Gentoo Archives: gentoo-dev

From: WANG Xuerui <i.gentoo@×××××.name>
To: gentoo-dev@l.g.o
Subject: Re: [gentoo-dev] [RFC] Plans for a Gentoo/LoongArch port
Date: Thu, 07 Oct 2021 12:46:51
Message-Id: dd1a91ef-4487-9acc-9667-df391ec6be40@xen0n.name
In Reply to: [gentoo-dev] [RFC] Plans for a Gentoo/LoongArch port by WANG Xuerui
1 On 2021/8/12 00:39, WANG Xuerui wrote:
2 > Hi everyone,
3 >
4 > <snip>
5 >
6 > ## Gentoo porting plans
7 >
8 > I'm planning to take ARCH=loongarch for the port; and support the LP64
9 > ABI
10 > first. I'd like to support both LP64 and ILP32 ABIs, but that's not a
11 > priority.
12 >
13 > The ABI flag might be named "ABI_LOONGARCH" but that's IMO a bit long
14 > (pun
15 > semi-intended); ARCH=loong and ABI_LOONG might be better, I'm open to
16 > suggestions.
17 >
18 > Because much of the ABI and even some toolchain internals are going
19 > through
20 > VERY fierce debate and rework, obviously the port will remain
21 > experimental
22 > for a long time. Some minimal support should get in tree though; doing so
23 > would ease a lot of pain for experimentation. I already hacked my way to
24 > generate working crossdev toolchains, and is halfway towards a rootfs
25 > with
26 > working Python (and Portage). I've already independently ported
27 > strace, and
28 > plan to do the same to libffi in the coming days which would give me
29 > Python.
30 >
31 > I'll do all work in my own loongson-overlay first, and upstream these
32 > when
33 > appropriate. Eventually I hope to have working crossdev, qemu-user
34 > emulation
35 > and proper catalyst support.
36 >
37 Some kind of "progress update": I've successfully built stage3 and
38 minimal installation CD for ARCH=loong; ABI_LOONG is not implemented due
39 to Loongson retracting 32-bit support themselves, this port is 64-bit
40 only for now, and may remain so forever (depends on Loongson's
41 intentions). All modifications are made in loongson-overlay[1]; no
42 ad-hoc patches to main Portage tree is needed after all. My used
43 catalyst specs are put at [2]; currently all hard-coded and no automation.
44
45 I'm going to send patches for review gradually in the following days, as
46 the holiday here is over; however I'd like to know whether we need to
47 wait for the upstream merge of toolchain/kernel support before we can
48 take patches for ARCH=loong, or if there's more prerequisites.
49
50
51 [1]: https://github.com/xen0n/loongson-overlay
52
53 [2]: https://github.com/xen0n/releng/tree/loong