STM32F4_HAL_MMC erase command check is wrong. == has higher precedence than &

[janjongboom]

Maybe I'm wrong, but I think the check here is incorrect. This gets evaluated as:

if((hmmc->MmcCard.Class) & (SDIO_CCCC_ERASE == 0U))

Because == has higher precedence than &.

[0xc0170]

cc @bcostm @adustm @LMESTM @jeromecoutant

[adustm]

Hi @janjongboom
Let me rephrase your pull request just to be sure.
According to you:
The current behavior of if((hmmc->MmcCard.Class) & SDIO_CCCC_ERASE == 0U) is :
if((hmmc->MmcCard.Class) & (SDIO_CCCC_ERASE == 0U))

and the expected behavior should be
if(((hmmc->MmcCard.Class) & SDIO_CCCC_ERASE) == 0U)

If my rephrasing is correct, I can check internally within ST team, but you are probably right.
Kind regards
Armelle
ST_INTERNAL_REF 36423

[janjongboom]

@adustm Yes, your rephrasing is correct.

[0xc0170]

@adustm Let us know the feedback from the team. As this changes their driver code, if we accept this change now, and later with the new update will be overwritten?

[jeromecoutant]

Hi

if we accept this change now, and later with the new update will be overwritten?

It should not :-)
We take into account MBED current code during out HAL code update.

[0xc0170]

/morph test

[adustm]

Hello @0xc0170 @janjongboom
I just got confirmation from our internal team. You can proceed with this PR.

We'll manage the merge with the official fix when it will be available in a later STM32F4 Cube FW release.
Thank you for the fix. Kind regards
Armelle

[mbed-bot]

Result: FAILURE

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 952

Build failed!

[0xc0170]

Restarting, master included one bug that was fixed

/morph test

[mbed-bot]

Result: SUCCESS

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 962

All builds and test passed!