Re: [PATCH, v4] PR 20569, segv in follow_exec

[Luis Machado <lgustavo@codesourcery.com>]

On 10/25/2016 01:30 PM, Luis Machado wrote:
> On 10/25/2016 01:20 PM, Pedro Alves wrote:
>> On 10/25/2016 07:15 PM, Luis Machado wrote:
>>
>>> Before i go ahead and adjust this even more, what's your plan and ETA
>>> for the above?
>>
>> I'll try to post this today.
>>
>
> Sounds good.
>
>>> This is disturbing more code as we try to consolidade
>>> slightly different functions into a single one in order to make things a
>>> bit more clean. But i'm afraid this is besides the point of the original
>>> patch itself?
>>>
>>> I just want to understand what's the end goal, because the scope seems
>>> to be changing slightly with each iteration. :-)
>>
>> No, the scope has not changed at all.  Your original version duplicated
>> a large chunk of code, and then the attempt to refactor things did it
>> incorrectly.  Still the same scope, but the patch as is, is buggy.
>>
>> Between accepted duplicated code, and fixing the patch, I take the
>> latter.
>
> Fine by me.
>
> Thanks,
> Luis

Going back to the flags problem, maybe adding "add_flags" as argument to 
symbol_file_add_main_1, just like symbol_file_add, and appending those 
flags to the existing flag variable there would make things work?