linux-kernel - Re: [PATCH v6 16/26] huge_memory: Add vmf_insert_folio

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <yl437hg47vr7a7hc467eqag2imceyj5xnnf4hju2f27rvxo4ib@hb27ob6jhq3q>
Date: Wed, 15 Jan 2025 18:05:56 +1100
From: Alistair Popple <apopple@...dia.com>
To: Dan Williams <dan.j.williams@...el.com>
Cc: David Hildenbrand <david@...hat.com>, akpm@...ux-foundation.org, 
	linux-mm@...ck.org, alison.schofield@...el.com, lina@...hilina.net, 
	zhang.lyra@...il.com, gerald.schaefer@...ux.ibm.com, vishal.l.verma@...el.com, 
	dave.jiang@...el.com, logang@...tatee.com, bhelgaas@...gle.com, jack@...e.cz, 
	jgg@...pe.ca, catalin.marinas@....com, will@...nel.org, mpe@...erman.id.au, 
	npiggin@...il.com, dave.hansen@...ux.intel.com, ira.weiny@...el.com, 
	willy@...radead.org, djwong@...nel.org, tytso@....edu, linmiaohe@...wei.com, 
	peterx@...hat.com, linux-doc@...r.kernel.org, linux-kernel@...r.kernel.org, 
	linux-arm-kernel@...ts.infradead.org, linuxppc-dev@...ts.ozlabs.org, nvdimm@...ts.linux.dev, 
	linux-cxl@...r.kernel.org, linux-fsdevel@...r.kernel.org, linux-ext4@...r.kernel.org, 
	linux-xfs@...r.kernel.org, jhubbard@...dia.com, hch@....de, david@...morbit.com, 
	chenhuacai@...nel.org, kernel@...0n.name, loongarch@...ts.linux.dev
Subject: Re: [PATCH v6 16/26] huge_memory: Add vmf_insert_folio_pmd()

On Tue, Jan 14, 2025 at 09:22:00AM -0800, Dan Williams wrote:
> David Hildenbrand wrote:
> > > +vm_fault_t vmf_insert_folio_pmd(struct vm_fault *vmf, struct folio *folio, bool write)
> > > +{
> > > +	struct vm_area_struct *vma = vmf->vma;
> > > +	unsigned long addr = vmf->address & PMD_MASK;
> > > +	struct mm_struct *mm = vma->vm_mm;
> > > +	spinlock_t *ptl;
> > > +	pgtable_t pgtable = NULL;
> > > +
> > > +	if (addr < vma->vm_start || addr >= vma->vm_end)
> > > +		return VM_FAULT_SIGBUS;
> > > +
> > > +	if (WARN_ON_ONCE(folio_order(folio) != PMD_ORDER))
> > > +		return VM_FAULT_SIGBUS;
> > > +
> > > +	if (arch_needs_pgtable_deposit()) {
> > > +		pgtable = pte_alloc_one(vma->vm_mm);
> > > +		if (!pgtable)
> > > +			return VM_FAULT_OOM;
> > > +	}
> > 
> > This is interesting and nasty at the same time (only to make ppc64 boo3s 
> > with has tables happy). But it seems to be the right thing to do.
> > 
> > > +
> > > +	ptl = pmd_lock(mm, vmf->pmd);
> > > +	if (pmd_none(*vmf->pmd)) {
> > > +		folio_get(folio);
> > > +		folio_add_file_rmap_pmd(folio, &folio->page, vma);
> > > +		add_mm_counter(mm, mm_counter_file(folio), HPAGE_PMD_NR);
> > > +	}
> > > +	insert_pfn_pmd(vma, addr, vmf->pmd, pfn_to_pfn_t(folio_pfn(folio)),
> > > +		       vma->vm_page_prot, write, pgtable);
> > > +	spin_unlock(ptl);
> > > +	if (pgtable)
> > > +		pte_free(mm, pgtable);
> > 
> > Ehm, are you unconditionally freeing the pgtable, even if consumed by 
> > insert_pfn_pmd() ?
> > 
> > Note that setting pgtable to NULL in insert_pfn_pmd() when consumed will 
> > not be visible here.
> > 
> > You'd have to pass a pointer to the ... pointer (&pgtable).
> > 
> > ... unless I am missing something, staring at the diff.
> 
> In fact I glazed over the fact that this has been commented on before
> and assumed it was fixed:
> 
> http://lore.kernel.org/66f61ce4da80_964f2294fb@dwillia2-xfh.jf.intel.com.notmuch
> 
> So, yes, insert_pfn_pmd needs to take &pgtable to report back if the
> allocation got consumed.
> 
> Good catch.

Yes, thanks Dave and Dan and apologies for missing that originally. Looking
at the thread I suspect I went down the rabbit hole of trying to implement
vmf_insert_folio() and when that wasn't possible forgot to come back and fix
this up. I have added a return code to insert_pfn_pmd() to indicate whether
or not the pgtable was consumed. I have also added a comment in the commit log
explaining why a vmf_insert_folio() isn't useful.

 - Alistair