STM32F407 SPI reads wrong data

void spi_init(void) {

      hspi.Instance = SPI1;

      hspi.Init.Mode = SPI_MODE_MASTER;

      hspi.Init.Direction = SPI_DIRECTION_2LINES;

      hspi.Init.DataSize = SPI_DATASIZE_8BIT;

      hspi.Init.CLKPolarity = SPI_POLARITY_LOW;

      hspi.Init.CLKPhase = SPI_PHASE_1EDGE;

      hspi.Init.NSS = SPI_NSS_SOFT;

      hspi.Init.BaudRatePrescaler = SPI_BAUDRATEPRESCALER_256;

      hspi.Init.FirstBit = SPI_FIRSTBIT_MSB;

      hspi.Init.TIMode = SPI_TIMODE_DISABLE;

      hspi.Init.CRCCalculation = SPI_CRCCALCULATION_DISABLE;

      hspi.Init.CRCPolynomial = 10;

      HAL_SPI_Init(&hspi);

}

uint8_t transceive_byte(uint8_t data) {

    uint8_t recv = 0x00;

    HAL_SPI_TransmitReceive(&hspi, &data, &recv, 1, HAL_MAX_DELAY);

    return recv;

}

uint32_t transceive_packet(uint8_t mode, uint8_t addr, uint8_t* data) {

    HAL_GPIO_WritePin(GPIO_IMU_SS_GPIO_Port, GPIO_IMU_SS_Pin, GPIO_PIN_RESET);

    delay(10); // delay is in uS

    transceive_byte(mode);

    delay(20);

    transceive_byte(addr);

    uint32_t recv_data = 0;

    for(int i = 0; i < 4; i++) {

        delay(20);

        recv_data |= transceive_byte(data[i]) << (i * 8);

    }

    HAL_GPIO_WritePin(GPIO_IMU_SS_GPIO_Port, GPIO_IMU_SS_Pin, GPIO_PIN_SET);

    return recv_data;

}

void imu_spi_test() {

    delay_init();

    spi_init();

    uint32_t cmd = 0;

    char fw_version[5] = {0};

    *((uint32_t*)fw_version)= transceive_packet(CMD_WRITE, 0xAA, &cmd);

    trace("FW: %s\n", fw_version);

}