Add CRC checks to Svc.PrmDb by odacindy-fprime · Pull Request #4586 · nasa/fprime

odacindy-fprime · 2025-12-20T00:18:21Z


*Related Issue(s)*
*Has Unit Tests (y/n)*
*Documentation Included (y/n)*
*Generative AI was used in this contribution (y/n)*

Change Description

Save and read a CRC to validate the contents of the PrmDb file.

Rationale

Provides protection against corruption

Testing/Review Recommendations

made minor updates to TestFile::open, seek, read, and write in order track to the StaticData::data.positionResult, which is used by TestFile::position() to return the current position of the pointer in the paramFile.
added file CRC to the beginning of the paramFile
added a few unit test cases to check CRC_SIZE, CRC, and CRC_BUFFER on PrmFileReadError, and CRC_PLACE on PrmWriteError; and PrmFileBadCrc

Future Work

none

AI Usage (see policy)

…uns, but without changing the existing unit tests, when I run them, some are now breaking, so will checkpoint now before making changes to existing unit tests

- Change the type for the buff argument in the computeCrc() declaration from "co nst U8*" to "const BYTE*" Svc/PrmDb/test/ut/PrmDbTester.cpp - Update PrmDbTester::runFileReadError() for the following: - add a test case for CRC_SIZE when reading the paramFile. - add test cases for CRC and CRC_BUFFER for a failed paramFile read due to the file not existing and the file not being opened - add a test case for FILE_DATA_ERROR having a bad file CRC when reading a paramFile - Update PrmDbTester::runFileWriteError() for the following: - add test cases for a write failures for CRC when sending the PRM_SAVE_FILE command. - Update PrmDbTester::PrmDbTestFile::read()'s FILE_SIZE_ERROR case, so that it only increments size (we increment the size to introduce a size error on a paramFile read, which is used for several unit test cases) if size is not zero. If we don't do this, then the unit test will go into an infinite loop, since we may loop multiple times when reading the CRC buffer (which is part of the paramFile upon which we calculate the CRC. In other words the CRC buffer is the paramFile minus the CRC). The CRC buffer contents contain the param database entries.

- add CURR_POSITION, SEEK_ZERO, and SEEK_POSITION to PrmWriteError enum - added return status checking for paramFile.seek() and paramFile.position() calls in PrmDbImpl::PRM_SAVE_FILE_cmdHandler()

Svc/PrmDb/PrmDbImpl.cpp


+U32 PrmDbImpl::computeCrc(U32 crc, const BYTE* buff, FwSizeType size) {
+    for (FwSizeType byte = 0; byte < size; byte++) {
+        crc = static_cast<U32>(update_crc_32(crc, static_cast<char>(buff[byte])));


Svc/PrmDb/PrmDbImpl.cpp

+    for (FwSizeType byte = 0; byte < size; byte++) {
+        crc = static_cast<U32>(update_crc_32(crc, static_cast<char>(buff[byte])));
+    }   
+    return crc;


Svc/PrmDb/PrmDbImpl.cpp

+    stat = paramFile.write(reinterpret_cast<const U8*>(&crc), writeSize, Os::File::WaitType::WAIT);
+    if (stat != Os::File::OP_OK) {
+        this->log_WARNING_HI_PrmFileWriteError(PrmWriteError::CRC_REAL, 0, stat);
+        this->cmdResponse_out(opCode, cmdSeq, Fw::CmdResponse::EXECUTION_ERROR);


Svc/PrmDb/PrmDbImpl.cpp

+    paramFile.seek(static_cast<FwSignedSizeType>(currPosInParamFile), Os::File::SeekType::ABSOLUTE);
+    if (stat != Os::File::OP_OK) {
+        this->log_WARNING_HI_PrmFileWriteError(PrmWriteError::SEEK_POSITION, 0, stat);
+        this->cmdResponse_out(opCode, cmdSeq, Fw::CmdResponse::EXECUTION_ERROR);


Svc/PrmDb/PrmDbImpl.cpp

+    paramFile.seek(static_cast<FwSignedSizeType>(currPosInParamFile), Os::File::SeekType::ABSOLUTE);
+    if (stat != Os::File::OP_OK) {
+        this->log_WARNING_HI_PrmFileWriteError(PrmWriteError::SEEK_POSITION, 0, stat);
+        this->cmdResponse_out(opCode, cmdSeq, Fw::CmdResponse::EXECUTION_ERROR);


Svc/PrmDb/PrmDbImpl.cpp

+    // seek to beginning and write CRC value
+    paramFile.seek(0, Os::File::SeekType::ABSOLUTE);
+    if (stat != Os::File::OP_OK) {
+        this->log_WARNING_HI_PrmFileWriteError(PrmWriteError::SEEK_ZERO, 0, stat);


Svc/PrmDb/PrmDbImpl.cpp

+    // seek to beginning and write CRC value
+    paramFile.seek(0, Os::File::SeekType::ABSOLUTE);
+    if (stat != Os::File::OP_OK) {
+        this->log_WARNING_HI_PrmFileWriteError(PrmWriteError::SEEK_ZERO, 0, stat);


LeStarch

Minor change to the CRC calculation. Hopefully the tests will make this switch easy.

Svc/PrmDb/PrmDbImpl.cpp

Svc/PrmDb/PrmDbImpl.hpp

…G printfs

Svc/PrmDb/PrmDbImpl.cpp

+    // Check for null pointer before dereferencing
+    if (buff == nullptr) {
+        // Return the input CRC unchanged if buffer is null
+        return crc;


Svc/PrmDb/PrmDbImpl.cpp

+        return crc;
+    }
+
+    // Check for zero size to avoid unnecessary processing


Svc/PrmDb/PrmDbImpl.cpp

+        return crc;
+    }
+
+    // Check for zero size to avoid unnecessary processing


Svc/PrmDb/PrmDbImpl.cpp

+
+    // Check for zero size to avoid unnecessary processing
+    if (size == 0) {
+        return crc;


Svc/PrmDb/PrmDbImpl.cpp

 }

+
+U32 PrmDbImpl::computeCrc(U32 crc, const BYTE* buff, FwSizeType size) {


…FER_SIZE. Also, reformatted code

thomas-bc · 2026-01-13T17:37:03Z

To fix the formatting, you can run this command:

git diff --name-only devel...HEAD | fprime-util format --stdin

default/config/PrmDbImplCfg.hpp

Svc/PrmDb/PrmDbImpl.cpp

odacindy-fprime added 4 commits December 5, 2025 17:56

issue-4270 Updated Svc/PrmDb to implement CRC. Currently builds and r…

95247fb

…uns, but without changing the existing unit tests, when I run them, some are now breaking, so will checkpoint now before making changes to existing unit tests

removed a status check since the current call doesn't returnn a failure

a513e73

fprime/Svc/PrmDb/PrmDb.fpp:

46a7f6c

- add CURR_POSITION, SEEK_ZERO, and SEEK_POSITION to PrmWriteError enum - added return status checking for paramFile.seek() and paramFile.position() calls in PrmDbImpl::PRM_SAVE_FILE_cmdHandler()

github-advanced-security bot found potential problems Dec 20, 2025

View reviewed changes

ran the code formatter on Svc/PrmDb

8a33f34

github-advanced-security bot found potential problems Dec 20, 2025

View reviewed changes

LeStarch requested changes Dec 20, 2025

View reviewed changes

Svc/PrmDb/PrmDbImpl.cpp Outdated Show resolved Hide resolved

Svc/PrmDb/PrmDbImpl.cpp Show resolved Hide resolved

Svc/PrmDb/PrmDbImpl.hpp Outdated Show resolved Hide resolved

odacindy-fprime added 3 commits January 11, 2026 16:31

fix some pull-request issues, UTs pass, but still need to remove DEBU…

abb65bc

…G printfs

remove debug prints

ca98ad8

remove a few more DEBUG prints

cae1f33

github-advanced-security bot found potential problems Jan 12, 2026

View reviewed changes

odacindy-fprime added 2 commits January 11, 2026 17:17

ran the code formatter

d93d832

Fixed deadcode caught by static analyzer for readSize = PRMDB_CRC_BUF…

732a753

…FER_SIZE. Also, reformatted code

fixed clang-format issue in default/config/PrmDbImplCfg.hpp

288f807

thomas-bc changed the title ~~Issue 4270 add crc chk to prm db~~ Add CRC checks to Svc.PrmDb Jan 22, 2026

LeStarch reviewed Feb 2, 2026

View reviewed changes

default/config/PrmDbImplCfg.hpp Outdated Show resolved Hide resolved

Remove unused PrmDb Config

2e9b29f

LeStarch previously approved these changes Feb 2, 2026

View reviewed changes

Fix formatting and one comment

ed52b62

thomas-bc dismissed LeStarch’s stale review via ed52b62 February 2, 2026 18:37

thomas-bc approved these changes Feb 2, 2026

View reviewed changes

Svc/PrmDb/PrmDbImpl.cpp Outdated Show resolved Hide resolved

LeStarch approved these changes Feb 4, 2026

View reviewed changes

LeStarch merged commit e9dfd3a into nasa:devel Feb 4, 2026
55 of 61 checks passed

		}


		U32 PrmDbImpl::computeCrc(U32 crc, const BYTE* buff, FwSizeType size) {

Conversation

odacindy-fprime commented Dec 20, 2025

Change Description

Rationale

Testing/Review Recommendations

Future Work

AI Usage (see policy)

Uh oh!

Check warning

Uh oh!

Check warning

Uh oh!

Uh oh!

Check warning

Uh oh!

Check warning

Check warning

Uh oh!

Check warning

Check warning

LeStarch left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Check warning

Check warning

Check warning

Check warning

Check notice

thomas-bc commented Jan 13, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants