On Mar 9, 2011, at 2:58 PM, Dimitry Andric wrote:

> On 2011-03-09 18:21, Argyrios Kyrtzidis wrote:
>> URL: http://llvm.org/viewvc/llvm-project?rev=127339&view=rev
>> Log:
>> Introduce '-chain-include' option to specify headers that will be converted 
>> to chained PCHs in memory
>> without having to use multiple runs and intermediate files.
> ...
>> Added: cfe/trunk/lib/Serialization/ChainedIncludesSource.cpp
>> URL: 
>> http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Serialization/ChainedIncludesSource.cpp?rev=127339&view=auto
> 
> ...
>> +  for (unsigned i = 0, e = includes.size(); i != e; ++i) {
>> +    bool firstInclude = (i == 0);
> ...
>> +      for (unsigned i = 0, e = serialBufs.size(); i != e; ++i) {
>> +        bufs.push_back(llvm::MemoryBuffer::getMemBufferCopy(
>> +                              
>> llvm::StringRef(serialBufs[i]->getBufferStart(),
>> +                                              
>> serialBufs[i]->getBufferSize())));
>> +      }
>> +      std::string pchName = includes[i-1];
>> +      llvm::raw_string_ostream os(pchName);
>> +      os<<  ".pch"<<  i-1;
> 
> These nested for loops, both having 'unsigned i' variables, make g++
> complain:
> 
> tools/clang/lib/Serialization/ChainedIncludesSource.cpp: In static member 
> function 'static clang::ChainedIncludesSource* 
> clang::ChainedIncludesSource::create(clang::CompilerInstance&)':
> tools/clang/lib/Serialization/ChainedIncludesSource.cpp:122: warning: name 
> lookup of 'i' changed
> tools/clang/lib/Serialization/ChainedIncludesSource.cpp:67: warning:   
> matches this 'i' under ISO standard rules
> tools/clang/lib/Serialization/ChainedIncludesSource.cpp:117: warning:   
> matches this 'i' under old rules
> 
> If the "std::string pchName = includes[i-1];" and "os<< ".pch"<< i-1;"
> lines are supposed to use the 'outer' i, could you maybe just rename the
> 'inner' i, to shut up the warning?

Done in r127370.


_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

Reply via email to