From: Sam James <sam@gentoo.org>
To: gentoo-dev@lists.gentoo.org
Subject: Re: [gentoo-dev] New bootstrap-prefix global USE-flag and patch to llvm.eclass
Date: Wed, 13 Sep 2023 07:22:30 +0100 [thread overview]
Message-ID: <87ledavbev.fsf@gentoo.org> (raw)
In-Reply-To: <ZQFTZBP1NQ46c9tE@gentoo.org>
Fabian Groffen <grobian@gentoo.org> writes:
> [[PGP Signed Part:Undecided]]
> On 12-09-2023 20:32:19 +0100, Alexey Sokolov wrote:
>> Bug: https://bugs.gentoo.org/758167
>> Full PR is at https://github.com/gentoo/gentoo/pull/32730
>>
>> Several LLVM packages require this early return, otherwise they fail to
>> build on Darwin. I'll also need this USE-flag for
>> sys-devel/clang-common, to distinguish between stage2 and stage3 of
>> bootstrap-prefix.sh to configure clang differently.
>>
>> --
>> Best regards,
>> Alexey "DarthGandalf" Sokolov
>
>> From de2bd1abc3e5c7607413633d132c604c6a801802 Mon Sep 17 00:00:00 2001
>> From: Alexey Sokolov <alexey+gentoo@asokolov.org>
>> Date: Mon, 11 Sep 2023 23:26:49 +0100
>> Subject: [PATCH] llvm.eclass: add global USE flag bootstrap-prefix
>>
>> Mask it everywhere except for prefix profiles
>>
>> Without this, stage2's LLVM packages fail to build.
>>
>> Bug: https://bugs.gentoo.org/758167
>> Signed-off-by: Alexey Sokolov <alexey+gentoo@asokolov.org>
>> ---
>> eclass/llvm.eclass | 7 +++++++
>> profiles/base/use.mask | 4 ++++
>> profiles/features/prefix/use.mask | 4 ++++
>> profiles/use.desc | 1 +
>> 4 files changed, 16 insertions(+)
>>
>> diff --git a/eclass/llvm.eclass b/eclass/llvm.eclass
>> index 8198650aad9a7..87c2cedb3a376 100644
>> --- a/eclass/llvm.eclass
>> +++ b/eclass/llvm.eclass
>> @@ -64,6 +64,8 @@ esac
>> if [[ ! ${_LLVM_ECLASS} ]]; then
>> _LLVM_ECLASS=1
>>
>> +IUSE="bootstrap-prefix"
>> +
>> # make sure that the versions installing straight into /usr/bin
>> # are uninstalled
>> DEPEND="!!sys-devel/llvm:0"
>> @@ -242,6 +244,11 @@ llvm_fix_tool_path() {
>> llvm_pkg_setup() {
>> debug-print-function ${FUNCNAME} "${@}"
>>
>> + if use bootstrap-prefix; then
>> + # AppleClang has unparseable version numbers, but it's irrelevant anyway
>> + return
>> + fi
>> +
>
> I might misunderstand this, but is this USE-flag supposed to be set only
> during bootstrap, e.g. when host-provided Clang is used? If so, would
> it be possible to use has_version or something instead?
Another option is something I think we've done in the past - check
for use prefix and then some extra env var we set in the bootstrap
script.
I think I'd prefer either your idea or the one I just mention
to a USE, but I don't think I feel very strongly between any of it.
(but given mgorny isn't keen on the USE in the PR at
https://github.com/gentoo/gentoo/pull/32730,
that's a vote against it)
>
> Thanks,
> Fabian
>
next prev parent reply other threads:[~2023-09-13 6:23 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-12 19:32 [gentoo-dev] New bootstrap-prefix global USE-flag and patch to llvm.eclass Alexey Sokolov
2023-09-13 6:15 ` Fabian Groffen
2023-09-13 6:22 ` Sam James [this message]
2023-09-13 7:41 ` Alexey Sokolov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=87ledavbev.fsf@gentoo.org \
--to=sam@gentoo.org \
--cc=gentoo-dev@lists.gentoo.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox