ARM Toolchain update to ARM Compiler 6.11(ARMC6) #9888

SenRamakri · 2019-02-28T14:22:48Z

Description

This is the PR to merge feature-armc6 branch to master.
With this PR we will be switching to ARMC6 as the default compiler for most targets.
Note that some targets which are incompatible with ARMC6 will stay with ARMC5,
but this is temporary and we expect to move all targets to ARMC6 soon. The tools changes are in
place to use the right toolchains based on target toolchain support. Also see related
documentation PR(ARMmbed/mbed-os-5-docs#945) which captures
this.

Pull request type

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

Release notes

Mbed OS currently supports three compilers - GCC, IAR and ARM. Mbed OS 5.11 and previous versions support ARM Compiler 5 but this doesn't support the new ARM v8-M architecture
and has been superseded by ARM Compiler 6. ARMC6 is built on Clang and LLVM technology and enables support for ARMv8-M architecture. With Mbed-OS 5.12 release, ARMC6
will be the default ARM toolchain for all Mbed-OS targets. Although most of the source code will be backward compatible with ARMC5 it's highly encouraged that new/existing
Mbed-OS targets/developers use ARM Compiler 6 for further development.

For more info on ARM Compiler 6 please refer - https://developer.arm.com/products/software-development-tools/compilers/arm-compiler/documentation/

If you are a developer needing access to ARM Compiler 6, it's available at https://developer.arm.com/products/software-development-tools/compilers/arm-compiler/downloads/version-6.

SenRamakri · 2019-02-28T14:24:03Z

@ARMmbed/mbed-os-maintainers - There is merge conflict with this PR. I'll be working on rebasing/resolving the conflicts.

SenRamakri · 2019-02-28T14:38:25Z

targets/targets.json

@@ -1325,7 +1325,7 @@
    "KW24D": {
        "supported_form_factors": ["ARDUINO"],
        "core": "Cortex-M4",
-        "supported_toolchains": ["ARM", "GCC_ARM", "IAR"],
+        "supported_toolchains": ["ARMC5", "GCC_ARM", "IAR"],


We are seeing runtime failures with this target with ARMC6. This needs further investigation/partner support, so currently KW24D will stay with ARMC5.

Can we have an issue to track this?

@0xc0170 - Please see ARM Internal Ref: IOTCORE-1059

SenRamakri · 2019-02-28T14:40:05Z

targets/targets.json

@@ -7692,7 +7694,7 @@
    "FVP_MPS2": {
        "inherits": ["ARM_FM"],
        "public": false,
-        "supported_toolchains": ["GCC_ARM", "ARM", "IAR"],
+        "supported_toolchains": ["GCC_ARM", "ARMC5", "IAR"],


FVP targets are failing with ARMC6 compiler, it appears the code itself is ok but the compiler fails with internal error/fault.
At this point, we dont know whats causing this and may need escalation to Compiler team and while we explore that we are using ARMC5 for FVP target.

This needs tracking issue then

Yes, ARM Compiler team has been contacted. We will continue to follow up with them.

SenRamakri · 2019-02-28T14:41:10Z

targets/targets.json

@@ -7878,7 +7880,7 @@
        "inherits": ["Target"],
        "macros": ["MBED_MPU_CUSTOM"],
        "default_toolchain": "GCC_ARM",
-        "supported_toolchains": ["GCC_ARM", "IAR", "ARM"],


All Cypress targets are failing ARMC6 compilation. This needs investigation from partners, so we are currently binding Cypress targets to ARMC5.

Tracking issue #9830

0xc0170 · 2019-02-28T14:57:48Z

Functionality change

Needs release notes section addition

cmonr · 2019-02-28T16:18:30Z

Needs release notes section additio

And a rebase

Assembler in rtx_core_ca.h is on the whole unified syntax such that it should work fine if assembled as either ARM or Thumb-2. The exception was __get_PSP, which uses STM^, which is only available in ARM state. Flagging for this function was incorrect, except for IAR - it switched assembler state without telling the compiler, meaning that the resulting assembler output could be incorrect, and that the function itself would not be correctly marked as an ARM function - the CPU would enter in Thumb state. Alternative fix would be to switch to System mode, which would work as either ARM or Thumb-2 assembler, like the rest of the file, but this is the minimal change. Fixes #526.

…ARM Compiler 6

SenRamakri · 2019-03-01T19:47:03Z

@cmonr @ARMmbed/mbed-os-maintainers - Please restart CI on this. I have pushed 2 commits - One to fix the exporter issues, and other to fix the FVP_MPS2 targets failing ARMC6 builds, and the fix/workaround was recommended by compiler team.

cmonr · 2019-03-01T20:04:59Z

CI started

SenRamakri · 2019-03-01T21:47:00Z

@ARMmbed/mbed-os-maintainers -
Im seeing the error - UndefinedParameter: Attempt to override undefined parameter 'nsapi.socket-stats-enabled' in 'application[*]'

I think we need this PR to fix it - #9067 ?

cmonr · 2019-03-01T22:58:52Z

@SenRamakri Correct.

Now that it's in, will retest.

cmonr · 2019-03-01T22:59:08Z

CI started

mbed-ci · 2019-03-01T23:29:11Z

Test run: FAILED

Summary: 3 of 9 test jobs failed
Build number : 2
Build artifacts

Failed test jobs:

jenkins-ci/mbed-os-ci_build-ARMC6
jenkins-ci/mbed-os-ci_build-IAR8
jenkins-ci/mbed-os-ci_build-GCC_ARM

cmonr · 2019-03-01T23:44:11Z

git checkout -b master 7c9a718
Commit message: "Merge pull request #9469 from kjbracey-arm/spi_muxing"

Found the problem. The job used a slightly older version of master.

~~Restarting CI.~~

No need. The above comment was from the second run. The third run with the updated master branch is still running.

mbed-ci · 2019-03-02T03:11:54Z

Test run: SUCCESS

Summary: 13 of 13 test jobs passed
Build number : 3
Build artifacts

SenRamakri · 2019-03-02T05:09:54Z

@ARMmbed/mbed-os-maintainers - Looks like all tests are passing now on this PR. But note that if we merge this PR before #9908 then #9908 will need additional changes in targets.json. Or if we merge #9908 first, this PR will need additional changes.
If we are expecting more changes in #9908 irrespective of this, maybe its ok to merge this before #9908 and have that communicated through review comments for #9908 as the changes are small/easy.

cmonr · 2019-03-02T05:26:29Z

@SenRamakri I think the original reason/plan as to why we wanted to hold off on merging this until Sunday was to limit the possible risk that bringing this PR in could have against other PRs that need CI.

The request has been fwd to other maintainers and PM.

SenRamakri · 2019-03-02T05:31:22Z

@cmonr - I agree with your comment, but I just wanted to let maintainers know about the implications of merging sequence. The point is, one of these PRs will need additional changes depending on which goes first.

0xc0170 · 2019-03-04T09:59:51Z

If we are expecting more changes in #9908 irrespective of this, maybe its ok to merge this before #9908 and have that communicated through review comments for #9908 as the changes are small/easy.

Status for today, 9908 still needs some changes, this should go in first.

Update: the PR is targeting rc2. this one is in rc1

0xc0170 · 2019-03-04T11:25:50Z

@SenRamakri I scheduled nightly job now, will check the progress

In the Building section in README.md, Arm Compiler 5 is mentioned. However, ARM Compiler 6 is default from Mbed OS 5.12. ARMmbed/mbed-os#9888

SenRamakri requested review from theotherjimmy, cmonr, a team, 0xc0170 and bulislaw February 28, 2019 14:24

bulislaw approved these changes Feb 28, 2019

View reviewed changes

SenRamakri commented Feb 28, 2019

View reviewed changes

0xc0170 added the needs: work label Feb 28, 2019

theotherjimmy approved these changes Feb 28, 2019

View reviewed changes

cmonr added the release-version: 5.12.0-rc1 label Feb 28, 2019

SenRamakri and others added 16 commits February 28, 2019 11:53

RealTek post script fix for ARMC6

3956969

ARMc6 build fixed for Nordic targets

988aba4

Check for __ARMCC_VERSION only

84785ae

Temporarily disable RDA targets to pass CI

7b3f30c

Enable RDA targets

0b54ea2

Temporarily disabling KW24D for AC6

5632b94

GR_LYCHEE,RZ_A1H,VK_RZ_A1H: Update to fix ARMC6 build failures

7aebee0

Back to the original : "rtx_core_ca.h"

0ef1620

Add ARMC6 binaries for MTB Cypress targets

76f90c2

Build tools changes to support AC6 and AC5

50eb4f9

Api tests modification and uLib related fixes

7490b1c

Build metadata check added

8624fdf

Fix configuration_store.h to use right toolchain check flags

65e7651

targets.json changes to select ARMC5 for specific targets

41e3329

Remove unwanted spaces

2008295

SenRamakri added 2 commits March 1, 2019 13:21

Fixes for exporters failing with wrong compiler settings

e1e4456

Workaround(recommended by compiler team) for FVP_MPS2 targets to use …

45c727e

…ARM Compiler 6

ghost added the PM_ACCEPTED label Mar 1, 2019

cmonr added needs: CI and removed needs: work labels Mar 1, 2019

0xc0170 added ready for merge and removed needs: CI labels Mar 4, 2019

0xc0170 mentioned this pull request Mar 4, 2019

Pr/cy mbed os 5.12.0 #9908

Merged

0xc0170 approved these changes Mar 4, 2019

View reviewed changes

0xc0170 merged commit e75794e into master Mar 4, 2019

0xc0170 removed the ready for merge label Mar 4, 2019

0xc0170 mentioned this pull request Mar 5, 2019

make armc6: error 8 for assembly file #9938

Closed

bridadan mentioned this pull request Mar 11, 2019

[OS ERROR] Command: armclang --vsn : Unable to build (with latest mbed-os) application and tests with ARM compiler 5 ARMmbed/mbed-os-tools#137

Closed

coisme mentioned this pull request Jun 11, 2019

[OoB 5.13] Arm Compiler 5 is not default ARMmbed/mbed-os-example-socket-stats#24

Merged

adbridge deleted the feature-armc6 branch August 20, 2019 10:50

ARM Toolchain update to ARM Compiler 6.11(ARMC6) #9888

ARM Toolchain update to ARM Compiler 6.11(ARMC6) #9888

Uh oh!

Conversation

SenRamakri commented Feb 28, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Pull request type

Release notes

Uh oh!

SenRamakri commented Feb 28, 2019

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

SenRamakri Feb 28, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

SenRamakri Feb 28, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

0xc0170 commented Feb 28, 2019

Uh oh!

cmonr commented Feb 28, 2019

Uh oh!

SenRamakri commented Mar 1, 2019

Uh oh!

cmonr commented Mar 1, 2019

Uh oh!

SenRamakri commented Mar 1, 2019

Uh oh!

cmonr commented Mar 1, 2019

Uh oh!

cmonr commented Mar 1, 2019

Uh oh!

mbed-ci commented Mar 1, 2019

Uh oh!

cmonr commented Mar 1, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mbed-ci commented Mar 2, 2019

Uh oh!

SenRamakri commented Mar 2, 2019

Uh oh!

cmonr commented Mar 2, 2019

Uh oh!

SenRamakri commented Mar 2, 2019

Uh oh!

0xc0170 commented Mar 4, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

0xc0170 commented Mar 4, 2019

Uh oh!

Uh oh!

SenRamakri commented Feb 28, 2019 •

edited

Loading

SenRamakri Feb 28, 2019 •

edited

Loading

SenRamakri Feb 28, 2019 •

edited

Loading

cmonr commented Mar 1, 2019 •

edited

Loading

0xc0170 commented Mar 4, 2019 •

edited

Loading