A potential bug of memory leak

[ash1852]

hello,I found a potential bug in source code,would you help me check if this bug is true?thank you very much.

==============================================================================

step 1 :
In file pccts/support/set/set.c , function set_pdq line 786 :
Allocate memory to q

step 2 :
In file pccts/support/set/set.c , function set_pdq line 787 :
Select the false branch at this point (q==null is false)

step 3 :
In file pccts/antlr/mrhoist.c , function MR_dumpRuleSet line 3016 :
Function set_pdq executes and stores the return value to cursor

step 4 :
In file pccts/antlr/mrhoist.c , function MR_dumpRuleSet line 3018 :
Select the true branch at this point (cursor!=null is true)

step 5 :
In file pccts/antlr/mrhoist.c , function MR_dumpRuleSet line 3020 :
Select the true branch at this point (RulePtr==null is true), the allocated memory is leaked

[sarnold]

Sounds like some valgrind might help, although the vendored pccts is quite stale :/

Did you happen to check if there's a more recent source available? I've been extra(extra) busy for a while, so I may not get back to this very soon. That said, as long as pccts doesn't hork up the build I wouldn't worry too much.

[sarnold]

So, I do have a really old branch with the last pccts update available at the time - MR33, which I just tried rebasing on top of master (and it came out fairly broken => pccts-MR33-rebase branch).

Sadly that's about all the time I have for now, but feel free to poke at it. If you can find an error that propagates to cccc, then please open a new bug. Thanks!