* [PATCH 0/2] x86/hvm: Improvements to HAP superpage feature detection
@ 2015-09-22 17:33 Andrew Cooper
2015-09-22 17:33 ` [PATCH 1/2] x86/hvm: Refine hap_has_{2mb,1gb} checks Andrew Cooper
` (2 more replies)
0 siblings, 3 replies; 6+ messages in thread
From: Andrew Cooper @ 2015-09-22 17:33 UTC (permalink / raw)
To: Xen-devel; +Cc: George Dunlap, Andrew Cooper, Jan Beulich
Two small patches which reduce the runtime overhead of detecting HAP
superpages by moving more initialisation into __init.
Noticed while reviewing "x86/p2m: use large pages for MMIO mappings".
Andrew Cooper (2):
x86/hvm: Refine hap_has_{2mb,1gb} checks
x86/hvm: Fold opt_hap_{2mb,1gb} into hap_capabilities
xen/arch/x86/hvm/hvm.c | 8 ++++++++
xen/arch/x86/mm/p2m-ept.c | 4 ++--
xen/arch/x86/mm/p2m.c | 10 ++++------
xen/include/asm-x86/hvm/hvm.h | 6 ++----
4 files changed, 16 insertions(+), 12 deletions(-)
--
1.7.10.4
^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH 1/2] x86/hvm: Refine hap_has_{2mb,1gb} checks
2015-09-22 17:33 [PATCH 0/2] x86/hvm: Improvements to HAP superpage feature detection Andrew Cooper
@ 2015-09-22 17:33 ` Andrew Cooper
2015-09-23 8:47 ` [PATCH 1/2] x86/hvm: Refine hap_has_{2mb, 1gb} checks George Dunlap
2015-09-22 17:33 ` [PATCH 2/2] x86/hvm: Fold opt_hap_{2mb, 1gb} into hap_capabilities Andrew Cooper
2015-09-23 17:53 ` [PATCH 0/2] x86/hvm: Improvements to HAP superpage feature detection Konrad Rzeszutek Wilk
2 siblings, 1 reply; 6+ messages in thread
From: Andrew Cooper @ 2015-09-22 17:33 UTC (permalink / raw)
To: Xen-devel; +Cc: George Dunlap, Andrew Cooper, Jan Beulich
HAP superpages are a host property and not dependent on domain configuration.
Drop the domain paramter (which was only used in one of the two callsites),
and drop the redundant hvm_ prefix to mirror the cpu_has_* style of feature
detection.
Finally, convert the checks to being proper booleans rather than just non-zero
integers.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
CC: Jan Beulich <JBeulich@suse.com>
CC: George Dunlap <george.dunlap@eu.citrix.com>
---
xen/arch/x86/mm/p2m-ept.c | 4 ++--
xen/arch/x86/mm/p2m.c | 4 ++--
xen/include/asm-x86/hvm/hvm.h | 6 ++----
3 files changed, 6 insertions(+), 8 deletions(-)
diff --git a/xen/arch/x86/mm/p2m-ept.c b/xen/arch/x86/mm/p2m-ept.c
index 58db34e..dde242e 100644
--- a/xen/arch/x86/mm/p2m-ept.c
+++ b/xen/arch/x86/mm/p2m-ept.c
@@ -696,8 +696,8 @@ bool_t ept_handle_misconfig(uint64_t gpa)
if ( ret > 0 )
needs_sync = sync_on;
- ASSERT((target == 2 && hvm_hap_has_1gb()) ||
- (target == 1 && hvm_hap_has_2mb()) ||
+ ASSERT((target == 2 && hap_has_1gb) ||
+ (target == 1 && hap_has_2mb) ||
(target == 0));
ASSERT(!p2m_is_foreign(p2mt) || target == 0);
diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c
index e1d930a..9cf7a71 100644
--- a/xen/arch/x86/mm/p2m.c
+++ b/xen/arch/x86/mm/p2m.c
@@ -452,9 +452,9 @@ int p2m_set_entry(struct p2m_domain *p2m, unsigned long gfn, mfn_t mfn,
{
if ( hap_enabled(d) )
order = ( (((gfn | mfn_x(mfn) | todo) & ((1ul << PAGE_ORDER_1G) - 1)) == 0) &&
- hvm_hap_has_1gb(d) && opt_hap_1gb ) ? PAGE_ORDER_1G :
+ hap_has_1gb && opt_hap_1gb) ? PAGE_ORDER_1G :
((((gfn | mfn_x(mfn) | todo) & ((1ul << PAGE_ORDER_2M) - 1)) == 0) &&
- hvm_hap_has_2mb(d) && opt_hap_2mb) ? PAGE_ORDER_2M : PAGE_ORDER_4K;
+ hap_has_2mb && opt_hap_2mb) ? PAGE_ORDER_2M : PAGE_ORDER_4K;
else
order = 0;
diff --git a/xen/include/asm-x86/hvm/hvm.h b/xen/include/asm-x86/hvm/hvm.h
index 9f49e6d..2c3192c 100644
--- a/xen/include/asm-x86/hvm/hvm.h
+++ b/xen/include/asm-x86/hvm/hvm.h
@@ -278,10 +278,8 @@ int vmsi_deliver(
(!!((v)->arch.hvm_vcpu.guest_efer & EFER_NX))
/* Can we use superpages in the HAP p2m table? */
-#define hvm_hap_has_1gb(d) \
- (hvm_funcs.hap_capabilities & HVM_HAP_SUPERPAGE_1GB)
-#define hvm_hap_has_2mb(d) \
- (hvm_funcs.hap_capabilities & HVM_HAP_SUPERPAGE_2MB)
+#define hap_has_1gb !!(hvm_funcs.hap_capabilities & HVM_HAP_SUPERPAGE_1GB)
+#define hap_has_2mb !!(hvm_funcs.hap_capabilities & HVM_HAP_SUPERPAGE_2MB)
/* Can the guest use 1GB superpages in its own pagetables? */
#define hvm_pse1gb_supported(d) \
--
1.7.10.4
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH 2/2] x86/hvm: Fold opt_hap_{2mb, 1gb} into hap_capabilities
2015-09-22 17:33 [PATCH 0/2] x86/hvm: Improvements to HAP superpage feature detection Andrew Cooper
2015-09-22 17:33 ` [PATCH 1/2] x86/hvm: Refine hap_has_{2mb,1gb} checks Andrew Cooper
@ 2015-09-22 17:33 ` Andrew Cooper
2015-09-23 8:49 ` George Dunlap
2015-09-23 17:53 ` [PATCH 0/2] x86/hvm: Improvements to HAP superpage feature detection Konrad Rzeszutek Wilk
2 siblings, 1 reply; 6+ messages in thread
From: Andrew Cooper @ 2015-09-22 17:33 UTC (permalink / raw)
To: Xen-devel; +Cc: George Dunlap, Andrew Cooper, Jan Beulich
This allows all runtime users to simply check hap_has_{2mb,1gb} rather than
having to check opt_hap_{2mb,1gb} as well.
As a result, opt_hap_{2mb,1gb} can move into __initdata.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
CC: Jan Beulich <JBeulich@suse.com>
CC: George Dunlap <george.dunlap@eu.citrix.com>
---
xen/arch/x86/hvm/hvm.c | 8 ++++++++
xen/arch/x86/mm/p2m.c | 10 ++++------
2 files changed, 12 insertions(+), 6 deletions(-)
diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
index f082b86..6afc344 100644
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -158,9 +158,17 @@ static int __init hvm_enable(void)
printk("HVM: Hardware Assisted Paging (HAP) detected\n");
printk("HVM: HAP page sizes: 4kB");
if ( fns->hap_capabilities & HVM_HAP_SUPERPAGE_2MB )
+ {
printk(", 2MB%s", opt_hap_2mb ? "" : " [disabled]");
+ if ( !opt_hap_2mb )
+ hvm_funcs.hap_capabilities &= ~HVM_HAP_SUPERPAGE_2MB;
+ }
if ( fns->hap_capabilities & HVM_HAP_SUPERPAGE_1GB )
+ {
printk(", 1GB%s", opt_hap_1gb ? "" : " [disabled]");
+ if ( !opt_hap_1gb )
+ hvm_funcs.hap_capabilities &= ~HVM_HAP_SUPERPAGE_1GB;
+ }
printk("\n");
}
diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c
index 9cf7a71..5d2a532 100644
--- a/xen/arch/x86/mm/p2m.c
+++ b/xen/arch/x86/mm/p2m.c
@@ -41,11 +41,9 @@
#include "mm-locks.h"
-/* turn on/off 1GB host page table support for hap, default on */
-bool_t __read_mostly opt_hap_1gb = 1;
+/* turn on/off host superpage page table support for hap, default on */
+bool_t __initdata opt_hap_1gb = 1, __initdata opt_hap_2mb = 1;
boolean_param("hap_1gb", opt_hap_1gb);
-
-bool_t __read_mostly opt_hap_2mb = 1;
boolean_param("hap_2mb", opt_hap_2mb);
@@ -452,9 +450,9 @@ int p2m_set_entry(struct p2m_domain *p2m, unsigned long gfn, mfn_t mfn,
{
if ( hap_enabled(d) )
order = ( (((gfn | mfn_x(mfn) | todo) & ((1ul << PAGE_ORDER_1G) - 1)) == 0) &&
- hap_has_1gb && opt_hap_1gb) ? PAGE_ORDER_1G :
+ hap_has_1gb) ? PAGE_ORDER_1G :
((((gfn | mfn_x(mfn) | todo) & ((1ul << PAGE_ORDER_2M) - 1)) == 0) &&
- hap_has_2mb && opt_hap_2mb) ? PAGE_ORDER_2M : PAGE_ORDER_4K;
+ hap_has_2mb) ? PAGE_ORDER_2M : PAGE_ORDER_4K;
else
order = 0;
--
1.7.10.4
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH 1/2] x86/hvm: Refine hap_has_{2mb, 1gb} checks
2015-09-22 17:33 ` [PATCH 1/2] x86/hvm: Refine hap_has_{2mb,1gb} checks Andrew Cooper
@ 2015-09-23 8:47 ` George Dunlap
0 siblings, 0 replies; 6+ messages in thread
From: George Dunlap @ 2015-09-23 8:47 UTC (permalink / raw)
To: Andrew Cooper; +Cc: Jan Beulich, Xen-devel
On Tue, Sep 22, 2015 at 6:33 PM, Andrew Cooper
<andrew.cooper3@citrix.com> wrote:
> HAP superpages are a host property and not dependent on domain configuration.
> Drop the domain paramter (which was only used in one of the two callsites),
> and drop the redundant hvm_ prefix to mirror the cpu_has_* style of feature
> detection.
>
> Finally, convert the checks to being proper booleans rather than just non-zero
> integers.
>
> Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: George Dunlap <george.dunlap@citrix.com>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH 2/2] x86/hvm: Fold opt_hap_{2mb, 1gb} into hap_capabilities
2015-09-22 17:33 ` [PATCH 2/2] x86/hvm: Fold opt_hap_{2mb, 1gb} into hap_capabilities Andrew Cooper
@ 2015-09-23 8:49 ` George Dunlap
0 siblings, 0 replies; 6+ messages in thread
From: George Dunlap @ 2015-09-23 8:49 UTC (permalink / raw)
To: Andrew Cooper; +Cc: Jan Beulich, Xen-devel
On Tue, Sep 22, 2015 at 6:33 PM, Andrew Cooper
<andrew.cooper3@citrix.com> wrote:
> This allows all runtime users to simply check hap_has_{2mb,1gb} rather than
> having to check opt_hap_{2mb,1gb} as well.
>
> As a result, opt_hap_{2mb,1gb} can move into __initdata.
>
> Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: George Dunlap <george.dunlap@citrix.com>
> ---
> CC: Jan Beulich <JBeulich@suse.com>
> CC: George Dunlap <george.dunlap@eu.citrix.com>
> ---
> xen/arch/x86/hvm/hvm.c | 8 ++++++++
> xen/arch/x86/mm/p2m.c | 10 ++++------
> 2 files changed, 12 insertions(+), 6 deletions(-)
>
> diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
> index f082b86..6afc344 100644
> --- a/xen/arch/x86/hvm/hvm.c
> +++ b/xen/arch/x86/hvm/hvm.c
> @@ -158,9 +158,17 @@ static int __init hvm_enable(void)
> printk("HVM: Hardware Assisted Paging (HAP) detected\n");
> printk("HVM: HAP page sizes: 4kB");
> if ( fns->hap_capabilities & HVM_HAP_SUPERPAGE_2MB )
> + {
> printk(", 2MB%s", opt_hap_2mb ? "" : " [disabled]");
> + if ( !opt_hap_2mb )
> + hvm_funcs.hap_capabilities &= ~HVM_HAP_SUPERPAGE_2MB;
> + }
> if ( fns->hap_capabilities & HVM_HAP_SUPERPAGE_1GB )
> + {
> printk(", 1GB%s", opt_hap_1gb ? "" : " [disabled]");
> + if ( !opt_hap_1gb )
> + hvm_funcs.hap_capabilities &= ~HVM_HAP_SUPERPAGE_1GB;
> + }
> printk("\n");
> }
>
> diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c
> index 9cf7a71..5d2a532 100644
> --- a/xen/arch/x86/mm/p2m.c
> +++ b/xen/arch/x86/mm/p2m.c
> @@ -41,11 +41,9 @@
>
> #include "mm-locks.h"
>
> -/* turn on/off 1GB host page table support for hap, default on */
> -bool_t __read_mostly opt_hap_1gb = 1;
> +/* turn on/off host superpage page table support for hap, default on */
> +bool_t __initdata opt_hap_1gb = 1, __initdata opt_hap_2mb = 1;
> boolean_param("hap_1gb", opt_hap_1gb);
> -
> -bool_t __read_mostly opt_hap_2mb = 1;
> boolean_param("hap_2mb", opt_hap_2mb);
>
>
> @@ -452,9 +450,9 @@ int p2m_set_entry(struct p2m_domain *p2m, unsigned long gfn, mfn_t mfn,
> {
> if ( hap_enabled(d) )
> order = ( (((gfn | mfn_x(mfn) | todo) & ((1ul << PAGE_ORDER_1G) - 1)) == 0) &&
> - hap_has_1gb && opt_hap_1gb) ? PAGE_ORDER_1G :
> + hap_has_1gb) ? PAGE_ORDER_1G :
> ((((gfn | mfn_x(mfn) | todo) & ((1ul << PAGE_ORDER_2M) - 1)) == 0) &&
> - hap_has_2mb && opt_hap_2mb) ? PAGE_ORDER_2M : PAGE_ORDER_4K;
> + hap_has_2mb) ? PAGE_ORDER_2M : PAGE_ORDER_4K;
> else
> order = 0;
>
> --
> 1.7.10.4
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH 0/2] x86/hvm: Improvements to HAP superpage feature detection
2015-09-22 17:33 [PATCH 0/2] x86/hvm: Improvements to HAP superpage feature detection Andrew Cooper
2015-09-22 17:33 ` [PATCH 1/2] x86/hvm: Refine hap_has_{2mb,1gb} checks Andrew Cooper
2015-09-22 17:33 ` [PATCH 2/2] x86/hvm: Fold opt_hap_{2mb, 1gb} into hap_capabilities Andrew Cooper
@ 2015-09-23 17:53 ` Konrad Rzeszutek Wilk
2 siblings, 0 replies; 6+ messages in thread
From: Konrad Rzeszutek Wilk @ 2015-09-23 17:53 UTC (permalink / raw)
To: Andrew Cooper; +Cc: George Dunlap, Jan Beulich, Xen-devel
On Tue, Sep 22, 2015 at 06:33:01PM +0100, Andrew Cooper wrote:
> Two small patches which reduce the runtime overhead of detecting HAP
> superpages by moving more initialisation into __init.
>
> Noticed while reviewing "x86/p2m: use large pages for MMIO mappings".
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
for both patches.
>
> Andrew Cooper (2):
> x86/hvm: Refine hap_has_{2mb,1gb} checks
> x86/hvm: Fold opt_hap_{2mb,1gb} into hap_capabilities
>
> xen/arch/x86/hvm/hvm.c | 8 ++++++++
> xen/arch/x86/mm/p2m-ept.c | 4 ++--
> xen/arch/x86/mm/p2m.c | 10 ++++------
> xen/include/asm-x86/hvm/hvm.h | 6 ++----
> 4 files changed, 16 insertions(+), 12 deletions(-)
>
> --
> 1.7.10.4
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2015-09-23 17:53 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-09-22 17:33 [PATCH 0/2] x86/hvm: Improvements to HAP superpage feature detection Andrew Cooper
2015-09-22 17:33 ` [PATCH 1/2] x86/hvm: Refine hap_has_{2mb,1gb} checks Andrew Cooper
2015-09-23 8:47 ` [PATCH 1/2] x86/hvm: Refine hap_has_{2mb, 1gb} checks George Dunlap
2015-09-22 17:33 ` [PATCH 2/2] x86/hvm: Fold opt_hap_{2mb, 1gb} into hap_capabilities Andrew Cooper
2015-09-23 8:49 ` George Dunlap
2015-09-23 17:53 ` [PATCH 0/2] x86/hvm: Improvements to HAP superpage feature detection Konrad Rzeszutek Wilk
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).