linux-kernel - Re: [PATCH v2 1/1] iommu/sva: Invalidate KVA range on kernel TLB flush

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: <20250711115441.GV1599700@nvidia.com>
Date: Fri, 11 Jul 2025 08:54:41 -0300
From: Jason Gunthorpe <jgg@...dia.com>
To: Baolu Lu <baolu.lu@...ux.intel.com>
Cc: Peter Zijlstra <peterz@...radead.org>, Joerg Roedel <joro@...tes.org>,
	Will Deacon <will@...nel.org>, Robin Murphy <robin.murphy@....com>,
	Kevin Tian <kevin.tian@...el.com>, Jann Horn <jannh@...gle.com>,
	Vasant Hegde <vasant.hegde@....com>,
	Dave Hansen <dave.hansen@...el.com>,
	Alistair Popple <apopple@...dia.com>,
	Uladzislau Rezki <urezki@...il.com>,
	Jean-Philippe Brucker <jean-philippe@...aro.org>,
	Andy Lutomirski <luto@...nel.org>,
	"Tested-by : Yi Lai" <yi1.lai@...el.com>, iommu@...ts.linux.dev,
	security@...nel.org, linux-kernel@...r.kernel.org,
	stable@...r.kernel.org
Subject: Re: [PATCH v2 1/1] iommu/sva: Invalidate KVA range on kernel TLB
 flush

On Fri, Jul 11, 2025 at 11:00:06AM +0800, Baolu Lu wrote:
> > > +void iommu_sva_invalidate_kva_range(unsigned long start, unsigned long end)
> > > +{
> > > +	struct iommu_mm_data *iommu_mm;
> > > +
> > > +	if (!static_branch_unlikely(&iommu_sva_present))
> > > +		return;
> > > +
> > > +	guard(spinlock_irqsave)(&iommu_mms_lock);
> > > +	list_for_each_entry(iommu_mm, &iommu_sva_mms, mm_list_elm)
> > > +		mmu_notifier_arch_invalidate_secondary_tlbs(iommu_mm->mm, start, end);
> > > +}
> > 
> > This is absolutely the wrong way to use static_branch. You want them in
> > inline functions guarding the function call, not inside the function
> > call.
> 
> I don't think a static branch is desirable here, as we have no idea how
> often the condition will switch in real-world scenarios. I will remove
> it in the next version if there are no objections.

The point of the static branch was to make the 99% normal case where
SVA has never been used have no-cost in the core MM. I think we should
keep that idea

Jason