M2351: Pre-build secure image/lib to favor pelion application #11594

ccli8 · 2019-09-30T06:57:12Z

Description

In this PR, default secure program is re-built most simplified so that non-secure program can make most use of memory for its large application like pelion. The secure program will partition memory:

Flash (512KiB in total): 64KiB for secure and 448KiB for nonsecure.
SRAM (96KiB in total): 8KiB for secure and 88KiB for nonsecure.

Note for running greentea on secure target

With secure program most simplified, there's no memory available to run greentea on secure target NU_PFM_M2351_NPSA_S. If it is necessary, re-partition memory in mbed_app.json:

{
    "target_overrides": {
    ....
        "NU_PFM_M2351_NPSA_S": {
            "target.mbed_rom_start"             : "0x0",
            "target.mbed_rom_size"              : "0x40000",
            "target.mbed_ram_start"             : "0x20000000",
            "target.mbed_ram_size"              : "0x8000"
        },
    ....
    }
}

Related PR

ARMmbed/mbed-os-example-pelion#25 (comment)

Pull request type

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

ciarmcom · 2019-09-30T07:00:23Z

@ccli8, thank you for your changes.
@Ronny-Liu @ARMmbed/mbed-os-maintainers please review.

0xc0170

Just a check: This change meets the min requirement for pelion - changing the memory sizes for secure/non-secure?

Is this just a fix for a patch release, or rather functional change ?

ccli8 · 2019-09-30T08:41:02Z

This change meets the min requirement for pelion - changing the memory sizes for secure/non-secure?

Yes, allocate more memory for non-secure program to meet min requirement of pelion.

Is this just a fix for a patch release, or rather functional change ?

Just patch release

But for CI/Greentea, it is necessary to re-partition memory for building secure target NU_PFM_M2351_NPSA_S, commented as above.

ccli8 · 2019-10-02T08:58:31Z

Update?

bulislaw

@Patater please have a look

Patater · 2019-10-02T13:28:30Z

...N/TARGET_M2351/TARGET_M23_NS/TARGET_NU_PFM_M2351_NPSA_NS/TARGET_NU_PREBUILD_SECURE/README.md

@@ -19,13 +20,51 @@ The pre-built secure code has the following hardware partition:
    - **PDMA0** hardwired to secure. Implements secure asynchronous transfer.
    - **PDMA1** configured to nonsecure. Implements nonsecure asynchronous transfer.

-## Pre-built secure code files
+In this memory partition, secure program is the most simplified.
+Then non-secure program can make use of the most memory available for its large application like pelion.


"pelion" should be "Pelion"

Patater · 2019-10-02T13:30:10Z

...N/TARGET_M2351/TARGET_M23_NS/TARGET_NU_PFM_M2351_NPSA_NS/TARGET_NU_PREBUILD_SECURE/README.md

+        }
+    }
+}
+```


Why is it necessary to change mbed_app.json to build the default image? Shouldn't the default targets.json for NU_PFM_M2351_NPSA_S be modified to have these values as the default instead?

Modify README.md and Non-PSA secure code. The default secure program can re-generate without modification.

Patater · 2019-10-02T13:31:17Z

targets/targets.json

        "mbed_ram_start"        : "0x20000000",
-        "mbed_ram_size"         : "0x8000"
+        "mbed_ram_size"         : "0x2000"


Should we add "target.boot-stack-size" : "0x600" here and also "target.device_has_remove": ["SERIAL", "SERIAL_ASYNCH", "SERIAL_FC", "STDIO_MESSAGES"],?

Yes, added.

ccli8 · 2019-10-03T07:02:18Z

Make modification:

Default non-PSA secure program can re-generate without modification in Non-PSA secure program.

Patater

Changes look fine to me. I think we need some more thought put into how we'd maintain this binary.

Patater · 2019-10-03T07:04:19Z

..._NS/TARGET_NU_PFM_M2351_NPSA_NS/TARGET_NU_PREBUILD_SECURE/NuMaker-mbed-TZ-secure-example.hex

-:104710008F46C04600F026B800F027B800F028B851
-:1047200000F029B800F02AB800F02BB800F02CB83F
-:1047300000F02DB800F02FB800F02DB800F02BB825
-:1047400000F029B800F027B800F025B8


What is the maintenance plan for this binary? When will it need to be rebuilt? When will we notice that it is broken and requires rebuilding? Maybe some sort of automated test is needed.

@0xc0170 How do we indicate the license this binary is provided under?

What is the maintenance plan for this binary? When will it need to be rebuilt? When will we notice that it is broken and requires rebuilding? Maybe some sort of automated test is needed.

Currently, it is re-built by Nuvoton manually. When there's any of memory re-partition, change of IP security attribute, or change of secure function I/F, it will be re-built.

@0xc0170 How do we indicate the license this binary is provided under?

license file located in the folder

bulislaw · 2019-10-07T15:27:57Z

General question, why does this need to be a binary rather than source?

ccli8 · 2019-10-08T01:41:03Z

General question, why does this need to be a binary rather than source?

@bulislaw This is well-defined build flow for TrustZone targets. TF-M targets like MUSCA_A1 also follows it.

ccli8 · 2019-10-08T09:55:36Z

Other concern?

ccli8 · 2019-10-18T10:58:22Z

Update?

0xc0170 · 2019-10-18T11:11:58Z

I am reviewing now, sorry for the delay. Looks like all reviewers approved (not explicitly though).

0xc0170 · 2019-10-18T11:13:56Z

CI started

mbed-ci · 2019-10-18T13:59:16Z

Test run: FAILED

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

Failed test jobs:

jenkins-ci/mbed-os-ci_build-ARM

ccli8 · 2019-10-21T02:19:25Z

To achieve minimum memory footprint on secure target, this PR re-partitions memory and makes related modifications for non-secure target to have maximum memory to run Pelion. To run CI on secure target, the above configuration must change back to normal use for secure target via mbed_app.json:

        "NU_PFM_M2351_NPSA_S": {
            "target.boot-stack-size"                : "0x1000",
            "target.device_has_add"                 : ["SERIAL", "SERIAL_ASYNCH", "SERIAL_FC", "STDIO_MESSAGES"],
            "target.mbed_rom_start"                 : "0x0",
            "target.mbed_rom_size"                  : "0x40000",
            "target.mbed_ram_start"                 : "0x20000000",
            "target.mbed_ram_size"                  : "0x8000"
        },

0xc0170 · 2019-10-23T13:58:37Z

@Patater Can you re-review please?

0xc0170 · 2019-10-23T13:59:20Z

@ccli8 The latest comment was regarding CI failures? Or not? How do we fix them?

ccli8 · 2019-10-28T02:45:12Z

Make modifications:

Do rebase
Fix conflict

0xc0170 · 2019-10-28T09:35:14Z

CI restarted, might still fails, will follow up what to do with the build with @ARMmbed/mbed-os-test

mbed-ci · 2019-10-28T10:47:55Z

Test run: FAILED

Summary: 1 of 4 test jobs failed
Build number : 2
Build artifacts

Failed test jobs:

jenkins-ci/mbed-os-ci_build-ARM

0xc0170 · 2019-10-28T11:06:49Z

Expected failure.

CI restarted, might still fails, will follow up what to do with the build with @ARMmbed/mbed-os-test

Please help us reviewing the failures and what can we do to pass the build for this target.

ccli8 · 2019-10-29T02:05:12Z

On NU_PFM_M2351_NPSA_S target, not only disabled serial but also just 8KiB SRAM can cause CI build failure. My suggestion: exclude NU_PFM_M2351_NPSA_S from CI or follow #11594 (comment)

0xc0170 · 2019-11-26T15:32:04Z

@ccli8 We will have the NU_PFM_M2351_NPSA_S disabled for building, so will restart CI once it is integrated

cyliangtw · 2019-12-02T12:12:01Z

@0xc0170 For restart CI, do you need any help on the integration of "NU_PFM_M2351_NPSA_S disabled for building" ?

0xc0170 · 2019-12-02T12:31:34Z

@cyliangtw I'll restart CI (should be fixed) later today

0xc0170 · 2019-12-02T12:42:25Z

CI restarted

mbed-ci · 2019-12-02T14:19:07Z

Test run: SUCCESS

Summary: 11 of 11 test jobs passed
Build number : 3
Build artifacts

ccli8 · 2019-12-04T05:11:44Z

What's the schedule for 6.0.0? 5.14.3 or 5.15.0 in between?

0xc0170 · 2020-01-14T15:09:10Z

What's the schedule for 6.0.0? 5.14.3 or 5.15.0 in between?

We almost missed this one. 6.0.0-alpha1 tomorrow. 5.15 was released back in December

ciarmcom requested review from Ronny-Liu and a team September 30, 2019 07:00

ciarmcom added the needs: review label Sep 30, 2019

0xc0170 reviewed Sep 30, 2019

View reviewed changes

ccli8 mentioned this pull request Oct 2, 2019

Support Nuvoton targets ARMmbed/mbed-os-example-pelion#25

Closed

bulislaw reviewed Oct 2, 2019

View reviewed changes

Patater reviewed Oct 2, 2019

View reviewed changes

ccli8 force-pushed the nuvoton_m2351_partition-memory-pelion branch from 61a231b to f406f5a Compare October 3, 2019 06:52

Patater reviewed Oct 3, 2019

View reviewed changes

ccli8 mentioned this pull request Oct 4, 2019

Support NU_PFM_M2351_NPSA_NS ARMmbed/mbed-os-example-pelion#42

Closed

0xc0170 approved these changes Oct 18, 2019

View reviewed changes

0xc0170 added needs: CI and removed needs: review labels Oct 18, 2019

0xc0170 added needs: work and removed needs: CI labels Oct 18, 2019

ccli8 mentioned this pull request Oct 21, 2019

M2351: Support TFM level 1 #10959

Merged

ccli8 force-pushed the nuvoton_m2351_partition-memory-pelion branch from 45547e4 to 26e6b15 Compare October 28, 2019 02:43

0xc0170 added needs: CI and removed needs: work labels Oct 28, 2019

0xc0170 added needs: work and removed needs: CI labels Oct 29, 2019

0xc0170 requested a review from a team October 29, 2019 09:09

Patater mentioned this pull request Oct 29, 2019

Test frameworks included on targets that can't run tests #11766

Closed

VeliMattiLahtela approved these changes Nov 26, 2019

View reviewed changes

0xc0170 added needs: CI and removed needs: work labels Dec 2, 2019

0xc0170 requested a review from Patater December 3, 2019 08:05

0xc0170 added the release-version: 6.0.0-alpha-1 First pre-release version of 6.0.0 label Dec 3, 2019

0xc0170 merged commit a8ee2d8 into ARMmbed:master Dec 3, 2019

0xc0170 removed the needs: CI label Dec 3, 2019

ccli8 mentioned this pull request Jan 13, 2020

M2351: Remove LPTICKER to spare memory for non-PSA minimal secure build #12241

Merged

ccli8 mentioned this pull request Apr 14, 2020

NU_PFM_M2351_NPSA_S: failing nightly in ATHandler - not supporting serial #12803

Closed

ccli8 mentioned this pull request May 18, 2020

M2351: Enable non-PSA minimal secure build (5.15) #12986

Merged

cyliangtw deleted the nuvoton_m2351_partition-memory-pelion branch March 9, 2023 05:29

M2351: Pre-build secure image/lib to favor pelion application #11594

M2351: Pre-build secure image/lib to favor pelion application #11594

Uh oh!

Conversation

ccli8 commented Sep 30, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Note for running greentea on secure target

Related PR

Pull request type

Uh oh!

ciarmcom commented Sep 30, 2019

Uh oh!

0xc0170 left a comment • 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!

ccli8 commented Sep 30, 2019

Uh oh!

ccli8 commented Oct 2, 2019

Uh oh!

bulislaw left a comment

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!

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!

Choose a reason for hiding this comment

Uh oh!

ccli8 commented Oct 3, 2019

Uh oh!

Patater left a comment

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!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

bulislaw commented Oct 7, 2019

Uh oh!

ccli8 commented Oct 8, 2019

Uh oh!

ccli8 commented Oct 8, 2019

Uh oh!

ccli8 commented Oct 18, 2019

Uh oh!

0xc0170 commented Oct 18, 2019

Uh oh!

0xc0170 commented Oct 18, 2019

Uh oh!

mbed-ci commented Oct 18, 2019

Uh oh!

ccli8 commented Oct 21, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

0xc0170 commented Oct 23, 2019

Uh oh!

0xc0170 commented Oct 23, 2019

Uh oh!

ccli8 commented Oct 28, 2019

Uh oh!

0xc0170 commented Oct 28, 2019

Uh oh!

mbed-ci commented Oct 28, 2019

Uh oh!

0xc0170 commented Oct 28, 2019

Uh oh!

ccli8 commented Oct 29, 2019

Uh oh!

ccli8 commented Sep 30, 2019 •

edited

Loading

0xc0170 left a comment •

edited

Loading

ccli8 commented Oct 21, 2019 •

edited

Loading