STM32F4: sending data to 2 chained 74HCT595 shift registers

main.c

#include <stm32f4xx.h>
#include "stm32f4xx_conf.h"
#include "stm32f4_discovery.h"
#include "math.h"
#include "arm_math.h"

uint16_t counter = 0;

void sendCounter();

__attribute__ ((noreturn)) void main(void) {
    //enable FPU
    SCB->CPACR |= 0b000000000111100000000000000000000UL;

    STM_EVAL_LEDInit(LED3);
    STM_EVAL_LEDInit(LED4);
    STM_EVAL_LEDInit(LED5);
    STM_EVAL_LEDInit(LED6);
    STM_EVAL_LEDOff(LED3);
    STM_EVAL_LEDOff(LED4);
    STM_EVAL_LEDOff(LED5);
    STM_EVAL_LEDOff(LED6);

    RCC_APB1PeriphClockCmd(RCC_APB1Periph_SPI2, ENABLE);
    RCC_AHB1PeriphClockCmd(RCC_AHB1Periph_GPIOB, ENABLE);

    GPIO_PinAFConfig(GPIOB, GPIO_PinSource12, GPIO_AF_SPI2);
    GPIO_PinAFConfig(GPIOB, GPIO_PinSource13, GPIO_AF_SPI2);
    GPIO_PinAFConfig(GPIOB, GPIO_PinSource14, GPIO_AF_SPI2);
    GPIO_PinAFConfig(GPIOB, GPIO_PinSource15, GPIO_AF_SPI2);

    GPIO_Init(GPIOB, &(GPIO_InitTypeDef) {
            .GPIO_Pin =  GPIO_Pin_12 | GPIO_Pin_13 | GPIO_Pin_15,
            .GPIO_Mode = GPIO_Mode_AF,
            .GPIO_Speed = GPIO_Speed_50MHz,
            .GPIO_OType = GPIO_OType_PP,
            .GPIO_PuPd = GPIO_PuPd_DOWN});
    GPIO_Init(GPIOB, &(GPIO_InitTypeDef) {
            .GPIO_Pin =GPIO_Pin_14,
            .GPIO_Mode = GPIO_Mode_AF,
            .GPIO_Speed = GPIO_Speed_50MHz,
            .GPIO_OType = GPIO_OType_OD});
    GPIO_Init(GPIOB, &(GPIO_InitTypeDef) {
            .GPIO_Pin = GPIO_Pin_11,
            .GPIO_Mode = GPIO_Mode_OUT,
            .GPIO_Speed = GPIO_Speed_50MHz,
            .GPIO_OType = GPIO_OType_PP});
    SPI_I2S_DeInit(SPI2);
    SPI_Init(SPI2, &(SPI_InitTypeDef) {
            .SPI_Mode = SPI_Mode_Master,
            .SPI_Direction = SPI_Direction_1Line_Tx,
            .SPI_DataSize = SPI_DataSize_8b,
            .SPI_CPOL = SPI_CPOL_Low,
            .SPI_CPHA = SPI_CPHA_1Edge,
            .SPI_NSS = SPI_NSS_Hard,
            .SPI_BaudRatePrescaler = SPI_BaudRatePrescaler_128,
            .SPI_FirstBit = SPI_FirstBit_MSB,
            .SPI_CRCPolynomial = 7
    });
    SPI_SSOutputCmd(SPI2, ENABLE);
    SPI_TIModeCmd(SPI2, DISABLE);
    SPI_Cmd(SPI2, ENABLE);
    STM_EVAL_PBInit(BUTTON_USER, BUTTON_MODE_EXTI);

#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wmissing-noreturn"
    while (1) {
        counter++;
        sendCounter();
        for (uint32_t i = 0; i < 1000000; i++)
                __asm__("");
    }
#pragma clang diagnostic pop
}

void EXTI0_IRQHandler(void) {
    if (EXTI_GetITStatus(EXTI_Line0) != RESET) {
        ++counter;
        STM_EVAL_LEDToggle(LED4);
        sendCounter();
        EXTI_ClearITPendingBit(EXTI_Line0);
    }
}

void sendCounter() {
    GPIO_ResetBits(GPIOB, GPIO_Pin_11);
    SPI_I2S_SendData(SPI2, counter >> 8);
    while ((SPI2->SR & SPI_SR_TXE) == 0)
            __asm__("");
    SPI_I2S_SendData(SPI2, counter);
    while ((SPI2->SR & SPI_SR_TXE) == 0)
            __asm__("");
    while (SPI2->SR & SPI_SR_BSY)
            __asm__("");
    GPIO_SetBits(GPIOB, GPIO_Pin_11);
}