mm: Make SPLIT_PTE_PTLOCKS depend on SMP

[PATCH v3] mm: Make SPLIT_PTE_PTLOCKS depend on SMP

Posted by Guenter Roeck 2 months ago

SPLIT_PTE_PTLOCKS depends on "NR_CPUS >= 4". Unfortunately, that evaluates
to true if there is no NR_CPUS configuration option. This results in
CONFIG_SPLIT_PTE_PTLOCKS=y for mac_defconfig. This in turn causes the m68k
"q800" and "virt" machines to crash in qemu if debugging options are
enabled.

Making CONFIG_SPLIT_PTE_PTLOCKS dependent on the existence of NR_CPUS
does not work since a dependency on the existence of a numeric Kconfig
entry always evaluates to false. Example:

config HAVE_NO_NR_CPUS
       def_bool y
       depends on !NR_CPUS

After adding this to a Kconfig file, "make defconfig" includes:
$ grep NR_CPUS .config
CONFIG_NR_CPUS=64
CONFIG_HAVE_NO_NR_CPUS=y

Defining NR_CPUS for m68k does not help either since many architectures
define NR_CPUS only for SMP configurations.

Make SPLIT_PTE_PTLOCKS depend on SMP instead to solve the problem.

Fixes: 394290cba966 ("mm: turn USE_SPLIT_PTE_PTLOCKS / USE_SPLIT_PTE_PTLOCKS into Kconfig options")
Cc: David Hildenbrand <david@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
---
v3: Make SPLIT_PTE_PTLOCKS depend on SMP instead of defining NR_CPUS
    for m68k.
    Drop David's Acked-by: tag.

v2: Instead of trying to make SPLIT_PTE_PTLOCKS depend on the 
    existence of NR_CPUS, define NR_CPUS for m68k.
    https://lore.kernel.org/lkml/20240923235617.1584056-1-linux@roeck-us.net/

v1: https://lore.kernel.org/lkml/20240923142533.1197982-1-linux@roeck-us.net/

 mm/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mm/Kconfig b/mm/Kconfig
index e85164c2812b..33fa51d608dc 100644
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -595,6 +595,7 @@ config ARCH_MHP_MEMMAP_ON_MEMORY_ENABLE
 config SPLIT_PTE_PTLOCKS
 	def_bool y
 	depends on MMU
+	depends on SMP
 	depends on NR_CPUS >= 4
 	depends on !ARM || CPU_CACHE_VIPT
 	depends on !PARISC || PA20
-- 
2.45.2

Re: [PATCH v3] mm: Make SPLIT_PTE_PTLOCKS depend on SMP

Posted by Geert Uytterhoeven 2 months ago

On Tue, Sep 24, 2024 at 5:42 PM Guenter Roeck <linux@roeck-us.net> wrote:
> SPLIT_PTE_PTLOCKS depends on "NR_CPUS >= 4". Unfortunately, that evaluates
> to true if there is no NR_CPUS configuration option. This results in
> CONFIG_SPLIT_PTE_PTLOCKS=y for mac_defconfig. This in turn causes the m68k
> "q800" and "virt" machines to crash in qemu if debugging options are
> enabled.
>
> Making CONFIG_SPLIT_PTE_PTLOCKS dependent on the existence of NR_CPUS
> does not work since a dependency on the existence of a numeric Kconfig
> entry always evaluates to false. Example:
>
> config HAVE_NO_NR_CPUS
>        def_bool y
>        depends on !NR_CPUS
>
> After adding this to a Kconfig file, "make defconfig" includes:
> $ grep NR_CPUS .config
> CONFIG_NR_CPUS=64
> CONFIG_HAVE_NO_NR_CPUS=y
>
> Defining NR_CPUS for m68k does not help either since many architectures
> define NR_CPUS only for SMP configurations.
>
> Make SPLIT_PTE_PTLOCKS depend on SMP instead to solve the problem.
>
> Fixes: 394290cba966 ("mm: turn USE_SPLIT_PTE_PTLOCKS / USE_SPLIT_PTE_PTLOCKS into Kconfig options")
> Cc: David Hildenbrand <david@redhat.com>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Cc: Geert Uytterhoeven <geert@linux-m68k.org>
> Signed-off-by: Guenter Roeck <linux@roeck-us.net>

Reviewed-by: Geert Uytterhoeven <geert@linux-m68k.org>
Tested-by: Geert Uytterhoeven <geert@linux-m68k.org>

> --- a/mm/Kconfig
> +++ b/mm/Kconfig
> @@ -595,6 +595,7 @@ config ARCH_MHP_MEMMAP_ON_MEMORY_ENABLE
>  config SPLIT_PTE_PTLOCKS
>         def_bool y
>         depends on MMU
> +       depends on SMP
>         depends on NR_CPUS >= 4

Nit: please combine these two lines, to make it clear they are related.

>         depends on !ARM || CPU_CACHE_VIPT
>         depends on !PARISC || PA20

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 v3] mm: Make SPLIT_PTE_PTLOCKS depend on SMP

Posted by David Hildenbrand 2 months ago

On 24.09.24 17:42, Guenter Roeck wrote:
> SPLIT_PTE_PTLOCKS depends on "NR_CPUS >= 4". Unfortunately, that evaluates
> to true if there is no NR_CPUS configuration option. This results in
> CONFIG_SPLIT_PTE_PTLOCKS=y for mac_defconfig. This in turn causes the m68k
> "q800" and "virt" machines to crash in qemu if debugging options are
> enabled.
> 
> Making CONFIG_SPLIT_PTE_PTLOCKS dependent on the existence of NR_CPUS
> does not work since a dependency on the existence of a numeric Kconfig
> entry always evaluates to false. Example:
> 
> config HAVE_NO_NR_CPUS
>         def_bool y
>         depends on !NR_CPUS
> 
> After adding this to a Kconfig file, "make defconfig" includes:
> $ grep NR_CPUS .config
> CONFIG_NR_CPUS=64
> CONFIG_HAVE_NO_NR_CPUS=y
> 
> Defining NR_CPUS for m68k does not help either since many architectures
> define NR_CPUS only for SMP configurations.
> 
> Make SPLIT_PTE_PTLOCKS depend on SMP instead to solve the problem.
> 
> Fixes: 394290cba966 ("mm: turn USE_SPLIT_PTE_PTLOCKS / USE_SPLIT_PTE_PTLOCKS into Kconfig options")
> Cc: David Hildenbrand <david@redhat.com>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Cc: Geert Uytterhoeven <geert@linux-m68k.org>
> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> ---

Acked-by: David Hildenbrand <david@redhat.com>

-- 
Cheers,

David / dhildenb