Svadev mailing list

[Baozeng <sploving1 AT gmail.com>]

Dear John,

When I tested BBC (mainline) as the following:
 clang -fmemsafety -bbc test/core/buffer-002.c -o buffer

It reported:

'common' global must have a zero initializer!
{ [1024 x i8], [1016 x i8], { i32, i32* } }* @array

I think that is because the linkage is not set correctly.

The related code is in lib/BaggyBoundsChecks/BaggyBoundsChecks.cpp


GlobalVariable *GV_new = new GlobalVariable (*(GV->getParent()),
                                                 newType,
                                                 GV->isConstant(),
                                                 GV->getLinkage(),
                                                 c,
                                                 "baggy." + GV->getName());

Which kind of linkage should be used?  The mainline LLVM has strict type 
checks.

2012/9/23 Baozeng 
<sploving1 AT gmail.com>:
> 2012/9/22 John Criswell 
> <criswell AT illinois.edu>:
>> On 9/21/12 9:58 PM, Baozeng wrote:
>>> 2012/9/18 John Criswell 
>>> <criswell AT illinois.edu>:
>>>> Dear All,
>>>>
>>>> I'm writing to let everyone know that SAFECode mainline (which works with
>>>> LLVM mainline) is now ready for development.  All SAFECode developers 
>>>> should
>>>> submit new patches and commits to mainline SAFECode; we're freezing the
>>>> release_30 branch of SAFECode.
>>>>
>>>> We'll be doing periodic merges with Clang and LLVM mainline to make sure 
>>>> we
>>>> stay working with the latest and greatest LLVM.  The current plan is to 
>>>> make
>>>> a release_32 branch of SAFECode when LLVM 3.2 is released.
>>>>
>>>> Because work on LLVM mainline is on-going, I won't update the web docs to
>>>> tell people to use mainline yet as we'll get occasional breakage.  Users 
>>>> who
>>>> want something solid should continue using the release_30 branch for now;
>>>> use mainline SAFECode at your own risk.
>>>>
>>>> Baozeng, I haven't tested the Baggy Bounds Checking functionality you've
>>>> added.  When you get a chance, can you please look at it and make sure 
>>>> that
>>>> the changes are correct?
>>>>
>>> Okay.
>>> I download maintain llvm, poolalloc, and SAFECode
>>> and when compiling it, it shows the following error:
>>> GCall.cpp:1046: error：‘constructAlignmentFromInt’ is not a member of
>>> ‘llvm::Attribute’.
>>>
>>> I think it should be changed to llvm::Attributes, am I right?
>>
>> There was a change to the attributes API that broke compilation of the
>> Baggy Bounds instrumentation pass. I fixed it so that it compiles again,
>> but I haven't tested it. Can you test it and make sure I didn't break
>> anything?
>>
> Okay.
>> I also merged in mainline Clang to bring in fixes to use the new
>> attributes API. The build now works for me on Linux.
>>
>> -- John T.
>>
>>>
>>>> Ott, I've been holding off on your patches until mainline SAFECode clang 
>>>> was
>>>> ready.  Now that it is, we can begin reviewing your optimization patches
>>>> again if you'd like.  I think there's one optimization that I need to 
>>>> review
>>>> since you've submitted changes. I'll take a look at that either today or
>>>> tomorrow.
>>>>
>>>> Thanks to John Mainzer for his work on integrating the new mainline 
>>>> version
>>>> of Clang into SAFECode.
>>>>
>>>> If anyone has any questions or comments, please feel free to email the 
>>>> list.
>>>>
>>>> -- John T.
>>>>
>>>
>>>
>>
>
>
>
> --
>      Best Regards,
>                                                                  Baozeng 
> Ding
>                                                                  
> OSTG,NFS,ISCAS



-- 
     Best Regards,
                                                                 Baozeng Ding
                                                                 
OSTG,NFS,ISCAS