1 2 Known problems in GDB 6.1 3 4 See also: http://www.gnu.org/software/gdb/bugs/ 5 6 7*** Build problems 8 9build/1458: comple failed on hpux11 10 11GDB 6.1 is known to have build problems on HP/UX 11.00 using the 12vendor supplied compilers (GDB does build on HP/UX 11.11, and using 13GCC). 14 15*** Misc 16 17gdb/1560: Control-C does not always interrupt GDB. 18 19When GDB is busy processing a command which takes a long time to 20complete, hitting Control-C does not have the expected effect. 21The command execution is not aborted, and the "QUIT" message confirming 22the abortion is displayed only after the command has been completed. 23 24*** C++ support 25 26gdb/931: GDB could be more generous when reading types C++ templates on input 27 28When the user types a template, GDB frequently requires the type to be 29typed in a certain way (e.g. "const char*" as opposed to "const char *" 30or "char const *" or "char const*"). 31 32gdb/1512: no canonical way to output names of C++ types 33 34We currently don't have any canonical way to output names of C++ types. 35E.g. "const char *" versus "char const *"; more subtleties arise when 36dealing with templates. 37 38gdb/1516: [regression] local classes, gcc 2.95.3, dwarf-2 39 40With gcc 2.95.3 and the dwarf-2 debugging format, classes which are 41defined locally to a function include the demangled name of the function 42as part of their name. For example, if a function "foobar" contains a 43local class definition "Local", gdb will say that the name of the class 44type is "foobar__Fi.0:Local". 45 46This applies only to classes where the class type is defined inside a 47function, not to variables defined with types that are defined somewhere 48outside any function (which most types are). 49 50gdb/1588: names of c++ nested types in casts must be enclosed in quotes 51 52You must type 53 (gdb) print ('Foo::Bar') x 54or 55 (gdb) print ('Foo::Bar' *) y 56instead of 57 (gdb) print (Foo::Bar) x 58or 59 (gdb) print (Foo::Bar *) y 60respectively. 61 62gdb/1091: Constructor breakpoints ignored 63gdb/1193: g++ 3.3 creates multiple constructors: gdb 5.3 can't set breakpoints 64 65When gcc 3.x compiles a C++ constructor or C++ destructor, it generates 662 or 3 different versions of the object code. These versions have 67unique mangled names (they have to, in order for linking to work), but 68they have identical source code names, which leads to a great deal of 69confusion. Specifically, if you set a breakpoint in a constructor or a 70destructor, gdb will put a breakpoint in one of the versions, but your 71program may execute the other version. This makes it impossible to set 72breakpoints reliably in constructors or destructors. 73 74gcc 3.x generates these multiple object code functions in order to 75implement virtual base classes. gcc 2.x generated just one object code 76function with a hidden parameter, but gcc 3.x conforms to a multi-vendor 77ABI for C++ which requires multiple object code functions. 78 79*** Stack backtraces 80 81GDB's core code base has been updated to use a new backtrace 82mechanism. This mechanism makes it possible to support new features 83such DWARF 2 Call Frame Information (which in turn makes possible 84backtraces through optimized code). 85 86Since this code is new, it is known to still have a few problems: 87 88gdb/1505: [regression] gdb prints a bad backtrace for a thread 89 90When backtracing a thread, gdb does not stop when it reaches the 91outermost frame, instead continuing until it hits garbage. This is 92sensitive to the operating system and thread library. 93 94hppa*-*-* 95mips*-*-* 96 97The MIPS and HPPA backtrace code has only very recently been updated 98to use GDB's new frame mechanism. At present there are still a few 99problems, in particular backtraces through signal handlers do not 100work. 101 102People encountering problems with these architectures should consult 103GDB's web pages and mailing lists (http://www.gnu.org/software/gdb/) 104to see if there are updates. 105 106powerpc*-*-* 107 108PowerPC architecture support, in 6.1, does not use the new frame code. 109 110Fortunately, PowerPC architecture support, in GDB's mainline sources, 111have been updated. People encountering problems should consider 112downloading a more current snapshot of GDB 113(http://www.gnu.org/software/gdb/current/). 114