1 |
> For a transition we can probably build everything with -fPIE but not |
2 |
> link with -pie. If we want that to happen fast, gcc-6 might do that and |
3 |
> gcc-7 add the -pie option. |
4 |
|
5 |
I am not entirely convinced that a transition period of one gcc version |
6 |
is enough for a smooth transition [1]. |
7 |
|
8 |
It might be better to go through a quick transition process that |
9 |
requires a world rebuild. - In particular we already forced everyone on |
10 |
~amd64 to play beta tester in this regard [2,3]. |
11 |
|
12 |
Anyway the current use flag situation is a mess and has to be cleaned |
13 |
up asap. |
14 |
|
15 |
So, dos anyone recall why USE=pie was masked for >gcc-6.2 for everyone |
16 |
except amd64? |
17 |
|
18 |
Related to that |
19 |
|
20 |
- for which architectures shall we unmask the use flag? |
21 |
|
22 |
- shall we use.force a certain behavior per profile, or keep the flag |
23 |
unpinned? |
24 |
|
25 |
|
26 |
After having thought about the issue for a bit I still want to propose |
27 |
what we have already accidentally done - switch to USE=pie per default |
28 |
for gcc-6. |
29 |
|
30 |
Best, |
31 |
Matthias |
32 |
|
33 |
|
34 |
[1] Indeed *every* major linux distribution for which I have an lxc |
35 |
container has -pie enabled. If we decide on some slow transition we |
36 |
risk to be late to the party by quite a bit. |
37 |
|
38 |
[2] Which is extremely unfortunate. |
39 |
|
40 |
[3] The fallout I currently see due to enabled USE=pie is noticeably but |
41 |
by no stretch crazy bad. After all, static linkage is rarely used |
42 |
(with the exception of some languages). |