ARM: shmobile: smp: Enforce shmobile_smp_* alignment

[PATCH] ARM: shmobile: smp: Enforce shmobile_smp_* alignment

Posted by Geert Uytterhoeven 11 months ago

When the addresses of the shmobile_smp_mpidr, shmobile_smp_fn, and
shmobile_smp_arg variables are not multiples of 4 bytes, secondary CPU
bring-up fails:

    smp: Bringing up secondary CPUs ...
    CPU1: failed to come online
    CPU2: failed to come online
    CPU3: failed to come online
    smp: Brought up 1 node, 1 CPU

Fix this by adding the missing alignment directive.

Fixes: 4e960f52fce16a3b ("ARM: shmobile: Move shmobile_smp_{mpidr, fn, arg}[] from .text to .bss")
Closes: https://lore.kernel.org/r/CAMuHMdU=QR-JLgEHKWpsr6SbaZRc-Hz9r91JfpP8c3n2G-OjqA@mail.gmail.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
To be queued in renesas-fixes-for-v6.14.
---
 arch/arm/mach-shmobile/headsmp.S | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm/mach-shmobile/headsmp.S b/arch/arm/mach-shmobile/headsmp.S
index a956b489b6ea12ca..2bc7e73a8582d2b3 100644
--- a/arch/arm/mach-shmobile/headsmp.S
+++ b/arch/arm/mach-shmobile/headsmp.S
@@ -136,6 +136,7 @@ ENDPROC(shmobile_smp_sleep)
 	.long	shmobile_smp_arg - 1b
 
 	.bss
+	.align	2
 	.globl	shmobile_smp_mpidr
 shmobile_smp_mpidr:
 	.space	NR_CPUS * 4
-- 
2.43.0

Re: [PATCH] ARM: shmobile: smp: Enforce shmobile_smp_* alignment

Posted by Lad, Prabhakar 11 months ago

Hi Geert,

Thank you for the patch.

On Mon, Mar 10, 2025 at 1:14 PM Geert Uytterhoeven
<geert+renesas@glider.be> wrote:
>
> When the addresses of the shmobile_smp_mpidr, shmobile_smp_fn, and
> shmobile_smp_arg variables are not multiples of 4 bytes, secondary CPU
> bring-up fails:
>
>     smp: Bringing up secondary CPUs ...
>     CPU1: failed to come online
>     CPU2: failed to come online
>     CPU3: failed to come online
>     smp: Brought up 1 node, 1 CPU
>
> Fix this by adding the missing alignment directive.
>
> Fixes: 4e960f52fce16a3b ("ARM: shmobile: Move shmobile_smp_{mpidr, fn, arg}[] from .text to .bss")
I wonder if this fixes tag should go back a bit far as I was able to
reproduce this on 5.10-cip BSP kernel on RZ/G1E this was only seen
when CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE was enabled,

RZ/G1E failure logs:
--------------------------
[    0.006719] smp: Bringing up secondary CPUs ...
[    1.040749] CPU1: failed to come online
[    1.041014] smp: Brought up 1 node, 1 CPU
[    1.041038] SMP: Total of 1 processors activated (65.00 BogoMIPS).
[    1.041063] CPU: All CPU(s) started in SVC mode.
[    1.041904] devtmpfs: initialized
[    1.050811] VFP support v0.3: implementor 41 architecture 2 part 30
variant 7 rev 5
[    1.051143] clocksource: jiffies: mask: 0xffffffff max_cycles:
0xffffffff, max_idle_ns: 19112604462750000 ns
[    1.051197] futex hash table entries: 512 (order: 3, 32768 bytes, linear)
[    1.060510] pinctrl core: initialized pinctrl subsystem
[    1.063139] NET: Registered protocol family 16

$ grep shmobile_smp_ System.map-notworking
c021caa0 t shmobile_smp_apmu_cpu_kill
c021cb20 t shmobile_smp_apmu_enter_suspend
c021cb54 t shmobile_smp_apmu_boot_secondary
c021cbc0 t shmobile_smp_apmu_cpu_shutdown
c021cc18 t shmobile_smp_apmu_do_suspend
c021cc60 t shmobile_smp_apmu_cpu_die
c021cc84 T shmobile_smp_hook
c021ccd8 T shmobile_smp_cpu_can_disable
c021d050 t shmobile_smp_continue_gen2
c021d0c4 T shmobile_smp_boot
c021d0e4 t shmobile_smp_boot_find_mpidr
c021d0fc t shmobile_smp_boot_next
c021d10c t shmobile_smp_boot_found
c021d114 T shmobile_smp_sleep
c021d380 T shmobile_smp_scu_cpu_die
c021d3b4 T shmobile_smp_scu_cpu_kill
c120a87c t shmobile_smp_apmu_prepare_cpus_dt
c120aab8 T shmobile_smp_apmu_suspend_init
c120aad0 T shmobile_smp_init_fallback_ops
c120b0f0 T shmobile_smp_scu_prepare_cpus
c1262778 t __cpu_method_of_table_shmobile_smp_apmu
c154d5fd B shmobile_smp_mpidr
c154d61d B shmobile_smp_fn
c154d63d B shmobile_smp_arg


> Closes: https://lore.kernel.org/r/CAMuHMdU=QR-JLgEHKWpsr6SbaZRc-Hz9r91JfpP8c3n2G-OjqA@mail.gmail.com
> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
> ---
> To be queued in renesas-fixes-for-v6.14.
> ---
>  arch/arm/mach-shmobile/headsmp.S | 1 +
>  1 file changed, 1 insertion(+)
>
Tested-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>

Cheers,
Prabhakar

> diff --git a/arch/arm/mach-shmobile/headsmp.S b/arch/arm/mach-shmobile/headsmp.S
> index a956b489b6ea12ca..2bc7e73a8582d2b3 100644
> --- a/arch/arm/mach-shmobile/headsmp.S
> +++ b/arch/arm/mach-shmobile/headsmp.S
> @@ -136,6 +136,7 @@ ENDPROC(shmobile_smp_sleep)
>         .long   shmobile_smp_arg - 1b
>
>         .bss
> +       .align  2
>         .globl  shmobile_smp_mpidr
>  shmobile_smp_mpidr:
>         .space  NR_CPUS * 4
> --
> 2.43.0
>
>

Re: [PATCH] ARM: shmobile: smp: Enforce shmobile_smp_* alignment

Posted by Geert Uytterhoeven 11 months ago

Hi Prabhakar,

On Tue, 11 Mar 2025 at 16:11, Lad, Prabhakar <prabhakar.csengg@gmail.com> wrote:
> On Mon, Mar 10, 2025 at 1:14 PM Geert Uytterhoeven
> > When the addresses of the shmobile_smp_mpidr, shmobile_smp_fn, and
> > shmobile_smp_arg variables are not multiples of 4 bytes, secondary CPU
> > bring-up fails:
> >
> >     smp: Bringing up secondary CPUs ...
> >     CPU1: failed to come online
> >     CPU2: failed to come online
> >     CPU3: failed to come online
> >     smp: Brought up 1 node, 1 CPU
> >
> > Fix this by adding the missing alignment directive.
> >
> > Fixes: 4e960f52fce16a3b ("ARM: shmobile: Move shmobile_smp_{mpidr, fn, arg}[] from .text to .bss")
>
> I wonder if this fixes tag should go back a bit far as I was able to
> reproduce this on 5.10-cip BSP kernel on RZ/G1E this was only seen
> when CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE was enabled,

4e960f52fce16a3b is in v4.5, so it covers 5.10-cip.

> RZ/G1E failure logs:
> --------------------------
> [    0.006719] smp: Bringing up secondary CPUs ...
> [    1.040749] CPU1: failed to come online

> c154d5fd B shmobile_smp_mpidr
> c154d61d B shmobile_smp_fn
> c154d63d B shmobile_smp_arg

Bang!

> > Closes: https://lore.kernel.org/r/CAMuHMdU=QR-JLgEHKWpsr6SbaZRc-Hz9r91JfpP8c3n2G-OjqA@mail.gmail.com
> > Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
> > ---
> > To be queued in renesas-fixes-for-v6.14.
> > ---
> >  arch/arm/mach-shmobile/headsmp.S | 1 +
> >  1 file changed, 1 insertion(+)
> >
> Tested-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>

Thanks!

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH] ARM: shmobile: smp: Enforce shmobile_smp_* alignment

Posted by Lad, Prabhakar 11 months ago

Hi Geert,

On Tue, Mar 11, 2025 at 3:20 PM Geert Uytterhoeven <geert@linux-m68k.org> wrote:
>
> Hi Prabhakar,
>
> On Tue, 11 Mar 2025 at 16:11, Lad, Prabhakar <prabhakar.csengg@gmail.com> wrote:
> > On Mon, Mar 10, 2025 at 1:14 PM Geert Uytterhoeven
> > > When the addresses of the shmobile_smp_mpidr, shmobile_smp_fn, and
> > > shmobile_smp_arg variables are not multiples of 4 bytes, secondary CPU
> > > bring-up fails:
> > >
> > >     smp: Bringing up secondary CPUs ...
> > >     CPU1: failed to come online
> > >     CPU2: failed to come online
> > >     CPU3: failed to come online
> > >     smp: Brought up 1 node, 1 CPU
> > >
> > > Fix this by adding the missing alignment directive.
> > >
> > > Fixes: 4e960f52fce16a3b ("ARM: shmobile: Move shmobile_smp_{mpidr, fn, arg}[] from .text to .bss")
> >
> > I wonder if this fixes tag should go back a bit far as I was able to
> > reproduce this on 5.10-cip BSP kernel on RZ/G1E this was only seen
> > when CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE was enabled,
>
> 4e960f52fce16a3b is in v4.5, so it covers 5.10-cip.
>
Thanks for pointing.

Cheers,
Prabhakar