| 1 |
On 09/22/2012 09:08 AM, Ambroz Bizjak wrote: |
| 2 |
> Yes, I think this is a good idea, it would allow the dependencies to |
| 3 |
> be expressed nicely as conditions. |
| 4 |
> |
| 5 |
> But I'm not sure how this would be a USE flag. It should behave like |
| 6 |
> one during the build, but it would be best if it was not written into |
| 7 |
> the VDB as such, at least in a way that would be considered by |
| 8 |
> --newuse. It don't want "emerge -unD" on the booted system want to |
| 9 |
> reinstall all packages because the current ones were cross-compiled. |
| 10 |
> Does the test flag already behave nicely like that? In that case, all |
| 11 |
> is good, and I can try to implement this. |
| 12 |
|
| 13 |
Simply add your special flag to the _feature_flags variable in |
| 14 |
config.py, and it will be exempt from --newuse. See this commit: |
| 15 |
|
| 16 |
http://git.overlays.gentoo.org/gitweb/?p=proj/portage.git;a=commit;h=6b19f71b39b6af43307abf20654511bace041217 |
| 17 |
|
| 18 |
> On a slightly different subject: I've been porting some packages to |
| 19 |
> HDEPEND and I've seen problems with packages that want to use the |
| 20 |
> programs they build during the build (or in postinst). Of couse this |
| 21 |
> works for native builds, and it can be fixed to work for cross-compile |
| 22 |
> builds (build native version or HDEPEND on host package). |
| 23 |
> |
| 24 |
> But what do we do with the strange case where ROOT!=/ but |
| 25 |
> --crosscompile/FEATURES=crosscompile is not in affect? Can we expect |
| 26 |
> that we will be able to run the programs that were built? What if they |
| 27 |
> link to libraries only available in ROOT? |
| 28 |
> |
| 29 |
> So, I think it would make sense for a lot of packages to treat ROOT!=/ |
| 30 |
> equivalently to cross-compilation, i.e. require host tool to be |
| 31 |
> present. But with what has currently been proposed there is no |
| 32 |
> conditional dependency on ROOT!=/, so a package cannot demand that a |
| 33 |
> tool be present on the host. Then, it may be a good idea to add a |
| 34 |
> conditional dependency on ROOT!=/. |
| 35 |
|
| 36 |
If I understand correctly, that would be like a CROSS_TDEPEND? If we |
| 37 |
translate that to a conditional, it would become DEPEND="crosscompile? ( |
| 38 |
foo )", since our plan was to make DEPEND apply to ROOT!=/ and HDEPEND |
| 39 |
apply to ROOT=/, right? |
| 40 |
|
| 41 |
> In fact, I think that --crosscompile or FEATURES=crosscompile could |
| 42 |
> actually be abolished and only this condition would be available. It's |
| 43 |
> true that some packages would only use the host dependency if there's |
| 44 |
> actual cross-compilation going on, but nothing will break. This would |
| 45 |
> ease configuration and reduce the number of cases to be tested. |
| 46 |
|
| 47 |
Yeah, the split between HDEPEND and DEPEND might be enough so that you |
| 48 |
don't need these conditionals. If you're not really sure that the |
| 49 |
conditionals are needed, then maybe it's better to eliminate them for now. |
| 50 |
-- |
| 51 |
Thanks, |
| 52 |
Zac |
Archives