qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Christian Borntraeger <borntraeger@de.ibm.com>
To: Cornelia Huck <cornelia.huck@de.ibm.com>, qemu-devel@nongnu.org
Cc: jfrei@linux.vnet.ibm.com, agraf@suse.de,
	Thomas Huth <thuth@linux.vnet.ibm.com>
Subject: Re: [Qemu-devel] [PATCH 09/10] s390x/mmu: Use ioctl for reading and writing from/to guest memory
Date: Mon, 20 Apr 2015 16:55:34 +0200	[thread overview]
Message-ID: <55351366.9000600@de.ibm.com> (raw)
In-Reply-To: <1428569498-27393-10-git-send-email-cornelia.huck@de.ibm.com>

Am 09.04.2015 um 10:51 schrieb Cornelia Huck:
> From: Thomas Huth <thuth@linux.vnet.ibm.com>
> 
> Add code to make use of the new ioctl for reading from / writing to
> virtual guest memory. By using the ioctl, the memory accesses are now
> protected with the so-called ipte-lock in the kernel.
> 
> Signed-off-by: Thomas Huth <thuth@linux.vnet.ibm.com>
> Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>

Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>

> ---
>  target-s390x/cpu.h        |  7 +++++++
>  target-s390x/kvm.c        | 35 +++++++++++++++++++++++++++++++++++
>  target-s390x/mmu_helper.c |  9 +++++++++
>  3 files changed, 51 insertions(+)
> 
> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
> index 79bc80b..9c42743 100644
> --- a/target-s390x/cpu.h
> +++ b/target-s390x/cpu.h
> @@ -401,6 +401,8 @@ void kvm_s390_vcpu_interrupt(S390CPU *cpu, struct kvm_s390_irq *irq);
>  void kvm_s390_floating_interrupt(struct kvm_s390_irq *irq);
>  int kvm_s390_inject_flic(struct kvm_s390_irq *irq);
>  void kvm_s390_access_exception(S390CPU *cpu, uint16_t code, uint64_t te_code);
> +int kvm_s390_mem_op(S390CPU *cpu, vaddr addr, void *hostbuf, int len,
> +                    bool is_write);
>  int kvm_s390_get_clock(uint8_t *tod_high, uint64_t *tod_clock);
>  int kvm_s390_set_clock(uint8_t *tod_high, uint64_t *tod_clock);
>  #else
> @@ -418,6 +420,11 @@ static inline int kvm_s390_set_clock(uint8_t *tod_high, uint64_t *tod_low)
>  {
>      return -ENOSYS;
>  }
> +static inline int kvm_s390_mem_op(S390CPU *cpu, vaddr addr, void *hostbuf,
> +                                  int len, bool is_write)
> +{
> +    return -ENOSYS;
> +}
>  static inline void kvm_s390_access_exception(S390CPU *cpu, uint16_t code,
>                                               uint64_t te_code)
>  {
> diff --git a/target-s390x/kvm.c b/target-s390x/kvm.c
> index 619684b..e2ae2b3 100644
> --- a/target-s390x/kvm.c
> +++ b/target-s390x/kvm.c
> @@ -123,6 +123,7 @@ const KVMCapabilityInfo kvm_arch_required_capabilities[] = {
> 
>  static int cap_sync_regs;
>  static int cap_async_pf;
> +static int cap_mem_op;
> 
>  static void *legacy_s390_alloc(size_t size, uint64_t *align);
> 
> @@ -247,6 +248,7 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
>  {
>      cap_sync_regs = kvm_check_extension(s, KVM_CAP_SYNC_REGS);
>      cap_async_pf = kvm_check_extension(s, KVM_CAP_ASYNC_PF);
> +    cap_mem_op = kvm_check_extension(s, KVM_CAP_S390_MEM_OP);
> 
>      kvm_s390_enable_cmma(s);
> 
> @@ -550,6 +552,39 @@ int kvm_s390_set_clock(uint8_t *tod_high, uint64_t *tod_low)
>      return kvm_vm_ioctl(kvm_state, KVM_SET_DEVICE_ATTR, &attr);
>  }
> 
> +/**
> + * kvm_s390_mem_op:
> + * @addr:      the logical start address in guest memory
> + * @hostbuf:   buffer in host memory. NULL = do only checks w/o copying
> + * @len:       length that should be transfered
> + * @is_write:  true = write, false = read
> + * Returns:    0 on success, non-zero if an exception or error occured
> + *
> + * Use KVM ioctl to read/write from/to guest memory. An access exception
> + * is injected into the vCPU in case of translation errors.
> + */
> +int kvm_s390_mem_op(S390CPU *cpu, vaddr addr, void *hostbuf, int len,
> +                    bool is_write)
> +{
> +    struct kvm_s390_mem_op mem_op = {
> +        .gaddr = addr,
> +        .flags = KVM_S390_MEMOP_F_INJECT_EXCEPTION,
> +        .size = len,
> +        .op = is_write ? KVM_S390_MEMOP_LOGICAL_WRITE
> +                       : KVM_S390_MEMOP_LOGICAL_READ,
> +        .buf = (uint64_t)hostbuf,
> +    };
> +
> +    if (!cap_mem_op) {
> +        return -ENOSYS;
> +    }
> +    if (!hostbuf) {
> +        mem_op.flags |= KVM_S390_MEMOP_F_CHECK_ONLY;
> +    }
> +
> +    return kvm_vcpu_ioctl(CPU(cpu), KVM_S390_MEM_OP, &mem_op);
> +}
> +
>  /*
>   * Legacy layout for s390:
>   * Older S390 KVM requires the topmost vma of the RAM to be
> diff --git a/target-s390x/mmu_helper.c b/target-s390x/mmu_helper.c
> index 9b88498..1e7412c 100644
> --- a/target-s390x/mmu_helper.c
> +++ b/target-s390x/mmu_helper.c
> @@ -450,6 +450,15 @@ int s390_cpu_virt_mem_rw(S390CPU *cpu, vaddr laddr, void *hostbuf,
>      target_ulong *pages;
>      int ret;
> 
> +    if (kvm_enabled()) {
> +        ret = kvm_s390_mem_op(cpu, laddr, hostbuf, len, is_write);
> +        if (ret >= 0) {
> +            return ret;
> +        } else if (ret != -ENOSYS) {
> +            error_printf("kvm_s390_mem_op() failed: %s\n", strerror(-ret));
> +        }
> +    }
> +
>      nr_pages = (((laddr & ~TARGET_PAGE_MASK) + len - 1) >> TARGET_PAGE_BITS)
>                 + 1;
>      pages = g_malloc(nr_pages * sizeof(*pages));
> 

  reply	other threads:[~2015-04-20 14:55 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-04-09  8:51 [Qemu-devel] [PATCH 00/10] first pile of s390x patches for 2.4 Cornelia Huck
2015-04-09  8:51 ` [Qemu-devel] [PATCH 01/10] s390x/kvm: Remove message about "switch_amode" kernel parameter Cornelia Huck
2015-04-23 12:26   ` Christian Borntraeger
2015-04-23 12:33     ` Cornelia Huck
2015-04-09  8:51 ` [Qemu-devel] [PATCH 02/10] virtio-ccw: sort into categories Cornelia Huck
2015-04-20 14:47   ` Christian Borntraeger
2015-04-09  8:51 ` [Qemu-devel] [PATCH 03/10] s390-virtio: " Cornelia Huck
2015-04-20 14:47   ` Christian Borntraeger
2015-04-09  8:51 ` [Qemu-devel] [PATCH 04/10] sclp: " Cornelia Huck
2015-04-20 14:48   ` Christian Borntraeger
2015-04-09  8:51 ` [Qemu-devel] [PATCH 05/10] s390x/ipl: " Cornelia Huck
2015-04-20 14:48   ` Christian Borntraeger
2015-04-09  8:51 ` [Qemu-devel] [PATCH 06/10] s390x/mmu: Use access type definitions instead of magic values Cornelia Huck
2015-04-09  8:51 ` [Qemu-devel] [PATCH 07/10] linux-headers: update Cornelia Huck
2015-04-24  9:30   ` Paolo Bonzini
2015-04-24 10:23     ` Cornelia Huck
2015-04-09  8:51 ` [Qemu-devel] [PATCH 08/10] s390x/kvm: Put vm name, extended name and UUID into STSI322 SYSIB Cornelia Huck
2015-04-09  8:51 ` [Qemu-devel] [PATCH 09/10] s390x/mmu: Use ioctl for reading and writing from/to guest memory Cornelia Huck
2015-04-20 14:55   ` Christian Borntraeger [this message]
2015-04-21 19:12     ` Alexander Graf
2015-04-21 19:34       ` Christian Borntraeger
2015-04-21 19:36         ` Alexander Graf
2015-04-21 19:48           ` Christian Borntraeger
2015-04-22  8:37             ` Cornelia Huck
2015-04-09  8:51 ` [Qemu-devel] [PATCH 10/10] s390x/kvm: Support access register mode for KVM_S390_MEM_OP ioctl Cornelia Huck

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=55351366.9000600@de.ibm.com \
    --to=borntraeger@de.ibm.com \
    --cc=agraf@suse.de \
    --cc=cornelia.huck@de.ibm.com \
    --cc=jfrei@linux.vnet.ibm.com \
    --cc=qemu-devel@nongnu.org \
    --cc=thuth@linux.vnet.ibm.com \
    /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).