| Message ID | 1467820195-13302-1-git-send-email-czuzu@bitdefender.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 7/6/2016 6:49 PM, Corneliu ZUZU wrote: > Minor optimization @ vmx_update_guest_cr: checks if v->arch.hvm_vmx.exec_control > was modified before actually calling vmx_update_cpu_exec_control(v). > > Signed-off-by: Corneliu ZUZU <czuzu@bitdefender.com> > --- > Changed since v2: <nothing> > --- > xen/arch/x86/hvm/vmx/vmx.c | 5 ++++- > 1 file changed, 4 insertions(+), 1 deletion(-) > > diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c > index df19579..8ab074f 100644 > --- a/xen/arch/x86/hvm/vmx/vmx.c > +++ b/xen/arch/x86/hvm/vmx/vmx.c > @@ -1434,8 +1434,10 @@ static void vmx_update_guest_cr(struct vcpu *v, unsigned int cr) > if ( paging_mode_hap(v->domain) ) > { > /* Manage GUEST_CR3 when CR0.PE=0. */ > + uint32_t old_ctls = v->arch.hvm_vmx.exec_control; > uint32_t cr3_ctls = (CPU_BASED_CR3_LOAD_EXITING | > CPU_BASED_CR3_STORE_EXITING); > + > v->arch.hvm_vmx.exec_control &= ~cr3_ctls; > if ( !hvm_paging_enabled(v) && !vmx_unrestricted_guest(v) ) > v->arch.hvm_vmx.exec_control |= cr3_ctls; > @@ -1445,7 +1447,8 @@ static void vmx_update_guest_cr(struct vcpu *v, unsigned int cr) > monitor_ctrlreg_bitmask(VM_EVENT_X86_CR3) ) > v->arch.hvm_vmx.exec_control |= CPU_BASED_CR3_LOAD_EXITING; > > - vmx_update_cpu_exec_control(v); > + if ( old_ctls != v->arch.hvm_vmx.exec_control ) > + vmx_update_cpu_exec_control(v); > } > > if ( !nestedhvm_vcpu_in_guestmode(v) ) Jan, I'm wondering if you could ack this as well (if there's nothing wrong w/ it) and push it to be done with it. :-) Thanks, Corneliu.
>>> On 08.07.16 at 13:39, <czuzu@bitdefender.com> wrote: > On 7/6/2016 6:49 PM, Corneliu ZUZU wrote: >> Minor optimization @ vmx_update_guest_cr: checks if > v->arch.hvm_vmx.exec_control >> was modified before actually calling vmx_update_cpu_exec_control(v). >> >> Signed-off-by: Corneliu ZUZU <czuzu@bitdefender.com> >> --- >> Changed since v2: <nothing> >> --- >> xen/arch/x86/hvm/vmx/vmx.c | 5 ++++- >> 1 file changed, 4 insertions(+), 1 deletion(-) >> >> diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c >> index df19579..8ab074f 100644 >> --- a/xen/arch/x86/hvm/vmx/vmx.c >> +++ b/xen/arch/x86/hvm/vmx/vmx.c >> @@ -1434,8 +1434,10 @@ static void vmx_update_guest_cr(struct vcpu *v, > unsigned int cr) >> if ( paging_mode_hap(v->domain) ) >> { >> /* Manage GUEST_CR3 when CR0.PE=0. */ >> + uint32_t old_ctls = v->arch.hvm_vmx.exec_control; >> uint32_t cr3_ctls = (CPU_BASED_CR3_LOAD_EXITING | >> CPU_BASED_CR3_STORE_EXITING); >> + >> v->arch.hvm_vmx.exec_control &= ~cr3_ctls; >> if ( !hvm_paging_enabled(v) && !vmx_unrestricted_guest(v) ) >> v->arch.hvm_vmx.exec_control |= cr3_ctls; >> @@ -1445,7 +1447,8 @@ static void vmx_update_guest_cr(struct vcpu *v, > unsigned int cr) >> monitor_ctrlreg_bitmask(VM_EVENT_X86_CR3) ) >> v->arch.hvm_vmx.exec_control |= CPU_BASED_CR3_LOAD_EXITING; >> >> - vmx_update_cpu_exec_control(v); >> + if ( old_ctls != v->arch.hvm_vmx.exec_control ) >> + vmx_update_cpu_exec_control(v); >> } >> >> if ( !nestedhvm_vcpu_in_guestmode(v) ) > > I'm wondering if you could ack this as well (if there's nothing wrong w/ > it) and push it to be done with it. :-) Well, if you had pinged the patch at least once, I probably would. I don't, however, recall having seen any such ping (only resends). Jan
On 7/8/2016 2:48 PM, Jan Beulich wrote: >>>> On 08.07.16 at 13:39, <czuzu@bitdefender.com> wrote: >> On 7/6/2016 6:49 PM, Corneliu ZUZU wrote: >>> Minor optimization @ vmx_update_guest_cr: checks if >> v->arch.hvm_vmx.exec_control >>> was modified before actually calling vmx_update_cpu_exec_control(v). >>> >>> Signed-off-by: Corneliu ZUZU <czuzu@bitdefender.com> >>> --- >>> Changed since v2: <nothing> >>> --- >>> xen/arch/x86/hvm/vmx/vmx.c | 5 ++++- >>> 1 file changed, 4 insertions(+), 1 deletion(-) >>> >>> diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c >>> index df19579..8ab074f 100644 >>> --- a/xen/arch/x86/hvm/vmx/vmx.c >>> +++ b/xen/arch/x86/hvm/vmx/vmx.c >>> @@ -1434,8 +1434,10 @@ static void vmx_update_guest_cr(struct vcpu *v, >> unsigned int cr) >>> if ( paging_mode_hap(v->domain) ) >>> { >>> /* Manage GUEST_CR3 when CR0.PE=0. */ >>> + uint32_t old_ctls = v->arch.hvm_vmx.exec_control; >>> uint32_t cr3_ctls = (CPU_BASED_CR3_LOAD_EXITING | >>> CPU_BASED_CR3_STORE_EXITING); >>> + >>> v->arch.hvm_vmx.exec_control &= ~cr3_ctls; >>> if ( !hvm_paging_enabled(v) && !vmx_unrestricted_guest(v) ) >>> v->arch.hvm_vmx.exec_control |= cr3_ctls; >>> @@ -1445,7 +1447,8 @@ static void vmx_update_guest_cr(struct vcpu *v, >> unsigned int cr) >>> monitor_ctrlreg_bitmask(VM_EVENT_X86_CR3) ) >>> v->arch.hvm_vmx.exec_control |= CPU_BASED_CR3_LOAD_EXITING; >>> >>> - vmx_update_cpu_exec_control(v); >>> + if ( old_ctls != v->arch.hvm_vmx.exec_control ) >>> + vmx_update_cpu_exec_control(v); >>> } >>> >>> if ( !nestedhvm_vcpu_in_guestmode(v) ) >> I'm wondering if you could ack this as well (if there's nothing wrong w/ >> it) and push it to be done with it. :-) > Well, if you had pinged the patch at least once, I probably would. > I don't, however, recall having seen any such ping (only resends). > > Jan So is a 'ping' (haven't done that before) still necessary? Is a ping a simple 'reply-to-all' including 'Ping: ' in the subject? :-) Thanks, Corneliu.
>>> On 08.07.16 at 13:55, <czuzu@bitdefender.com> wrote: > On 7/8/2016 2:48 PM, Jan Beulich wrote: >>>>> On 08.07.16 at 13:39, <czuzu@bitdefender.com> wrote: >>> On 7/6/2016 6:49 PM, Corneliu ZUZU wrote: >>>> Minor optimization @ vmx_update_guest_cr: checks if >>> v->arch.hvm_vmx.exec_control >>>> was modified before actually calling vmx_update_cpu_exec_control(v). >>>> >>>> Signed-off-by: Corneliu ZUZU <czuzu@bitdefender.com> >>>> --- >>>> Changed since v2: <nothing> >>>> --- >>>> xen/arch/x86/hvm/vmx/vmx.c | 5 ++++- >>>> 1 file changed, 4 insertions(+), 1 deletion(-) >>>> >>>> diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c >>>> index df19579..8ab074f 100644 >>>> --- a/xen/arch/x86/hvm/vmx/vmx.c >>>> +++ b/xen/arch/x86/hvm/vmx/vmx.c >>>> @@ -1434,8 +1434,10 @@ static void vmx_update_guest_cr(struct vcpu *v, >>> unsigned int cr) >>>> if ( paging_mode_hap(v->domain) ) >>>> { >>>> /* Manage GUEST_CR3 when CR0.PE=0. */ >>>> + uint32_t old_ctls = v->arch.hvm_vmx.exec_control; >>>> uint32_t cr3_ctls = (CPU_BASED_CR3_LOAD_EXITING | >>>> CPU_BASED_CR3_STORE_EXITING); >>>> + >>>> v->arch.hvm_vmx.exec_control &= ~cr3_ctls; >>>> if ( !hvm_paging_enabled(v) && !vmx_unrestricted_guest(v) ) >>>> v->arch.hvm_vmx.exec_control |= cr3_ctls; >>>> @@ -1445,7 +1447,8 @@ static void vmx_update_guest_cr(struct vcpu *v, >>> unsigned int cr) >>>> monitor_ctrlreg_bitmask(VM_EVENT_X86_CR3) ) >>>> v->arch.hvm_vmx.exec_control |= CPU_BASED_CR3_LOAD_EXITING; >>>> >>>> - vmx_update_cpu_exec_control(v); >>>> + if ( old_ctls != v->arch.hvm_vmx.exec_control ) >>>> + vmx_update_cpu_exec_control(v); >>>> } >>>> >>>> if ( !nestedhvm_vcpu_in_guestmode(v) ) >>> I'm wondering if you could ack this as well (if there's nothing wrong w/ >>> it) and push it to be done with it. :-) >> Well, if you had pinged the patch at least once, I probably would. >> I don't, however, recall having seen any such ping (only resends). > > So is a 'ping' (haven't done that before) still necessary? Is a ping a > simple 'reply-to-all' including 'Ping: ' in the subject? :-) It's not very much formalized. What I do in such a case is, as you say, prefix the subject with Ping:, but change addressing so that the people expected to reply end up in To:, while everyone else in the original To: list (i.e. including xen-devel) would get moved to Cc:. And while, with the adjustment to addressees in my earlier reply, I already tried to do kind of a ping to them, I think it wouldn't hurt if you did another explicit one. Jan
On 7/6/2016 6:49 PM, Corneliu ZUZU wrote: > Minor optimization @ vmx_update_guest_cr: checks if v->arch.hvm_vmx.exec_control > was modified before actually calling vmx_update_cpu_exec_control(v). > > Signed-off-by: Corneliu ZUZU <czuzu@bitdefender.com> > --- > Changed since v2: <nothing> > --- > xen/arch/x86/hvm/vmx/vmx.c | 5 ++++- > 1 file changed, 4 insertions(+), 1 deletion(-) > > diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c > index df19579..8ab074f 100644 > --- a/xen/arch/x86/hvm/vmx/vmx.c > +++ b/xen/arch/x86/hvm/vmx/vmx.c > @@ -1434,8 +1434,10 @@ static void vmx_update_guest_cr(struct vcpu *v, unsigned int cr) > if ( paging_mode_hap(v->domain) ) > { > /* Manage GUEST_CR3 when CR0.PE=0. */ > + uint32_t old_ctls = v->arch.hvm_vmx.exec_control; > uint32_t cr3_ctls = (CPU_BASED_CR3_LOAD_EXITING | > CPU_BASED_CR3_STORE_EXITING); > + > v->arch.hvm_vmx.exec_control &= ~cr3_ctls; > if ( !hvm_paging_enabled(v) && !vmx_unrestricted_guest(v) ) > v->arch.hvm_vmx.exec_control |= cr3_ctls; > @@ -1445,7 +1447,8 @@ static void vmx_update_guest_cr(struct vcpu *v, unsigned int cr) > monitor_ctrlreg_bitmask(VM_EVENT_X86_CR3) ) > v->arch.hvm_vmx.exec_control |= CPU_BASED_CR3_LOAD_EXITING; > > - vmx_update_cpu_exec_control(v); > + if ( old_ctls != v->arch.hvm_vmx.exec_control ) > + vmx_update_cpu_exec_control(v); > } > > if ( !nestedhvm_vcpu_in_guestmode(v) )
> From: Corneliu ZUZU [mailto:czuzu@bitdefender.com] > Sent: Wednesday, July 06, 2016 11:50 PM > > Minor optimization @ vmx_update_guest_cr: checks if v->arch.hvm_vmx.exec_control > was modified before actually calling vmx_update_cpu_exec_control(v). > > Signed-off-by: Corneliu ZUZU <czuzu@bitdefender.com> Acked-by: Kevin Tian <kevin.tian@intel.com>
diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index df19579..8ab074f 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -1434,8 +1434,10 @@ static void vmx_update_guest_cr(struct vcpu *v, unsigned int cr) if ( paging_mode_hap(v->domain) ) { /* Manage GUEST_CR3 when CR0.PE=0. */ + uint32_t old_ctls = v->arch.hvm_vmx.exec_control; uint32_t cr3_ctls = (CPU_BASED_CR3_LOAD_EXITING | CPU_BASED_CR3_STORE_EXITING); + v->arch.hvm_vmx.exec_control &= ~cr3_ctls; if ( !hvm_paging_enabled(v) && !vmx_unrestricted_guest(v) ) v->arch.hvm_vmx.exec_control |= cr3_ctls; @@ -1445,7 +1447,8 @@ static void vmx_update_guest_cr(struct vcpu *v, unsigned int cr) monitor_ctrlreg_bitmask(VM_EVENT_X86_CR3) ) v->arch.hvm_vmx.exec_control |= CPU_BASED_CR3_LOAD_EXITING; - vmx_update_cpu_exec_control(v); + if ( old_ctls != v->arch.hvm_vmx.exec_control ) + vmx_update_cpu_exec_control(v); } if ( !nestedhvm_vcpu_in_guestmode(v) )
Minor optimization @ vmx_update_guest_cr: checks if v->arch.hvm_vmx.exec_control was modified before actually calling vmx_update_cpu_exec_control(v). Signed-off-by: Corneliu ZUZU <czuzu@bitdefender.com> --- Changed since v2: <nothing> --- xen/arch/x86/hvm/vmx/vmx.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)