[Lldb-commits] [PATCH 3/8] Test if an ELF object is executable by checking if an entry point is defined.

[Stephen Wilson]

Greg Clayton <gclayton at apple.com> writes:

> Could a bootloader ELF file have an entry point of zero? Should this be:

Interesting question.  I have never seen one myself (but that does not
mean they do not exist).  

However the spec specifically says that if an ELF does not have an entry
point then this field must be zero.  For that reason I think this should
be a relatively safe test.


> bool
> ObjectFileELF::IsExecutable() const
> {
>     return m_header.e_entry != LLDB_INVALID_ADDRESS;
> }
>
> And then you would need to make sure that m_header.e_entry always gets initialized to LLDB_INVALID_ADDRESS?
>
>
> On Jan 14, 2011, at 1:12 PM, Stephen Wilson wrote:
>
>> The previous check on header type ET_EXEC is not general enough.  Position
>> independent executables have type ET_DYN.
>> ---
>> source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp |    2 +-
>> 1 files changed, 1 insertions(+), 1 deletions(-)
>> 
>> diff --git a/source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp b/source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp
>> index 5b99d01..f58c779 100644
>> --- a/source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp
>> +++ b/source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp
>> @@ -159,7 +159,7 @@ ObjectFileELF::~ObjectFileELF()
>> bool
>> ObjectFileELF::IsExecutable() const
>> {
>> -    return m_header.e_type == ET_EXEC;
>> +    return m_header.e_entry != 0;
>> }
>> 
>> Address
>> -- 
>> 1.7.3.5
>> 
>> _______________________________________________
>> lldb-commits mailing list
>> lldb-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
>



--
steve