1 |
On Wed, 2009-07-22 at 09:59 +0200, Fabian Groffen wrote: |
2 |
> On 22-07-2009 08:15:31 +0200, Markus Duft wrote: |
3 |
> > > ok - i missed some NEEDED.PFX.1 -> NEEDED.PECOFF.1 renames :) that of |
4 |
> > > course caused some problems... the attached patch has all this fixed and |
5 |
> > > is confirmed to work on both interix and winnt now... |
6 |
> > > |
7 |
> > > how should i proceed? i have a commit ready to fire to commit to the |
8 |
> > > current portage-13830 ebuild (also includes prefix-chaining update). |
9 |
> > |
10 |
> > still waiting for your ok to commit :) |
11 |
> |
12 |
> ah. Well, I meant to look at the patch in detail last night, commit it |
13 |
> and make a new snapshot ... but something came inbetween. |
14 |
|
15 |
ah, well - no problem :) |
16 |
|
17 |
> |
18 |
> I have to look into detail, because it seems your PECOFF actually has |
19 |
> exactly the same behaviour as ELF, so I'd like to see if we can make the |
20 |
> LinkageMapPeCoff a shallow wrapper around the LinkageMap(ELF) one, to |
21 |
> avoid duplication. Also, I'm not sure if scanpecoff should be in |
22 |
> Portage, or in a bumped version of pax-utils. |
23 |
|
24 |
Hm. The LinkageMapPeCoff inherits from LinkageMap, and thus already has |
25 |
as low code duplication as possible. i _have_ to implement the parsing |
26 |
(if you can call simple reading from stdout "parsing" ;)) in a separate |
27 |
"rebuild" method, and i _have_ to implement the _ObjectKey myself, since |
28 |
inodes are no good on windows... |
29 |
|
30 |
scanpecoff's location... hmmm... there is no "really good" place for it. |
31 |
since it handles both interix and winnt, it is not suited to be put into |
32 |
parity, nor into binutils or something thelike. i think portage is the |
33 |
best place, since only portage will ever use it (i hope so - nobody else |
34 |
will have fun with it, since it only produces _exactly_ the format |
35 |
required by portage...). However, i'm open to location suggestions :) |
36 |
|
37 |
> |
38 |
> At the moment I'm trying to open a docx file from Microsoft that should |
39 |
> be the specification of PECOFF. |
40 |
|
41 |
uh... you won't have fun with that document - i know it by rote ;) it |
42 |
does _not_ specify how both interix and winnt handle runpaths, needed |
43 |
shared libs, etc. |
44 |
|
45 |
how interix handles those things is completely hidden somewhere in the |
46 |
interix loader/binary format/binutils patches at M$. (However the |
47 |
current binutils 2.17 patches are open, so i could have a look at it - |
48 |
the patch is about 300.000 lines long ;*( ...) |
49 |
|
50 |
The format used by winnt to model shared library features (rpath, |
51 |
needed, etc.) is defined by parity (native win32 does not support those |
52 |
out of the box...) - it builds upon the PE/COFF specs, but of course it |
53 |
does it's own things :) if you want to know what it does exactly -> ask |
54 |
me, as i wrote it ;) |
55 |
|
56 |
Cheers, Markus |
57 |
|
58 |
> |
59 |
> |