Summary: | sys-devel/clang-3.1 - Updated ebuild with +python and +doc | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Alex Turbov <I.zaufi> |
Component: | New packages | Assignee: | Gentoo Clang Team (OBSOLETE) <clang+obsolete> |
Status: | RESOLVED FIXED | ||
Severity: | enhancement | Keywords: | EBUILD |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
new ebuild w/ +python and +doc USE flags
new ebuild w/ +python and +doc USE flags "clean" patch by diff -ub |
Description
Alex Turbov
2012-11-16 20:32:25 UTC
Created attachment 329672 [details]
new ebuild w/ +python and +doc USE flags
Created attachment 329674 [details]
new ebuild w/ +python and +doc USE flags
sorry, the previous one was wrong. here is a better one.
Can you provide a clean minimal diff? I just ran a diff against the latest revision in portage and it's hard to see the changes from the diff, did you change a lot of indentation by accident? Created attachment 329688 [details, diff]
"clean" patch by diff -ub
(In reply to comment #3) > Can you provide a clean minimal diff? I just ran a diff against the latest > revision in portage and it's hard to see the changes from the diff, did you > change a lot of indentation by accident? not by accedent... my favorite editor configured that way... btw, you may use `diff -b` to avoid spaces counting... anyway I've attached a 'clean' patch Thanks, haven't considered that and will do so in the future; I usually attach a diff to ease the job. Assigned the bug to the maintainers. Thanks for the patch. I have some doubts, however. Firstly, do you defer documentation building to src_install()? It should happen in src_compile(), and src_install() should just install the files. Secondly, would you mind if we deferred the Python-related changes for a few days? I'd like to make it use pythonr-r1, and for that I'd need a few more days to get feedback. (In reply to comment #7) > Thanks for the patch. I have some doubts, however. > > Firstly, do you defer documentation building to src_install()? It should > happen in src_compile(), and src_install() should just install the files. > yep, I did that in my first version... but unfortunately 'install-doxygen' target declared in the clang/docs/Makefile depends on another one called 'doxygen'. The latter is declared as PHONEY, so if you'll put `make doxygen` into src_compile() it will build API dox (wasting noticable time), and then on `make install` it will rebuild all (same) docs again... that behaviour is definitely silly, but its in upstream :( ... > Secondly, would you mind if we deferred the Python-related changes for a few > days? I'd like to make it use pythonr-r1, and for that I'd need a few more > days to get feedback. yeah, sure Cleaning old llvm/clang bugs, USE=doc on recent llvm ebuilds also installs clang documentation, and python parts in clang are installed (not really worth patching to disable the install). So everything done for this bug :) Unfortunately setting USE+=doc on llvm do not produce clang API documentation :( Do you mean the doxygen doc? If it is different from the sphinx doc, we can enable it (the ebuilds have hardcoded -DLLVM_ENABLE_DOXYGEN=OFF) If that is the case, can you open a new bug for this? Thanks! |