Suspicious code in revision 867140 (Part 2)
mcguire at kde.org
Sun Oct 5 18:14:22 BST 2008
On Saturday 04 October 2008 12:44:10 Christoph Bartoschek wrote:
> - kdepim/kmail/kmcomposewin.cpp:2918
> autoDelete is always true from line 2897 on.
Looks like a false positive. I guess the code is there so that autoDelete can
be changed to false if the need arises.
> - kdepim/kmail/actionscheduler.cpp:731
> Line 723 indicates that msg can be NULL here.
> - kdepim/kmail/kmfolderdir.cpp:288
> A final else after line 281 would prevent uninitialized folderType here.
> - kdepim/kmail/kmfoldermaildir.cpp:954
> The comma-operator here? Maybe the ,0 should be one parenthesis to the
Fixed (by removing ,0, I have no idea why that was added at all, the function
doesn't accept a second parameter)
> - kdepim/kmail/objecttreeparser.cpp:559
> Line 551 indicates that cleartextData might be NULL here.
> - kdepim/kmail/objecttreeparser.cpp:491
> Line 391 indicates that cryptProto might be NULL here.
I don't understand writeOpaqueOrMultipartSignedData() good enough to judge if
these are indeed problems, and I can't fix them.
> - kdepim/kmail/partNode.cpp:332
> otherState has type KMMsgSignatureState. KMMsgEncryptionProblematic
> however is of type KMMsgEncryptionState.
> - kdepim/kmail/kmcommands.cpp:1925 and others
> Line 1879 indicates that mDestFolder can be NULL here.
> - kdepim/kmail/antispamwizard.cpp:167
> Line 153 indicates that mVirusRulesPage can be NULL here.
False positive, the code here is in "if ( mMode == AntiVirus )" of line 163,
and line 127 indicates that mVirusRulesPage is not 0 in this case.
> - kdepim/kmail/recipientseditor.cpp:625
> If there is no it.hasNext() then line is uninitialized. maybe line 617
> should initialize with NULL.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 194 bytes
Desc: This is a digitally signed message part.
More information about the kde-core-devel