Add mcuxpresso exporter #4916

JojoS62 · 2017-08-15T15:57:13Z

Description

Add an exporter for MCUXpresso (NXP), successor of LPCXpresso

Status

IN DEVELOPMENT
needs work

base is gnuarmecplipse exporter
Works for the targets that LPCXpresso supported + LPC54114
LPC54608 compiles as well, but the MCU settings for debugging are not yet available.
Supports debug and release builds, develop not yet working
flags are read from mbed build system and converted to eclipse tools settings

Migrations

If this PR changes any APIs or behaviors, give a short description of what API users should do when this PR is merged.

NO APIs changed

Related PRs

#4819

Todos

Tests
Documentation

Deploy notes

needs installation of MCUXpresso, tested with version 10.0.2

Steps to test or reproduce

create a project file with 'mbed export -i mcuxpresso'
import the created project.zip
** Note: this version always creates a zip file for faster testing. Needs a clean solution for creating a zip file as an option.

theotherjimmy · 2017-08-15T16:24:35Z

Awesome work @JojoS62

It looks like your rebase included some commits from master somehow. Could you remove those duplicated commits from this PR?

theotherjimmy

It looks like you copied the contents of the gnuarmeclipse exporter. To prevent duplication of fixes, could you import it and override things instead?

theotherjimmy · 2017-08-15T16:28:53Z

Needs a clean solution for creating a zip file as an option.

I'll get that for you as another PR real quick.

theotherjimmy · 2017-08-15T16:29:58Z

@studavekar Can we get the deploy notes installed in CI? Could you let us know when that's Done?

theotherjimmy · 2017-08-15T16:34:23Z

@JojoS62 Let me know if you would like some help with this. @mmahadevan108 FYI.

JojoS62 · 2017-08-15T20:38:04Z

@theotherjimmy I tried git rebase origin but this returned errors:

Applying: Renamed so that we have one configuration for all STM32F439 targets.
Using index info to reconstruct a base tree...
A       features/mbedtls/targets/TARGET_STM/TARGET_STM32F4/TARGET_NUCLEO_F439ZI/mbedtls_device.h
.git/rebase-apply/patch:43: trailing whitespace.
 *  mbedtls_device.h
warning: 1 line adds whitespace errors.
Falling back to patching base and 3-way merge...
CONFLICT (rename/rename): Rename "features/mbedtls/targets/TARGET_STM/TARGET_STM32F4/TARGET_NUCLEO_F439ZI/mbedtls_device
.h"->"features/mbedtls/targets/TARGET_STM/TARGET_STM32F7/TARGET_NUCLEO_F756ZG/mbedtls_device.h" in branch "HEAD" rename
"features/mbedtls/targets/TARGET_STM/TARGET_STM32F4/TARGET_NUCLEO_F439ZI/mbedtls_device.h"->"features/mbedtls/targets/TA
RGET_STM/TARGET_STM32F4/TARGET_STM32F439xI/mbedtls_device.h" in "Renamed so that we have one configuration for all STM32
F439 targets."
error: Failed to merge in the changes.
Patch failed at 0013 Renamed so that we have one configuration for all STM32F439 targets.
The copy of the patch that failed is found in: .git/rebase-apply/patch

When you have resolved this problem, run "git rebase --continue".
If you prefer to skip this patch, run "git rebase --skip" instead.
To check out the original branch and stop rebasing, run "git rebase --abort".

What do have I to do with this patch file or to resolve the problem?

patch.zip

Signed-off-by: Mahadevan Mahesh <[email protected]>

JojoS62 · 2017-08-16T14:08:40Z

@theotherjimmy I'm working on deriving from the gae class, but I will have some questions...
POST_BINARY_WHITELIST : for what is this used?

theotherjimmy · 2017-08-16T14:14:36Z

I'm working on deriving from the gae class

+💯

POST_BINARY_WHITELIST : for what is this used?

That used to help generate a list of supported targets. When using a fixed list of supported targets, you should not have to worry about it. The export API now uses the class method is_target_supported and the class method all_supported_targets to examine support. This elides lots of unused computation (it speed up exporting by a factor of 2 on my machine), and simplifies most exporters supported check implementation. The whitelist is used by apply_supported_whitelist and that method is used by exporters who's supported targets list grows with every new target. Exporters with fixed supported lists should not have to worry about it.

theotherjimmy · 2017-08-16T14:15:22Z

@JojoS62 I forgot to respond to your prior comment. I'm sorry. I'll get right on that.

theotherjimmy · 2017-08-16T14:27:21Z

Oh Dang, your history looks pretty duplicated. It looks like you rebased then tried to git push but git told you "Your history does not match, do this pull thing first" which you should NOT do. I understand that you may not have known that at the time, so I'm going to see what I can do to help you correct this mishap.

BACK UP THE CURRENT BRANCH This is important because we are going to rewrite history (we're the winners) git checkout -b old-add_MCUXpresso_exporter
Checkout the commit before the merge (It should be HEAD^) git checkout HEAD^
Delete the branch git branch -D add_MCUXpresso_exporter
Recreate the branch on this commit git checkout -b add_MCUXpresso_exporter
Update this PR via force push git push -u <your-remote-here > -f

If something goes horribly wrong from steps 2-5, you can restore your branch to it's former glory with:

Change to the backup branch git checkout old-add_MCUXpresso_exporter
Delete the current monstrosity git branch -D add_MCUXpresso_exporter
Restore former glory from backup git checkout -b add_MCUXpresso_exporer

theotherjimmy · 2017-08-16T14:29:18Z

BTW, I tried the steps 1-4, and the history looks good.

JojoS62 · 2017-08-16T15:12:28Z

Thanks, ok, I'll try to clean up. I made my old mistake again to merge updates in my branch because the zip export failed. So I'd better use two branches, one with only my work on something and another one for merging this and other stuff?
I have read your answer about the whitelist after I implemented the template file exist method. This works and the list of supported targets can be extended without changing the exporters code. But maybe the filesystem checks are too expensive for the use in the online compiler?

theotherjimmy · 2017-08-16T15:14:20Z

Correctness over liveness, every time. I'll take a look.

theotherjimmy · 2017-08-16T15:15:58Z

Don't worry about checking for existence, that's pretty cheap compared to scanning all repos that make up the project and zipping all of that. I bet you would have trouble measuring the difference anyway.

JojoS62 · 2017-08-16T15:43:28Z

´(It should be HEAD^) : has ^to be replaced by something? The command line aks for 'more?' after this command.

theotherjimmy · 2017-08-16T15:44:52Z

Hmm... That works for me. Try the exact hash.

theotherjimmy · 2017-08-16T15:45:21Z

For example git checkout 3aafb2d

theotherjimmy · 2017-08-16T15:46:55Z

It looks like you're still adding commits. You may have to cherry pick them after you are done with the process above.

JojoS62 · 2017-08-16T15:55:24Z

Now the PR contains only the one from 'git checkout sha'. I have added only the one commit to get the working stage clean.

theotherjimmy · 2017-08-16T15:56:39Z

Looks like a good start. Are you planning on cherry-picking the other commits and pushing them up now?

JojoS62 · 2017-08-16T15:59:19Z

I'll try

- used LPC4088 for testing - split template for LPC54114 into common and Cortex-Mx dependet parts like in old LPCXpresso exporter - changed fixed linker script name to name from options Todo: use flags and configurations from options (like in gae exporter)

creates output for debug, develop and release builds TODO. replaces some fixed flags, develop build not working , linker has no input files

change backslash to forward slash in excluded_folders set

project is read, but still fixes necessary: - archticture is fixed entry, read from config - linker options - linker script file ? - linker settings 'managed build' page causes error - develop still not working

target family, fpu removed 'develop' proffile, it causes errors in import

target specific templates use a common template .cproject.tmpl and extend it by adding target information. This is necessary for the flash and debugging tools

mmahadevan108 · 2017-08-31T19:03:41Z

@JojoS62 Below is a patch for platforms (e.g: K64F, LPC54608) that are supported via the SDK being installed in MCUXpresso. Without this change, build will work but failures will be seen during launch. Can this be included. This is the last patch from me. Thanks.
0002-Add-support-to-specify-SDK-name-as-this-is-needed-fo.zip

- fixes invalid char '#' in linker script - added 'mbedclean'

added template for K64F

some targets need an additional SDK.

JojoS62 · 2017-09-01T07:49:35Z

@mmahadevan108

Can this be included. This is the last patch from me. Thanks.

no problem, you are welcome. I'm happy when this exporter gets more attention and support.
I just haven't used the Kinetis boards yet, its great when the exporter works for them too.
Thanks.

0xc0170 · 2017-09-04T15:17:09Z

@JojoS62 What is the status of this patch?

JojoS62 · 2017-09-04T15:32:57Z

Compiling the K64F target works, thats what I could test.
The review marker is still set because @theotherjimmy wanted to check something in the test procedure, the test for this exporter did not run.

theotherjimmy · 2017-09-05T14:37:42Z

@0xc0170 @JojoS62 Sorry about the delay. as the saying goes: "when you're up to your neck in alligators, it's hard to remember that your initial objective was to drain the swamp" That's been dealing with CI this past week or two.

maclobdell · 2017-09-05T15:15:08Z

@theotherjimmy - pretty please make this a top priority to test and merge when ready. We would like this to be part of the next mbed OS release (in 2 weeks?), so we can use it in internal training in 3 weeks. Thanks a lot @JojoS62 and @mmahadevan108 for all your work on this!

theotherjimmy · 2017-09-06T16:48:06Z

@JojoS62 @mmahadevan108 @0xc0170 I'm going to run what the CI job would do on my machine, and mark this as ready for merge if that passes.

theotherjimmy · 2017-09-06T22:38:40Z

Hey guys,

I ran this through a local version of CI on my machine, and it looks good. I had to modify the success check and add some missing imports. I'm going to post the results shortly, and we can elide an export build on this one until we have this eclipse variant installed in CI.

theotherjimmy · 2017-09-06T22:50:21Z

Examples that passed:

mbed-os-example-ble-BatteryLevel
mbed-os-example-ble-Beacon
mbed-os-example-ble-Button
mbed-os-example-ble-EddystoneObserver
mbed-os-example-ble-EddystoneService
mbed-os-example-ble-GAPButton
mbed-os-example-ble-HeartRate
mbed-os-example-ble-LED
mbed-os-example-ble-LEDBlinker
mbed-os-example-ble-Thermometer
mbed-os-example-blinky
mbed-os-example-bootloader
mbed-os-example-cellular
mbed-os-example-client
mbed-os-example-fat-filesystem
mbed-os-example-mesh-minimal
mbed-os-example-sockets
mbed-os-example-tls-authcrypt
mbed-os-example-tls-benchmark
mbed-os-example-tls-hashing
mbed-os-example-tls-tls-client
mbed-os-example-uvisor
mbed-os-example-uvisor-thread
mbed-os-example-wifi
nanostack-border-router

No examples failed to build.

Tested with:

K64F
LPC54114

theotherjimmy · 2017-09-06T22:51:28Z

@0xc0170 marking this as ready for merge speculatively. Please check that the default checks have run on this PR, and then it's ready.

0xc0170 · 2017-09-07T11:45:15Z

@theotherjimmy This needs exporters tests again (new commits)

theotherjimmy · 2017-09-07T13:54:31Z

Uh, nope. This exporter is not yet it CI, and I posted the test report above.

theotherjimmy · 2017-09-07T16:47:53Z

There is no benefit to running an export-build here, as this exporter is not covered.

theotherjimmy suggested changes Aug 15, 2017

View reviewed changes

theotherjimmy added the needs: work label Aug 15, 2017

Add support for MCUXpresso inside mbed exporter

8ae5eda

Signed-off-by: Mahadevan Mahesh <[email protected]>

JojoS62 force-pushed the add_MCUXpresso_exporter branch from febfe6d to 8ae5eda Compare August 16, 2017 15:49

JojoS62 added 6 commits August 16, 2017 18:07

changed template to gae exporter style

af177c5

creates output for debug, develop and release builds TODO. replaces some fixed flags, develop build not working , linker has no input files

make excluded_folders POSIX conform

544fd2a

change backslash to forward slash in excluded_folders set

created project compiles in MCUX

41c5fdb

project is read, but still fixes necessary: - archticture is fixed entry, read from config - linker options - linker script file ? - linker settings 'managed build' page causes error - develop still not working

added generated flags

16c4150

target family, fpu removed 'develop' proffile, it causes errors in import

added templates for LPC4088 and LPC11H37

2bd7b4d

target specific templates use a common template .cproject.tmpl and extend it by adding target information. This is necessary for the flash and debugging tools

mmahadevan108 mentioned this pull request Aug 31, 2017

Add support for MCUXpresso inside mbed exporter #4819

Closed

JojoS62 added 3 commits September 1, 2017 09:41

applied patch from mmahadevan108

2615982

- fixes invalid char '#' in linker script - added 'mbedclean'

add support for K64F target

d00d6f4

added template for K64F

added support for targets with extra SDK

32c56b2

some targets need an additional SDK.

Add imports for build

9d51b1a

Allow eclipse to fail for mysteriously in build

a3bb8e3

theotherjimmy approved these changes Sep 6, 2017

View reviewed changes

theotherjimmy added ready for merge and removed needs: work labels Sep 6, 2017

0xc0170 added needs: CI and removed ready for merge labels Sep 7, 2017

theotherjimmy added ready for merge and removed needs: CI labels Sep 7, 2017

theotherjimmy merged commit a108adf into ARMmbed:master Sep 8, 2017

theotherjimmy added release-version: 5.6.0 and removed ready for merge labels Sep 8, 2017

JojoS62 deleted the add_MCUXpresso_exporter branch March 6, 2020 13:25

Add mcuxpresso exporter #4916

Add mcuxpresso exporter #4916

Uh oh!

Conversation

JojoS62 commented Aug 15, 2017

Description

Status

Migrations

Related PRs

Todos

Deploy notes

Steps to test or reproduce

Uh oh!

theotherjimmy commented Aug 15, 2017

Uh oh!

theotherjimmy left a comment

Choose a reason for hiding this comment

Uh oh!

theotherjimmy commented Aug 15, 2017

Uh oh!

theotherjimmy commented Aug 15, 2017

Uh oh!

theotherjimmy commented Aug 15, 2017

Uh oh!

JojoS62 commented Aug 15, 2017

Uh oh!

JojoS62 commented Aug 16, 2017

Uh oh!

theotherjimmy commented Aug 16, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

theotherjimmy commented Aug 16, 2017

Uh oh!

theotherjimmy commented Aug 16, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

theotherjimmy commented Aug 16, 2017

Uh oh!

JojoS62 commented Aug 16, 2017

Uh oh!

theotherjimmy commented Aug 16, 2017

Uh oh!

theotherjimmy commented Aug 16, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

JojoS62 commented Aug 16, 2017

Uh oh!

theotherjimmy commented Aug 16, 2017

Uh oh!

theotherjimmy commented Aug 16, 2017

Uh oh!

theotherjimmy commented Aug 16, 2017

Uh oh!

JojoS62 commented Aug 16, 2017

Uh oh!

theotherjimmy commented Aug 16, 2017

Uh oh!

JojoS62 commented Aug 16, 2017

Uh oh!

mmahadevan108 commented Aug 31, 2017

Uh oh!

JojoS62 commented Sep 1, 2017

Uh oh!

0xc0170 commented Sep 4, 2017

Uh oh!

JojoS62 commented Sep 4, 2017

Uh oh!

theotherjimmy commented Sep 5, 2017

Uh oh!

maclobdell commented Sep 5, 2017

Uh oh!

theotherjimmy commented Sep 6, 2017

Uh oh!

theotherjimmy commented Sep 6, 2017

Uh oh!

theotherjimmy commented Sep 6, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

theotherjimmy commented Sep 6, 2017

Uh oh!

theotherjimmy commented Aug 16, 2017 •

edited

Loading

theotherjimmy commented Aug 16, 2017 •

edited

Loading

theotherjimmy commented Aug 16, 2017 •

edited

Loading

theotherjimmy commented Sep 6, 2017 •

edited

Loading