CubeMX 4.24 and 4.25 issues

[Markus Schwarzl]

Posted on February 19, 2018 at 10:22

Dear Team,

I've migrated my CubeMX project for a STM32F412 MCU today and there are some changes which I would classify as bug.

1) I've configured TIM6 as System-Timer (Timebase-Source) in the SYS-category. Since V4.24.0 and still in V2.25.0 the Interrupt functions are not generated anymore for TIM6!?

2) I've configured TIM2 for Input Capture on Ch1 with an 'Input Filter (4 bits value)' of 15. Since version V4.24 and still in V4.25.0 the generated code contains 0 instead of 15 in the init-code.

There are also 2 warnings in my SystemWorkbench project, which were already there in older versions:

'No return, in function returning non-void' twice in queue.c (line 723 and 1237). I've FreeRtos enabled.

Since it is a customer project I cannot attach my *.ioc file.

Best regards,

Markus

[Khouloud GARSI]

Posted on April 19, 2018 at 17:11

Hello

Schwarzl.Markus

,

This is reported internally for further check. We will get back to you with any update on this point ASAP.

Best regards,

Khouloud.

[Nawres GHARBI]

Posted on April 23, 2018 at 16:29

Hi Markus,

Could you please detail the migration steps you did (did you type on migrate or continue ? )

If possible could you create another ioc with the 4.24 with a basic configuration which with we can reproduce this issue. 

Thanks

[Markus Schwarzl]

Posted on April 23, 2018 at 17:18

Dear Nawres,

yes, I clicked on migrate!

I just tried to reproduce everything in a small configuration...

...but I failed to reproduce issue 1 there. The ISR is correctly generated. But I compared the NVIC-configurations of both ioc-files and found out that in the test-project the ''Enable'' checkbox is checked but in my project it is not... but since it is disabled (greyed out) I cannot change it mayself.

NVIC-settings of test-project:

vs. NVIC-settings of my project:

...for issue 2 I made a closer compare and found out that ''sConfigIC.ICFilter = 15;'' is set correctly but ''sSlaveConfig.TriggerFilter = 0;'' has changed to 0 with an upgrade. In older version both were set to the same, since no aditional field is available in Cube, but I'm not sure if both have to be set. I just noticed the change with the upgrade some time ago.

But I see another issue with newly enabled TIM7 on this micro... can you confirm that the init-code tries to configure MasterConfigSynchronization and that call fails in an assert_param!?

/* TIM7 init function */

void MX_TIM7_Init(void){  TIM_MasterConfigTypeDef sMasterConfig;

  htim7.Instance = TIM7;

  htim7.Init.Prescaler = 9999;  htim7.Init.CounterMode = TIM_COUNTERMODE_UP;  htim7.Init.Period = 65535;  if (HAL_TIM_Base_Init(&htim7) != HAL_OK)  {    _Error_Handler(__FILE__, __LINE__);  }

  sMasterConfig.MasterOutputTrigger = TIM_TRGO_RESET;

  sMasterConfig.MasterSlaveMode = TIM_MASTERSLAVEMODE_DISABLE;  if (HAL_TIMEx_MasterConfigSynchronization(&htim7, &sMasterConfig) != HAL_OK)  {    _Error_Handler(__FILE__, __LINE__);  }

}

Thanks for your support!

Markus

[Nawres GHARBI]

Posted on May 03, 2018 at 15:09

Hi

Schwarzl.Markus

‌

Thanks for attaching the ioc

For Issue 1 I unfortunately could not reproduce, Could you attach the original ioc made by the 4.24

Issue 2 will be fixed next release

Thanks for your interest in STM32CubeMX

[Markus Schwarzl]

Posted on May 03, 2018 at 15:19

Dear Nawres,

unfortunately the original ioc is from a customer project that I’m currently not allowed to share.

Thanks for your effort!

Markus Schwarzl

Von: Nawres GHARBI <st-microelectronics@jiveon.com>

Gesendet: Donnerstag, 03. Mai 2018 15:10

An: Markus Schwarzl <m.schwarzl@meds.at>

Betreff: Re: - Re: CubeMX 4.24 and 4.25 issues

STMicroelectronics Community <https://community.st.com/?et=watches.email.thread>

Re: CubeMX 4.24 and 4.25 issues

reply from Nawres GHARBI<https://community.st.com/0D70X000006Sz8bSAC

[Markus Schwarzl]

Posted on May 25, 2018 at 14:33

Hi all!

I could solve issue 1 myself by adding the following line in the ioc-file with a text-editor:

NVIC.TIM6_IRQn=true\:0\:0\:false\:false\:true\:false\:false

(copied from the sample ioc-file I attached here)

BR,

Markus