| 1 |
> there should be a way of turning these off systematically. the |
| 2 |
> advantage of the current hardened gcc specs is that one can switch |
| 3 |
> between them using gcc-config. if these are forced on for the default |
| 4 |
> profile then there will be no easy way to systematically turn them off. |
| 5 |
|
| 6 |
No - there won't be an easy way for systematically turning off |
| 7 |
SSP and PIE in 17.0 profiles [1,2]. |
| 8 |
|
| 9 |
The hardened toolchain with its different gcc profiles came from a time |
| 10 |
where SSP and PIE were relatively new security features and a certain |
| 11 |
amount of fine-grained control was needed. Further, at that time we were |
| 12 |
talking about external patches against gcc. Nowadays everything is |
| 13 |
upstreamed and (almost) no patches to gcc for hardened profiles are |
| 14 |
applied any more. |
| 15 |
|
| 16 |
Given the fact that all major linux distributions are following the path |
| 17 |
of improved default hardening features (see for example [1]) and that we |
| 18 |
have been using ssp/pie in hardened profiles for years now the purpose |
| 19 |
of fine-grained control over ssp/pie is also highly questionable. |
| 20 |
|
| 21 |
The consensus at the moment is that PIE and SSP (as well as stricter |
| 22 |
linker flags) will soon be standard (or, actually *are* already |
| 23 |
standard) compilation options. A per-package override (if absoluetely |
| 24 |
needed) is fine - and, in fact, already in place everywhere where |
| 25 |
needed. |
| 26 |
|
| 27 |
Thus, we should go with the time and simply force these well tested |
| 28 |
hardening features on platforms that support it. |
| 29 |
|
| 30 |
Best, |
| 31 |
Matthias |
| 32 |
|
| 33 |
[1] for amd64/x86 and well supported profiles |
| 34 |
|
| 35 |
[2] there is always the possibility to override forced use flags |
| 36 |
|
| 37 |
[1] https://wiki.debian.org/Hardening/PIEByDefaultTransition |
Archives