Add flash metadata #308
Add flash metadata #308
Conversation
Pull reviewers statsStats of the last 120 days for UWOrbital:
|
| #define MAX_FNAME_LINENUM_SIZE 128U | ||
| #define MAX_FNAME_LINENUM_SIZE 150U |
There was a problem hiding this comment.
not sure where this is from
There was a problem hiding this comment.
I think this was from one of Natvaj's logging prs
There was a problem hiding this comment.
was the logging pr merged? if so, try pulling from main
…mware into kashifb/flash-metadata
| #define MAX_FNAME_LINENUM_SIZE 128U | ||
| #define MAX_FNAME_LINENUM_SIZE 150U |
There was a problem hiding this comment.
I think this was from one of Natvaj's logging prs
| int32_t blUartWriteBufferLen = | ||
| snprintf(blUartWriteBuffer, BL_MAX_MSG_SIZE, "Failed to init flash, error code: %d\r\n", errCode); | ||
| if (blUartWriteBufferLen < 0) { | ||
| blUartWriteBytes(BL_UART_SCIREG, strlen("Error with processing message buffer length\r\n"), |
There was a problem hiding this comment.
bl error logging should be identical to obc err logging (LOG_ERR_CODE(x)), but ill let @ThenujaL handle this in his PR since hes already porting the logging stuff over
obc/bl/bl_main.c
Outdated
|
|
||
| blFlashFapiInitBank(0U); | ||
|
|
||
| errCode = blFlashFapiBlockErase(METADATA_START_ADDRESS, METADATA_SIZE_BYTES - 1); |
There was a problem hiding this comment.
why - 1? I dont think we do that anywhere else
There was a problem hiding this comment.
metadata is from 13fff8 to 13ffff (8). currently the erase function deletes the entire sector of the address. 13fff8 + 8 = 140000, so without the -1 it would delete the entire sector starting at 140000 (undefined behavior)
| #include <stdint.h> | ||
|
|
||
| typedef enum { | ||
| BL_UART_SCIREG_1 = 0, | ||
| BL_UART_SCIREG_2 = 1, | ||
| BL_UART_SCIREG = 0, |
There was a problem hiding this comment.
I'm confused why this enum needs to exist anymore since we only have (and will ever have?) 1 value
There was a problem hiding this comment.
yea i just added it from the old asynch logging PR. i can remove it
| #define MAX_FNAME_LINENUM_SIZE 128U | ||
| #define MAX_FNAME_LINENUM_SIZE 150U |
There was a problem hiding this comment.
was the logging pr merged? if so, try pulling from main
obc/bl/source/bl_uart.c
Outdated
There was a problem hiding this comment.
why change this? this layer supports both uart ports. why change it to only support 1?
There was a problem hiding this comment.
i think the synchronous logging PR implemented this change, it was reverted but we found that the BL didn't work properly on the Rev1 without the port changes from that PR. i could use #if directive to check if board type is OBC_REVISION_1 nd then only use the one port from rev1
Purpose
Read below
Link to notion task: https://www.notion.so/uworbital/Modify-flash-layout-to-store-app-metadata-b4c9aab740cc4f65be1b6852e9623190?pvs=4
New Changes
Testing
-With BL, and app sent over UART using the metadata test example:
-Without BL, app used over uniflash using the metadata test example
-Erase test (flashing 2 binaries using BL consecutively), the two bin sizes should be different
Outstanding Changes
-Unneeded currently IMO unless we plan on storing more than checksum + some cmake definitions later on
NOTE: Flashing the BL and then the app binary using uniflash (instead of sending the app binary through serial) will cause the readAppMetadata function to segfault if DENABLE_BL_BYPASS is set to false. This is because the memory addresses will not have anything in them, however since DENABLE_BL_BYPASS will be set to 0, the function will not return no metadata err code and instead try to access the memory