By default the ARM compiler is very chatty and it is hard to see "the wood for the trees". I would propose to disable the following ARM warnings (as far as I can tell none of these warnings are enabled on most ports): #68-D: integer conversion resulted in a change of sign #111-D: statement is unreachable #177-D: variable XXX was declared but never referenced #368-D: class XXX defines no constructor to initialize the following: YYY #830-D: function XXX "XXX::operator new" has no corresponding operator delete (to be called if an exception is thrown during initialization of an allocated object) #1293-D: assignment in condition
Created attachment 39268 [details] proposed path Also did a bit of refactoring so that the rules don't repeat for all ARM compiler targets.
Maybe we should add a comment in the build file about the warnings you are disabling? Or at least in the ChangeLog?
(In reply to comment #2) > Maybe we should add a comment in the build file about the warnings you are > disabling? Or at least in the ChangeLog? Eric, I will add the list of warnings to the ChangeLog when I commit. Thanks !
Created attachment 39272 [details] Updated ChangeLog Incorporating Eric's comment.
Do you really want to disable all these? #68-D: integer conversion resulted in a change of sign #111-D: statement is unreachable #177-D: variable XXX was declared but never referenced #368-D: class XXX defines no constructor to initialize the following: YYY #830-D: function XXX "XXX::operator new" has no corresponding operator delete (to be called if an exception is thrown during initialization of an allocated object) #1293-D: assignment in condition All of those sound useful (and likely real bugs), no?
(In reply to comment #5) > Do you really want to disable all these? > > #68-D: integer conversion resulted in a change of sign > #111-D: statement is unreachable > #177-D: variable XXX was declared but never referenced > #368-D: class XXX defines no constructor to initialize the following: YYY > #830-D: function XXX "XXX::operator new" has no corresponding operator delete > (to be called if an exception is thrown during initialization of an allocated > object) > #1293-D: assignment in condition > > All of those sound useful (and likely real bugs), no? Eric, I scanned trough the warnings and the corresponding code and I could not spot a real bug. The direction is definitively to get rid of all the warnings gradually (and take them out from this list), but restricting the list help to focus on the remaining warnings first.
Committed as http://trac.webkit.org/changeset/48250