Hi,
On Mon, 2024-11-04 at 17:08 +0900, Hajime Tazaki wrote:
> this is a great clean up I think.
> I've also applied my tentative nommu patch and works fine with minor
> modifications.
>
> On Mon, 04 Nov 2024 00:05:34 +0900,
> Benjamin Berg wrote:
>
> > diff --git a/arch/um/include/asm/current.h
> > b/arch/um/include/asm/current.h
> > new file mode 100644
> > index 000000000000..93ee89e56372
> > --- /dev/null
> > +++ b/arch/um/include/asm/current.h
> > @@ -0,0 +1,24 @@
> > +/* SPDX-License-Identifier: GPL-2.0 */
> > +#ifndef __ASM_CURRENT_H
> > +#define __ASM_CURRENT_H
> > +
> > +#include <linux/compiler.h>
> > +
> > +#ifndef __ASSEMBLY__
> > +
> > +#include <as-layout.h>
>
> with this include, I guess <generated/asm-offsets.h> is now in a
> circular dependency.
>
> after only applying this patch onto uml/next, and make mrproper
> ARCH=um, I cannot build UML kernel with defconfig.
Oh. Probably the easiest is to just get rid of "struct cpu_task" and
then copy the definition of cpu_tasks over. i.e.
extern void * cpu_tasks[];
at that point.
> > diff --git a/arch/um/os-Linux/signal.c b/arch/um/os-Linux/signal.c
> > index 1978eaa557e9..87d31a0672c0 100644
> > --- a/arch/um/os-Linux/signal.c
> > +++ b/arch/um/os-Linux/signal.c
> > @@ -186,47 +186,14 @@ static void (*handlers[_NSIG])(int sig,
> > struct siginfo *si, mcontext_t *mc) = {
> > [SIGUSR1] = sigusr1_handler,
> > };
> >
> > +extern char cpu0_irqstack[];
> > +
>
> maybe this line isn't needed ?
Oops, yes. Initially I thought I still needed to detect nested signals.
That is a leftover of doing that by inspecting the stack pointer in the
mcontext register set.
Benjamin
Intel Deutschland GmbH
Registered Address: Am Campeon 10, 85579 Neubiberg, Germany
Tel: +49 89 99 8853-0, www.intel.de
Managing Directors: Sean Fennelly, Jeffrey Schneiderman, Tiffany Doon Silva
Chairperson of the Supervisory Board: Nicole Lau
Registered Office: Munich
Commercial Register: Amtsgericht Muenchen HRB 186928
