1 |
On Sat, 26 Jan 2013 16:06:35 +0100 |
2 |
Michał Górny <mgorny@g.o> wrote: |
3 |
|
4 |
> On Sat, 26 Jan 2013 11:51:22 -0300 |
5 |
> Alexis Ballier <aballier@g.o> wrote: |
6 |
> |
7 |
> > On Sat, 26 Jan 2013 13:11:41 +0100 |
8 |
> > Michał Górny <mgorny@g.o> wrote: |
9 |
> > |
10 |
> > > On Wed, 23 Jan 2013 21:40:13 -0300 |
11 |
> > > Alexis Ballier <aballier@g.o> wrote: |
12 |
> > > |
13 |
> > > > On Thu, 24 Jan 2013 00:23:57 +0100 |
14 |
> > > > Michał Górny <mgorny@g.o> wrote: |
15 |
> > > > |
16 |
> > > > > This is mostly a proof-of-concept. If approved, I will work on |
17 |
> > > > > moving the code into a separate eclass, possibly named |
18 |
> > > > > 'multilib-build' ;). --- |
19 |
> > > > > gx86/eclass/autotools-multilib.eclass | 24 |
20 |
> > > > > +++++++++++++++++++++--- 1 file changed, 21 insertions(+), 3 |
21 |
> > > > > deletions(-) |
22 |
> > > > > |
23 |
> > > > > diff --git a/gx86/eclass/autotools-multilib.eclass |
24 |
> > > > > b/gx86/eclass/autotools-multilib.eclass index 7c8697a..eef7bcc |
25 |
> > > > > 100644 --- a/gx86/eclass/autotools-multilib.eclass |
26 |
> > > > > +++ b/gx86/eclass/autotools-multilib.eclass |
27 |
> > > > > @@ -32,7 +32,23 @@ inherit autotools-utils multilib |
28 |
> > > > > |
29 |
> > > > > EXPORT_FUNCTIONS src_configure src_compile src_test |
30 |
> > > > > src_install |
31 |
> > > > > -IUSE=multilib |
32 |
> > > > > +# Declare all of them, profiles will control their |
33 |
> > > > > visibility. +IUSE='abi_x86_32 abi_x86_64' |
34 |
> > > > > + |
35 |
> > > > > +# @FUNCTION: _autotools-multilib_get_enabled_abis |
36 |
> > > > > +# @DESCRIPTION: |
37 |
> > > > > +# Get the list of enabled ABIs. The returned names are |
38 |
> > > > > suitable for use +# with multilib.eclass. |
39 |
> > > > > +# |
40 |
> > > > > +# If multilib is not enabled or not supported, returns an |
41 |
> > > > > empty list. + |
42 |
> > > > > + debug-print-function ${FUNCNAME} "${@}" |
43 |
> > > > > + |
44 |
> > > > > + if use amd64; then |
45 |
> > > > > + use abi_x86_64 && echo amd64 |
46 |
> > > > > + use abi_x86_32 && echo x86 |
47 |
> > > > > + fi |
48 |
> > > > > +} |
49 |
> > > > |
50 |
> > > > I would rather iterate over a variable than hardcoding and |
51 |
> > > > duplicating it here: |
52 |
> > > > |
53 |
> > > > MULTILIB_ABIS='abi_x86_32:x86 abi_x86_64:amd64' |
54 |
> > > > IUSE="" |
55 |
> > > > for i in $MULTILIB_ABIS ; do |
56 |
> > > > IUSE+=" ${i%:*}" |
57 |
> > > > done |
58 |
> > > > |
59 |
> > > > _autotools-multilib_get_enabled_abis() { |
60 |
> > > > for i in $MULTILIB_ABIS ; do |
61 |
> > > > use ${i%:*} && echo ${i#*:} |
62 |
> > > > done |
63 |
> > > > } |
64 |
> > > |
65 |
> > > What are the advantages? I feel like the explicit solution is much |
66 |
> > > more readable and obvious at the first glance. |
67 |
> > |
68 |
> > yes it is more readable but IMHO it's better to avoid to have to |
69 |
> > touch multiple places when adding a new ABI: you only have to |
70 |
> > document that adding a new ABI consists simply in adding it to this |
71 |
> > list (and document the useflag) instead of 'add the useflag, add |
72 |
> > support for it to function foo and bar, etc.' |
73 |
> > your call in the end, but I fear not trying to separate code from |
74 |
> > data could make the eclass harder to maintain. |
75 |
> |
76 |
> Ok, separating data from code seems a valid concern. However, I feel |
77 |
> like this is not really clear concept in bash, and the intent is that |
78 |
> the ABI list will be used only in that function. |
79 |
|
80 |
makes sense; the foo:bar trick is what I've came up in trying to emulate |
81 |
python dictionaries :) |
82 |
|
83 |
|
84 |
> > also, IMHO you shouldn't use arch to guess what useflag to check or |
85 |
> > not: they'll all be in IUSE in the end and the profiles should be |
86 |
> > the one deciding what to mask or not (which you'll have to do |
87 |
> > anyway), not the eclass. |
88 |
> |
89 |
> Well, I think I've stated that already. That 'use foo &&' is just |
90 |
> an optimization hack -- to not check all the irrelevant flags all |
91 |
> the time. |
92 |
|
93 |
I'm not sure 'optimizations' are relevant when we're going to build |
94 |
twice the package here :P |
95 |
Also, # of arches is about the same order of magnitude than # of abis, |
96 |
thus checking # of arches time use $arch or # of abis times use $abi is, |
97 |
in the end, comparable while checking arch adds the constraint that |
98 |
the eclass and the profiles have to be in sync. |
99 |
|
100 |
Alexis. |