[lldb-dev] UnicodeDecodeError for serialize SBValue description

Tamas Berghammer via lldb-dev lldb-dev at lists.llvm.org
Thu Apr 7 02:45:41 PDT 2016


LLDB supports adding data formatters without modifying the source code and
I would strongly prefer to go that way as we don't want each user of LLDB
to start adding data formatters to their own custom types. We have a pretty
detailed (but possible a bit outdated) description about how they work and
how you can add a new one here: http://lldb.llvm.org/varformats.html

Enrico: Is there any reason you suggested the data formatters written
inside LLDB over the python based ones?

On Thu, Apr 7, 2016 at 3:31 AM Jeffrey Tan via lldb-dev <
lldb-dev at lists.llvm.org> wrote:

> Thanks Enrico. This is very detailed! I will take a look.
> Btw: originally, I was hoping that data formatter can be added without
> changing the source code. Like giving a xml/json format file telling lldb
> the memory layout/structure of the data structure, lldb can parse the
> xml/json and deduce the formatting. This is approach used by data
> visualizer in VS debugger:
> https://msdn.microsoft.com/en-us/library/jj620914.aspx
> This will make adding data formatter more extensible/flexible. Any reason
> we did not take this approach?
>
> Jeffrey
>
> On Wed, Apr 6, 2016 at 11:49 AM, Enrico Granata <egranata at apple.com>
> wrote:
>
>>
>> On Apr 5, 2016, at 2:42 PM, Jeffrey Tan <jeffrey.fudan at gmail.com> wrote:
>>
>> Hi Enrico,
>>
>> Any suggestion/example how to add a data formatter for our own STL
>> string? From the output below I can see we are using our own "
>> *fbstring_core*" which I assume I need to write a type summary for this
>> type:
>>
>> frame variable corpus -T
>> (const string &const) corpus = error: summary string parsing error: {
>>   (std::*fbstring_core*<char>) store_ = {
>>     (std::*fbstring_core*<char>::(anonymous union))  = {
>>       (char [24]) small_ = "www"
>>       (std::fbstring_core<char>::MediumLarge) ml_ = {
>>         (char *) data_ = 0x0000000000777777
>> "H\x89U\xa8H\x89M\xa0L\x89E\x98H\x8bE\xa8H\x89��_U��D\x88e�H\x8bE\xa0H\x89��]U��H\x89�H\x8dE�H\x89�H\x89���
>> ��L\x8dm�H\x8bE\x98H\x89��IU��\x88]�L\x8be\xb0L\x89��
>>         (std::size_t) size_ = 0
>>         (std::size_t) capacity_ = 1441151880758558720
>>       }
>>     }
>>   }
>> }
>>
>>
>> Admittedly, this is going to be a little vague since I haven’t really
>> seen your code and I am only working off of one sample
>>
>> There’s going to be two parts to getting this to work:
>>
>> *Part 1 - Formatting fbstring_core*
>>
>> At a glance, an fbstring_core<char> can be backed by two representations.
>> A “small” representation (a char array), and a “medium/large"
>> representation (a char* + a size)
>> I assume that the way you tell one from the other is
>>
>> if (size == 0) small
>> else medium-large
>>
>> If my assumption is not correct, you’ll need to discover what the correct
>> discriminator logic is - the class has to know, and so do you :-)
>>
>> Armed with that knowledge, look in lldb
>> source/Plugins/Language/CPlusPlus/Formatters/LibCxx.cpp
>> There’s a bunch of code that deals with formatting llvm’s libc++
>> std::string - which follows a very similar logic to your class
>>
>> ExtractLibcxxStringInfo() is the function that handles discovering which
>> layout the string uses - where the data lives - and how much data there is
>>
>> Once you have told yourself how much data there is (the size) and where
>> it lives (array or pointer), LibcxxStringSummaryProvider() has the easy
>> task - it sets up a StringPrinter, tells it how much data to print, where
>> to get it from, and then delegates the StringPrinter to do the grunt work
>> StringPrinter is a nifty little tool - it can handle generating summaries
>> for different kinds of strings (UTF8? UTF16? we got it - is a \0 a
>> terminator? what quote character would you like? …) - you point it at some
>> data, set up a few options, and it will generate a printable representation
>> for you - if your string type is doing anything out of the ordinary, let’s
>> talk - I am definitely open to extending StringPrinter to handle even more
>> magic
>>
>> *Part 2 - Teaching std::string that it can be backed by an fbstring_core*
>>
>> At the end of part 1, you’ll probably end up with a
>> FBStringCoreSummaryProvider() - now you need to teach LLDB about it
>> The obvious thing you could do would be to go in CPlusPlusLanguage
>> ::GetFormatters() add a LoadFBStringFormatter(g_category) to it - and
>> then imitate - say - LoadLibCxxFormatters()
>>
>>     AddCXXSummary(cpp_category_sp, lldb_private::formatters::
>> FBStringCoreSummaryProvider, “fbstringcore summary provider", ConstString
>> (“std::fbstring_core<.+>"), stl_summary_flags, true);
>>
>> That will work - but what you would see is:
>>
>> (const string &const) corpus = error: summary string parsing error: {
>>   (std::*fbstring_core*<char>) store_ = “www"
>>
>>
>> You wanna do
>>
>> (lldb) log enable lldb formatters
>> (lldb) frame variable -T corpus
>>
>> It will list one or more typenames - the most specific one is the one you
>> like (e.g. for libc++ we get std::__1::string - this is how we tell
>> ourselves this is the std::string from libc++)
>> Once you find that typename, you’ll make a new formatter -
>> FBStringSummaryProvider() - and register that formatter with that very
>> specific typename
>>
>> All that FBStringSummaryProvider() has to do is get the “store_” member
>> (ValueObject::GetChildMemberWithName() is your friend) - and pass it down
>> to FBStringCoreSummaryProvider()
>>
>>
>> I understand this may seem a little convoluted and arcane at first - but
>> feel free to ask more questions, and I’ll try to help out!
>>
>> Thanks.
>> Jeffrey
>>
>> On Mon, Mar 28, 2016 at 11:38 AM, Enrico Granata <egranata at apple.com>
>> wrote:
>>
>>> This is kind of orthogonal to your problem, but the reason why you are
>>> not seeing the kind of simplified printing Greg is suggesting, is because
>>> your std::string doesn’t look like any of the kinds we recognize
>>>
>>> Specifically, LLDB data formatters work by matching against type names,
>>> and once they recognize a typename, then they try to inspect the variable
>>> in order to grab a summary
>>> In your example, your std::string exposes a layout that we are not
>>> handling - hence we bail out of the formatter and we fall back to the raw
>>> view
>>>
>>> If you want pretty printing to work, you’ll need to write a data
>>> formatter
>>>
>>> There are a few avenues. The obvious easy one is to extend the existing
>>> std::string formatter to recognize your type’s internal layout.
>>> If one were signing up for more infrastructure work, they could decide
>>> to try and detect shared library loads and load formatters that match with
>>> whatever libraries are being loaded.
>>>
>>> On Mar 28, 2016, at 9:47 AM, Greg Clayton via lldb-dev <
>>> lldb-dev at lists.llvm.org> wrote:
>>>
>>> So you need to be prepared to escape any text that can have special
>>> characters. A "std::string" or any container can contain special
>>> characters. If you are encoding stuff into JSON, you will either need to
>>> escape any special characters, or hex encode the string into ASCII hex
>>> bytes.
>>>
>>> In debuggers we often get bogus data because variables are not
>>> initialized, but the compiler tells us that a variable is valid in address
>>> range [0x1000-0x2000), but it actually is [0x1200-0x2000). If we read a
>>> variable in this case, a std::string might contain bogus data and the bytes
>>> might not make sense. So you always have to be prepared for bad data.
>>>
>>> If we look at:
>>>
>>>  store_ = {
>>>     = {
>>>      small_ = "www"
>>>      ml_ = (data_ =
>>>
>>> "��UH\x89�H�}�H\x8bE�]ÐUH\x89�H��H\x89}�H\x8bE�H\x89��~\xb4��\x90��UH\x89�SH\x83�H\x89}�H�u�H�E�H���\x9e���H\x8b\x18H\x8bE�H���O\xb4��H\x89ƿ\b",
>>> size_ = 0, capacity_ = 1441151880758558720)
>>>    }
>>>  }
>>> }
>>>
>>> We can see the "size_" is zero, and capacity_ is 1441151880758558720
>>> (which is 0x1400000000000000). "data_" seems to be some random pointer.
>>>
>>> On MacOSX, we have a special formatting code that displays std::string
>>> in CPlusPlusLanguage.cpp that gets installed in the LoadLibCxxFormatters()
>>> or LoadLibStdcppFormatters() functions with code like:
>>>
>>>    lldb::TypeSummaryImplSP std_string_summary_sp(new
>>> CXXFunctionSummaryFormat(stl_summary_flags,
>>> lldb_private::formatters::LibcxxStringSummaryProvider, "std::string summary
>>> provider"));
>>>    cpp_category_sp->GetTypeSummariesContainer()->Add(ConstString("std::__1::string"),
>>> std_string_summary_sp);
>>>
>>> Special flags are set on std::string to say "don't show children of this
>>> and just show a summary" So if a std::string contained "hello". So for the
>>> following code:
>>>
>>> std::string h ("hello");
>>>
>>> You should just see:
>>>
>>> (lldb) fr var h
>>> (std::__1::string) h = "hello"
>>>
>>> If you take a look at the normal value in the raw we see:
>>>
>>> (lldb) fr var --raw h
>>> (std::__1::string) h = {
>>>  __r_ = {
>>>    std::__1::__libcpp_compressed_pair_imp<std::__1::basic_string<char,
>>> std::__1::char_traits<char>, std::__1::allocator<char> >::__rep,
>>> std::__1::allocator<char>, 2> = {
>>>      __first_ = {
>>>         = {
>>>          __l = {
>>>            __cap_ = 122511465736202
>>>            __size_ = 0
>>>            __data_ = 0x0000000000000000
>>>          }
>>>          __s = {
>>>             = {
>>>              __size_ = '\n'
>>>              __lx = '\n'
>>>            }
>>>            __data_ = {
>>>              [0] = 'h'
>>>              [1] = 'e'
>>>              [2] = 'l'
>>>              [3] = 'l'
>>>              [4] = 'o'
>>>              [5] = '\0'
>>>              [6] = '\0'
>>>              [7] = '\0'
>>>              [8] = '\0'
>>>              [9] = '\0'
>>>              [10] = '\0'
>>>              [11] = '\0'
>>>              [12] = '\0'
>>>              [13] = '\0'
>>>              [14] = '\0'
>>>              [15] = '\0'
>>>              [16] = '\0'
>>>              [17] = '\0'
>>>              [18] = '\0'
>>>              [19] = '\0'
>>>              [20] = '\0'
>>>              [21] = '\0'
>>>              [22] = '\0'
>>>            }
>>>          }
>>>          __r = {
>>>            __words = {
>>>              [0] = 122511465736202
>>>              [1] = 0
>>>              [2] = 0
>>>            }
>>>          }
>>>        }
>>>      }
>>>    }
>>>  }
>>> }
>>>
>>> So the main question is why are our "std::string" formatters not kicking
>>> in for you. That comes down to a typename match, or the format of the
>>> string isn't what the formatter is expecting.
>>>
>>> But again, since you std::string can contain anything, you will need to
>>> escape any and all text that is encoded into JSON to ensure it doesn't
>>> contain anything JSON can't deal with.
>>>
>>> On Mar 27, 2016, at 9:20 PM, Jeffrey Tan via lldb-dev <
>>> lldb-dev at lists.llvm.org> wrote:
>>>
>>> Thanks Siva. All the DW_TAG_member related errors seems to go away after
>>> patching with your fix. The current problem is handling the decoding.
>>>
>>> Here is the correct decoding from gdb whic might be useful:
>>> (gdb) p corpus
>>> $3 = (const std::string &) @0x7fd133cfb888: {
>>>  static npos = 18446744073709551615, store_ = {
>>>    static kIsLittleEndian = <optimized out>,
>>>    static kIsBigEndian = <optimized out>, {
>>>      small_ = "www", '\000' <repeats 20 times>, "\024", ml_ = {
>>>        data_ = 0x777777 <std::_Any_data::_M_access<void
>>> folly::fibers::Baton::waitFiber<folly::fibers::FirstArgOf<facebook::servicerouter::RequestDispatcherBase<facebook::servicerouter::ThriftDispatcher>::prepareForSelection(facebook::servicerouter::DispatchContext&)::{lambda(folly::fibers::Promise<facebook::servicerouter::RequestDispatcherBase<facebook::servicerouter::ThriftDispatcher>::prepareForSelection(facebook::servicerouter::DispatchContext&)::SelectionResult>)#1},
>>> void>::type::value_type
>>> folly::fibers::await<facebook::servicerouter::RequestDispatcherBase<facebook::servicerouter::ThriftDispatcher>::prepareForSelection(facebook::servicerouter::DispatchContext&)::{lambda(folly::fibers::Promise<facebook::servicerouter::RequestDispatcherBase<facebook::servicerouter::ThriftDispatcher>::prepareForSelection(facebook::servicerouter::DispatchContext&)::SelectionResult>)#1}>(folly::fibers::FirstArgOf&&)::{lambda()#1}>(folly::fibers::FiberManager&,
>>> folly::fibers::FirstArgOf<folly::fibers::FirstArgOf<facebook::servicerouter::RequestDispatcherBase<facebook::servicerouter::ThriftDispatcher>::prepareForSelection(facebook::servicerouter::DispatchContext&)::{lambda(folly::fibers::Promise<facebook::servicerouter::RequestDispatcherBase<facebook::servicerouter::ThriftDispatcher>::prepareForSelection(facebook::servicerouter::DispatchContext&)::SelectionResult>)#1},
>>> void>::type::value_type
>>> folly::fibers::await<facebook::servicerouter::RequestDispatcherBase<facebook::servicerouter::ThriftDispatcher>::prepareForSelection(facebook::servicerouter::DispatchContext&)::{lambda(folly::fibers::Promise<facebook::servicerouter::RequestDispatcherBase<facebook::servicerouter::ThriftDispatcher>::prepareForSelection(facebook::servicerouter::DispatchContext&)::SelectionResult>)#1}>(folly::fibers::FirstArgOf&&)::{lambda()#1},
>>> void>::type::value_type)::{lambda(folly::fibers::Fiber&)#1}*>() const+25>
>>> "\311\303UH\211\345H\211}\370H\213E\370]ÐUH\211\345H\203\354\020H\211}\370H\213E\370H\211\307\350~\264\312\377\220\311\303UH\211\345SH\203\354\030H\211}\350H\211u\340H\213E\340H\211\307\350\236\377\377\377H\213\030H\213E\350H\211\307\350O\264\312\377H\211ƿ\b",
>>> size_ = 0,
>>>        capacity_ = 1441151880758558720}}}}
>>>
>>> Utf-16 does not seem to decode it, while 'latin-1' does:
>>>
>>> '\xc9'.decode('utf-16')
>>>
>>> Traceback (most recent call last):
>>>  File "<stdin>", line 1, in <module>
>>>  File
>>> "/mnt/gvfs/third-party2/python/55c1fd79d91c77c95932db31a4769919611c12bb/2.7.8/centos6-native/da39a3e/lib/python2.7/encodings/utf_16.py",
>>> line 16, in decode
>>>    return codecs.utf_16_decode(input, errors, True)
>>> UnicodeDecodeError: 'utf16' codec can't decode byte 0xc9 in position 0:
>>> truncated data
>>>
>>> '\xc9'.decode('latin-1')
>>>
>>> u'\xc9'
>>>
>>> Instead of guessing what kind of decoding I should use, I would use
>>> 'ensure_ascii=False' to prevent the crash for now.
>>>
>>> I tried to reproduce this crash, but it seems that the crash might be
>>> related with some internal stl implementation we are using. I will see if I
>>> can narrow down to a small repro later.
>>>
>>> Thanks
>>> Jeffrey
>>>
>>> On Sun, Mar 27, 2016 at 2:49 PM, Siva Chandra <sivachandra at gmail.com>
>>> wrote:
>>> On Sat, Mar 26, 2016 at 11:58 PM, Jeffrey Tan <jeffrey.fudan at gmail.com>
>>> wrote:
>>>
>>> Btw: after patching with Siva's fix http://reviews.llvm.org/D18008, the
>>> first field 'small_' is fixed, however the second field 'ml_' still emits
>>> garbage:
>>>
>>> (lldb) fr v corpus
>>> (const string &const) corpus = error: summary string parsing error: {
>>>  store_ = {
>>>     = {
>>>      small_ = "www"
>>>      ml_ = (data_ =
>>>
>>> "��UH\x89�H�}�H\x8bE�]ÐUH\x89�H��H\x89}�H\x8bE�H\x89��~\xb4��\x90��UH\x89�SH\x83�H\x89}�H�u�H�E�H���\x9e���H\x8b\x18H\x8bE�H���O\xb4��H\x89ƿ\b",
>>> size_ = 0, capacity_ = 1441151880758558720)
>>>    }
>>>  }
>>> }
>>>
>>>
>>> Do you still see the DW_TAG_member related error?
>>>
>>> A wild (and really wild at that) guess: Is it utf16 data that is being
>>> decoded as utf8?
>>>
>>> As David Blaikie mentioned on the other thread, it would really help
>>> if you provide us with a minimal example to repro this. Atleast, repro
>>> instructions.
>>>
>>> _______________________________________________
>>> lldb-dev mailing list
>>> lldb-dev at lists.llvm.org
>>> http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev
>>>
>>>
>>> _______________________________________________
>>> lldb-dev mailing list
>>> lldb-dev at lists.llvm.org
>>> http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev
>>>
>>>
>>>
>>> Thanks,
>>> *- Enrico*
>>> 📩 egranata@.com ☎️ 27683
>>>
>>>
>>
>>
>> Thanks,
>> *- Enrico*
>> 📩 egranata@.com ☎️ 27683
>>
>>
> _______________________________________________
> lldb-dev mailing list
> lldb-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/lldb-dev/attachments/20160407/cbd168a0/attachment-0001.html>


More information about the lldb-dev mailing list