[nrf fromlist] soc: add ironside boot report #2752
Conversation
|
|
||
| cpuflpr_code_partition: partition@f4000 { | ||
| reg = <0xf4000 DT_SIZE_K(48)>; | ||
| }; |
There was a problem hiding this comment.
FYI, I intend to introduce a "memory map" DTS file that both APP and RAD use soon.
This is so that the app can send the correct vector table address to IRONside SE.
We need to make sure we don't conflict with each other more than necessary somehow.
There was a problem hiding this comment.
I suggest your change goes in first, then I adapt to that. I need this for the random numbers anyway.
soc/nordic/nrf54h/ironside_boot_report/include/nrf/ironside_boot_report.h
Outdated
Show resolved
Hide resolved
soc/nordic/nrf54h/ironside_boot_report/include/nrf/ironside_boot_report.h
Outdated
Show resolved
Hide resolved
hakonfam
force-pushed
the
boot-report-ds
branch
2 times, most recently
from
233d4c6 to
ba5eedb
Compare
hakonfam
force-pushed
the
boot-report-ds
branch
2 times, most recently
from
8c81397 to
5cf2744
Compare
| /** See @ref ironside_se_boot_report_uicr_error */ | ||
| struct ironside_se_boot_report_uicr_error uicr_error_description; | ||
| /** Data passed from booting local domain to local domain being booted */ | ||
| uint8_t local_domain_context[IRONSIDE_SE_BOOT_REPORT_LOCAL_DOMAIN_CONTEXT_SIZE]; |
There was a problem hiding this comment.
I see that it's optional for this to be populated.
So I assume we need a local_domain_context_len to communicate how many bytes were written?
(0 when cpusec boots cpuapp for instance)
There was a problem hiding this comment.
Alternatively we could make this mandatory.
There was a problem hiding this comment.
IMO there is a contract between the image being booted and the image that boots in this situation. For instance, for sec -> app we know that there is no local domain context.
For app -> rad there would be a struct that would communicate within in whether or not there is data of interest within the local domain context.
soc/nordic/nrf54h/ironside_se/include/nrf/ironside_se_boot_report.h
Outdated
Show resolved
Hide resolved
soc/nordic/nrf54h/ironside_se/include/nrf/ironside_se_boot_report.h
Outdated
Show resolved
Hide resolved
| config SOC_NRF54H20_IRONSIDE_SE_BOOT_REPORT | ||
| bool "Nordic IRONside SE boot report" | ||
| default y if SOC_NRF54H20_CPUAPP | ||
| depends on SOC_NRF54H20_IRON |
There was a problem hiding this comment.
Right now we depend on building for cpuapp as well - I guess there is no other core that sets the IRON config until cpurad gets added, but should we add a depends on regardless for correctness? Better to see it fail at build time than run time if/when cpurad gets added
There was a problem hiding this comment.
We use this for CPSEC, I will see if there is some other way to get around that.
There was a problem hiding this comment.
did we decide on ironside_se vs. ironside/se for the directory layout?
There was a problem hiding this comment.
Should be ironside/se IIRC
There was a problem hiding this comment.
I see the point for consistency. But we are under a 'nrf54h' folder here, which dictates the "ironside" variant being used. I will update it anyway
hakonfam
force-pushed
the
boot-report-ds
branch
4 times, most recently
from
91ce46f to
c83c88c
Compare
hakonfam
force-pushed
the
boot-report-ds
branch
from
c83c88c to
14dc342
Compare
hakonfam
changed the title
hakonfam
force-pushed
the
boot-report-ds
branch
from
14dc342 to
1fae501
Compare
Also update memory map to leverage unused MRAM and move sysctrl IPC to RAM20 to free global RAM. Upstream PR #: 88932 Signed-off-by: Håkon Amundsen <[email protected]>
hakonfam
force-pushed
the
boot-report-ds
branch
from
1fae501 to
3228627
Compare
|
non-empty PR description