Re: [PATCH v8 11/15] s390: Distinguish hardware and software PTEs

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] s390/mm: set_pte() discards _PAGE_UNUSED clearance
--

commit a9d06a7ccc3dae14aa4eb7e5a46302fce251265b
Author: Alexander Gordeev <agordeev@linux.ibm.com>
s390: Distinguish hardware and software PTEs

This commit introduces the hw_pte_t type for hardware PTEs on s390,
distinguishing them from software PTEs which continue to use the pte_t type.
It updates various memory management functions and callbacks to enforce this
type separation, similar to ARM64. It also enables the ARCH_HAS_HW_PTE_T
configuration for the architecture.

> diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
> index e882663a58e77..c47264f3abf21 100644
> --- a/arch/s390/include/asm/pgtable.h
> +++ b/arch/s390/include/asm/pgtable.h

[ ... ]

> @@ -978,17 +978,19 @@ static inline void set_pmd(pmd_t *pmdp, pmd_t pmd)
>  	WRITE_ONCE(*pmdp, pmd);
>  }
>  
> -static inline void set_pte(pte_t *ptep, pte_t pte)
> +static inline void set_pte(hw_pte_t *ptep, pte_t pte)
>  {
> +	hw_pte_t hwpte = (hw_pte_t) { (pte) };
> +
>  	if (pte_present(pte))
>  		pte = clear_pte_bit(pte, __pgprot(_PAGE_UNUSED));
> -	WRITE_ONCE(*ptep, pte);
> +	WRITE_ONCE(*ptep, hwpte);

[Severity: High]
Does this discard the _PAGE_UNUSED bit clearance?

Because hwpte is initialized before pte is potentially modified, the
WRITE_ONCE call writes the original snapshot to the page table, ignoring
the update from clear_pte_bit.

If this _PAGE_UNUSED bit incorrectly remains set on an active page, could the
hypervisor (z/VM or KVM) reclaim its backing physical memory and cause silent
guest data loss or memory corruption?

>  }
>  
>  #define ptep_get ptep_get

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1791365932.git.agordeev@linux.ibm.com?part=11