| 1 |
Wow, what a lengthy email :) |
| 2 |
|
| 3 |
On Tue, 1 Dec 2015 22:58:55 +0000 |
| 4 |
James Le Cuirot <chewi@g.o> wrote: |
| 5 |
[...] |
| 6 |
> |
| 7 |
> I raised one further point with mgorny that he feels could potentially |
| 8 |
> go into EAPI 7 but I think could remain an implementation detail. In |
| 9 |
> cases #3 and #4 (basically when ROOT != / and |
| 10 |
> PORTAGE_CONFIGROOT != /), the toolchain needs to know how to source |
| 11 |
> headers and libraries from ROOT instead of /. |
| 12 |
|
| 13 |
Use SYSROOT. ROOT has nothing to do with building. This should be |
| 14 |
defined in PMS though. |
| 15 |
|
| 16 |
|
| 17 |
> gcc and friends support a --sysroot argument. It used to be the case |
| 18 |
> that this didn't work on a toolchain configured without a sysroot, |
| 19 |
> possibly creating issues for #4, but I've tested and it now works |
| 20 |
> regardless. |
| 21 |
|
| 22 |
$ gcc --sysroot=/tmp/foo -I/usr/include foo.c |
| 23 |
/usr/lib/gcc/x86_64-pc-linux-gnu/5.2.0/../../../../x86_64-pc-linux-gnu/bin/ld: |
| 24 |
this linker was not configured to use sysroots |
| 25 |
collect2: error: ld returned 1 exit status |
| 26 |
|
| 27 |
Two problems here: |
| 28 |
1. link fails |
| 29 |
2. gcc doesn't warn for -I/usr/include with sysroot (iirc crossdev's |
| 30 |
gcc do) |
| 31 |
|
| 32 |
[...] |
| 33 |
> CMake supports a CMAKE_SYSROOT argument that you set in the toolchain |
| 34 |
> file. We currently don't set this but it could easily be added to |
| 35 |
> cmake-utils.eclass as well as (or instead of?) the similar variables |
| 36 |
> that we already do set. |
| 37 |
|
| 38 |
iirc, CMAKE_SYSROOT is only for your case #4 and behaves more like |
| 39 |
'build as if I had chrooted in CMAKE_SYSROOT' and thus doesn't help at |
| 40 |
all for cross-compilation |
| 41 |
|
| 42 |
> The vast majority of configure scripts (using libtool and not ancient) |
| 43 |
> support a --with-sysroot argument that can easily be detected by |
| 44 |
> grepping for lt_sysroot. The behaviour of econf is largely defined by |
| 45 |
> PMS, hence why mgorny thinks this particular detail could go into |
| 46 |
> EAPI 7, but it only says which arguments must be passed and doesn't |
| 47 |
> say that the PM can't pass additional ones. I don't wish this force |
| 48 |
> this stuff upon Paludis and pkgcore unnecessarily. |
| 49 |
|
| 50 |
|
| 51 |
This could help your case #4. Or, e.g., stage building: |
| 52 |
https://bugs.gentoo.org/show_bug.cgi?id=563034 |
| 53 |
|
| 54 |
toolchain's sysroot seems way more reliable though |
| 55 |
|
| 56 |
[...] |
| 57 |
|
| 58 |
Alexis. |
Archives