[PATCH] Enable standalone-debug by default on FreeBSD

David Blaikie dblaikie at gmail.com
Thu May 8 14:29:58 PDT 2014


On Thu, May 8, 2014 at 2:05 PM, Eric Christopher <echristo at gmail.com> wrote:
> On Thu, May 8, 2014 at 12:20 PM, Ed Maste <emaste at freebsd.org> wrote:
>> On 8 May 2014 15:10, Eric Christopher <echristo at gmail.com> wrote:
>>> That's a pretty serious lldb bug. Have you filed it? What's the case where
>>> it shows up?
>>
>> It's PR 19676 - it's in the summary of D3654 :-)
>> http://llvm.org/bugs/show_bug.cgi?id=19676
>
> Aha. I saw a radar entry which is why I was curious.
>
>>
>> A thread on the issue is here:
>> http://lists.cs.uiuc.edu/pipermail/lldb-dev/2014-April/003800.html
>
> *sigh* I thought they'd fixed that. It's pretty bad.

Ugh. Didn't realize LLDB still had that assert text there. That's
grossly misleading.

>
>>
>> As it happens I was debugging clang's debuginfo generation when I
>> encountered it:
>>
>>  * thread #1: tid = 104525, 0x000000081191ff6e
>> libclangCodeGen.so`clang::CodeGen::CGDebugInfo::CreateType(this=0x0000000815c0a000,
>> BT=0x0000000815c3d400) + 30 at CGDebugInfo.cpp:391, stop reason =
>> breakpoint 1.1
>>     frame #0: 0x000000081191ff6e
>> libclangCodeGen.so`clang::CodeGen::CGDebugInfo::CreateType(this=0x0000000815c0a000,
>> BT=0x0000000815c3d400) + 30 at CGDebugInfo.cpp:391
>>    388  /// CreateType - Get the Basic type from the cache or create a new
>>    389  /// one if necessary.
>>    390  llvm::DIType CGDebugInfo::CreateType(const BuiltinType *BT) {
>> -> 391    unsigned Encoding = 0;
>>    392    StringRef BTName;
>>    393    switch (BT->getKind()) {
>>    394  #define BUILTIN_TYPE(Id, SingletonId)
>> (lldb) p BT
>> Assertion failed: (D && "Cannot get layout of forward declarations!"),
>> function getASTRecordLayout, file
>> ../tools/clang/lib/AST/RecordLayoutBuilder.cpp, line 2783.
>>
>> At the very least we need to have LLDB return an error to the user
>> rather than tripping the assertion, but it really needs to find the
>> definition debuginfo elsewhere.
>
> It does. It has the logic for it.
>
> -eric
> _______________________________________________
> cfe-commits mailing list
> cfe-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits



More information about the cfe-commits mailing list