1 |
On Tue, Aug 10, 2010 at 4:45 AM, Brian Harring wrote: |
2 |
> On Mon, Aug 09, 2010 at 07:05:11PM -0400, Mike Frysinger wrote: |
3 |
>> On Mon, Aug 9, 2010 at 7:03 PM, Markos Chandras wrote: |
4 |
>> > On Sat, Aug 07, 2010 at 10:16:24PM -0400, Mike Frysinger wrote: |
5 |
>> >> obviously you only mean linux x86/amd64 dev profiles. i dont have a strong |
6 |
>> >> opinion on that small subset in either direction. |
7 |
>> > |
8 |
>> > So do you agree to make this linker option default to linux x86/amd64 dev/ |
9 |
>> > profiles? |
10 |
>> |
11 |
>> add them or dont add them, "i dont have a [...] opinion [...] in |
12 |
>> either direction". if put to a vote, i'd abstain. |
13 |
> |
14 |
> Possibly a stupid question, but any reason we've not looked at |
15 |
> injecting something that has lower actual affect but can still be used |
16 |
> for a canary? I'm thinking of --build-id specifically... |
17 |
|
18 |
my gut reaction there is now you're requiring even newer versions of |
19 |
binutils than before, and not just to find ones that support |
20 |
--build-id, but do so without bugs (that's my vague recollection of |
21 |
things; perhaps i'm wrong). and you still wouldnt pass the "not safe |
22 |
outside of Gentoo Linux profiles". |
23 |
|
24 |
also, although the overhead is minor, the build id section would serve |
25 |
no useful purpose that i can think once it has been merged. gnu hash |
26 |
however is always used at runtime. |
27 |
-mike |