STM32 EMAC configuration update

[jeromecoutant]

Description

Hi

Issue comes from discussion in #11506 (comment)

It looks not easy to remove the default Ethernet capability for 1 board...

So

• TARGET_STM_EMAC directory has been renamed into TARGET_STM
  (which doesn't need any specific extra label addition)
• network-default-interface-type is set in the mbed_lib json file

Now removing the EMAC device_has feature is the only parameter to set or not

Pull request type

[ ] Fix
[x] Refactor
[ ] Target update
[ ] Functionality change
[ ] Docs update
[ ] Test update
[ ] Breaking change

Reviewers

@ARMmbed/team-st-mcd
@ARMmbed/team-ublox

Release Notes

[ciarmcom]

@jeromecoutant, thank you for your changes.
@ARMmbed/mbed-os-maintainers @ARMmbed/mbed-os-ipcore please review.

[0xc0170]

Refactor

This touches only target code, shall it go to 5.15 rather?

[jeromecoutant]

shall it go to 5.15 rather?

I prefer next 5.14 ...

[SeppoTakalo]

This feels more like a workaround, than fix.

It should be found out why those buffers mentioned in #11506 are not dropped by the linker.

[mbed-ci]

Test run: FAILED

Summary: 3 of 4 test jobs failed
Build number : 1
Build artifacts

Failed test jobs:

• jenkins-ci/mbed-os-ci_build-GCC_ARM
• jenkins-ci/mbed-os-ci_build-ARM
• jenkins-ci/mbed-os-ci_build-IAR

[0xc0170]

It should be found out why those buffers mentioned in #11506 are not dropped by the linker.

@jeromecoutant Any update?

[jeromecoutant]

Build issue should be corrected now.

Compilation ex with the echo test:

• NUCLEO_F429ZI default configuration:
  | name | target | toolchain | static_ram | total_flash |
  | echo | NUCLEO_F429ZI | ARMC6 | 26967 | 65556 |

• NUCLEO_F429ZI with local configuration "target.device_has_remove": ["EMAC"] :
  | echo | NUCLEO_F429ZI | ARMC6 | 11126 | 55085 |

[jeromecoutant]

Thanks to start CI

[0xc0170]

It should be found out why those buffers mentioned in #11506 are not dropped by the linker.

@jeromecoutant Is this fixing 11506 or ? It really does not look like a proper fix. However if this is a simplification work, then should be fine.

[jeromecoutant]

I don't plan any more work on this topic...

[0xc0170]

This feels more like a workaround, than fix.

It should be found out why those buffers mentioned in #11506 are not dropped by the linker.

@jeromecoutant This does not fix 11506 completely, so that issue would still be opened (it hides it that I do not like much as stated in the quote above)..
The changes look good to me as they are here, removing STM EMAC label

[0xc0170]

CI started

[mbed-ci]

Test run: SUCCESS

Summary: 12 of 12 test jobs passed
Build number : 2
Build artifacts

[0xc0170]

@SeppoTakalo Happy with this (considering this is not the fix for the issue, but rather refactor) ?

[0xc0170]

@jeromecoutant This does not fix 11506 completely, so that issue would still be opened (it hides it that I do not like much as stated in the quote above)..

Taking this in as it is, #11506 stays opened