thanks<br><br><div class="gmail_quote">On Mon, Sep 27, 2010 at 2:00 AM, Michael Spencer <span dir="ltr"><<a href="mailto:bigcheesegs@gmail.com">bigcheesegs@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
On Sun, Sep 26, 2010 at 6:47 PM, Nathan Jeffords<br>
<div><div></div><div class="h5"><<a href="mailto:blunted2night@gmail.com">blunted2night@gmail.com</a>> wrote:<br>
> On Sun, Sep 26, 2010 at 3:28 PM, Michael Spencer <<a href="mailto:bigcheesegs@gmail.com">bigcheesegs@gmail.com</a>><br>
> wrote:<br>
>><br>
>> On Sun, Sep 26, 2010 at 5:27 PM, Nathan Jeffords<br>
>> <<a href="mailto:blunted2night@gmail.com">blunted2night@gmail.com</a>> wrote:<br>
>> > Hi guys,<br>
>> > While trying to get dwarf debugging information to work with Win32 COFF<br>
>> > targets, I came across a couple of issues with the current<br>
>> > implementation<br>
>> > of WinCOFFObjectWriter. Emitting empty section causes debug information<br>
>> > to<br>
>> > invalid, as the presence of certain debug section implies available<br>
>> > information, and emission of labels as symbols confused gdb about the<br>
>> > structure of the program.<br>
>> > The attached patch allows the WinCOFFObjectWriter to drop empty<br>
>> > sections,<br>
>> > and label symbols. It converts relocations targeted at symbols into<br>
>> > relocations relative to the containing section.<br>
>> > I have not run the tests in the '/test/MC/COFF', but have run through<br>
>> > the<br>
>> > unit-test framework for my compiler and generated some text executables<br>
>> > with<br>
>> > clang and everything looks to be in order. It may change the output of<br>
>> > the '/test/MC/COFF' if the string references are emitted as<br>
>> > temporaries.<br>
>> > -Nathan<br>
>><br>
>> The patch looks good, but please follow the LLVM coding conventions.<br>
>> Also, clang emits warnings on all of the signed-unsigned comparisons,<br>
>> and the order of initialization for COFFSymbol::MCData.<br>
>><br>
>> The test fails, but that's only because it is looking for binary<br>
>> equivalence, just llc the test and run coff-dump to re update it. I<br>
>> ran the test-suite with clang -integrated-as with and without your<br>
>> patch and I got identical results.<br>
>><br>
>> So fix the coding conventions, warnings, and update the tests and it's<br>
>> good. I don't mind doing this if you would prefer.<br>
>><br>
>> - Michael Spencer<br>
><br>
> If you don't mind then please make the appropriate changes. Thanks.<br>
> -Nathan<br>
<br>
</div></div>Committed in r114823.<br>
<font color="#888888"><br>
- Michael Spencer<br>
</font></blockquote></div><br>