[cfe-commits] r148392 - in /cfe/trunk/test: CodeGen/char-literal.c CodeGen/string-literal-short-wstring.c Lexer/char-literal-encoding-error.c Lexer/char-literal.cpp Lexer/constants.c Lexer/utf8-char-literal.cpp Lexer/wchar.c

[Nico Weber]

Thanks! Looks good.

On Thu, Jan 19, 2012 at 7:34 PM, Seth Cantrell <seth.cantrell at gmail.com> wrote:
> Here's a patch that fixes that typo and adds a few tests
>
>
> On Jan 18, 2012, at 10:37 AM, Nico Weber wrote:
>
>> On Wed, Jan 18, 2012 at 4:27 AM, Seth Cantrell <seth.cantrell at gmail.com> wrote:
>>> +  // Should get concatonated characters
>>
>> typo 'concatonated'
>>
>>> -  // Should pick last character and store its lowest byte.
>>> -  // This does not match gcc, which takes the last character, converts it to
>>> -  // utf8, and then picks the second-lowest byte of that (they probably store
>>> -  // the utf8 in uint16_ts internally and take the lower byte of that).
>>> -  // CHECK-C: store i8 48
>>> -  // CHECK-CPP0X: store i8 48
>>> -  char c = '\u1120\u0220\U00102030';
>>> -
>>
>> You're removing these codegen tests but aren't them replacing with
>> anything that tests this input. Is this intentional?
>>