Re: [dev] sta.li progress

I'd never used jam before, it really is a hideous beast. I was only
messing with it to figure out what exactly needs to happen to get
bionic to compile so it could be translated into a working mkfile.

And I've hit some degree of success. I've just gotten bionic to
compile. Some things apparently didn't go right. I had to write my own
_start(), and for some reason my hello.c produced a 220K executable.
But I'll keep poking it.

I'd volunteer but I'm really not very familiar with build systems at all.

On 13 October 2010 08:37, Anselm R Garbe <garbeam_AT_gmail.com> wrote:
> Anyone volunteering to create an mkfile for bionic, this jam looks
> like jam and should be avoided.
> I volunteer to review it and to bring it in good shape.
>
> -Anselm
>
> On 13 October 2010 14:34, Jens Staal <staal1978_AT_gmail.com> wrote:
>> I have no idea what the jam stuff is but I found this after some googling too
>>
>> https://www.metasploit.com/redmine/projects/framework/repository/revisions/10202/entry/external/source/meterpreter/source/bionic/libc/Jamfile
>>
>> 2010/10/13 Corey Thomasson <cthom.lists_AT_gmail.com>:
>>> Excellent! I've been searching for something like this. I'll check it out.
>>>
>>> On 13 October 2010 08:29, Jens Staal <staal1978_AT_gmail.com> wrote:
>>>> Are those issues already solved by
>>>>
>>>> http://www.metasploit.com/redmine/attachments/433/get_bionic_working.diff
>>>>
>>>> ?
>>>>
>>>> 2010/10/13 Corey Thomasson <cthom.lists_AT_gmail.com>:
>>>>> On 12 October 2010 20:58, Wolf Tivy <wtivy1_AT_my.bcit.ca> wrote:
>>>>>>>I've managed to make it compile a good chunk of the object files,
>>>>>>>but not malloc/free so its somewhat wasted.
>>>>>>
>>>>>> It'll talk eventually, keep up the pressure.
>>>>>>
>>>>>>> When I get a chance to go at it again I believe the android distribution
>>>>>>>has some "clean" kernel headers included. I may try to move those to
>>>>>>>wherever its looking now.
>>>>>>
>>>>>> Yes it does, in fact I think it shouldn't need any system headers at all,
>>>>>> if you point it to the paths in OVERVIEW.TXT. Is there some
>>>>>> "include_path" environment variable you can set, or do we have to hack
>>>>>> the jamfile? Or we could do it your way and move them. Worst case is a new
>>>>>> makefile.
>>>>>
>>>>> there's a variable in the Jamfile INCLUDES_x86, the included header
>>>>> files dont seem to do the trick though. Running into syntax errors.
>>>>>
>>>>>> Surely someone else must have dealt with this. Metasploit has been mentioned
>>>>>> a few times, but I couldn't find any thing more than the blog post
>>>>>> (issue report, whatever) that jens linked. Anyone have a link to more info?
>>>>>>
>>>>>> About uClibc, it's LGPL, so isn't static linking a bit marginal? For
>>>>>> GPL-compatible stuff it's ok, but 9base is incompatible, so it may actually
>>>>>> need to be ported to bionic. Have I got this right?
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
Received on Wed Oct 13 2010 - 16:21:36 CEST