Re: [RFC PATCH 05/12] fs/proc/task_mmu: refactor pagemap_pmd_range()

[Gregory Price]

On Fri, Oct 24, 2025 at 08:41:21AM +0100, Lorenzo Stoakes wrote:
> Separate out THP logic so we can drop an indentation level and reduce the
> amount of noise in this function.
> 
> We add pagemap_pmd_range_thp() for this purpose.
> 
> While we're here, convert the VM_BUG_ON() to a VM_WARN_ON_ONCE() at the
> same time.
> 
> Signed-off-by: Lorenzo Stoakes <lorenzo.stoakes@xxxxxxxxxx>
... >8
> +static int pagemap_pmd_range(pmd_t *pmdp, unsigned long addr, unsigned long end,
> +			     struct mm_walk *walk)
> +{
> +	struct vm_area_struct *vma = walk->vma;
> +	struct pagemapread *pm = walk->private;
> +	spinlock_t *ptl;
> +	pte_t *pte, *orig_pte;
> +	int err = 0;
> +
> +#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> +	ptl = pmd_trans_huge_lock(pmdp, vma);
> +	if (ptl)
> +		return pagemap_pmd_range_thp(pmdp, addr, end, vma, pm, ptl);
> +#endif

Maybe something like...

#ifdef CONFIG_TRANSPARENT_HUGEPAGE
ptl = pmd_trans_huge_lock(pmdp, vma);
if (ptl) {
	err = pagemap_pmd_range_thp(pmdp, addr, end, vma, pm, ptl);
	spin_unlock(ptl);
	return err;
}
#endif

and drop the spin_unlock(ptl) calls from pagemap_pmd_range_thp?

Makes it easier to understand the locking semantics.

Might be worth adding a comment to pagemap_pmd_range_thp that callers
must hold the ptl lock.

~Gregory

P.S. This patch set made my day, the whole non-swap-swap thing has
always broken my brain.  <3