD_MEASURE: {
+ uint16_t peak_u16 = 0;
+ float sum_rms_squared = 0.0f;
+ uint8_t nb_valid_meas = 0;
+ /* reset measurements if last measurement was too long ago (which implies stored
+ * measurements aren't relevant anymore and shouldn't bias the new one) */
+ const int32_t delay_ms = visualizer_get_delta_time_ms_from_updated_time(visu_ctxt);
+ if (delay_ms > DISCARD_MEASUREMENTS_TIME_MS) {
+ uint32_t i;
+ ALOGV("Discarding measurements, last measurement is %dms old", delay_ms);
+ for (i=0 ; i<visu_ctxt->meas_wndw_size_in_buffers ; i++) {
+ visu_ctxt->past_meas[i].is_valid = false;
+ visu_ctxt->past_meas[i].peak_u16 = 0;
+ visu_ctxt->past_meas[i].rms_squared = 0;
+ }
+ visu_ctxt->meas_buffer_idx = 0;
+ } else {
+ /* only use actual measurements, otherwise the first RMS measure happening before
+ * MEASUREMENT_WINDOW_MAX_SIZE_IN_BUFFERS have been played will always be artificially
+ * low */
+ uint32_t i;
+ for (i=0 ; i < visu_ctxt->meas_wndw_size_in_buffers ; i++) {
+ if (visu_ctxt->past_meas[i].is_valid) {
+ if (visu_ctxt->past_meas[i].peak_u16 > peak_u16) {
+ peak_u16 = visu_ctxt->past_meas[i].peak_u16;
+ }
+ sum_rms_squared += visu_ctxt->past_meas[i].rms_squared;
+ nb_valid_meas++;
+ }
+ }
+ }
+ float rms = nb_valid_meas == 0 ? 0.0f : sqrtf(sum_rms_squared / nb_valid_meas);
+ int32_t* p_int_reply_data = (int32_t*)pReplyData;
+ /* convert from I16 sample values to mB and write results */
+ if (rms < 0.000016f) {
+ p_int_reply_data[MEASUREMENT_IDX_RMS] = -9600; //-96dB
+ } else {
+ p_int_reply_data[MEASUREMENT_IDX_RMS] = (int32_t) (2000 * log10(rms / 32767.0f));
+ }
+ if (peak_u16 == 0) {
+ p_int_reply_data[MEASUREMENT_IDX_PEAK] = -9600; //-96dB
+ } else {
+ p_int_reply_data[MEASUREMENT_IDX_PEAK] = (int32_t) (2000 * log10(peak_u16 / 32767.0f));
+ }
+ ALOGV("VISUALIZER_CMD_MEASURE peak=%d (%dmB), rms=%.1f (%dmB)",
+ peak_u16, p_int_reply_data[MEASUREMENT_IDX_PEAK],
+ rms, p_int_reply_data[MEASUREMENT_IDX_RMS]);
+ }
+ break;
+
+ default:
+ ALOGW("%s invalid command %d", __func__, cmdCode);
+ return -EINVAL;
+ }
+ return 0;
+}
+
+
+/*
+ * Effect Library Interface Implementation
+ */
+
+int effect_lib_create(const effect_uuid_t *uuid,
+ int32_t sessionId,
+ int32_t ioId,
+ effect_handle_t *pHandle) {
+ int ret;
+ int i;
+
+ if (lib_init() != 0)
+ return init_status;
+
+ if (pHandle == NULL || uuid == NULL)
+ return -EINVAL;
+
+ for (i = 0; descriptors[i] != NULL; i++) {
+ if (memcmp(uuid, &descriptors[i]->uuid, sizeof(effect_uuid_t)) == 0)
+ break;
+ }
+
+ if (descriptors[i] == NULL)
+ return -EINVAL;
+
+ effect_context_t *context;
+ if (memcmp(uuid, &visualizer_descriptor.uuid, sizeof(effect_uuid_t)) == 0) {
+ visualizer_context_t *visu_ctxt = (visualizer_context_t *)calloc(1,
+ sizeof(visualizer_context_t));
+ if (visu_ctxt == NULL) {
+ ALOGE("%s fail to allocate memory", __func__);
+ return -ENOMEM;
+ }
+ context = (effect_context_t *)visu_ctxt;
+ context->ops.init = visualizer_init;
+ context->ops.reset = visualizer_reset;
+ context->ops.process = visualizer_process;
+ context->ops.set_parameter = visualizer_set_parameter;
+ context->ops.get_parameter = visualizer_get_parameter;
+ context->ops.command = visualizer_command;
+ context->desc = &visualizer_descriptor;
+ } else {
+ return -EINVAL;
+ }
+
+ context->itfe = &effect_interface;
+ context->state = EFFECT_STATE_UNINITIALIZED;
+ context->out_handle = (audio_io_handle_t)ioId;
+
+ ret = context->ops.init(context);
+ if (ret < 0) {
+ ALOGW("%s init failed", __func__);
+ free(context);
+ return ret;
+ }
+
+ context->state = EFFECT_STATE_INITIALIZED;
+
+ pthread_mutex_lock(&lock);
+ list_add_tail(&created_effects_list, &context->effects_list_node);
+ output_context_t *out_ctxt = get_output(ioId);
+ if (out_ctxt != NULL)
+ add_effect_to_output(out_ctxt, context);
+ pthread_mutex_unlock(&lock);
+
+ *pHandle = (effect_handle_t)context;
+
+ ALOGV("%s created context %p", __func__, context);
+
+ return 0;
+
+}
+
+int effect_lib_release(effect_handle_t handle) {
+ effect_context_t *context = (effect_context_t *)handle;
+ int status;
+
+ if (lib_init() != 0)
+ return init_status;
+
+ ALOGV("%s context %p", __func__, handle);
+ pthread_mutex_lock(&lock);
+ status = -EINVAL;
+ if (effect_exists(context)) {
+ output_context_t *out_ctxt = get_output(context->out_handle);
+ if (out_ctxt != NULL)
+ remove_effect_from_output(out_ctxt, context);
+ list_remove(&context->effects_list_node);
+ if (context->ops.release)
+ context->ops.release(context);
+ free(context);
+ status = 0;
+ }
+ pthread_mutex_unlock(&lock);
+
+ return status;
+}
+
+int effect_lib_get_descriptor(const effect_uuid_t *uuid,
+ effect_descriptor_t *descriptor) {
+ int i;
+
+ if (lib_init() != 0)
+ return init_status;
+
+ if (descriptor == NULL || uuid == NULL) {
+ ALOGV("%s called with NULL pointer", __func__);
+ return -EINVAL;
+ }
+
+ for (i = 0; descriptors[i] != NULL; i++) {
+ if (memcmp(uuid, &descriptors[i]->uuid, sizeof(effect_uuid_t)) == 0) {
+ *descriptor = *descriptors[i];
+ return 0;
+ }
+ }
+
+ return -EINVAL;
+}
+
+/*
+ * Effect Control Interface Implementation
+ */
+
+ /* Stub function for effect interface: never called for offloaded effects */
+int effect_process(effect_handle_t self,
+ audio_buffer_t *inBuffer,
+ audio_buffer_t *outBuffer)
+{
+ effect_context_t * context = (effect_context_t *)self;
+ int status = 0;
+
+ ALOGW("%s Called ?????", __func__);
+
+ pthread_mutex_lock(&lock);
+ if (!effect_exists(context)) {
+ status = -EINVAL;
+ goto exit;
+ }
+
+ if (context->state != EFFECT_STATE_ACTIVE) {
+ status = -EINVAL;
+ goto exit;
+ }
+
+exit:
+ pthread_mutex_unlock(&lock);
+ return status;
+}
+
+int effect_command(effect_handle_t self, uint32_t cmdCode, uint32_t cmdSize,
+ void *pCmdData, uint32_t *replySize, void *pReplyData)
+{
+
+ effect_context_t * context = (effect_context_t *)self;
+ int retsize;
+ int status = 0;
+
+ pthread_mutex_lock(&lock);
+
+ if (!effect_exists(context)) {
+ status = -EINVAL;
+ goto exit;
+ }
+
+ if (context == NULL || context->state == EFFECT_STATE_UNINITIALIZED) {
+ status = -EINVAL;
+ goto exit;
+ }
+
+// ALOGV_IF(cmdCode != VISUALIZER_CMD_CAPTURE,
+// "%s command %d cmdSize %d", __func__, cmdCode, cmdSize);
+
+ switch (cmdCode) {
+ case EFFECT_CMD_INIT:
+ if (pReplyData == NULL || *replySize != sizeof(int)) {
+ status = -EINVAL;
+ goto exit;
+ }
+ if (context->ops.init)
+ *(int *) pReplyData = context->ops.init(context);
+ else
+ *(int *) pReplyData = 0;
+ break;
+ case EFFECT_CMD_SET_CONFIG:
+ if (pCmdData == NULL || cmdSize != sizeof(effect_config_t)
+ || pReplyData == NULL || *replySize != sizeof(int)) {
+ status = -EINVAL;
+ goto exit;
+ }
+ *(int *) pReplyData = set_config(context, (effect_config_t *) pCmdData);
+ break;
+ case EFFECT_CMD_GET_CONFIG:
+ if (pReplyData == NULL ||
+ *replySize != sizeof(effect_config_t)) {
+ status = -EINVAL;
+ goto exit;
+ }
+ if (!context->offload_enabled) {
+ status = -EINVAL;
+ goto exit;
+ }
+
+ get_config(context, (effect_config_t *)pReplyData);
+ break;
+ case EFFECT_CMD_RESET:
+ if (context->ops.reset)
+ context->ops.reset(context);
+ break;
+ case EFFECT_CMD_ENABLE:
+ if (pReplyData == NULL || *replySize != sizeof(int)) {
+ status = -EINVAL;
+ goto exit;
+ }
+ if (context->state != EFFECT_STATE_INITIALIZED) {
+ status = -ENOSYS;
+ goto exit;
+ }
+ context->state = EFFECT_STATE_ACTIVE;
+ if (context->ops.enable)
+ context->ops.enable(context);
+ pthread_cond_signal(&cond);
+ ALOGV("%s EFFECT_CMD_ENABLE", __func__);
+ *(int *)pReplyData = 0;
+ break;
+ case EFFECT_CMD_DISABLE:
+ if (pReplyData == NULL || *replySize != sizeof(int)) {
+ status = -EINVAL;
+ goto exit;
+ }
+ if (context->state != EFFECT_STATE_ACTIVE) {
+ status = -ENOSYS;
+ goto exit;
+ }
+ context->state = EFFECT_STATE_INITIALIZED;
+ if (context->ops.disable)
+ context->ops.disable(context);
+ pthread_cond_signal(&cond);
+ ALOGV("%s EFFECT_CMD_DISABLE", __func__);
+ *(int *)pReplyData = 0;
+ break;
+ case EFFECT_CMD_GET_PARAM: {
+ if (pCmdData == NULL ||
+ cmdSize != (int)(sizeof(effect_param_t) + sizeof(uint32_t)) ||
+ pReplyData == NULL ||
+ *replySize < (int)(sizeof(effect_param_t) + sizeof(uint32_t) + sizeof(uint32_t))) {
+ status = -EINVAL;
+ goto exit;
+ }
+ if (!context->offload_enabled) {
+ status = -EINVAL;
+ goto exit;
+ }
+ memcpy(pReplyData, pCmdData, sizeof(effect_param_t) + sizeof(uint32_t));
+ effect_param_t *p = (effect_param_t *)pReplyData;
+ if (context->ops.get_parameter)
+ context->ops.get_parameter(context, p, replySize);
+ } break;
+ case EFFECT_CMD_SET_PARAM: {
+ if (pCmdData == NULL ||
+ cmdSize != (int)(sizeof(effect_param_t) + sizeof(uint32_t) + sizeof(uint32_t)) ||
+ pReplyData == NULL || *replySize != sizeof(int32_t)) {
+ status = -EINVAL;
+ goto exit;
+ }
+ *(int32_t *)pReplyData = 0;
+ effect_param_t *p = (effect_param_t *)pCmdData;
+ if (context->ops.set_parameter)
+ *(int32_t *)pReplyData = context->ops.set_parameter(context, p, *replySize);
+
+ } break;
+ case EFFECT_CMD_SET_DEVICE:
+ case EFFECT_CMD_SET_VOLUME:
+ case EFFECT_CMD_SET_AUDIO_MODE:
+ break;
+
+ case EFFECT_CMD_OFFLOAD: {
+ output_context_t *out_ctxt;
+
+ if (cmdSize != sizeof(effect_offload_param_t) || pCmdData == NULL
+ || pReplyData == NULL || *replySize != sizeof(int)) {
+ ALOGV("%s EFFECT_CMD_OFFLOAD bad format", __func__);
+ status = -EINVAL;
+ break;
+ }
+
+ effect_offload_param_t* offload_param = (effect_offload_param_t*)pCmdData;
+
+ ALOGV("%s EFFECT_CMD_OFFLOAD offload %d output %d",
+ __func__, offload_param->isOffload, offload_param->ioHandle);
+
+ *(int *)pReplyData = 0;
+
+ context->offload_enabled = offload_param->isOffload;
+ if (context->out_handle == offload_param->ioHandle)
+ break;
+
+ out_ctxt = get_output(context->out_handle);
+ if (out_ctxt != NULL)
+ remove_effect_from_output(out_ctxt, context);
+
+ context->out_handle = offload_param->ioHandle;
+ out_ctxt = get_output(offload_param->ioHandle);
+ if (out_ctxt != NULL)
+ add_effect_to_output(out_ctxt, context);
+
+ } break;
+
+
+ default:
+ if (cmdCode >= EFFECT_CMD_FIRST_PROPRIETARY && context->ops.command)
+ status = context->ops.command(context, cmdCode, cmdSize,
+ pCmdData, replySize, pReplyData);
+ else {
+ ALOGW("%s invalid command %d", __func__, cmdCode);
+ status = -EINVAL;
+ }
+ break;
+ }
+
+exit:
+ pthread_mutex_unlock(&lock);
+
+// ALOGV_IF(cmdCode != VISUALIZER_CMD_CAPTURE,"%s DONE", __func__);
+ return status;
+}
+
+/* Effect Control Interface Implementation: get_descriptor */
+int effect_get_descriptor(effect_handle_t self,
+ effect_descriptor_t *descriptor)
+{
+ effect_context_t *context = (effect_context_t *)self;
+
+ if (!effect_exists(context))
+ return -EINVAL;
+
+ if (descriptor == NULL)
+ return -EINVAL;
+
+ *descriptor = *context->desc;
+
+ return 0;
+}
+
+/* effect_handle_t interface implementation for visualizer effect */
+const struct effect_interface_s effect_interface = {
+ effect_process,
+ effect_command,
+ effect_get_descriptor,
+ NULL,
+};
+
+__attribute__ ((visibility ("default")))
+audio_effect_library_t AUDIO_EFFECT_LIBRARY_INFO_SYM = {
+ tag : AUDIO_EFFECT_LIBRARY_TAG,
+ version : EFFECT_LIBRARY_API_VERSION,
+ name : "Visualizer Library",
+ implementor : "The Android Open Source Project",
+ create_effect : effect_lib_create,
+ release_effect : effect_lib_release,
+ get_descriptor : effect_lib_get_descriptor,
+};