[lld] r349979 - gdb-index: Handle errors when parsing ranges

Rui Ueyama via llvm-commits llvm-commits at lists.llvm.org
Thu Jan 3 14:55:38 PST 2019


Sure. Here is my follow-up: https://reviews.llvm.org/rL350070

On Fri, Dec 28, 2018 at 10:16 PM David Blaikie <dblaikie at gmail.com> wrote:

>
>
> On Wed, Dec 26, 2018 at 11:11 AM Rui Ueyama <ruiu at google.com> wrote:
>
>> On Fri, Dec 21, 2018 at 4:34 PM David Blaikie via llvm-commits <
>> llvm-commits at lists.llvm.org> wrote:
>>
>>> Author: dblaikie
>>> Date: Fri Dec 21 16:31:05 2018
>>> New Revision: 349979
>>>
>>> URL: http://llvm.org/viewvc/llvm-project?rev=349979&view=rev
>>> Log:
>>> gdb-index: Handle errors when parsing ranges
>>>
>>> When parsing CU ranges for gdb-index, handle the error (now propagated
>>> up though the API lld is calling here - previously the error was
>>> printed within the libDebugInfo API, not allowing lld to format or
>>> handle the message at all) - including information about the object and
>>> archive name, as well as failing the link.
>>>
>>> Added:
>>>     lld/trunk/test/ELF/Inputs/gdb-index-invalid-ranges.obj.s
>>>     lld/trunk/test/ELF/gdb-index-invalid-ranges.s
>>> Modified:
>>>     lld/trunk/ELF/SyntheticSections.cpp
>>>
>>> Modified: lld/trunk/ELF/SyntheticSections.cpp
>>> URL:
>>> http://llvm.org/viewvc/llvm-project/lld/trunk/ELF/SyntheticSections.cpp?rev=349979&r1=349978&r2=349979&view=diff
>>>
>>> ==============================================================================
>>> --- lld/trunk/ELF/SyntheticSections.cpp (original)
>>> +++ lld/trunk/ELF/SyntheticSections.cpp Fri Dec 21 16:31:05 2018
>>> @@ -2408,17 +2408,18 @@ static std::vector<GdbIndexSection::CuEn
>>>    return Ret;
>>>  }
>>>
>>> -static std::vector<GdbIndexSection::AddressEntry>
>>> +static Expected<std::vector<GdbIndexSection::AddressEntry>>
>>>  readAddressAreas(DWARFContext &Dwarf, InputSection *Sec) {
>>>    std::vector<GdbIndexSection::AddressEntry> Ret;
>>>
>>>    uint32_t CuIdx = 0;
>>>    for (std::unique_ptr<DWARFUnit> &Cu : Dwarf.compile_units()) {
>>> -    DWARFAddressRangesVector Ranges;
>>> -    Cu->collectAddressRanges(Ranges);
>>> +    Expected<DWARFAddressRangesVector> Ranges =
>>> Cu->collectAddressRanges();
>>> +    if (!Ranges)
>>> +      return Ranges.takeError();
>>>
>>>      ArrayRef<InputSectionBase *> Sections = Sec->File->getSections();
>>> -    for (DWARFAddressRange &R : Ranges) {
>>> +    for (DWARFAddressRange &R : *Ranges) {
>>>        InputSectionBase *S = Sections[R.SectionIndex];
>>>        if (!S || S == &InputSection::Discarded || !S->Live)
>>>          continue;
>>> @@ -2431,7 +2432,8 @@ readAddressAreas(DWARFContext &Dwarf, In
>>>      }
>>>      ++CuIdx;
>>>    }
>>> -  return Ret;
>>> +
>>> +  return std::move(Ret);
>>>  }
>>>
>>>  template <class ELFT>
>>> @@ -2563,7 +2565,17 @@ template <class ELFT> GdbIndexSection *G
>>>
>>>      Chunks[I].Sec = Sections[I];
>>>      Chunks[I].CompilationUnits = readCuList(Dwarf);
>>> -    Chunks[I].AddressAreas = readAddressAreas(Dwarf, Sections[I]);
>>> +    Expected<std::vector<GdbIndexSection::AddressEntry>> AddressAreas =
>>> +        readAddressAreas(Dwarf, Sections[I]);
>>> +    if (!AddressAreas) {
>>> +      std::string Msg = File->getName();
>>> +      Msg += ": ";
>>> +      if (!File->ArchiveName.empty())
>>> +        Msg += "in archive " + File->ArchiveName + ": ";
>>> +      Msg += toString(AddressAreas.takeError());
>>> +      fatal(Msg);
>>> +    }
>>>
>>
>> fatal() immediately terminates the process, and it is preferred to use
>> error() which does return whenever possible.
>>
>
> Fair enough - this was a "first pass" sort of approach.
>
>
>> Also I think you can use toString(Sec) to stringize sections (including
>> concatenating an archive file name) easily. I'll address them in a
>> follow-up patch.
>>
>
> Oh, nice - thanks! I went looking for it through test cases, etc, and the
> example for duplicate symbols was the codepath I found, which used some
> more manual logic like this.
>
> Could you reply here with the revisions of the followups, so I can take a
> look/understand?
>
> Thanks again,
> - Dave
>
>
>>
>> +    Chunks[I].AddressAreas = *AddressAreas;
>>>      NameAttrs[I] = readPubNamesAndTypes<ELFT>(
>>>          static_cast<const LLDDwarfObj<ELFT> &>(Dwarf.getDWARFObj()),
>>>          Chunks[I].CompilationUnits);
>>>
>>> Added: lld/trunk/test/ELF/Inputs/gdb-index-invalid-ranges.obj.s
>>> URL:
>>> http://llvm.org/viewvc/llvm-project/lld/trunk/test/ELF/Inputs/gdb-index-invalid-ranges.obj.s?rev=349979&view=auto
>>>
>>> ==============================================================================
>>> --- lld/trunk/test/ELF/Inputs/gdb-index-invalid-ranges.obj.s (added)
>>> +++ lld/trunk/test/ELF/Inputs/gdb-index-invalid-ranges.obj.s Fri Dec 21
>>> 16:31:05 2018
>>> @@ -0,0 +1,2 @@
>>> +main:
>>> +  callq f1
>>>
>>> Added: lld/trunk/test/ELF/gdb-index-invalid-ranges.s
>>> URL:
>>> http://llvm.org/viewvc/llvm-project/lld/trunk/test/ELF/gdb-index-invalid-ranges.s?rev=349979&view=auto
>>>
>>> ==============================================================================
>>> --- lld/trunk/test/ELF/gdb-index-invalid-ranges.s (added)
>>> +++ lld/trunk/test/ELF/gdb-index-invalid-ranges.s Fri Dec 21 16:31:05
>>> 2018
>>> @@ -0,0 +1,42 @@
>>> +# REQUIRES: x86
>>> +# RUN: llvm-mc -filetype=obj -triple=x86_64-unknown-linux %s -o %t.o
>>> +# RUN: not ld.lld --gdb-index -e main %t.o -o %t 2>&1 | FileCheck %s
>>> +# RUN: llvm-mc -filetype=obj -triple=x86_64-pc-linux
>>> %p/Inputs/gdb-index-invalid-ranges.obj.s -o %t2.o
>>> +# RUN: llvm-ar rc %t.a %t.o
>>> +# RUN: not ld.lld --gdb-index -e main %t2.o %t.a -o %t 2>&1 | FileCheck
>>> --check-prefix=ARCHIVE %s
>>> +
>>> +# CHECK: ld.lld: error: {{.*}}gdb-index-invalid-ranges.s.tmp.o:
>>> decoding address ranges: invalid range list entry at offset 0x10
>>> +# ARCHIVE: ld.lld: error: gdb-index-invalid-ranges.s.tmp.o: in archive
>>> {{.*}}gdb-index-invalid-ranges.s.tmp.a: decoding address ranges: invalid
>>> range list entry at offset 0x10
>>> +
>>> +.section .text.foo1,"ax", at progbits
>>> +.globl f1
>>> +.Lfunc_begin0:
>>> +f1:
>>> + nop
>>> +.Lfunc_end0:
>>> +
>>> +.section .debug_abbrev,"", at progbits
>>> +.byte 1                       # Abbreviation Code
>>> +.byte 17                      # DW_TAG_compile_unit
>>> +.byte 0                       # DW_CHILDREN_no
>>> +.byte 85                      # DW_AT_ranges
>>> +.byte 23                      # DW_FORM_sec_offset
>>> +.byte 0                       # EOM(1)
>>> +.byte 0                       # EOM(2)
>>> +.byte 0                       # EOM(3)
>>> +
>>> +.section .debug_info,"", at progbits
>>> +.Lcu_begin0:
>>> +.long .Lunit_end0-.Lunit_begin0 # Length of Unit
>>> +.Lunit_begin0:
>>> +.short 4                      # DWARF version number
>>> +.long .debug_abbrev           # Offset Into Abbrev. Section
>>> +.byte 8                       # Address Size (in bytes)
>>> +.byte 1                       # Abbrev [1] 0xb:0x1f DW_TAG_compile_unit
>>> +.long .Ldebug_ranges0         # DW_AT_ranges
>>> +.Lunit_end0:
>>> +
>>> +.section .debug_ranges,"", at progbits
>>> +.Ldebug_ranges0:
>>> +.quad .Lfunc_begin0
>>> +.quad .Lfunc_end0
>>>
>>>
>>> _______________________________________________
>>> llvm-commits mailing list
>>> llvm-commits at lists.llvm.org
>>> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190103/543859fc/attachment.html>


More information about the llvm-commits mailing list