From: David Gibson <david@gibson.dropbear.id.au>
To: Paul Mackerras <paulus@samba.org>
Cc: Alexey Kardashevskiy <aik@ozlabs.ru>,
Gavin Shan <gwshan@linux.vnet.ibm.com>,
linux-kernel@vger.kernel.org,
Alex Williamson <alex.williamson@redhat.com>,
linuxppc-dev@lists.ozlabs.org
Subject: Re: [PATCH kernel v9 28/32] powerpc/mmu: Add userspace-to-physical addresses translation cache
Date: Fri, 1 May 2015 13:39:53 +1000 [thread overview]
Message-ID: <20150501033953.GJ24886@voom.redhat.com> (raw)
In-Reply-To: <20150430082525.GA22373@iris.ozlabs.ibm.com>
[-- Attachment #1: Type: text/plain, Size: 3815 bytes --]
On Thu, Apr 30, 2015 at 06:25:25PM +1000, Paul Mackerras wrote:
> On Thu, Apr 30, 2015 at 04:34:55PM +1000, David Gibson wrote:
> > On Sat, Apr 25, 2015 at 10:14:52PM +1000, Alexey Kardashevskiy wrote:
> > > We are adding support for DMA memory pre-registration to be used in
> > > conjunction with VFIO. The idea is that the userspace which is going to
> > > run a guest may want to pre-register a user space memory region so
> > > it all gets pinned once and never goes away. Having this done,
> > > a hypervisor will not have to pin/unpin pages on every DMA map/unmap
> > > request. This is going to help with multiple pinning of the same memory
> > > and in-kernel acceleration of DMA requests.
> > >
> > > This adds a list of memory regions to mm_context_t. Each region consists
> > > of a header and a list of physical addresses. This adds API to:
> > > 1. register/unregister memory regions;
> > > 2. do final cleanup (which puts all pre-registered pages);
> > > 3. do userspace to physical address translation;
> > > 4. manage a mapped pages counter; when it is zero, it is safe to
> > > unregister the region.
> > >
> > > Multiple registration of the same region is allowed, kref is used to
> > > track the number of registrations.
> >
> > [snip]
> > > +long mm_iommu_alloc(unsigned long ua, unsigned long entries,
> > > + struct mm_iommu_table_group_mem_t **pmem)
> > > +{
> > > + struct mm_iommu_table_group_mem_t *mem;
> > > + long i, j;
> > > + struct page *page = NULL;
> > > +
> > > + list_for_each_entry_rcu(mem, ¤t->mm->context.iommu_group_mem_list,
> > > + next) {
> > > + if ((mem->ua == ua) && (mem->entries == entries))
> > > + return -EBUSY;
> > > +
> > > + /* Overlap? */
> > > + if ((mem->ua < (ua + (entries << PAGE_SHIFT))) &&
> > > + (ua < (mem->ua + (mem->entries << PAGE_SHIFT))))
> > > + return -EINVAL;
> > > + }
> > > +
> > > + mem = kzalloc(sizeof(*mem), GFP_KERNEL);
> > > + if (!mem)
> > > + return -ENOMEM;
> > > +
> > > + mem->hpas = vzalloc(entries * sizeof(mem->hpas[0]));
> > > + if (!mem->hpas) {
> > > + kfree(mem);
> > > + return -ENOMEM;
> > > + }
> >
> > So, I've thought more about this and I'm really confused as to what
> > this is supposed to be accomplishing.
> >
> > I see that you need to keep track of what regions are registered, so
> > you don't double lock or unlock, but I don't see what the point of
> > actualy storing the translations in hpas is.
> >
> > I had assumed it was so that you could later on get to the
> > translations in real mode when you do in-kernel acceleration. But
> > that doesn't make sense, because the array is vmalloc()ed, so can't be
> > accessed in real mode anyway.
>
> We can access vmalloc'd arrays in real mode using real_vmalloc_addr().
Ah, ok.
> > I can't think of a circumstance in which you can use hpas where you
> > couldn't just walk the page tables anyway.
>
> The problem with walking the page tables is that there is no guarantee
> that the page you find that way is the page that was returned by the
> gup_fast() we did earlier. Storing the hpas means that we know for
> sure that the page we're doing DMA to is one that we have an elevated
> page count on.
>
> Also, there are various points where a Linux PTE is made temporarily
> invalid for a short time. If we happened to do a H_PUT_TCE on one cpu
> while another cpu was doing that, we'd get a spurious failure returned
> by the H_PUT_TCE.
I think we want this explanation in the commit message. Anr/or in a
comment somewhere, I'm not sure.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]
next prev parent reply other threads:[~2015-05-01 3:55 UTC|newest]
Thread overview: 110+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-04-25 12:14 [PATCH kernel v9 00/32] powerpc/iommu/vfio: Enable Dynamic DMA windows Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 01/32] powerpc/iommu: Split iommu_free_table into 2 helpers Alexey Kardashevskiy
2015-04-29 2:03 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 02/32] Revert "powerpc/powernv: Allocate struct pnv_ioda_pe iommu_table dynamically" Alexey Kardashevskiy
2015-04-27 21:05 ` Alex Williamson
2015-04-29 2:05 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 03/32] vfio: powerpc/spapr: Move page pinning from arch code to VFIO IOMMU driver Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 04/32] vfio: powerpc/spapr: Check that IOMMU page is fully contained by system page Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 05/32] vfio: powerpc/spapr: Use it_page_size Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 06/32] vfio: powerpc/spapr: Move locked_vm accounting to helpers Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 07/32] vfio: powerpc/spapr: Disable DMA mappings on disabled container Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 08/32] vfio: powerpc/spapr: Moving pinning/unpinning to helpers Alexey Kardashevskiy
2015-04-29 2:14 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 09/32] vfio: powerpc/spapr: Rework groups attaching Alexey Kardashevskiy
2015-04-29 2:16 ` David Gibson
2015-04-30 2:29 ` Alexey Kardashevskiy
2015-04-30 4:05 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 10/32] powerpc/powernv: Do not set "read" flag if direction==DMA_NONE Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 11/32] powerpc/iommu: Move tce_xxx callbacks from ppc_md to iommu_table Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 12/32] powerpc/spapr: vfio: Switch from iommu_table to new iommu_table_group Alexey Kardashevskiy
2015-04-29 2:49 ` David Gibson
2015-04-30 2:30 ` Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 13/32] vfio: powerpc/spapr/iommu/powernv/ioda2: Rework IOMMU ownership control Alexey Kardashevskiy
2015-04-29 3:02 ` David Gibson
2015-04-29 9:19 ` Alexey Kardashevskiy
2015-04-30 4:08 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 14/32] powerpc/iommu: Fix IOMMU ownership control functions Alexey Kardashevskiy
2015-04-29 3:08 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 15/32] powerpc/powernv/ioda/ioda2: Rework TCE invalidation in tce_build()/tce_free() Alexey Kardashevskiy
2015-04-29 3:18 ` David Gibson
2015-04-30 2:58 ` Alexey Kardashevskiy
2015-04-30 4:16 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 16/32] powerpc/powernv/ioda: Move TCE kill register address to PE Alexey Kardashevskiy
2015-04-27 21:05 ` Alex Williamson
2015-04-29 3:25 ` David Gibson
2015-04-29 9:00 ` Alexey Kardashevskiy
2015-04-30 4:18 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 17/32] powerpc/powernv: Implement accessor to TCE entry Alexey Kardashevskiy
2015-04-29 4:04 ` David Gibson
2015-04-29 9:02 ` Alexey Kardashevskiy
2015-04-30 0:13 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 18/32] powerpc/iommu/powernv: Release replaced TCE Alexey Kardashevskiy
2015-04-29 4:18 ` David Gibson
2015-04-29 9:51 ` Alexey Kardashevskiy
2015-04-30 4:21 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 19/32] powerpc/powernv/ioda2: Rework iommu_table creation Alexey Kardashevskiy
2015-04-29 4:27 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 20/32] powerpc/powernv/ioda2: Introduce pnv_pci_create_table/pnv_pci_free_table Alexey Kardashevskiy
2015-04-29 4:39 ` David Gibson
2015-04-29 9:12 ` Alexey Kardashevskiy
2015-04-30 4:24 ` David Gibson
2015-05-01 10:13 ` Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 21/32] powerpc/powernv/ioda2: Introduce pnv_pci_ioda2_set_window Alexey Kardashevskiy
2015-04-29 4:45 ` David Gibson
2015-04-29 9:26 ` Alexey Kardashevskiy
2015-04-30 4:32 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 22/32] powerpc/powernv: Implement multilevel TCE tables Alexey Kardashevskiy
2015-04-29 5:04 ` David Gibson
2015-05-01 9:48 ` Alexey Kardashevskiy
2015-05-05 12:05 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 23/32] powerpc/powernv/ioda: Define and implement DMA table/window management callbacks Alexey Kardashevskiy
2015-04-29 5:30 ` David Gibson
2015-04-29 9:44 ` Alexey Kardashevskiy
2015-04-30 4:37 ` David Gibson
2015-04-30 9:56 ` Alexey Kardashevskiy
2015-05-01 3:36 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 24/32] powerpc/powernv/ioda2: Use new helpers to do proper cleanup on PE release Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 25/32] vfio: powerpc/spapr: powerpc/powernv/ioda2: Rework ownership Alexey Kardashevskiy
2015-04-29 5:39 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 26/32] powerpc/iommu: Add userspace view of TCE table Alexey Kardashevskiy
2015-04-29 6:31 ` David Gibson
2015-05-01 4:01 ` Alexey Kardashevskiy
2015-05-01 4:23 ` David Gibson
2015-05-01 7:12 ` Alexey Kardashevskiy
2015-05-05 12:02 ` David Gibson
2015-05-11 2:11 ` Alexey Kardashevskiy
2015-05-11 4:52 ` Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 27/32] powerpc/iommu/ioda2: Add get_table_size() to calculate the size of future table Alexey Kardashevskiy
2015-04-29 6:40 ` David Gibson
2015-05-01 4:10 ` Alexey Kardashevskiy
2015-05-01 5:12 ` David Gibson
2015-05-01 6:53 ` Alexey Kardashevskiy
2015-05-05 11:58 ` David Gibson
2015-05-11 2:24 ` Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 28/32] powerpc/mmu: Add userspace-to-physical addresses translation cache Alexey Kardashevskiy
2015-04-29 7:01 ` David Gibson
2015-05-01 11:26 ` Alexey Kardashevskiy
2015-05-05 12:12 ` David Gibson
2015-04-30 6:34 ` David Gibson
2015-04-30 8:25 ` Paul Mackerras
2015-05-01 3:39 ` David Gibson [this message]
2015-04-25 12:14 ` [PATCH kernel v9 29/32] vfio: powerpc/spapr: Register memory and define IOMMU v2 Alexey Kardashevskiy
2015-04-30 6:55 ` David Gibson
2015-05-01 4:35 ` Alexey Kardashevskiy
2015-05-01 5:23 ` David Gibson
2015-05-01 6:27 ` Alexey Kardashevskiy
2015-05-05 11:53 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 30/32] vfio: powerpc/spapr: Use 32bit DMA window properties from table_group Alexey Kardashevskiy
2015-04-27 22:18 ` Alex Williamson
2015-04-30 6:58 ` David Gibson
2015-04-25 12:14 ` [PATCH kernel v9 31/32] vfio: powerpc/spapr: Support multiple groups in one container if possible Alexey Kardashevskiy
2015-04-30 7:22 ` David Gibson
2015-04-30 9:33 ` Alexey Kardashevskiy
2015-05-01 0:46 ` Benjamin Herrenschmidt
2015-05-01 4:44 ` David Gibson
2015-05-01 4:33 ` David Gibson
2015-05-01 6:05 ` Alexey Kardashevskiy
2015-05-05 11:50 ` David Gibson
2015-05-11 2:26 ` Alexey Kardashevskiy
2015-04-25 12:14 ` [PATCH kernel v9 32/32] vfio: powerpc/spapr: Support Dynamic DMA windows Alexey Kardashevskiy
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20150501033953.GJ24886@voom.redhat.com \
--to=david@gibson.dropbear.id.au \
--cc=aik@ozlabs.ru \
--cc=alex.williamson@redhat.com \
--cc=gwshan@linux.vnet.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=paulus@samba.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).