[HAL] Modified SPI to use shared mutex

[neilt6]

Modified the SPI class to use a shared mutex for all instances in response to #2276. This is consistent with I2C and others.

[sg-]

@c1728p9

[c1728p9]

Thanks for the PR @neilt6. This looks good to me.

[0xc0170]

@mbed-bot: TEST

HOST_OSES=windows
BUILD_TOOLCHAINS=GCC_ARM,ARM,IAR
TARGETS=K64F,NRF51_DK,NRF51_MICROBIT,NUCLEO_F411RE

[0xc0170]

/morph test

[anpilog]

Hi,

What about MCU with more than 1 SPI controller?
Does it mean that all of them share one mutex in RTOS build?

[0xc0170]

What about MCU with more than 1 SPI controller?
Does it mean that all of them share one mutex in RTOS build?

yes, same as there's only one owner (private member variable).
We shall have a look at it and provide better locking mechanism (per peripheral instance).

[mbed-bot]

[Build 739]
SUCCESS: Building succeeded and tests were run! Be sure to check the test results

[anpilog]

Then it probably should be:

class SPI {
public:
    virtual bool lock(uint32_t millisec=osWaitForever) {
        return _mutex->lock(millisec) == osOK;
    }
}

Otherwise any attempt to use SPI methods would lead to wait.

[0xc0170]

@anpilog Can you create an issue for this? this might not be just a case for SPI. Shall we discuss this aside from this PR?

[mbed-bot]

Result: SUCCESS

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

/morph test

Output

mbed Build Number: 560

All builds and test passed!

[anpilog]

Ok.
#2310

[0xc0170]

CI results are good !