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: <CANgfPd915=4=dYmxh=qn6Cu75S4uHL753h_x3JHG-mFyO0dggg@mail.gmail.com>
Date:   Thu, 13 May 2021 15:33:57 -0700
From:   Ben Gardon <bgardon@...gle.com>
To:     Axel Rasmussen <axelrasmussen@...gle.com>
Cc:     Aaron Lewis <aaronlewis@...gle.com>,
        Alexander Graf <graf@...zon.com>,
        Andrew Jones <drjones@...hat.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Emanuele Giuseppe Esposito <eesposit@...hat.com>,
        Eric Auger <eric.auger@...hat.com>,
        Jacob Xu <jacobhxu@...gle.com>,
        Makarand Sonare <makarandsonare@...gle.com>,
        Oliver Upton <oupton@...gle.com>,
        Paolo Bonzini <pbonzini@...hat.com>,
        Peter Xu <peterx@...hat.com>, Shuah Khan <shuah@...nel.org>,
        Yanan Wang <wangyanan55@...wei.com>, kvm <kvm@...r.kernel.org>,
        LKML <linux-kernel@...r.kernel.org>,
        linux-kselftest@...r.kernel.org
Subject: Re: [PATCH 3/5] KVM: selftests: create alias mappings when using
 shared memory

On Wed, May 12, 2021 at 2:45 PM Axel Rasmussen <axelrasmussen@...gle.com> wrote:
>
> When a memory region is added with a src_type specifying that it should
> use some kind of shared memory, also create an alias mapping to the same
> underlying physical pages.
>
> And, add an API so tests can get access to these alias addresses.
> Basically, for a guest physical address, let us look up the analogous
> host *alias* address.
>
> In a future commit, we'll modify the demand paging test to take
> advantage of this to exercise UFFD minor faults. The idea is, we
> pre-fault the underlying pages *via the alias*. When the *guest*
> faults, it gets a "minor" fault (PTEs don't exist yet, but a page is
> already in the page cache). Then, the userfaultfd theads can handle the
> fault: they could potentially modify the underlying memory *via the
> alias* if they wanted to, and then they install the PTEs and let the
> guest carry on via a UFFDIO_CONTINUE ioctl.
>
> Signed-off-by: Axel Rasmussen <axelrasmussen@...gle.com>

Reviewed-by: Ben Gardon <bgardon@...gle.com>

> ---
>  .../testing/selftests/kvm/include/kvm_util.h  |  1 +
>  tools/testing/selftests/kvm/lib/kvm_util.c    | 50 +++++++++++++++++++
>  .../selftests/kvm/lib/kvm_util_internal.h     |  2 +
>  3 files changed, 53 insertions(+)
>
> diff --git a/tools/testing/selftests/kvm/include/kvm_util.h b/tools/testing/selftests/kvm/include/kvm_util.h
> index a8f022794ce3..0624f25a6803 100644
> --- a/tools/testing/selftests/kvm/include/kvm_util.h
> +++ b/tools/testing/selftests/kvm/include/kvm_util.h
> @@ -146,6 +146,7 @@ void virt_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr,
>  void *addr_gpa2hva(struct kvm_vm *vm, vm_paddr_t gpa);
>  void *addr_gva2hva(struct kvm_vm *vm, vm_vaddr_t gva);
>  vm_paddr_t addr_hva2gpa(struct kvm_vm *vm, void *hva);
> +void *addr_gpa2alias(struct kvm_vm *vm, vm_paddr_t gpa);
>
>  /*
>   * Address Guest Virtual to Guest Physical
> diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
> index 6fbe124e0e16..838d58633f7e 100644
> --- a/tools/testing/selftests/kvm/lib/kvm_util.c
> +++ b/tools/testing/selftests/kvm/lib/kvm_util.c
> @@ -809,6 +809,19 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm,
>
>         /* Add to linked-list of memory regions. */
>         list_add(&region->list, &vm->userspace_mem_regions);
> +
> +       /* If shared memory, create an alias. */
> +       if (region->fd >= 0) {
> +               region->mmap_alias = mmap(NULL, region->mmap_size,
> +                                         PROT_READ | PROT_WRITE,
> +                                         vm_mem_backing_src_alias(src_type)->flag,
> +                                         region->fd, 0);
> +               TEST_ASSERT(region->mmap_alias != MAP_FAILED,
> +                           "mmap of alias failed, errno: %i", errno);
> +
> +               /* Align host alias address */
> +               region->host_alias = align(region->mmap_alias, alignment);
> +       }
>  }
>
>  /*
> @@ -1237,6 +1250,43 @@ vm_paddr_t addr_hva2gpa(struct kvm_vm *vm, void *hva)
>         return -1;
>  }
>
> +/*
> + * Address VM physical to Host Virtual *alias*.
> + *
> + * Input Args:
> + *   vm - Virtual Machine
> + *   gpa - VM physical address
> + *
> + * Output Args: None
> + *
> + * Return:
> + *   Equivalent address within the host virtual *alias* area, or NULL
> + *   (without failing the test) if the guest memory is not shared (so
> + *   no alias exists).
> + *
> + * When vm_create() and related functions are called with a shared memory
> + * src_type, we also create a writable, shared alias mapping of the
> + * underlying guest memory. This allows the host to manipulate guest memory,
> + * e.g. to implement demand paging.

I would amend this to: "This allows the host to manipulate guest
memory without mapping that memory in the guest's address space," or
something to that effect.

> + */
> +void *addr_gpa2alias(struct kvm_vm *vm, vm_paddr_t gpa)
> +{
> +       struct userspace_mem_region *region;
> +
> +       list_for_each_entry(region, &vm->userspace_mem_regions, list) {
> +               if (!region->host_alias)
> +                       continue;
> +
> +               if ((gpa >= region->region.guest_phys_addr)
> +                       && (gpa <= (region->region.guest_phys_addr
> +                               + region->region.memory_size - 1)))
> +                       return (void *) ((uintptr_t) region->host_alias
> +                               + (gpa - region->region.guest_phys_addr));
> +       }
> +
> +       return NULL;
> +}
> +
>  /*
>   * VM Create IRQ Chip
>   *
> diff --git a/tools/testing/selftests/kvm/lib/kvm_util_internal.h b/tools/testing/selftests/kvm/lib/kvm_util_internal.h
> index 91ce1b5d480b..a25af33d4a9c 100644
> --- a/tools/testing/selftests/kvm/lib/kvm_util_internal.h
> +++ b/tools/testing/selftests/kvm/lib/kvm_util_internal.h
> @@ -16,7 +16,9 @@ struct userspace_mem_region {
>         int fd;
>         off_t offset;
>         void *host_mem;
> +       void *host_alias;
>         void *mmap_start;
> +       void *mmap_alias;
>         size_t mmap_size;
>         struct list_head list;
>  };
> --
> 2.31.1.607.g51e8a6a459-goog
>

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ