At the moment the real mode handler of H_PUT_TCE calls iommu_tce_xchg_rm()
which in turn reads the old TCE and if it was a valid entry, marks
the physical page dirty if it was mapped for writing. Since it is in
real mode, realmode_pfn_to_page() is used instead of pfn_to_page()
to get the page struct. However SetPageDirty() itself reads the compound
page head and returns a virtual address for the head page struct and
setting dirty bit for that kills the system.
This adds additional dirty bit tracking into the MM/IOMMU API for use
in the real mode. Note that this does not change how VFIO and
KVM (in virtual mode) set this bit. The KVM (real mode) changes include:
- use the lowest bit of the cached host phys address to carry
the dirty bit;
- mark pages dirty when they are unpinned which happens when
the preregistered memory is released which always happens in virtual
mode;
- add mm_iommu_ua_mark_dirty_rm() helper to set delayed dirty bit;
- change iommu_tce_xchg_rm() to take the kvm struct for the mm to use
in the new mm_iommu_ua_mark_dirty_rm() helper;
- move iommu_tce_xchg_rm() to book3s_64_vio_hv.c (which is the only
caller anyway) to reduce the real mode KVM and IOMMU knowledge
across different subsystems.
This removes realmode_pfn_to_page() as it is not used anymore.
While we at it, remove some EXPORT_SYMBOL_GPL() as that code is for
the real mode only and modules cannot call it anyway.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Commit 76fa4975f3 ("KVM: PPC: Check if IOMMU page is contained in
the pinned physical page", 2018-07-17) added some checks to ensure
that guest DMA mappings don't attempt to map more than the guest is
entitled to access. However, errors in the logic mean that legitimate
guest requests to map pages for DMA are being denied in some
situations. Specifically, if the first page of the range passed to
mm_iommu_get() is mapped with a normal page, and subsequent pages are
mapped with transparent huge pages, we end up with mem->pageshift ==
0. That means that the page size checks in mm_iommu_ua_to_hpa() and
mm_iommu_up_to_hpa_rm() will always fail for every page in that
region, and thus the guest can never map any memory in that region for
DMA, typically leading to a flood of error messages like this:
qemu-system-ppc64: VFIO_MAP_DMA: -22
qemu-system-ppc64: vfio_dma_map(0x10005f47780, 0x800000000000000, 0x10000, 0x7fff63ff0000) = -22 (Invalid argument)
The logic errors in mm_iommu_get() are:
(a) use of 'ua' not 'ua + (i << PAGE_SHIFT)' in the find_linux_pte()
call (meaning that find_linux_pte() returns the pte for the
first address in the range, not the address we are currently up
to);
(b) use of 'pageshift' as the variable to receive the hugepage shift
returned by find_linux_pte() - for a normal page this gets set
to 0, leading to us setting mem->pageshift to 0 when we conclude
that the pte returned by find_linux_pte() didn't match the page
we were looking at;
(c) comparing 'compshift', which is a page order, i.e. log base 2 of
the number of pages, with 'pageshift', which is a log base 2 of
the number of bytes.
To fix these problems, this patch introduces 'cur_ua' to hold the
current user address and uses that in the find_linux_pte() call;
introduces 'pteshift' to hold the hugepage shift found by
find_linux_pte(); and compares 'pteshift' with 'compshift +
PAGE_SHIFT' rather than 'compshift'.
The patch also moves the local_irq_restore to the point after the PTE
pointer returned by find_linux_pte() has been dereferenced because
otherwise the PTE could change underneath us, and adds a check to
avoid doing the find_linux_pte() call once mem->pageshift has been
reduced to PAGE_SHIFT, as an optimization.
Fixes: 76fa4975f3 ("KVM: PPC: Check if IOMMU page is contained in the pinned physical page")
Cc: stable@vger.kernel.org # v4.12+
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
A VM which has:
- a DMA capable device passed through to it (eg. network card);
- running a malicious kernel that ignores H_PUT_TCE failure;
- capability of using IOMMU pages bigger that physical pages
can create an IOMMU mapping that exposes (for example) 16MB of
the host physical memory to the device when only 64K was allocated to the VM.
The remaining 16MB - 64K will be some other content of host memory, possibly
including pages of the VM, but also pages of host kernel memory, host
programs or other VMs.
The attacking VM does not control the location of the page it can map,
and is only allowed to map as many pages as it has pages of RAM.
We already have a check in drivers/vfio/vfio_iommu_spapr_tce.c that
an IOMMU page is contained in the physical page so the PCI hardware won't
get access to unassigned host memory; however this check is missing in
the KVM fastpath (H_PUT_TCE accelerated code). We were lucky so far and
did not hit this yet as the very first time when the mapping happens
we do not have tbl::it_userspace allocated yet and fall back to
the userspace which in turn calls VFIO IOMMU driver, this fails and
the guest does not retry,
This stores the smallest preregistered page size in the preregistered
region descriptor and changes the mm_iommu_xxx API to check this against
the IOMMU page size.
This calculates maximum page size as a minimum of the natural region
alignment and compound page size. For the page shift this uses the shift
returned by find_linux_pte() which indicates how the page is mapped to
the current userspace - if the page is huge and this is not a zero, then
it is a leaf pte and the page is mapped within the range.
Fixes: 121f80ba68 ("KVM: PPC: VFIO: Add in-kernel acceleration for VFIO")
Cc: stable@vger.kernel.org # v4.12+
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
No allocation callback is using this argument anymore. new_page_node
used to use this parameter to convey node_id resp. migration error up
to move_pages code (do_move_page_to_node_array). The error status never
made it into the final status field and we have a better way to
communicate node id to the status field now. All other allocation
callbacks simply ignored the argument so we can drop it finally.
[mhocko@suse.com: fix migration callback]
Link: http://lkml.kernel.org/r/20180105085259.GH2801@dhcp22.suse.cz
[akpm@linux-foundation.org: fix alloc_misplaced_dst_page()]
[mhocko@kernel.org: fix build]
Link: http://lkml.kernel.org/r/20180103091134.GB11319@dhcp22.suse.cz
Link: http://lkml.kernel.org/r/20180103082555.14592-3-mhocko@kernel.org
Signed-off-by: Michal Hocko <mhocko@suse.com>
Reviewed-by: Zi Yan <zi.yan@cs.rutgers.edu>
Cc: Andrea Reale <ar@linux.vnet.ibm.com>
Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
alloc_contig_range() initiates compaction and eventual migration for the
purpose of either CMA or HugeTLB allocations. At present, the reason
code remains the same MR_CMA for either of these cases. Let's make it
MR_CONTIG_RANGE which will appropriately reflect the reason code in both
these cases.
Link: http://lkml.kernel.org/r/20180202091518.18798-1-khandual@linux.vnet.ibm.com
Signed-off-by: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The CMA pages migration code does not support compound pages at
the moment so it performs few tests before proceeding to actual page
migration.
One of the tests - PageTransHuge() - has VM_BUG_ON_PAGE(PageTail()) as
it is designed to be called on head pages only. Since we also test for
PageCompound(), and it contains PageTail() and PageHead(), we can
simplify the check by leaving just PageCompound() and therefore avoid
possible VM_BUG_ON_PAGE.
Fixes: 2e5bbb5461 ("KVM: PPC: Book3S HV: Migrate pinned pages out of CMA")
Cc: stable@vger.kernel.org # v4.9+
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Acked-by: Balbir Singh <bsingharora@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This makes mm_iommu_lookup() able to work in realmode by replacing
list_for_each_entry_rcu() (which can do debug stuff which can fail in
real mode) with list_for_each_entry_lockless().
This adds realmode version of mm_iommu_ua_to_hpa() which adds
explicit vmalloc'd-to-linear address conversion.
Unlike mm_iommu_ua_to_hpa(), mm_iommu_ua_to_hpa_rm() can fail.
This changes mm_iommu_preregistered() to receive @mm as in real mode
@current does not always have a correct pointer.
This adds realmode version of mm_iommu_lookup() which receives @mm
(for the same reason as for mm_iommu_preregistered()) and uses
lockless version of list_for_each_entry_rcu().
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
We are going to split <linux/sched/signal.h> out of <linux/sched.h>, which
will have to be picked up from other headers and a couple of .c files.
Create a trivial placeholder <linux/sched/signal.h> file that just
maps to <linux/sched.h> to make this patch obviously correct and
bisectable.
Include the new header in the files that are going to need it.
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Currently the build breaks if CMA=n and SPAPR_TCE_IOMMU=y:
arch/powerpc/mm/mmu_context_iommu.c: In function ‘mm_iommu_get’:
arch/powerpc/mm/mmu_context_iommu.c:193:42: error: ‘MIGRATE_CMA’ undeclared (first use in this function)
if (get_pageblock_migratetype(page) == MIGRATE_CMA) {
^~~~~~~~~~~
Fix it by using the existing is_migrate_cma_page(), which evaulates to
false when CMA=n.
Fixes: 2e5bbb5461 ("KVM: PPC: Book3S HV: Migrate pinned pages out of CMA")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
At the moment the userspace tool is expected to request pinning of
the entire guest RAM when VFIO IOMMU SPAPR v2 driver is present.
When the userspace process finishes, all the pinned pages need to
be put; this is done as a part of the userspace memory context (MM)
destruction which happens on the very last mmdrop().
This approach has a problem that a MM of the userspace process
may live longer than the userspace process itself as kernel threads
use userspace process MMs which was runnning on a CPU where
the kernel thread was scheduled to. If this happened, the MM remains
referenced until this exact kernel thread wakes up again
and releases the very last reference to the MM, on an idle system this
can take even hours.
This moves preregistered regions tracking from MM to VFIO; insteads of
using mm_iommu_table_group_mem_t::used, tce_container::prereg_list is
added so each container releases regions which it has pre-registered.
This changes the userspace interface to return EBUSY if a memory
region is already registered in a container. However it should not
have any practical effect as the only userspace tool available now
does register memory region once per container anyway.
As tce_iommu_register_pages/tce_iommu_unregister_pages are called
under container->lock, this does not need additional locking.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This changes mm_iommu_xxx helpers to take mm_struct as a parameter
instead of getting it from @current which in some situations may
not have a valid reference to mm.
This changes helpers to receive @mm and moves all references to @current
to the caller, including checks for !current and !current->mm;
checks in mm_iommu_preregistered() are removed as there is no caller
yet.
This moves the mm_iommu_adjust_locked_vm() call to the caller as
it receives mm_iommu_table_group_mem_t but it needs mm.
This should cause no behavioral change.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
We are going to get rid of @current references in mmu_context_boos3s64.c
and cache mm_struct in the VFIO container. Since mm_context_t does not
have reference counting, we will be using mm_struct which does have
the reference counter.
This changes mm_iommu_init/mm_iommu_cleanup to receive mm_struct rather
than mm_context_t (which is embedded into mm).
This should not cause any behavioral change.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When PCI Device pass-through is enabled via VFIO, KVM-PPC will
pin pages using get_user_pages_fast(). One of the downsides of
the pinning is that the page could be in CMA region. The CMA
region is used for other allocations like the hash page table.
Ideally we want the pinned pages to be from non CMA region.
This patch (currently only for KVM PPC with VFIO) forcefully
migrates the pages out (huge pages are omitted for the moment).
There are more efficient ways of doing this, but that might
be elaborate and might impact a larger audience beyond just
the kvm ppc implementation.
The magic is in new_iommu_non_cma_page() which allocates the
new page from a non CMA region.
I've tested the patches lightly at my end. The full solution
requires migration of THP pages in the CMA region. That work
will be done incrementally on top of this.
Signed-off-by: Balbir Singh <bsingharora@gmail.com>
Acked-by: Alexey Kardashevskiy <aik@ozlabs.ru>
[mpe: Merged via powerpc tree as that's where the changes are]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
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.
Another use of it is in-kernel real mode (mmu off) acceleration of
DMA requests where real time translation of guest physical to host
physical addresses is non-trivial and may fail as linux ptes may be
temporarily invalid. Also, having cached host physical addresses
(compared to just pinning at the start and then walking the page table
again on every H_PUT_TCE), we can be sure that the addresses which we put
into TCE table are the ones we already pinned.
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 usage counters; multiple registration of the same memory
is allowed (once per container).
This implements 2 counters per registered memory region:
- @mapped: incremented on every DMA mapping; decremented on unmapping;
initialized to 1 when a region is just registered; once it becomes zero,
no more mappings allowe;
- @used: incremented on every "register" ioctl; decremented on
"unregister"; unregistration is allowed for DMA mapped regions unless
it is the very last reference. For the very last reference this checks
that the region is still mapped and returns -EBUSY so the userspace
gets to know that memory is still pinned and unregistration needs to
be retried; @used remains 1.
Host physical addresses are stored in vmalloc'ed array. In order to
access these in the real mode (mmu off), there is a real_vmalloc_addr()
helper. In-kernel acceleration patchset will move it from KVM to MMU code.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Alexey Kardashevskiy