AudioEffect JNI: use new max preprocessing constant

android_media_AudioEffect_native_queryPreProcessings(JNIEnv *env, jclass clazz __unused,

                                                     jint audioSession)

{

    // kDefaultNumEffects is a "reasonable" value ensuring that only one query will be enough on

    // most devices to get all active audio pre processing on a given session.

    static const uint32_t kDefaultNumEffects = 5;

    effect_descriptor_t *descriptors = new effect_descriptor_t[kDefaultNumEffects];

    uint32_t numEffects = kDefaultNumEffects;

    effect_descriptor_t *descriptors = new effect_descriptor_t[AudioEffect::kMaxPreProcessing];

    uint32_t numEffects = AudioEffect::kMaxPreProcessing;

    status_t status = AudioEffect::queryDefaultPreProcessing(audioSession,

                                           descriptors,

                                           &numEffects);

    if ((status != NO_ERROR && status != NO_MEMORY) ||

            numEffects == 0) {

        delete[] descriptors;

        return NULL;

    }

    if (status == NO_MEMORY) {

        delete [] descriptors;

        descriptors = new effect_descriptor_t[numEffects];

        status = AudioEffect::queryDefaultPreProcessing(audioSession,

                                               descriptors,

                                               &numEffects);

    }

    if (status != NO_ERROR || numEffects == 0) {

        delete[] descriptors;

        return NULL;