[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1598945161.7wv2e8mu4h.astroid@bobo.none>
Date: Tue, 01 Sep 2020 17:50:24 +1000
From: Nicholas Piggin <npiggin@...il.com>
To: Geert Uytterhoeven <geert@...ux-m68k.org>
Cc: Arnd Bergmann <arnd@...db.de>,
Linux-Arch <linux-arch@...r.kernel.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
linux-m68k <linux-m68k@...ts.linux-m68k.org>,
Linux MM <linux-mm@...ck.org>
Subject: Re: [PATCH v2 09/23] m68k: use asm-generic/mmu_context.h for no-op
implementations
Excerpts from Geert Uytterhoeven's message of September 1, 2020 5:03 pm:
> Hi Nick,
>
> On Tue, Sep 1, 2020 at 8:23 AM Nicholas Piggin <npiggin@...il.com> wrote:
>> Excerpts from Geert Uytterhoeven's message of August 27, 2020 7:33 pm:
>> > On Wed, Aug 26, 2020 at 4:53 PM Nicholas Piggin <npiggin@...il.com> wrote:
>> >> Cc: Geert Uytterhoeven <geert@...ux-m68k.org>
>> >> Cc: linux-m68k@...ts.linux-m68k.org
>> >> Signed-off-by: Nicholas Piggin <npiggin@...il.com>
>> >
>> > With the below fixed:
>> > Acked-by: Geert Uytterhoeven <geert@...ux-m68k.org>
>> >
>> >> --- a/arch/m68k/include/asm/mmu_context.h
>> >> +++ b/arch/m68k/include/asm/mmu_context.h
>> >> @@ -79,19 +76,6 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
>> >> set_context(tsk->mm->context, next->pgd);
>> >> }
>> >>
>> >> -/*
>> >> - * After we have set current->mm to a new value, this activates
>> >> - * the context for the new mm so we see the new mappings.
>> >> - */
>> >> -static inline void activate_mm(struct mm_struct *active_mm,
>> >> - struct mm_struct *mm)
>> >> -{
>> >> - get_mmu_context(mm);
>> >> - set_context(mm->context, mm->pgd);
>> >> -}
>> >
>> > Assumed switch_mm() in [PATCH v2 01/23] is revived with the above body.
>>
>> I'm not sure what you mean here. We can remove this because it's a copy
>> of switch_mm above, and that's what the new header defaults to if you
>> don't provide an active_mm.
>
> IC. I thought it started relying on <asm-generic/mmu_context.h> for this,
> where you removed switch_mm().
>
> Seems I missed the definition above.
It's supposed to all build incrementally, I'll try to make sure it's
right...
>
>> Patch 1 should not have changed that, it should only affect the nommu
>> architectures (and actually didn't touch m68k because it was not using
>> the asm-generic/mmu_context.h header).
>
> OK. Sorry for the noise.
No problem thanks for looking at it.
Thanks,
Nick
Powered by blists - more mailing lists