Message-ID: <bc49d4df-5d64-4eaf-951c-37dc5d4389d4@arm.com>
Date:   Tue, 21 Nov 2023 15:14:55 +0000
From:   Ryan Roberts <ryan.roberts@....com>
To:     Alistair Popple <apopple@...dia.com>
Cc:     Catalin Marinas <catalin.marinas@....com>,
        Will Deacon <will@...nel.org>,
        Ard Biesheuvel <ardb@...nel.org>,
        Marc Zyngier <maz@...nel.org>,
        Oliver Upton <oliver.upton@...ux.dev>,
        James Morse <james.morse@....com>,
        Suzuki K Poulose <suzuki.poulose@....com>,
        Zenghui Yu <yuzenghui@...wei.com>,
        Andrey Ryabinin <ryabinin.a.a@...il.com>,
        Alexander Potapenko <glider@...gle.com>,
        Andrey Konovalov <andreyknvl@...il.com>,
        Dmitry Vyukov <dvyukov@...gle.com>,
        Vincenzo Frascino <vincenzo.frascino@....com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Anshuman Khandual <anshuman.khandual@....com>,
        Matthew Wilcox <willy@...radead.org>,
        Yu Zhao <yuzhao@...gle.com>,
        Mark Rutland <mark.rutland@....com>,
        David Hildenbrand <david@...hat.com>,
        Kefeng Wang <wangkefeng.wang@...wei.com>,
        John Hubbard <jhubbard@...dia.com>, Zi Yan <ziy@...dia.com>,
        linux-arm-kernel@...ts.infradead.org, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 12/14] arm64/mm: Wire up PTE_CONT for user mappings

On 21/11/2023 11:22, Alistair Popple wrote:
> 
> Ryan Roberts <ryan.roberts@....com> writes:
> 
> [...]
> 
>> +static void contpte_fold(struct mm_struct *mm, unsigned long addr,
>> +			pte_t *ptep, pte_t pte, bool fold)
>> +{
>> +	struct vm_area_struct vma = TLB_FLUSH_VMA(mm, 0);
>> +	unsigned long start_addr;
>> +	pte_t *start_ptep;
>> +	int i;
>> +
>> +	start_ptep = ptep = contpte_align_down(ptep);
>> +	start_addr = addr = ALIGN_DOWN(addr, CONT_PTE_SIZE);
>> +	pte = pfn_pte(ALIGN_DOWN(pte_pfn(pte), CONT_PTES), pte_pgprot(pte));
>> +	pte = fold ? pte_mkcont(pte) : pte_mknoncont(pte);
>> +
>> +	for (i = 0; i < CONT_PTES; i++, ptep++, addr += PAGE_SIZE) {
>> +		pte_t ptent = __ptep_get_and_clear(mm, addr, ptep);
>> +
>> +		if (pte_dirty(ptent))
>> +			pte = pte_mkdirty(pte);
>> +
>> +		if (pte_young(ptent))
>> +			pte = pte_mkyoung(pte);
>> +	}
>> +
>> +	__flush_tlb_range(&vma, start_addr, addr, PAGE_SIZE, true, 3);
>> +
>> +	__set_ptes(mm, start_addr, start_ptep, pte, CONT_PTES);
>> +}
>> +
>> +void __contpte_try_fold(struct mm_struct *mm, unsigned long addr,
>> +			pte_t *ptep, pte_t pte)
>> +{
>> +	/*
>> +	 * We have already checked that the virtual and pysical addresses are
>> +	 * correctly aligned for a contpte mapping in contpte_try_fold() so the
>> +	 * remaining checks are to ensure that the contpte range is fully
>> +	 * covered by a single folio, and ensure that all the ptes are valid
>> +	 * with contiguous PFNs and matching prots. We ignore the state of the
>> +	 * access and dirty bits for the purpose of deciding if its a contiguous
>> +	 * range; the folding process will generate a single contpte entry which
>> +	 * has a single access and dirty bit. Those 2 bits are the logical OR of
>> +	 * their respective bits in the constituent pte entries. In order to
>> +	 * ensure the contpte range is covered by a single folio, we must
>> +	 * recover the folio from the pfn, but special mappings don't have a
>> +	 * folio backing them. Fortunately contpte_try_fold() already checked
>> +	 * that the pte is not special - we never try to fold special mappings.
>> +	 * Note we can't use vm_normal_page() for this since we don't have the
>> +	 * vma.
>> +	 */
>> +
>> +	struct page *page = pte_page(pte);
>> +	struct folio *folio = page_folio(page);
>> +	unsigned long folio_saddr = addr - (page - &folio->page) * PAGE_SIZE;
>> +	unsigned long folio_eaddr = folio_saddr + folio_nr_pages(folio) * PAGE_SIZE;
>> +	unsigned long cont_saddr = ALIGN_DOWN(addr, CONT_PTE_SIZE);
>> +	unsigned long cont_eaddr = cont_saddr + CONT_PTE_SIZE;
>> +	unsigned long pfn;
>> +	pgprot_t prot;
>> +	pte_t subpte;
>> +	pte_t *orig_ptep;
>> +	int i;
>> +
>> +	if (folio_saddr > cont_saddr || folio_eaddr < cont_eaddr)
>> +		return;
>> +
>> +	pfn = pte_pfn(pte) - ((addr - cont_saddr) >> PAGE_SHIFT);
>> +	prot = pte_pgprot(pte_mkold(pte_mkclean(pte)));
>> +	orig_ptep = ptep;
>> +	ptep = contpte_align_down(ptep);
>> +
>> +	for (i = 0; i < CONT_PTES; i++, ptep++, pfn++) {
>> +		subpte = __ptep_get(ptep);
>> +		subpte = pte_mkold(pte_mkclean(subpte));
>> +
>> +		if (!pte_valid(subpte) ||
>> +		    pte_pfn(subpte) != pfn ||
>> +		    pgprot_val(pte_pgprot(subpte)) != pgprot_val(prot))
>> +			return;
>> +	}
>> +
>> +	contpte_fold(mm, addr, orig_ptep, pte, true);
>> +}
>> +EXPORT_SYMBOL(__contpte_try_fold);
>> +
>> +void __contpte_try_unfold(struct mm_struct *mm, unsigned long addr,
>> +			pte_t *ptep, pte_t pte)
>> +{
>> +	/*
>> +	 * We have already checked that the ptes are contiguous in
>> +	 * contpte_try_unfold(), so we can unfold unconditionally here.
>> +	 */
>> +
>> +	contpte_fold(mm, addr, ptep, pte, false);
> 
> I'm still working my way through the series but 

Thanks for taking the time to review!

> calling a fold during an
> unfold stood out as it seemed wrong. Obviously further reading revealed
> the boolean flag that changes the functions meaning but I think it would
> be better to refactor that.

Yes that sounds reasonable.

> 
> We could easily rename contpte_fold() to eg. set_cont_ptes() and factor
> the pte calculation loop into a separate helper
> (eg. calculate_contpte_dirty_young() or some hopefully better name)
> called further up the stack. That has an added benefit of providing a
> spot to add the nice comment for young/dirty rules you provided in the
> patch description ;-)
> 
> In other words we'd have something like:
> 
> void __contpte_try_unfold() {
>      pte = calculate_contpte_dirty_young(mm, addr, ptep, pte);
>      pte = pte_mknoncont(pte);
>      set_cont_ptes(mm, addr, ptep, pte);
> }

My concern with this approach is that calculate_contpte_dirty_young() has side
effects; it has to clear each PTE as it loops through it prevent a race between
our reading access/dirty and another thread causing access/dirty to be set. So
its not just a "calculation", its the teardown portion of the process too. I
guess its a taste thing, so happy for it to be argued the other way, but I would
prefer to keep it all together in one function.

How about renaming contpte_fold() to contpte_convert() or contpte_repaint()
(other suggestions welcome), and extracting the pte_mkcont()/pte_mknoncont()
part (so we can remove the bool param):

void __contpte_try_unfold() {
	pte = pte_mknoncont(pte);
	contpte_convert(mm, addr, ptep, pte);
}

Thanks,
Ryan

> 
> Which IMHO is more immediately understandable.
> 
>  - Alistair
> 

Powered by blists - more mailing lists