RTEMS | bsps/stm32h7: make errata configurable (!184)
Karel Gardas (@kgardas)
gitlab at rtems.org
Thu Aug 15 07:14:28 UTC 2024
Karel Gardas commented: https://gitlab.rtems.org/rtems/rtos/rtems/-/merge_requests/184#note_111068
Thanks for the review and suggestion, but even with this I'm still not able to make errata.yaml work as defined here:
```
SPDX-License-Identifier: CC-BY-SA-4.0 OR BSD-2-Clause
build-type: option
actions:
- get-string: null
- split: null
- env-append: ABI_FLAGS
copyrights:
- Copyright (C) 2024 Karel Gardas
default:
- enabled-by: STM32H7_ERRATA_ENABLE_FMC_282
value: -mno-unaligned-access
- enabled-by: true
value: ""
enabled-by: true
links: []
name: STM32H7_ERRATA
format: '{}'
prepare-build: null
prepare-configure: null
type: build
```
during the configure stage it simply does nothing.
Due to this and due to the fact that this whole is half-baked solution I'm considering abandoning this and even reverting original:
```
commit d81bf04501cb83de4586f431ae91123e4965d86a (origin/main, origin/HEAD, main)
Author: Karel Gardas <karel at functional.vision>
Date: Fri Aug 9 15:39:11 2024 +0200
bsps/stm32h7: Workaround STM32 H7 errata by enforcing aligned access only
The patch changes compiler ABI flags option and adds -mno-unaligned-access
there. Usage of this option is also officially recommended way
by the errata document how to work around the issue. For more details
about the errata please see referenced issue number.
Closes #5110.
```
the thing is that this issue is quite complex to fix as a whole and currently I lack a time for that. What do you prefer? Half-baked solution or rather revert completely? -- with perhaps some not to H7 BSP doc about a need to investigate erratas?
--
View it on GitLab: https://gitlab.rtems.org/rtems/rtos/rtems/-/merge_requests/184#note_111068
You're receiving this email because of your account on gitlab.rtems.org.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/bugs/attachments/20240815/d3554574/attachment-0001.htm>
More information about the bugs
mailing list