Update CMSIS to 5.8.0 #14900
Update CMSIS to 5.8.0 #14900
Conversation
| * | ||
| * ----------------------------------------------------------------------------- | ||
| */ | ||
|
|
||
|
|
||
| .syntax unified | ||
|
|
||
| #include "rtx_def.h" |
There was a problem hiding this comment.
this is causing a problem as we somehow get some C/C++ headers to this .S file and it just errors with strange errors.
We ignored the inclusion previously because of this, will need to check how to fix it this time.
2902af9 to
14beb29
Compare
| @@ -22,7 +22,7 @@ | |||
| * limitations under the License. | |||
| */ | |||
|
|
|||
| #if !FEATURE_TFM | |||
| #if !TARGET_TFM | |||
There was a problem hiding this comment.
this one should be reverted as well, will check how it got in
14beb29 to
551bceb
Compare
|
@0xc0170, thank you for your changes. |
|
CI started |
|
|
| * Title: RTX derived definitions | ||
| * | ||
| * ----------------------------------------------------------------------------- | ||
| */ |
There was a problem hiding this comment.
This file failed CI due to missing SPDX header
There was a problem hiding this comment.
I am reporting this upstream, it should be fixed asap. I assume as this is already released and new version might take some time, we can patch this file manually. I'll check.
There was a problem hiding this comment.
Fix for upstream ARM-software/CMSIS_5#1238
Jenkins CI Test : ❌ FAILEDBuild Number: 1 | 🔒 Jenkins CI Job | 🌐 Logs & ArtifactsCLICK for Detailed Summary
|
9639d83 to
2b3bb8b
Compare
|
I rebased instead of fixup 🤦 I patched rtx_def, we had similar patch in place but they introduced this via the new version. I provided details in commit msg. I already asked cmsis team for clarification and will create an issue once I get details what should be fixed to get this fix upstream or our codebase needs some refactoring. Commit 89d9642 |
|
@0xc0170, thank you for your changes. |
|
I started another round of tests. I can locally build Gcc Arm without errors (I assume TFM related things might have issues or FPU - both things were patched previously in Mbed OS, I reverted some patches so lets see what we still need or do not need). ARMClang reports an error with one overload, I am checking it now. |
|
I fixed also locally ARMClang (removing old compat header, we should use CMSIS functionality instead) |
|
Aborting the current build and reruning CI |
We use preprocessor for asm files even for Armcc. If symbol is defined it's replaced by preprocessor, asembler would just see 1 or 0 in this case and errors: TARGET_M33\\irq_armv8mml.S", line 31: Error: A1185E: Symbol missing Use preprocessor instead.
CMSIS 5.8.0 requires for armcc to include compat header prior any other header. See Known issues for the release.
For FPU, use armasm to select fpu selection. This will be fixed upstream in the next version of CMSIS. Meanwhile, we use our local patch.
Taken from --cpu selection for armasm:
--cpu | {FPU}
Cortex-M4.fp.sp | VFPv4_SP_D16
Cortex-M7.fp.sp | FPv5-SP
Cortex-M7.fp.dp | FPv5_D16
Tracking issue: ARM-software/CMSIS_5#1266
Fixes error osErrorParameter that triggers our assert. RTX delay ignored 0 previously, it now returns osErrorParameter instead.
188b73f to
5e29db6
Compare
|
Fixed both |
See specific SHA for details.
|
SHAs added to importer. I'll rerun CI |
Jenkins CI Test : ✔️ SUCCESSBuild Number: 12 | 🔒 Jenkins CI Job | 🌐 Logs & ArtifactsCLICK for Detailed Summary
|
|
Lets get this in and monitor nightly over the weekend. |
|
This PR has dropped IAR support.... again.... :-( |
Summary of changes
Update CMSIS to 5.8.0. It brings lot of fixes, see https://github.com/ARM-software/CMSIS_5/releases/tag/5.8.0
We again had some breakages, mainly it was arm compat (workaround implemented).
Impact of changes
Migration actions required
Documentation
Pull request type
Test results
Reviewers