[Greentea] Fix test code mbed-os-tests-mbed_hal-flash

[ccli8]

This PR fixes the test code mbed-os-tests-mbed_hal-flash.

[0xc0170]

This PR fixes the test code mbed-os-tests-mbed_hal-flash.

Thanks for the fix. how does this fixes it ? Does it fail ? With this patch it passes ?

It would be helpful to provide test results (prior and after this change). We can reproduce this locally and compare our results.

[ccli8]

@0xc0170 The issue is not met on K64F, but on my NUMAKER_PFM_NUC472 on which flash support is being added. In original mbed-os-tests-mbed_hal-flash > main.cpp > erase_range:

static void erase_range(flash_t *flash, uint32_t addr, uint32_t size)
{
    while (size > 0) {
        uint32_t sector_size = flash_get_sector_size(flash, addr);
        TEST_ASSERT_NOT_EQUAL(0, sector_size);
        int32_t ret = flash_erase_sector(flash, addr);
        TEST_ASSERT_EQUAL_INT32(0, ret);
        size = size > sector_size ? size - sector_size : 0;
    }
}

In the while loop, addr should be advanced with each iteration. On K64F, the passed in size is 0x1000 which equals its sector size, so there is only one iteration in the loop and the bug is not met. On my NUMAKER_PFM_NUC472 case, the passed in size is 0x2000 which is 4 times its sector size 0x800. There are 4 iterations in the while loop but addr is not updated.

[sg-]

/morph test-nightly

[mbed-bot]

Result: SUCCESS

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

/morph test-nightly

Output

mbed Build Number: 1716

All builds and test passed!

[0xc0170]

cc @c1728p9