From: Haozhong Zhang <haozhong.zhang@intel.com>
To: xen-devel@lists.xen.org
Cc: Haozhong Zhang <haozhong.zhang@intel.com>,
Kevin Tian <kevin.tian@intel.com>, Keir Fraser <keir@xen.org>,
Jan Beulich <jbeulich@suse.com>,
Andrew Cooper <andrew.cooper3@citrix.com>,
Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>,
Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>,
Boris Ostrovsky <boris.ostrovsky@oracle.com>
Subject: [PATCH v7 2/5] x86/hvm: Replace architecture TSC scaling by a common function
Date: Tue, 1 Mar 2016 09:39:32 +0800 [thread overview]
Message-ID: <1456796372-12655-1-git-send-email-haozhong.zhang@intel.com> (raw)
In-Reply-To: <1456664094-5161-3-git-send-email-haozhong.zhang@intel.com>
This patch implements a common function hvm_scale_tsc() to scale TSC by
using TSC scaling information collected by architecture code.
Signed-off-by: Haozhong Zhang <haozhong.zhang@intel.com>
Acked-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> for SVM bits
---
CC: Keir Fraser <keir@xen.org>
CC: Jan Beulich <jbeulich@suse.com>
CC: Andrew Cooper <andrew.cooper3@citrix.com>
CC: Boris Ostrovsky <boris.ostrovsky@oracle.com>
CC: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
CC: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>
CC: Kevin Tian <kevin.tian@intel.com>
---
Mixing named and positional operands is tested with gcc 4.1.2. It can
be compiled without any errors/warnings and provides the same result at
runtime as that compiled by newer version of gcc (v4.8.4).
Changes in v7:
For inline assembly in hvm_scale_tsc():
* Add '&' to the restriction of output operand 'dummy'.
---
xen/arch/x86/hvm/hvm.c | 21 +++++++++++++++++++--
xen/arch/x86/hvm/svm/svm.c | 8 --------
xen/arch/x86/time.c | 3 +--
xen/include/asm-x86/hvm/hvm.h | 3 +--
4 files changed, 21 insertions(+), 14 deletions(-)
diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
index 6c32e99..2d87de4 100644
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -333,6 +333,23 @@ u64 hvm_get_tsc_scaling_ratio(u32 gtsc_khz)
return ratio > max_ratio ? 0 : ratio;
}
+u64 hvm_scale_tsc(const struct domain *d, u64 tsc)
+{
+ u64 ratio = d->arch.hvm_domain.tsc_scaling_ratio;
+ u64 dummy;
+
+ if ( ratio == hvm_default_tsc_scaling_ratio )
+ return tsc;
+
+ /* tsc = (tsc * ratio) >> hvm_funcs.tsc_scaling.ratio_frac_bits */
+ asm ( "mulq %[ratio]; shrdq %[frac],%%rdx,%[tsc]"
+ : [tsc] "+a" (tsc), "=&d" (dummy)
+ : [frac] "c" (hvm_funcs.tsc_scaling.ratio_frac_bits),
+ [ratio] "rm" (ratio) );
+
+ return tsc;
+}
+
void hvm_set_guest_tsc_fixed(struct vcpu *v, u64 guest_tsc, u64 at_tsc)
{
uint64_t tsc;
@@ -347,7 +364,7 @@ void hvm_set_guest_tsc_fixed(struct vcpu *v, u64 guest_tsc, u64 at_tsc)
{
tsc = at_tsc ?: rdtsc();
if ( hvm_tsc_scaling_supported )
- tsc = hvm_funcs.tsc_scaling.scale_tsc(v, tsc);
+ tsc = hvm_scale_tsc(v->domain, tsc);
}
delta_tsc = guest_tsc - tsc;
@@ -379,7 +396,7 @@ u64 hvm_get_guest_tsc_fixed(struct vcpu *v, uint64_t at_tsc)
{
tsc = at_tsc ?: rdtsc();
if ( hvm_tsc_scaling_supported )
- tsc = hvm_funcs.tsc_scaling.scale_tsc(v, tsc);
+ tsc = hvm_scale_tsc(v->domain, tsc);
}
return tsc + v->arch.hvm_vcpu.cache_tsc_offset;
diff --git a/xen/arch/x86/hvm/svm/svm.c b/xen/arch/x86/hvm/svm/svm.c
index 7172f25..979d226 100644
--- a/xen/arch/x86/hvm/svm/svm.c
+++ b/xen/arch/x86/hvm/svm/svm.c
@@ -819,13 +819,6 @@ static uint64_t scale_tsc(uint64_t host_tsc, uint64_t ratio)
return scaled_host_tsc;
}
-static uint64_t svm_scale_tsc(const struct vcpu *v, uint64_t tsc)
-{
- ASSERT(cpu_has_tsc_ratio && !v->domain->arch.vtsc);
-
- return scale_tsc(tsc, hvm_tsc_scaling_ratio(v->domain));
-}
-
static uint64_t svm_get_tsc_offset(uint64_t host_tsc, uint64_t guest_tsc,
uint64_t ratio)
{
@@ -2291,7 +2284,6 @@ static struct hvm_function_table __initdata svm_function_table = {
.tsc_scaling = {
.max_ratio = ~TSC_RATIO_RSVD_BITS,
- .scale_tsc = svm_scale_tsc,
},
};
diff --git a/xen/arch/x86/time.c b/xen/arch/x86/time.c
index fda9692..687e39b 100644
--- a/xen/arch/x86/time.c
+++ b/xen/arch/x86/time.c
@@ -816,8 +816,7 @@ static void __update_vcpu_system_time(struct vcpu *v, int force)
{
if ( has_hvm_container_domain(d) && hvm_tsc_scaling_supported )
{
- tsc_stamp =
- hvm_funcs.tsc_scaling.scale_tsc(v, t->local_tsc_stamp);
+ tsc_stamp = hvm_scale_tsc(d, t->local_tsc_stamp);
_u.tsc_to_system_mul = d->arch.vtsc_to_ns.mul_frac;
_u.tsc_shift = d->arch.vtsc_to_ns.shift;
}
diff --git a/xen/include/asm-x86/hvm/hvm.h b/xen/include/asm-x86/hvm/hvm.h
index ddb1e33..c5c9328 100644
--- a/xen/include/asm-x86/hvm/hvm.h
+++ b/xen/include/asm-x86/hvm/hvm.h
@@ -231,8 +231,6 @@ struct hvm_function_table {
uint8_t ratio_frac_bits;
/* maximum-allowed TSC scaling ratio */
uint64_t max_ratio;
-
- uint64_t (*scale_tsc)(const struct vcpu *v, uint64_t tsc);
} tsc_scaling;
};
@@ -278,6 +276,7 @@ u64 hvm_get_guest_tsc_fixed(struct vcpu *v, u64 at_tsc);
#define hvm_tsc_scaling_ratio(d) \
((d)->arch.hvm_domain.tsc_scaling_ratio)
+u64 hvm_scale_tsc(const struct domain *d, u64 tsc);
u64 hvm_get_tsc_scaling_ratio(u32 gtsc_khz);
int hvm_set_mode(struct vcpu *v, int mode);
--
2.4.8
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
next prev parent reply other threads:[~2016-03-01 1:39 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-02-28 12:54 [PATCH v6 0/5] Add VMX TSC scaling support Haozhong Zhang
2016-02-28 12:54 ` [PATCH v6 1/5] x86/hvm: Setup TSC scaling ratio Haozhong Zhang
2016-02-29 13:41 ` Jan Beulich
2016-02-29 13:49 ` Boris Ostrovsky
2016-02-29 13:55 ` Haozhong Zhang
2016-02-29 13:55 ` Jan Beulich
2016-02-29 14:09 ` Boris Ostrovsky
2016-02-28 12:54 ` [PATCH v6 2/5] x86/hvm: Replace architecture TSC scaling by a common function Haozhong Zhang
2016-02-29 13:44 ` Jan Beulich
2016-02-29 14:03 ` Haozhong Zhang
2016-02-29 14:21 ` Boris Ostrovsky
2016-03-01 1:39 ` Haozhong Zhang [this message]
2016-03-01 10:13 ` [PATCH v7 " Jan Beulich
2016-02-28 12:54 ` [PATCH v6 3/5] x86/hvm: Move saving/loading vcpu's TSC to common code Haozhong Zhang
2016-02-28 12:54 ` [PATCH v6 4/5] vmx: Add VMX RDTSC(P) scaling support Haozhong Zhang
2016-02-28 12:54 ` [PATCH v6 5/5] docs: Add descriptions of TSC scaling in xl.cfg and tscmode.txt Haozhong Zhang
2016-03-01 14:09 ` Ian Jackson
2016-03-01 14:21 ` Jan Beulich
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=1456796372-12655-1-git-send-email-haozhong.zhang@intel.com \
--to=haozhong.zhang@intel.com \
--cc=Aravind.Gopalakrishnan@amd.com \
--cc=andrew.cooper3@citrix.com \
--cc=boris.ostrovsky@oracle.com \
--cc=jbeulich@suse.com \
--cc=keir@xen.org \
--cc=kevin.tian@intel.com \
--cc=suravee.suthikulpanit@amd.com \
--cc=xen-devel@lists.xen.org \
/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).