Re: [PATCH 02/12] mm: hugetlb_vmemmap: unify VMEMMAP_{SPLIT,REMAP}_NO_TLB_FLUSH

[Qi Zheng <qi.zheng@linux.dev>]

On 10/20/26 7:58 PM, Muchun Song wrote:
> VMEMMAP_SPLIT_NO_TLB_FLUSH and VMEMMAP_REMAP_NO_TLB_FLUSH do not
> describe fully independent choices. A split-only walk can defer the
> PMD-split flush to a later global flush, and a remap-only walk can
> defer the final remap flush. However, when a PMD split is followed
> by PTE remapping, the split flush must happen before the PTEs are
> changed, because stale large-page translations must not overlap with
> new small-page translations that map different physical pages.
> 
> Replace the two flags with a single VMEMMAP_REMAP_NO_TLB_FLUSH flag.
> Keep split-only walks able to defer the flush, but force the split
> flush whenever a remap_pte callback is present. Callers that batch
> HVO work can still defer the final remap flush and issue their own
> global flush.
> 
> Current callers keep the same flushing behavior.
> 
> Signed-off-by: Muchun Song <songmuchun@bytedance.com>
> ---
>   mm/hugetlb_vmemmap.c | 15 +++++++++------
>   1 file changed, 9 insertions(+), 6 deletions(-)
> 
> diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c
> index 1105aa925acc..8b3abb562599 100644
> --- a/mm/hugetlb_vmemmap.c
> +++ b/mm/hugetlb_vmemmap.c
> @@ -42,10 +42,8 @@ struct vmemmap_remap_walk {
>   	struct list_head	*vmemmap_pages;
>   
>   
> -/* Skip the TLB flush when we split the PMD */
> -#define VMEMMAP_SPLIT_NO_TLB_FLUSH	BIT(0)
> -/* Skip the TLB flush when we remap the PTE */
> -#define VMEMMAP_REMAP_NO_TLB_FLUSH	BIT(1)
> +/* Skip the TLB flush while updating the vmemmap page tables. */
> +#define VMEMMAP_REMAP_NO_TLB_FLUSH	BIT(0)

Since this flag is now also used by the split-only path, the
VMEMMAP_REMAP_NO_TLB_FLUSH name is somewhat misleading. Moreover, the
flag does not suppress every TLB flush, as the PMD-split flush is still
mandatory when remapping follows.

Would VMEMMAP_DEFER_TLB_FLUSH better describe its semantics? It
indicates that flushes which can be safely batched are deferred, without
implying that all TLB flushes are skipped.

Thanks,
Qi

>   	unsigned long		flags;
>   };
>   
> @@ -85,7 +83,12 @@ static int vmemmap_split_pmd(pmd_t *pmd, struct page *head, unsigned long start,
>   		/* Make pte visible before pmd. See comment in pmd_install(). */
>   		smp_wmb();
>   		pmd_populate_kernel(&init_mm, pmd, pgtable);
> -		if (!(walk->flags & VMEMMAP_SPLIT_NO_TLB_FLUSH))
> +		/*
> +		 * Modern MMUs allow overlapping TLB entries with different mapping
> +		 * granularities only if they map the same physical address. So
> +		 * remapping must not proceed without first flushing the TLB.
> +		 */
> +		if (walk->remap_pte || !(walk->flags & VMEMMAP_REMAP_NO_TLB_FLUSH))
>   			flush_tlb_kernel_range(start, start + PMD_SIZE);
>   	} else {
>   		pte_free_kernel(&init_mm, pgtable);
> @@ -272,7 +275,7 @@ static int vmemmap_remap_split(unsigned long start, unsigned long end)
>   {
>   	struct vmemmap_remap_walk walk = {
>   		.remap_pte	= NULL,
> -		.flags		= VMEMMAP_SPLIT_NO_TLB_FLUSH,
> +		.flags		= VMEMMAP_REMAP_NO_TLB_FLUSH,
>   	};
>   
>   	return vmemmap_remap_range(start, end, &walk);