| Message ID | 20200304025554.2159-1-jianjay.zhou@huawei.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | kvm: support to get/set dirty log initial-all-set capability | expand |
On Wed, Mar 04, 2020 at 10:55:54AM +0800, Jay Zhou wrote: > Since the new capability KVM_DIRTY_LOG_INITIALLY_SET of > KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 has been introduced in the > kernel, tweak the userspace side to detect and enable this > capability. > > Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com> > --- > accel/kvm/kvm-all.c | 21 ++++++++++++++------- > linux-headers/linux/kvm.h | 3 +++ > 2 files changed, 17 insertions(+), 7 deletions(-) > > diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c > index 439a4efe52..45ab25be63 100644 > --- a/accel/kvm/kvm-all.c > +++ b/accel/kvm/kvm-all.c > @@ -100,7 +100,7 @@ struct KVMState > bool kernel_irqchip_required; > OnOffAuto kernel_irqchip_split; > bool sync_mmu; > - bool manual_dirty_log_protect; > + uint64_t manual_dirty_log_protect; > /* The man page (and posix) say ioctl numbers are signed int, but > * they're not. Linux, glibc and *BSD all treat ioctl numbers as > * unsigned, and treating them as signed here can break things */ > @@ -1882,6 +1882,7 @@ static int kvm_init(MachineState *ms) > int ret; > int type = 0; > const char *kvm_type; > + uint64_t dirty_log_manual_caps; > > s = KVM_STATE(ms->accelerator); > > @@ -2007,14 +2008,20 @@ static int kvm_init(MachineState *ms) > s->coalesced_pio = s->coalesced_mmio && > kvm_check_extension(s, KVM_CAP_COALESCED_PIO); > > - s->manual_dirty_log_protect = > + dirty_log_manual_caps = > kvm_check_extension(s, KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2); > - if (s->manual_dirty_log_protect) { > - ret = kvm_vm_enable_cap(s, KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, 1); > + dirty_log_manual_caps &= (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE | > + KVM_DIRTY_LOG_INITIALLY_SET); > + s->manual_dirty_log_protect = dirty_log_manual_caps; > + if (dirty_log_manual_caps) { > + ret = kvm_vm_enable_cap(s, KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, > + dirty_log_manual_caps); > if (ret) { > - warn_report("Trying to enable KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 " > - "but failed. Falling back to the legacy mode. "); > - s->manual_dirty_log_protect = false; > + warn_report("Trying to enable capability %"PRIu64" of " > + "KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 but failed. " > + "Falling back to the legacy mode. ", > + dirty_log_manual_caps); > + s->manual_dirty_log_protect = 0; > } > } > > diff --git a/linux-headers/linux/kvm.h b/linux-headers/linux/kvm.h > index 265099100e..3cb71c2b19 100644 > --- a/linux-headers/linux/kvm.h > +++ b/linux-headers/linux/kvm.h > @@ -1628,4 +1628,7 @@ struct kvm_hyperv_eventfd { > #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff > #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) > > +#define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) > +#define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) > + The patch looks ok, though ideally I think we need to wait until the kernel patch got pushed then we'll be sure these macros won't be overwrite by other ./scripts/update-linux-headers.sh updates (or another patch to call the update script to fetch the macros...). Reviewed-by: Peter Xu <peterx@redhat.com> Thanks,
On Wed, 4 Mar 2020 10:55:54 +0800 Jay Zhou <jianjay.zhou@huawei.com> wrote: > Since the new capability KVM_DIRTY_LOG_INITIALLY_SET of > KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 has been introduced in the > kernel, tweak the userspace side to detect and enable this > capability. > > Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com> > --- > accel/kvm/kvm-all.c | 21 ++++++++++++++------- > linux-headers/linux/kvm.h | 3 +++ > 2 files changed, 17 insertions(+), 7 deletions(-) <standard message> Please do any linux-headers updates in a separate patch; that makes it easy to replace the manual update with a complete headers update. </standard message> :)
On 04/03/20 03:55, Jay Zhou wrote: > Since the new capability KVM_DIRTY_LOG_INITIALLY_SET of > KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 has been introduced in the > kernel, tweak the userspace side to detect and enable this > capability. > > Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com> > --- > accel/kvm/kvm-all.c | 21 ++++++++++++++------- > linux-headers/linux/kvm.h | 3 +++ > 2 files changed, 17 insertions(+), 7 deletions(-) > > diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c > index 439a4efe52..45ab25be63 100644 > --- a/accel/kvm/kvm-all.c > +++ b/accel/kvm/kvm-all.c > @@ -100,7 +100,7 @@ struct KVMState > bool kernel_irqchip_required; > OnOffAuto kernel_irqchip_split; > bool sync_mmu; > - bool manual_dirty_log_protect; > + uint64_t manual_dirty_log_protect; > /* The man page (and posix) say ioctl numbers are signed int, but > * they're not. Linux, glibc and *BSD all treat ioctl numbers as > * unsigned, and treating them as signed here can break things */ > @@ -1882,6 +1882,7 @@ static int kvm_init(MachineState *ms) > int ret; > int type = 0; > const char *kvm_type; > + uint64_t dirty_log_manual_caps; > > s = KVM_STATE(ms->accelerator); > > @@ -2007,14 +2008,20 @@ static int kvm_init(MachineState *ms) > s->coalesced_pio = s->coalesced_mmio && > kvm_check_extension(s, KVM_CAP_COALESCED_PIO); > > - s->manual_dirty_log_protect = > + dirty_log_manual_caps = > kvm_check_extension(s, KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2); > - if (s->manual_dirty_log_protect) { > - ret = kvm_vm_enable_cap(s, KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, 1); > + dirty_log_manual_caps &= (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE | > + KVM_DIRTY_LOG_INITIALLY_SET); > + s->manual_dirty_log_protect = dirty_log_manual_caps; > + if (dirty_log_manual_caps) { > + ret = kvm_vm_enable_cap(s, KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, > + dirty_log_manual_caps); > if (ret) { > - warn_report("Trying to enable KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 " > - "but failed. Falling back to the legacy mode. "); > - s->manual_dirty_log_protect = false; > + warn_report("Trying to enable capability %"PRIu64" of " > + "KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 but failed. " > + "Falling back to the legacy mode. ", > + dirty_log_manual_caps); > + s->manual_dirty_log_protect = 0; > } > } > > diff --git a/linux-headers/linux/kvm.h b/linux-headers/linux/kvm.h > index 265099100e..3cb71c2b19 100644 > --- a/linux-headers/linux/kvm.h > +++ b/linux-headers/linux/kvm.h > @@ -1628,4 +1628,7 @@ struct kvm_hyperv_eventfd { > #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff > #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) > > +#define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) > +#define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) > + > #endif /* __LINUX_KVM_H */ > Queued, thanks. Paolo
> -----Original Message----- > From: Paolo Bonzini [mailto:pbonzini@redhat.com] > Sent: Wednesday, March 18, 2020 6:48 PM > To: Zhoujian (jay) <jianjay.zhou@huawei.com>; qemu-devel@nongnu.org; > kvm@vger.kernel.org > Cc: mst@redhat.com; cohuck@redhat.com; peterx@redhat.com; wangxin (U) > <wangxinxin.wang@huawei.com>; Huangweidong (C) > <weidong.huang@huawei.com>; Liujinsong (Paul) <liu.jinsong@huawei.com> > Subject: Re: [PATCH] kvm: support to get/set dirty log initial-all-set capability > > On 04/03/20 03:55, Jay Zhou wrote: > > Since the new capability KVM_DIRTY_LOG_INITIALLY_SET of > > KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 has been introduced in the kernel, > > tweak the userspace side to detect and enable this capability. > > > > Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com> > > --- > > accel/kvm/kvm-all.c | 21 ++++++++++++++------- > > linux-headers/linux/kvm.h | 3 +++ > > 2 files changed, 17 insertions(+), 7 deletions(-) > > > > diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c index > > 439a4efe52..45ab25be63 100644 > > --- a/accel/kvm/kvm-all.c > > +++ b/accel/kvm/kvm-all.c > > @@ -100,7 +100,7 @@ struct KVMState > > bool kernel_irqchip_required; > > OnOffAuto kernel_irqchip_split; > > bool sync_mmu; > > - bool manual_dirty_log_protect; > > + uint64_t manual_dirty_log_protect; > > /* The man page (and posix) say ioctl numbers are signed int, but > > * they're not. Linux, glibc and *BSD all treat ioctl numbers as > > * unsigned, and treating them as signed here can break things */ > > @@ -1882,6 +1882,7 @@ static int kvm_init(MachineState *ms) > > int ret; > > int type = 0; > > const char *kvm_type; > > + uint64_t dirty_log_manual_caps; > > > > s = KVM_STATE(ms->accelerator); > > > > @@ -2007,14 +2008,20 @@ static int kvm_init(MachineState *ms) > > s->coalesced_pio = s->coalesced_mmio && > > kvm_check_extension(s, > KVM_CAP_COALESCED_PIO); > > > > - s->manual_dirty_log_protect = > > + dirty_log_manual_caps = > > kvm_check_extension(s, > KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2); > > - if (s->manual_dirty_log_protect) { > > - ret = kvm_vm_enable_cap(s, > KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, 1); > > + dirty_log_manual_caps &= > (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE | > > + KVM_DIRTY_LOG_INITIALLY_SET); > > + s->manual_dirty_log_protect = dirty_log_manual_caps; > > + if (dirty_log_manual_caps) { > > + ret = kvm_vm_enable_cap(s, > KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, > > + dirty_log_manual_caps); > > if (ret) { > > - warn_report("Trying to enable > KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 " > > - "but failed. Falling back to the legacy mode. "); > > - s->manual_dirty_log_protect = false; > > + warn_report("Trying to enable capability %"PRIu64" of " > > + "KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 > but failed. " > > + "Falling back to the legacy mode. ", > > + dirty_log_manual_caps); > > + s->manual_dirty_log_protect = 0; > > } > > } > > > > diff --git a/linux-headers/linux/kvm.h b/linux-headers/linux/kvm.h > > index 265099100e..3cb71c2b19 100644 > > --- a/linux-headers/linux/kvm.h > > +++ b/linux-headers/linux/kvm.h > > @@ -1628,4 +1628,7 @@ struct kvm_hyperv_eventfd { > > #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff > > #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) > > > > +#define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) > > +#define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) > > + > > #endif /* __LINUX_KVM_H */ > > > > Queued, thanks. > Hi Paolo, It seems that this patch isn't included in your last pull request... If there's something else to be done, please let me know. Regards, Jay Zhou
On 12/06/20 05:01, Zhoujian (jay) wrote: > > >> -----Original Message----- >> From: Paolo Bonzini [mailto:pbonzini@redhat.com] >> Sent: Wednesday, March 18, 2020 6:48 PM >> To: Zhoujian (jay) <jianjay.zhou@huawei.com>; qemu-devel@nongnu.org; >> kvm@vger.kernel.org >> Cc: mst@redhat.com; cohuck@redhat.com; peterx@redhat.com; wangxin (U) >> <wangxinxin.wang@huawei.com>; Huangweidong (C) >> <weidong.huang@huawei.com>; Liujinsong (Paul) <liu.jinsong@huawei.com> >> Subject: Re: [PATCH] kvm: support to get/set dirty log initial-all-set capability >> >> On 04/03/20 03:55, Jay Zhou wrote: >>> Since the new capability KVM_DIRTY_LOG_INITIALLY_SET of >>> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 has been introduced in the kernel, >>> tweak the userspace side to detect and enable this capability. >>> >>> Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com> >>> --- >>> accel/kvm/kvm-all.c | 21 ++++++++++++++------- >>> linux-headers/linux/kvm.h | 3 +++ >>> 2 files changed, 17 insertions(+), 7 deletions(-) >>> >>> diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c index >>> 439a4efe52..45ab25be63 100644 >>> --- a/accel/kvm/kvm-all.c >>> +++ b/accel/kvm/kvm-all.c >>> @@ -100,7 +100,7 @@ struct KVMState >>> bool kernel_irqchip_required; >>> OnOffAuto kernel_irqchip_split; >>> bool sync_mmu; >>> - bool manual_dirty_log_protect; >>> + uint64_t manual_dirty_log_protect; >>> /* The man page (and posix) say ioctl numbers are signed int, but >>> * they're not. Linux, glibc and *BSD all treat ioctl numbers as >>> * unsigned, and treating them as signed here can break things */ >>> @@ -1882,6 +1882,7 @@ static int kvm_init(MachineState *ms) >>> int ret; >>> int type = 0; >>> const char *kvm_type; >>> + uint64_t dirty_log_manual_caps; >>> >>> s = KVM_STATE(ms->accelerator); >>> >>> @@ -2007,14 +2008,20 @@ static int kvm_init(MachineState *ms) >>> s->coalesced_pio = s->coalesced_mmio && >>> kvm_check_extension(s, >> KVM_CAP_COALESCED_PIO); >>> >>> - s->manual_dirty_log_protect = >>> + dirty_log_manual_caps = >>> kvm_check_extension(s, >> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2); >>> - if (s->manual_dirty_log_protect) { >>> - ret = kvm_vm_enable_cap(s, >> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, 1); >>> + dirty_log_manual_caps &= >> (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE | >>> + KVM_DIRTY_LOG_INITIALLY_SET); >>> + s->manual_dirty_log_protect = dirty_log_manual_caps; >>> + if (dirty_log_manual_caps) { >>> + ret = kvm_vm_enable_cap(s, >> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, >>> + dirty_log_manual_caps); >>> if (ret) { >>> - warn_report("Trying to enable >> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 " >>> - "but failed. Falling back to the legacy mode. "); >>> - s->manual_dirty_log_protect = false; >>> + warn_report("Trying to enable capability %"PRIu64" of " >>> + "KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 >> but failed. " >>> + "Falling back to the legacy mode. ", >>> + dirty_log_manual_caps); >>> + s->manual_dirty_log_protect = 0; >>> } >>> } >>> >>> diff --git a/linux-headers/linux/kvm.h b/linux-headers/linux/kvm.h >>> index 265099100e..3cb71c2b19 100644 >>> --- a/linux-headers/linux/kvm.h >>> +++ b/linux-headers/linux/kvm.h >>> @@ -1628,4 +1628,7 @@ struct kvm_hyperv_eventfd { >>> #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff >>> #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) >>> >>> +#define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) >>> +#define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) >>> + >>> #endif /* __LINUX_KVM_H */ >>> >> >> Queued, thanks. >> > > Hi Paolo, > > It seems that this patch isn't included in your last pull request... > If there's something else to be done, please let me know. Sorry, I thought mistakenly that it was a 5.8 feature (so it would have to wait for the 5.8-rc1 release and header update). It's still queued though. Paolo
> -----Original Message----- > From: Paolo Bonzini [mailto:pbonzini@redhat.com] > Sent: Friday, June 12, 2020 5:28 PM > To: Zhoujian (jay) <jianjay.zhou@huawei.com>; qemu-devel@nongnu.org; > kvm@vger.kernel.org > Cc: mst@redhat.com; cohuck@redhat.com; peterx@redhat.com; Wangxin > (Alexander, Cloud Infrastructure Service Product Dept.) > <wangxinxin.wang@huawei.com>; Huangweidong (C) > <weidong.huang@huawei.com>; Liujinsong (Paul) <liu.jinsong@huawei.com> > Subject: Re: [PATCH] kvm: support to get/set dirty log initial-all-set capability > > On 12/06/20 05:01, Zhoujian (jay) wrote: > > > > > >> -----Original Message----- > >> From: Paolo Bonzini [mailto:pbonzini@redhat.com] > >> Sent: Wednesday, March 18, 2020 6:48 PM > >> To: Zhoujian (jay) <jianjay.zhou@huawei.com>; qemu-devel@nongnu.org; > >> kvm@vger.kernel.org > >> Cc: mst@redhat.com; cohuck@redhat.com; peterx@redhat.com; wangxin (U) > >> <wangxinxin.wang@huawei.com>; Huangweidong (C) > >> <weidong.huang@huawei.com>; Liujinsong (Paul) > >> <liu.jinsong@huawei.com> > >> Subject: Re: [PATCH] kvm: support to get/set dirty log > >> initial-all-set capability > >> > >> On 04/03/20 03:55, Jay Zhou wrote: > >>> Since the new capability KVM_DIRTY_LOG_INITIALLY_SET of > >>> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 has been introduced in the > kernel, > >>> tweak the userspace side to detect and enable this capability. > >>> > >>> Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com> > >>> --- > >>> accel/kvm/kvm-all.c | 21 ++++++++++++++------- > >>> linux-headers/linux/kvm.h | 3 +++ > >>> 2 files changed, 17 insertions(+), 7 deletions(-) > >>> > >>> diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c index > >>> 439a4efe52..45ab25be63 100644 > >>> --- a/accel/kvm/kvm-all.c > >>> +++ b/accel/kvm/kvm-all.c > >>> @@ -100,7 +100,7 @@ struct KVMState > >>> bool kernel_irqchip_required; > >>> OnOffAuto kernel_irqchip_split; > >>> bool sync_mmu; > >>> - bool manual_dirty_log_protect; > >>> + uint64_t manual_dirty_log_protect; > >>> /* The man page (and posix) say ioctl numbers are signed int, but > >>> * they're not. Linux, glibc and *BSD all treat ioctl numbers as > >>> * unsigned, and treating them as signed here can break things > >>> */ @@ -1882,6 +1882,7 @@ static int kvm_init(MachineState *ms) > >>> int ret; > >>> int type = 0; > >>> const char *kvm_type; > >>> + uint64_t dirty_log_manual_caps; > >>> > >>> s = KVM_STATE(ms->accelerator); > >>> > >>> @@ -2007,14 +2008,20 @@ static int kvm_init(MachineState *ms) > >>> s->coalesced_pio = s->coalesced_mmio && > >>> kvm_check_extension(s, > >> KVM_CAP_COALESCED_PIO); > >>> > >>> - s->manual_dirty_log_protect = > >>> + dirty_log_manual_caps = > >>> kvm_check_extension(s, > >> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2); > >>> - if (s->manual_dirty_log_protect) { > >>> - ret = kvm_vm_enable_cap(s, > >> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, 1); > >>> + dirty_log_manual_caps &= > >> (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE | > >>> + KVM_DIRTY_LOG_INITIALLY_SET); > >>> + s->manual_dirty_log_protect = dirty_log_manual_caps; > >>> + if (dirty_log_manual_caps) { > >>> + ret = kvm_vm_enable_cap(s, > >> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, > >>> + dirty_log_manual_caps); > >>> if (ret) { > >>> - warn_report("Trying to enable > >> KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 " > >>> - "but failed. Falling back to the legacy mode. > "); > >>> - s->manual_dirty_log_protect = false; > >>> + warn_report("Trying to enable capability %"PRIu64" of " > >>> + "KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 > >> but failed. " > >>> + "Falling back to the legacy mode. ", > >>> + dirty_log_manual_caps); > >>> + s->manual_dirty_log_protect = 0; > >>> } > >>> } > >>> > >>> diff --git a/linux-headers/linux/kvm.h b/linux-headers/linux/kvm.h > >>> index 265099100e..3cb71c2b19 100644 > >>> --- a/linux-headers/linux/kvm.h > >>> +++ b/linux-headers/linux/kvm.h > >>> @@ -1628,4 +1628,7 @@ struct kvm_hyperv_eventfd { > >>> #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff > >>> #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) > >>> > >>> +#define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) > >>> +#define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) > >>> + > >>> #endif /* __LINUX_KVM_H */ > >>> > >> > >> Queued, thanks. > >> > > > > Hi Paolo, > > > > It seems that this patch isn't included in your last pull request... > > If there's something else to be done, please let me know. > > Sorry, I thought mistakenly that it was a 5.8 feature (so it would have to wait for > the 5.8-rc1 release and header update). It's still queued though. Okay, never mind, :) Regards, Jay Zhou
diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c index 439a4efe52..45ab25be63 100644 --- a/accel/kvm/kvm-all.c +++ b/accel/kvm/kvm-all.c @@ -100,7 +100,7 @@ struct KVMState bool kernel_irqchip_required; OnOffAuto kernel_irqchip_split; bool sync_mmu; - bool manual_dirty_log_protect; + uint64_t manual_dirty_log_protect; /* The man page (and posix) say ioctl numbers are signed int, but * they're not. Linux, glibc and *BSD all treat ioctl numbers as * unsigned, and treating them as signed here can break things */ @@ -1882,6 +1882,7 @@ static int kvm_init(MachineState *ms) int ret; int type = 0; const char *kvm_type; + uint64_t dirty_log_manual_caps; s = KVM_STATE(ms->accelerator); @@ -2007,14 +2008,20 @@ static int kvm_init(MachineState *ms) s->coalesced_pio = s->coalesced_mmio && kvm_check_extension(s, KVM_CAP_COALESCED_PIO); - s->manual_dirty_log_protect = + dirty_log_manual_caps = kvm_check_extension(s, KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2); - if (s->manual_dirty_log_protect) { - ret = kvm_vm_enable_cap(s, KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, 1); + dirty_log_manual_caps &= (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE | + KVM_DIRTY_LOG_INITIALLY_SET); + s->manual_dirty_log_protect = dirty_log_manual_caps; + if (dirty_log_manual_caps) { + ret = kvm_vm_enable_cap(s, KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2, 0, + dirty_log_manual_caps); if (ret) { - warn_report("Trying to enable KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 " - "but failed. Falling back to the legacy mode. "); - s->manual_dirty_log_protect = false; + warn_report("Trying to enable capability %"PRIu64" of " + "KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 but failed. " + "Falling back to the legacy mode. ", + dirty_log_manual_caps); + s->manual_dirty_log_protect = 0; } } diff --git a/linux-headers/linux/kvm.h b/linux-headers/linux/kvm.h index 265099100e..3cb71c2b19 100644 --- a/linux-headers/linux/kvm.h +++ b/linux-headers/linux/kvm.h @@ -1628,4 +1628,7 @@ struct kvm_hyperv_eventfd { #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) +#define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) +#define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) + #endif /* __LINUX_KVM_H */
Since the new capability KVM_DIRTY_LOG_INITIALLY_SET of KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 has been introduced in the kernel, tweak the userspace side to detect and enable this capability. Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com> --- accel/kvm/kvm-all.c | 21 ++++++++++++++------- linux-headers/linux/kvm.h | 3 +++ 2 files changed, 17 insertions(+), 7 deletions(-)