1 |
On Thu, Jul 20, 2017 at 1:20 AM, Danny YUE <sheepduke@×××××.com> wrote: |
2 |
> |
3 |
> On 2017-07-20 05:59, R0b0t1 <r030t1@×××××.com> wrote: |
4 |
>> On Thu, Jul 20, 2017 at 12:42 AM, Danny YUE <sheepduke@×××××.com> wrote: |
5 |
>>> Hi guys, |
6 |
>>> |
7 |
>>> I am setting up cross compiling environment for my newly bought |
8 |
>>> Raspberry Pi 3, under the guide of: |
9 |
>>> https://wiki.gentoo.org/wiki/Raspberry_Pi |
10 |
>>> https://wiki.gentoo.org/wiki/Raspberry_Pi/Quick_Install_Guide |
11 |
>>> |
12 |
>>> My original idea was to use crossdev to cross compile packages on my PC |
13 |
>>> and install binaries on RPI. |
14 |
>>> However I found it really nasty because it kept giving me the error |
15 |
>>> message about "libintl: no such file or directory" during compilation of |
16 |
>>> packages such as attr, python etc. |
17 |
>>> |
18 |
>>> And if some (even very few) packages fail to build on the PC, it is |
19 |
>>> hardly possible to keep PC and RPI 'consistent'. |
20 |
>>> (Yes, I did setup the 'make.profile' symlink to the right place.) |
21 |
>>> |
22 |
>>> So my question is: |
23 |
>>> 1) If some packages are *doomed* to fail, how do you keep the |
24 |
>>> emerge world environment consistency between PC and RPI? |
25 |
>>> Or is my understanding of this method incorrect? |
26 |
>>> |
27 |
>> |
28 |
>> Match compilation and USE flags. |
29 |
>> |
30 |
>>> 2) If it is not really a good idea to use crossdev, which one do you |
31 |
>>> recommend between distcc and chroot method? |
32 |
>>> (I googled but did not really get one answer about compilation speed.) |
33 |
>>> |
34 |
>> |
35 |
>> You should be able to mix crossdev compiled packages freely with |
36 |
>> device compiled ones. Incompatible packages will be ignored and it |
37 |
>> will try to recompile a package with matching flags, so pay attention. |
38 |
>> If you need to compile something on device then I suspect you want to |
39 |
>> use distcc if at all possible. |
40 |
>> |
41 |
> Thanks for your reply. :-) |
42 |
> |
43 |
> Well, yes. But do you know how is distcc compared with chroot referring |
44 |
> to compilation speed? |
45 |
> |
46 |
|
47 |
Using qemu-user to emulate the target architecture and hosting the |
48 |
system in a chroot is generally slower than compiling on device, if |
49 |
that is what you are referring to. I've read of people who tested this |
50 |
with the RPi3 and some Hardkernel devices. |