1 |
-----BEGIN PGP SIGNED MESSAGE----- |
2 |
Hash: SHA512 |
3 |
|
4 |
On 06/15/2014 04:47 AM, Dale wrote: |
5 |
> John Campbell wrote: |
6 |
>> On 06/14/2014 10:10 PM, Dale wrote: |
7 |
>> |
8 |
>>> Well, I have 16Gbs here. I'm not lacking for memory. If |
9 |
>>> memory prices were to drop a bit, I could upgrade some more. |
10 |
>>> I'd have to swap out what I have tho. Old mobo would only take |
11 |
>>> 4GB sticks and this new one will take 8GB sticks. |
12 |
>> |
13 |
>> Unless you have a specific reason for keeping both x32 and 64 |
14 |
>> ABIs I'd suggest changing to ABI_X86="32 64" globally in |
15 |
>> /etc/make.conf (or /etc/portage/make.conf). It's a lot easier |
16 |
>> than waiting for the next conflict. Then do "emerge --new-use |
17 |
>> --deep @world" and you're done. There shouldn't bee too many |
18 |
>> packages that need rebuilding. |
19 |
>> |
20 |
>> |
21 |
> |
22 |
> I put that in make.conf and get this: |
23 |
> |
24 |
> # required by sys-fs/eudev-1.7[gudev] # required by @selected # |
25 |
> required by @world (argument) =dev-libs/glib-2.40.0 ~amd64 |
26 |
> |
27 |
> Does it make sense to keyword that? Isn't that the package that |
28 |
> once upgraded you can't go backward? I'm just double checking that |
29 |
> this all makes sense. < scratches head > |
30 |
> |
31 |
> Dale |
32 |
> |
33 |
> :-) :-) |
34 |
> |
35 |
|
36 |
The package you are thinking of that is a one-way upgrade is glibc, |
37 |
not glib. Glibc is the C Runtime Environment, absolutely everything |
38 |
written in C and C++ ends up linking against it (and things that |
39 |
aren't tend either to be compiled using something that is or |
40 |
interpreted by something that is). If glibc is downgraded, anything |
41 |
that uses something from a newer version (or something that *changed* |
42 |
in a newer version, due to how symbol versions work) will fail to run |
43 |
against the older version. |
44 |
|
45 |
For example, any program linked against glibc 2.14 or greater that |
46 |
uses memcpy ends up linking against the symbol "memcpy@@GLIBC_2.14", |
47 |
which is only in newer versions of glibc. Programs linked against |
48 |
older versions of glibc use "memcpy@GLIBC_2.2.5", which differs in |
49 |
some way (specifically in this case, the old memcpy always went in a |
50 |
particular direction, the new memcpy may be faster on some CPUs, but |
51 |
this broke old Adobe Flash). |
52 |
-----BEGIN PGP SIGNATURE----- |
53 |
Version: GnuPG v2 |
54 |
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ |
55 |
|
56 |
iQIcBAEBCgAGBQJTnqAdAAoJELHSF2kinlg4RwIQAIBA/fnW7hq6GUK3fQimcEIo |
57 |
RdeXK3RXy750T6XP5vo71NFhrJJmaROaNzSksKXfAtfJZt3PnpbIxFxtvKep7Xp3 |
58 |
bJnAOXHej+EndkSDXEoZTAepzEmqIj8V+Y4blSmRE25w+NLTe2Wgkjd92Y6ugroW |
59 |
Hk1A/nrGRNO5pBepSvIUOIn5GIoTMMH4485HBnwUdtt6+EQja50cfSMbTUoXqF38 |
60 |
GgkJkik28TOYSivn6b6VdzCvl+X8ZqFY0M+BhWSLLaN+7/BR1S5dI1Th1QNDaHSX |
61 |
KcCbtSIwqB1V5w5jSLb4JtU+Noi5nqXyy4ZnwR0wCN/MkhBFFQlRYE/sgOeYrOgw |
62 |
mxCj7442q0Ml1nv+4ASHdOsQ/M8VCoBR3TZhCYDU5wZB6Wb4AVY/tSxopDL54aqi |
63 |
4iutCTBn9fh7NIfasVjhvSdNMdglF/SHnHfiS9C94odFGvbZjHJd7ICwshSh9eNG |
64 |
/jwRRSrmdMK9IiqK9IkTYDXlCYkre2CtIt8urlypoWaQwD5IrkEIeE9PvXEUcDNU |
65 |
jw8S+3hhNkQ5/3MKErBwGOshdhoogTM03afbSg/M3DRn+B3GAo8Zi9i26wUPTMJo |
66 |
Y6Nm3xhPd5K5h9hmh4bTkG+eqIshtTpqC9cdzQ/xf1dA2X1yxvBdXt6B3cGQg/+1 |
67 |
kUyTwlCW3O11b9l0LzeH |
68 |
=C+Xr |
69 |
-----END PGP SIGNATURE----- |