[PATCH V3 1/4] riscv: mm: Fixup compat mode boot failure

[Charlie Jenkins]

On Sat, Dec 23, 2023 at 11:07:29AM +0800, Guo Ren wrote:
> On Sat, Dec 23, 2023 at 10:58 AM Leonardo Bras <leobras at redhat.com> wrote:
> >
> > On Fri, Dec 22, 2023 at 06:57:00AM -0500, guoren at kernel.org wrote:
> > > From: Guo Ren <guoren at linux.alibaba.com>
> > >
> > > In COMPAT mode, the STACK_TOP is DEFAULT_MAP_WINDOW (0x80000000), but
> > > the TASK_SIZE is 0x7fff000. When the user stack is upon 0x7fff000, it
> > > will cause a user segment fault. Sometimes, it would cause boot
> > > failure when the whole rootfs is rv32.
> > >
> > > Freeing unused kernel image (initmem) memory: 2236K
> > > Run /sbin/init as init process
> > > Starting init: /sbin/init exists but couldn't execute it (error -14)
> > > Run /etc/init as init process
> > > ...
> > >
> > > Increase the TASK_SIZE to cover STACK_TOP.
> > >
> > > Cc: stable at vger.kernel.org
> > > Fixes: add2cc6b6515 ("RISC-V: mm: Restrict address space for sv39,sv48,sv57")
> > > Signed-off-by: Guo Ren <guoren at linux.alibaba.com>
> > > Signed-off-by: Guo Ren <guoren at kernel.org>
> > > ---
> > >  arch/riscv/include/asm/pgtable.h | 2 +-
> > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > >
> > > diff --git a/arch/riscv/include/asm/pgtable.h b/arch/riscv/include/asm/pgtable.h
> > > index ab00235b018f..74ffb2178f54 100644
> > > --- a/arch/riscv/include/asm/pgtable.h
> > > +++ b/arch/riscv/include/asm/pgtable.h
> > > @@ -881,7 +881,7 @@ static inline pte_t pte_swp_clear_exclusive(pte_t pte)
> > >  #define TASK_SIZE_MIN        (PGDIR_SIZE_L3 * PTRS_PER_PGD / 2)
> > >
> > >  #ifdef CONFIG_COMPAT
> > > -#define TASK_SIZE_32 (_AC(0x80000000, UL) - PAGE_SIZE)
> > > +#define TASK_SIZE_32 (_AC(0x80000000, UL))
> > >  #define TASK_SIZE    (test_thread_flag(TIF_32BIT) ? \
> > >                        TASK_SIZE_32 : TASK_SIZE_64)
> > >  #else
> > > --
> > > 2.40.1
> > >
> >
> > I am not really involved in the issue this is solving, so I have no
> > technical opinion on the solution.
> >
> > IIUC there should always be (TASK_SIZE >= STACK_TOP), so by itself this
> > is fixing an issue.
> >
> > I have reviewed the code and it does exactly as stated into the commit
> > message, so FWIW:
> > Reviewed-by: Leonardo Bras <leobras at redhat.com>
> Thx,
> 
> I found this problem because it can't boot my rv32 buildroot-rootfs in
> v6.6. But it's okay in v6.5. So I used git bisect and found commit:
> add2cc6b6515 ("RISC-V: mm: Restrict address space for
> sv39,sv48,sv57"), which caused that.
> 
> Ping Charlie, I hope it can be fixed in the v6.6 long-term version.

I have looked at this more I do agree that this change makes sense. I
was thinking that it was valid to have STACK_TOP that was greater than
TASK_SIZE but I am no longer convinced that is true. Thank you for
looking into this.

Reviewed-by: Charlie Jenkins <charlie at rivosinc.com>

> 
> >
> 
> 
> 
> --
> Best Regards
>  Guo Ren