From: Han Gao <gaohan@iscas.ac.cn>
Allwinner currently offers d1(s)/v821/v861/v881 on RISC-V,
using different IPs.
d1(s): Xuantie C906
v821: Andes A27 + XuanTie E907
v861/v881: XuanTie C907
Signed-off-by: Han Gao <gaohan@iscas.ac.cn>
---
arch/riscv/Kconfig.socs | 22 +++++++++++++++++-----
1 file changed, 17 insertions(+), 5 deletions(-)
diff --git a/arch/riscv/Kconfig.socs b/arch/riscv/Kconfig.socs
index 848e7149e443..7cba5d6ec4c3 100644
--- a/arch/riscv/Kconfig.socs
+++ b/arch/riscv/Kconfig.socs
@@ -54,14 +54,26 @@ config SOC_STARFIVE
help
This enables support for StarFive SoC platform hardware.
-config ARCH_SUNXI
- bool "Allwinner sun20i SoCs"
+menuconfig ARCH_SUNXI
+ bool "Allwinner RISC-V SoCs"
+
+if ARCH_SUNXI
+
+config ARCH_SUNXI_XUANTIE
+ bool "Allwinner Xuantie IP SoCs"
depends on MMU && !XIP_KERNEL
- select ERRATA_THEAD
select SUN4I_TIMER
+ select ERRATA_THEAD
help
- This enables support for Allwinner sun20i platform hardware,
- including boards based on the D1 and D1s SoCs.
+ This enables support for Allwinner Xuantie IP SoCs.
+
+config ARCH_SUNXI_ANDES
+ bool "Allwinner Andes IP SoCs"
+ select ERRATA_ANDES
+ help
+ This enables support for Allwinner Andes IP SoCs.
+
+endif
config ARCH_THEAD
bool "T-HEAD RISC-V SoCs"
--
2.47.3
On 08/11/2025 09:20, gaohan@iscas.ac.cn wrote: > From: Han Gao <gaohan@iscas.ac.cn> > > Allwinner currently offers d1(s)/v821/v861/v881 on RISC-V, > using different IPs. > > d1(s): Xuantie C906 > v821: Andes A27 + XuanTie E907 > v861/v881: XuanTie C907 > > Signed-off-by: Han Gao <gaohan@iscas.ac.cn> > --- > arch/riscv/Kconfig.socs | 22 +++++++++++++++++----- > 1 file changed, 17 insertions(+), 5 deletions(-) > > diff --git a/arch/riscv/Kconfig.socs b/arch/riscv/Kconfig.socs > index 848e7149e443..7cba5d6ec4c3 100644 > --- a/arch/riscv/Kconfig.socs > +++ b/arch/riscv/Kconfig.socs > @@ -54,14 +54,26 @@ config SOC_STARFIVE > help > This enables support for StarFive SoC platform hardware. > > -config ARCH_SUNXI > - bool "Allwinner sun20i SoCs" > +menuconfig ARCH_SUNXI > + bool "Allwinner RISC-V SoCs" > + > +if ARCH_SUNXI > + > +config ARCH_SUNXI_XUANTIE You should not get multiple ARCHs. ARCH is only one. There is also not much rationale in commit msg for that. Best regards, Krzysztof
> -----Original Messages----- > From: "Krzysztof Kozlowski" <krzk@kernel.org> > Sent Time: 2025-11-08 19:29:07 (Saturday) > To: gaohan@iscas.ac.cn, "Paul Walmsley" <pjw@kernel.org>, "Palmer Dabbelt" <palmer@dabbelt.com>, "Albert Ou" <aou@eecs.berkeley.edu>, "Alexandre Ghiti" <alex@ghiti.fr>, "Rob Herring" <robh@kernel.org>, "Krzysztof Kozlowski" <krzk+dt@kernel.org>, "Conor Dooley" <conor+dt@kernel.org>, "Chen-Yu Tsai" <wens@csie.org>, "Jernej Skrabec" <jernej.skrabec@gmail.com>, "Samuel Holland" <samuel@sholland.org>, "Yixun Lan" <dlan@gentoo.org>, "Drew Fustini" <fustini@kernel.org>, "Geert Uytterhoeven" <geert+renesas@glider.be>, "Guodong Xu" <guodong@riscstar.com>, "Haylen Chu" <heylenay@4d2.org>, "Joel Stanley" <joel@jms.id.au> > Cc: linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-sunxi@lists.linux.dev, "Han Gao" <rabenda.cn@gmail.com> > Subject: Re: [PATCH 1/3] riscv: soc: re-organized allwinner menu > > On 08/11/2025 09:20, gaohan@iscas.ac.cn wrote: >> From: Han Gao <gaohan@iscas.ac.cn> >> >> Allwinner currently offers d1(s)/v821/v861/v881 on RISC-V, >> using different IPs. >> >> d1(s): Xuantie C906 >> v821: Andes A27 + XuanTie E907 >> v861/v881: XuanTie C907 >> >> Signed-off-by: Han Gao <gaohan@iscas.ac.cn> >> --- >> arch/riscv/Kconfig.socs | 22 +++++++++++++++++----- >> 1 file changed, 17 insertions(+), 5 deletions(-) >> >> diff --git a/arch/riscv/Kconfig.socs b/arch/riscv/Kconfig.socs >> index 848e7149e443..7cba5d6ec4c3 100644 >> --- a/arch/riscv/Kconfig.socs >> +++ b/arch/riscv/Kconfig.socs >> @@ -54,14 +54,26 @@ config SOC_STARFIVE >> help >> This enables support for StarFive SoC platform hardware. >> >> -config ARCH_SUNXI >> - bool "Allwinner sun20i SoCs" >> +menuconfig ARCH_SUNXI >> + bool "Allwinner RISC-V SoCs" >> + >> +if ARCH_SUNXI >> + >> +config ARCH_SUNXI_XUANTIE > > > You should not get multiple ARCHs. ARCH is only one. There is also not > much rationale in commit msg for that. The main goal is to avoid choosing multiple IP addresses for erreta. If using Andes IPs, I don't want to choose XuanTIe (T-Head) ERRETA. For example, v821 uses Andes ax27, but it used to select ERRATA_THEAD. > > Best regards, > Krzysztof
On 08/11/2025 14:59, revy wrote: > > > >> -----Original Messages----- >> From: "Krzysztof Kozlowski" <krzk@kernel.org> >> Sent Time: 2025-11-08 19:29:07 (Saturday) >> To: gaohan@iscas.ac.cn, "Paul Walmsley" <pjw@kernel.org>, "Palmer Dabbelt" <palmer@dabbelt.com>, "Albert Ou" <aou@eecs.berkeley.edu>, "Alexandre Ghiti" <alex@ghiti.fr>, "Rob Herring" <robh@kernel.org>, "Krzysztof Kozlowski" <krzk+dt@kernel.org>, "Conor Dooley" <conor+dt@kernel.org>, "Chen-Yu Tsai" <wens@csie.org>, "Jernej Skrabec" <jernej.skrabec@gmail.com>, "Samuel Holland" <samuel@sholland.org>, "Yixun Lan" <dlan@gentoo.org>, "Drew Fustini" <fustini@kernel.org>, "Geert Uytterhoeven" <geert+renesas@glider.be>, "Guodong Xu" <guodong@riscstar.com>, "Haylen Chu" <heylenay@4d2.org>, "Joel Stanley" <joel@jms.id.au> >> Cc: linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-sunxi@lists.linux.dev, "Han Gao" <rabenda.cn@gmail.com> >> Subject: Re: [PATCH 1/3] riscv: soc: re-organized allwinner menu >> >> On 08/11/2025 09:20, gaohan@iscas.ac.cn wrote: >>> From: Han Gao <gaohan@iscas.ac.cn> >>> >>> Allwinner currently offers d1(s)/v821/v861/v881 on RISC-V, >>> using different IPs. >>> >>> d1(s): Xuantie C906 >>> v821: Andes A27 + XuanTie E907 >>> v861/v881: XuanTie C907 >>> >>> Signed-off-by: Han Gao <gaohan@iscas.ac.cn> >>> --- >>> arch/riscv/Kconfig.socs | 22 +++++++++++++++++----- >>> 1 file changed, 17 insertions(+), 5 deletions(-) >>> >>> diff --git a/arch/riscv/Kconfig.socs b/arch/riscv/Kconfig.socs >>> index 848e7149e443..7cba5d6ec4c3 100644 >>> --- a/arch/riscv/Kconfig.socs >>> +++ b/arch/riscv/Kconfig.socs >>> @@ -54,14 +54,26 @@ config SOC_STARFIVE >>> help >>> This enables support for StarFive SoC platform hardware. >>> >>> -config ARCH_SUNXI >>> - bool "Allwinner sun20i SoCs" >>> +menuconfig ARCH_SUNXI >>> + bool "Allwinner RISC-V SoCs" >>> + >>> +if ARCH_SUNXI >>> + >>> +config ARCH_SUNXI_XUANTIE >> >> >> You should not get multiple ARCHs. ARCH is only one. There is also not >> much rationale in commit msg for that. > > The main goal is to avoid choosing multiple IP addresses for erreta. > If using Andes IPs, I don't want to choose XuanTIe (T-Head) ERRETA. Not explained in commit msg but anyway not a good argument. It is some sort of micro optimization and you completely miss the point we target multiarch kernels. Best regards, Krzysztof
On 08/11/2025 15:47, Krzysztof Kozlowski wrote: > On 08/11/2025 14:59, revy wrote: >> >> >> >>> -----Original Messages----- >>> From: "Krzysztof Kozlowski" <krzk@kernel.org> >>> Sent Time: 2025-11-08 19:29:07 (Saturday) >>> To: gaohan@iscas.ac.cn, "Paul Walmsley" <pjw@kernel.org>, "Palmer Dabbelt" <palmer@dabbelt.com>, "Albert Ou" <aou@eecs.berkeley.edu>, "Alexandre Ghiti" <alex@ghiti.fr>, "Rob Herring" <robh@kernel.org>, "Krzysztof Kozlowski" <krzk+dt@kernel.org>, "Conor Dooley" <conor+dt@kernel.org>, "Chen-Yu Tsai" <wens@csie.org>, "Jernej Skrabec" <jernej.skrabec@gmail.com>, "Samuel Holland" <samuel@sholland.org>, "Yixun Lan" <dlan@gentoo.org>, "Drew Fustini" <fustini@kernel.org>, "Geert Uytterhoeven" <geert+renesas@glider.be>, "Guodong Xu" <guodong@riscstar.com>, "Haylen Chu" <heylenay@4d2.org>, "Joel Stanley" <joel@jms.id.au> >>> Cc: linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-sunxi@lists.linux.dev, "Han Gao" <rabenda.cn@gmail.com> >>> Subject: Re: [PATCH 1/3] riscv: soc: re-organized allwinner menu >>> >>> On 08/11/2025 09:20, gaohan@iscas.ac.cn wrote: >>>> From: Han Gao <gaohan@iscas.ac.cn> >>>> >>>> Allwinner currently offers d1(s)/v821/v861/v881 on RISC-V, >>>> using different IPs. >>>> >>>> d1(s): Xuantie C906 >>>> v821: Andes A27 + XuanTie E907 >>>> v861/v881: XuanTie C907 >>>> >>>> Signed-off-by: Han Gao <gaohan@iscas.ac.cn> >>>> --- >>>> arch/riscv/Kconfig.socs | 22 +++++++++++++++++----- >>>> 1 file changed, 17 insertions(+), 5 deletions(-) >>>> >>>> diff --git a/arch/riscv/Kconfig.socs b/arch/riscv/Kconfig.socs >>>> index 848e7149e443..7cba5d6ec4c3 100644 >>>> --- a/arch/riscv/Kconfig.socs >>>> +++ b/arch/riscv/Kconfig.socs >>>> @@ -54,14 +54,26 @@ config SOC_STARFIVE >>>> help >>>> This enables support for StarFive SoC platform hardware. >>>> >>>> -config ARCH_SUNXI >>>> - bool "Allwinner sun20i SoCs" >>>> +menuconfig ARCH_SUNXI >>>> + bool "Allwinner RISC-V SoCs" >>>> + >>>> +if ARCH_SUNXI >>>> + >>>> +config ARCH_SUNXI_XUANTIE >>> >>> >>> You should not get multiple ARCHs. ARCH is only one. There is also not >>> much rationale in commit msg for that. >> >> The main goal is to avoid choosing multiple IP addresses for erreta. >> If using Andes IPs, I don't want to choose XuanTIe (T-Head) ERRETA. > > Not explained in commit msg but anyway not a good argument. It is some > sort of micro optimization and you completely miss the point we target > multiarch kernels. Heh, and I actually did not forbid or discourage choosing erratas per your soc. I said you only get one top level ARCH. Look at all arm64 platforms. How many ARCHs are there per one vendor? Best regards, Krzysztof
On Sat, Nov 08, 2025 at 03:48:18PM +0100, Krzysztof Kozlowski wrote: > On 08/11/2025 15:47, Krzysztof Kozlowski wrote: > > On 08/11/2025 14:59, revy wrote: > >> > >> > >> > >>> -----Original Messages----- > >>> From: "Krzysztof Kozlowski" <krzk@kernel.org> > >>> Sent Time: 2025-11-08 19:29:07 (Saturday) > >>> To: gaohan@iscas.ac.cn, "Paul Walmsley" <pjw@kernel.org>, "Palmer Dabbelt" <palmer@dabbelt.com>, "Albert Ou" <aou@eecs.berkeley.edu>, "Alexandre Ghiti" <alex@ghiti.fr>, "Rob Herring" <robh@kernel.org>, "Krzysztof Kozlowski" <krzk+dt@kernel.org>, "Conor Dooley" <conor+dt@kernel.org>, "Chen-Yu Tsai" <wens@csie.org>, "Jernej Skrabec" <jernej.skrabec@gmail.com>, "Samuel Holland" <samuel@sholland.org>, "Yixun Lan" <dlan@gentoo.org>, "Drew Fustini" <fustini@kernel.org>, "Geert Uytterhoeven" <geert+renesas@glider.be>, "Guodong Xu" <guodong@riscstar.com>, "Haylen Chu" <heylenay@4d2.org>, "Joel Stanley" <joel@jms.id.au> > >>> Cc: linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-sunxi@lists.linux.dev, "Han Gao" <rabenda.cn@gmail.com> > >>> Subject: Re: [PATCH 1/3] riscv: soc: re-organized allwinner menu > >>> > >>> On 08/11/2025 09:20, gaohan@iscas.ac.cn wrote: > >>>> From: Han Gao <gaohan@iscas.ac.cn> > >>>> > >>>> Allwinner currently offers d1(s)/v821/v861/v881 on RISC-V, > >>>> using different IPs. > >>>> > >>>> d1(s): Xuantie C906 > >>>> v821: Andes A27 + XuanTie E907 > >>>> v861/v881: XuanTie C907 > >>>> > >>>> Signed-off-by: Han Gao <gaohan@iscas.ac.cn> > >>>> --- > >>>> arch/riscv/Kconfig.socs | 22 +++++++++++++++++----- > >>>> 1 file changed, 17 insertions(+), 5 deletions(-) > >>>> > >>>> diff --git a/arch/riscv/Kconfig.socs b/arch/riscv/Kconfig.socs > >>>> index 848e7149e443..7cba5d6ec4c3 100644 > >>>> --- a/arch/riscv/Kconfig.socs > >>>> +++ b/arch/riscv/Kconfig.socs > >>>> @@ -54,14 +54,26 @@ config SOC_STARFIVE > >>>> help > >>>> This enables support for StarFive SoC platform hardware. > >>>> > >>>> -config ARCH_SUNXI > >>>> - bool "Allwinner sun20i SoCs" > >>>> +menuconfig ARCH_SUNXI > >>>> + bool "Allwinner RISC-V SoCs" > >>>> + > >>>> +if ARCH_SUNXI > >>>> + > >>>> +config ARCH_SUNXI_XUANTIE > >>> > >>> > >>> You should not get multiple ARCHs. ARCH is only one. There is also not > >>> much rationale in commit msg for that. > >> > >> The main goal is to avoid choosing multiple IP addresses for erreta. > >> If using Andes IPs, I don't want to choose XuanTIe (T-Head) ERRETA. > > > > Not explained in commit msg but anyway not a good argument. It is some > > sort of micro optimization and you completely miss the point we target > > multiarch kernels. > > Heh, and I actually did not forbid or discourage choosing erratas per > your soc. I said you only get one top level ARCH. Look at all arm64 > platforms. How many ARCHs are there per one vendor? Yeah, it only allows you to enable the errata, it doesn't force any of them to "y". Some will get enabled by default when ARCH_SUNXI is enabled, but if someone is only targeting on device they can just turn them off... I'm pretty inclined to just NAK this unless there's some actual value.
> On Nov 9, 2025, at 00:23, Conor Dooley <conor@kernel.org> wrote: > > On Sat, Nov 08, 2025 at 03:48:18PM +0100, Krzysztof Kozlowski wrote: >> On 08/11/2025 15:47, Krzysztof Kozlowski wrote: >>> On 08/11/2025 14:59, revy wrote: >>>> >>>> >>>> >>>>> -----Original Messages----- >>>>> From: "Krzysztof Kozlowski" <krzk@kernel.org> >>>>> Sent Time: 2025-11-08 19:29:07 (Saturday) >>>>> To: gaohan@iscas.ac.cn, "Paul Walmsley" <pjw@kernel.org>, "Palmer Dabbelt" <palmer@dabbelt.com>, "Albert Ou" <aou@eecs.berkeley.edu>, "Alexandre Ghiti" <alex@ghiti.fr>, "Rob Herring" <robh@kernel.org>, "Krzysztof Kozlowski" <krzk+dt@kernel.org>, "Conor Dooley" <conor+dt@kernel.org>, "Chen-Yu Tsai" <wens@csie.org>, "Jernej Skrabec" <jernej.skrabec@gmail.com>, "Samuel Holland" <samuel@sholland.org>, "Yixun Lan" <dlan@gentoo.org>, "Drew Fustini" <fustini@kernel.org>, "Geert Uytterhoeven" <geert+renesas@glider.be>, "Guodong Xu" <guodong@riscstar.com>, "Haylen Chu" <heylenay@4d2.org>, "Joel Stanley" <joel@jms.id.au> >>>>> Cc: linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-sunxi@lists.linux.dev, "Han Gao" <rabenda.cn@gmail.com> >>>>> Subject: Re: [PATCH 1/3] riscv: soc: re-organized allwinner menu >>>>> >>>>> On 08/11/2025 09:20, gaohan@iscas.ac.cn wrote: >>>>>> From: Han Gao <gaohan@iscas.ac.cn> >>>>>> >>>>>> Allwinner currently offers d1(s)/v821/v861/v881 on RISC-V, >>>>>> using different IPs. >>>>>> >>>>>> d1(s): Xuantie C906 >>>>>> v821: Andes A27 + XuanTie E907 >>>>>> v861/v881: XuanTie C907 >>>>>> >>>>>> Signed-off-by: Han Gao <gaohan@iscas.ac.cn> >>>>>> --- >>>>>> arch/riscv/Kconfig.socs | 22 +++++++++++++++++----- >>>>>> 1 file changed, 17 insertions(+), 5 deletions(-) >>>>>> >>>>>> diff --git a/arch/riscv/Kconfig.socs b/arch/riscv/Kconfig.socs >>>>>> index 848e7149e443..7cba5d6ec4c3 100644 >>>>>> --- a/arch/riscv/Kconfig.socs >>>>>> +++ b/arch/riscv/Kconfig.socs >>>>>> @@ -54,14 +54,26 @@ config SOC_STARFIVE >>>>>> help >>>>>> This enables support for StarFive SoC platform hardware. >>>>>> >>>>>> -config ARCH_SUNXI >>>>>> - bool "Allwinner sun20i SoCs" >>>>>> +menuconfig ARCH_SUNXI >>>>>> + bool "Allwinner RISC-V SoCs" >>>>>> + >>>>>> +if ARCH_SUNXI >>>>>> + >>>>>> +config ARCH_SUNXI_XUANTIE >>>>> >>>>> >>>>> You should not get multiple ARCHs. ARCH is only one. There is also not >>>>> much rationale in commit msg for that. >>>> >>>> The main goal is to avoid choosing multiple IP addresses for erreta. >>>> If using Andes IPs, I don't want to choose XuanTIe (T-Head) ERRETA. >>> >>> Not explained in commit msg but anyway not a good argument. It is some >>> sort of micro optimization and you completely miss the point we target >>> multiarch kernels. >> >> Heh, and I actually did not forbid or discourage choosing erratas per >> your soc. I said you only get one top level ARCH. Look at all arm64 >> platforms. How many ARCHs are there per one vendor? > > > Yeah, it only allows you to enable the errata, it doesn't force any of > them to "y". Some will get enabled by default when ARCH_SUNXI is > enabled, but if someone is only targeting on device they can just turn > them off... I'm pretty inclined to just NAK this unless there's some > actual value. I understand. I'm going to abandon this patch and plan to resubmit a patch that only modifies the description from sun20i to allwinnner.
© 2016 - 2025 Red Hat, Inc.