[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20260116065400-9deb9bae-1621-4e3a-b12a-22e34906cb85@linutronix.de>
Date: Fri, 16 Jan 2026 07:05:39 +0100
From: Thomas Weißschuh <thomas.weissschuh@...utronix.de>
To: Paul Walmsley <pjw@...nel.org>
Cc: Palmer Dabbelt <palmer@...belt.com>, Albert Ou <aou@...s.berkeley.edu>,
Alexandre Ghiti <alex@...ti.fr>, linux-riscv@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] riscv: Implement ARCH_HAS_CC_CAN_LINK
Hi Paul,
On Thu, Jan 15, 2026 at 09:50:20PM -0700, Paul Walmsley wrote:
> On Mon, 22 Dec 2025, Thomas Weißschuh wrote:
>
> > The generic CC_CAN_LINK detection relies on -m32/-m64 compiler flags.
> > These are not supported by riscv compilers.
> >
> > Use architecture-specific logic using -mabi instead.
> >
> > Signed-off-by: Thomas Weißschuh <thomas.weissschuh@...utronix.de>
> > ---
> > arch/riscv/Kconfig | 11 +++++++++++
> > 1 file changed, 11 insertions(+)
> >
> > diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
> > index 6b39f37f769a..b4c1f922eed5 100644
> > --- a/arch/riscv/Kconfig
> > +++ b/arch/riscv/Kconfig
> > @@ -25,6 +25,7 @@ config RISCV
> > select ARCH_ENABLE_SPLIT_PMD_PTLOCK if PGTABLE_LEVELS > 2
> > select ARCH_ENABLE_THP_MIGRATION if TRANSPARENT_HUGEPAGE
> > select ARCH_HAS_BINFMT_FLAT
> > + select ARCH_HAS_CC_CAN_LINK
> > select ARCH_HAS_CURRENT_STACK_POINTER
> > select ARCH_HAS_DEBUG_VIRTUAL if MMU
> > select ARCH_HAS_DEBUG_VM_PGTABLE
> > @@ -1351,6 +1352,16 @@ config PORTABLE
> > config ARCH_PROC_KCORE_TEXT
> > def_bool y
> >
> > +config ARCH_CC_CAN_LINK
> > + bool
> > + default $(cc_can_link_user,-mabi=lp64d) if 64BIT
> > + default $(cc_can_link_user,-mabi=ilp32d)
> > +
> > +config ARCH_USERFLAGS
> > + string
> > + default "-mabi=lp64d" if 64BIT
> > + default "-mabi=ilp32d"
> > +
>
> Any reason why this patch shouldn't use the non-d ABI variants, as the
> rest of arch/riscv/Kconfig does? I suspect this may not work for
> !CONFIG_FPU builds.
Toolchains tend to only provide a libc for the 'd' ABI variants.
At least the ones I checked in Debian, Arch Linux and
https://toolchains.bootlin.com/. So I'd like to prefer those to increase
the chance of CONFIG_CC_CAN_LINK=y.
I'll add dedicated clauses for FPU=n in the next revision.
It would also be possible to have non-d as a fallback for FPU=y, but it
is slightly more complex to represent and will probably never be used.
Thomas
Powered by blists - more mailing lists