Re: [PATCH] riscv: Fix wrong size passed to local_flush_tlb_range_asid()

[Palmer Dabbelt]

On Wed, 31 Jan 2024 19:45:27 PST (-0800), dennis@xxxxxxxxxx wrote:
> Hi Palmer,
>
> On Wed, Jan 31, 2024 at 12:34:40PM -0800, Palmer Dabbelt wrote:
> > On Mon, 29 Jan 2024 01:01:00 PST (-0800), dennis@xxxxxxxxxx wrote:
> > > Hi Alexandre,
> > >
> > > On Tue, Jan 23, 2024 at 02:27:30PM +0100, Alexandre Ghiti wrote:
> > > > local_flush_tlb_range_asid() takes the size as argument, not the end of
> > > > the range to flush, so fix this by computing the size from the end and
> > > > the start of the range.
> > > >
> > > > Fixes: 7a92fc8b4d20 ("mm: Introduce flush_cache_vmap_early()")
> > > > Signed-off-by: Alexandre Ghiti <alexghiti@xxxxxxxxxxxx>
> > > > ---
> > > >  arch/riscv/mm/tlbflush.c | 2 +-
> > > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > >
> > > > diff --git a/arch/riscv/mm/tlbflush.c b/arch/riscv/mm/tlbflush.c
> > > > index 8d12b26f5ac3..9619965f6501 100644
> > > > --- a/arch/riscv/mm/tlbflush.c
> > > > +++ b/arch/riscv/mm/tlbflush.c
> > > > @@ -68,7 +68,7 @@ static inline void local_flush_tlb_range_asid(unsigned long start,
> > > >
> > > >  void local_flush_tlb_kernel_range(unsigned long start, unsigned long end)
> > > >  {
> > > > -	local_flush_tlb_range_asid(start, end, PAGE_SIZE, FLUSH_TLB_NO_ASID);
> > > > +	local_flush_tlb_range_asid(start, end - start, PAGE_SIZE, FLUSH_TLB_NO_ASID);
> > > >  }
> > > >
> > > >  static void __ipi_flush_tlb_all(void *info)
> > > > --
> > > > 2.39.2
> > > >
> > >
> > > Sorry for the delay, I just pulled this into percpu#for-6.8-fixes. I'll
> > > send it to Linus this week.
> >
> > Do you mind if we do a shared tag or something?  It's going to conflict with
> > https://lore.kernel.org/all/20240117140333.2479667-1-vincent.chen@xxxxxxxxxx/
> > .  No big deal as it's a pretty trivial conflict, but they'll both need
> > stable backports.
>
> This alone won't need a stable backport, I merged the bug as part of
> enabling the percpu page allocator in the recent 6.8 merge window.
>
> That being said, this is the only patch I'm carrying for v6.8. I'm happy
> to drop it and have you pick it up instead. Saves me a tag and a PR.
> Lmk if that works for you.

Sorry I missed this, but that seems like the easiest way to go here -- 
the other patch fixes the same bug (and another one), so I think we're 
safe with just that (which I'm sending to Linus as part of my rc4 fixes 
now-ish, as I'd need to anyway).  It's d9807d60c145 ("riscv: mm: execute 
local TLB flush after populating vmemmap"), in case anyone's looking 
later...