[llvm] r221915 - Object, Mach-O: Refactor and clean code up

[Lang Hames]

Hi Guys,

Yeah - this is definitely an improvement. Tests would be great though. It
looks like we need a few malformed objects: (1) Duplicate SymtabLoadCmd,
(2) Duplicate DysymtabLoadCmd, (3) Duplicate DataInCodeLoadCmd, (4)
Duplicate DyldInfoLoadCmd, and (5) Duplicate UuidLoadCmd.

I've filed http://llvm.org/PR22343 - if anyone wants to dive on it feel
free, otherwise I'll try to get to it (time permitting) in the next few
days.

Cheers,
Lang.

On Fri, Jan 23, 2015 at 3:52 PM, Rafael Espíndola <
rafael.espindola at gmail.com> wrote:

>
>
> On 23 January 2015 at 17:14, Kevin Enderby <enderby at apple.com> wrote:
>
>> Added Lang to the To: line and he maybe better to comment.
>>
>> While the change looks cleaner I’m not sure what the error handling
>> should be for malformed Mach-O files in llvm’s libObject.  If malformed
>> Mach-O files cause asserts it is difficult to build tools to print and
>> diagnose bad files.  For example in Mac OS otool(1) is used both to display
>> information about good Mach-O files and do its best to display what it can
>> with warning about  malformed Mach-O files.  For the guy producing Mach-O
>> files which some times are broken this is a valuable approach for have
>> otool(1) be able so show what it can instead of just stopping.
>>
>
> I agree that it should not assert.
>
> The intention was to point out that the original change was (and still is)
> untested.
>
> Cheers,
> Rafael
>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150126/ba76355e/attachment.html>