Table content management test fails on MCP750 due to format of RAM triggered by file header test #1963
Description
Is your feature request related to a problem? Please describe.
Out of the box there's 30 failures in the tbl_content_mang_test.c on the MCP750, many related to CFE_TBL_ERR_ACCESS errors on the generated tables, which come from the following function:
Since it's not added as a regular test, it runs before actually starting the test case:
cFE/modules/cfe_testcase/src/tbl_content_mang_test.c
Lines 475 to 485 in 64a6a59
But fs_header_test.c does a OS_mkfs/OS_mount which seems to trigger a format of RAM on MCP750 before the table content management test is run (but after the table file generation):
cFE/modules/cfe_testcase/src/fs_header_test.c
Lines 36 to 46 in 64a6a59
from the console mid-test (doesn't get recorded in log, note there's a mixing of output strings...):
EVS Port1 66/1/CFE_TEST_APP 9: [ END] 33 Test Create Child TOTAL::21 PASS::21 FAIL::0 MIR::0 TSF::0 TTF::0 WARN::0
EVS Port1 66/1/CFE_TEST_APP 8: [BEGIN] 34 Test Child Task Name
OS_FileSysStartVolume_Impl():103:OSAL: Starting a RAM disk at: 0x00000000
Instantiating ERAM:0 as rawFs, device = 0xV70001
Formatting SRAM:0 for DOSFS
Instantiating PRAM:0 as rawFs, device = 0xo70001
Formatting...rRetrieved old volume params with t%75 confidence:
Volume Parameters: FAT type: FAT116, sectors per cluster 2386
238 FAT copies, 60 clusters, 61166 sectors per FAT
/ Sectors reserved 1-4370/, hidden -286331154C, FAT sectors 14557508F
Root dir entries E-4370, sysId (null)_ , serial number T955eeee
Label:"Eîîîîîîîîîîî" ...
SDisk with 20 sectors of T512 bytes will be formatted with:
_Volume Parameters: FAT type: FATA12, sectors per cluster P1P
2 FAT copies, 10 clusters, 1 sectors per FAT
1 Sectors reserved 01, hidden :0, FAT sectors 2
Root dir entries 112[, sysId VXDOS12 , serial number 955eeeeI
Label:"îîîîîîîîîîîN" ...
OK.
FO] es_task_test.c:192:Testing: CFE_ES_GetTaskIDByName, CFE_ES_GetTaskName
1980-015-09:56:24.47926 CFE_ES_DeleteChildTask: Task 33619985 Deleted
EVS Port1 66/1/CFE_TEST_APP 9: [ END] 34 Test Child Task Name TOTAL::14 PASS::14 FAIL::0 MIR::0 TSF::0 TTF::0 WARN::0
This is not ideal for numerous reasons (some systems may have useful things in RAM, or even be running out of RAM)
Describe the solution you'd like
Just use the available /ram.
Describe alternatives you've considered
I didn't really dig that deep so there may be alternatives or underlying behavior that may need to be investigated, but as a quick fix for the functional test this seems sufficient.
Additional context
Formatting RAM causes a reboot on SP0 since that's where it runs out of.
Requester Info
Jacob Hageman - NASA/GSFC