2023-03-22 08:25:23

by Mike Rapoport

Subject: [PATCH] loongarch: drop ranges for definition of ARCH_FORCE_MAX_ORDER

From: "Mike Rapoport (IBM)" <[email protected]>

LoongArch defines insane ranges for ARCH_FORCE_MAX_ORDER allowing
MAX_ORDER up to 63, which implies maximal contiguous allocation size of
2^63 pages.

Drop bogus definitions of ranges for ARCH_FORCE_MAX_ORDER and leave it a
simple integer with sensible defaults.

Users that *really* need to change the value of ARCH_FORCE_MAX_ORDER
will be able to do so but they won't be mislead by the bogus ranges.

Signed-off-by: Mike Rapoport (IBM) <[email protected]>
---

This applies to akpm/mm-unstable tree

arch/loongarch/Kconfig | 3 ---
1 file changed, 3 deletions(-)

diff --git a/arch/loongarch/Kconfig b/arch/loongarch/Kconfig
index 272a3a12c98d..e1e3a3828962 100644
--- a/arch/loongarch/Kconfig
+++ b/arch/loongarch/Kconfig
@@ -420,11 +420,8 @@ config NODES_SHIFT

config ARCH_FORCE_MAX_ORDER
int "Maximum zone order"
- range 13 63 if PAGE_SIZE_64KB
default "13" if PAGE_SIZE_64KB
- range 11 63 if PAGE_SIZE_16KB
default "11" if PAGE_SIZE_16KB
- range 10 63
default "10"
help
The kernel memory allocator divides physically contiguous memory
--
2.35.1

2023-03-22 10:09:52

by Kirill A. Shutemov

Subject: Re: [PATCH] loongarch: drop ranges for definition of ARCH_FORCE_MAX_ORDER

On Wed, Mar 22, 2023 at 10:17:27AM +0200, Mike Rapoport wrote:
> From: "Mike Rapoport (IBM)" <[email protected]>
>
> LoongArch defines insane ranges for ARCH_FORCE_MAX_ORDER allowing
> MAX_ORDER up to 63, which implies maximal contiguous allocation size of
> 2^63 pages.
>
> Drop bogus definitions of ranges for ARCH_FORCE_MAX_ORDER and leave it a
> simple integer with sensible defaults.
>
> Users that *really* need to change the value of ARCH_FORCE_MAX_ORDER
> will be able to do so but they won't be mislead by the bogus ranges.
>
> Signed-off-by: Mike Rapoport (IBM) <[email protected]>

Acked-by: Kirill A. Shutemov <[email protected]>

--
Kiryl Shutsemau / Kirill A. Shutemov

2023-03-22 17:19:15

by David Hildenbrand

Subject: Re: [PATCH] loongarch: drop ranges for definition of ARCH_FORCE_MAX_ORDER

On 22.03.23 09:17, Mike Rapoport wrote:
> From: "Mike Rapoport (IBM)" <[email protected]>
>
> LoongArch defines insane ranges for ARCH_FORCE_MAX_ORDER allowing
> MAX_ORDER up to 63, which implies maximal contiguous allocation size of
> 2^63 pages.
>
> Drop bogus definitions of ranges for ARCH_FORCE_MAX_ORDER and leave it a
> simple integer with sensible defaults.
>
> Users that *really* need to change the value of ARCH_FORCE_MAX_ORDER
> will be able to do so but they won't be mislead by the bogus ranges.
>
> Signed-off-by: Mike Rapoport (IBM) <[email protected]>
> ---
>
> This applies to akpm/mm-unstable tree
>
> arch/loongarch/Kconfig | 3 ---
> 1 file changed, 3 deletions(-)
>
> diff --git a/arch/loongarch/Kconfig b/arch/loongarch/Kconfig
> index 272a3a12c98d..e1e3a3828962 100644
> --- a/arch/loongarch/Kconfig
> +++ b/arch/loongarch/Kconfig
> @@ -420,11 +420,8 @@ config NODES_SHIFT
>
> config ARCH_FORCE_MAX_ORDER
> int "Maximum zone order"
> - range 13 63 if PAGE_SIZE_64KB
> default "13" if PAGE_SIZE_64KB
> - range 11 63 if PAGE_SIZE_16KB
> default "11" if PAGE_SIZE_16KB
> - range 10 63
> default "10"
> help
> The kernel memory allocator divides physically contiguous memory

Reviewed-by: David Hildenbrand <[email protected]>

--
Thanks,

David / dhildenb