[Xen-devel] [PATCH for-next v3 0/9] Port Xen to Hyper-V

[Xen-devel] [PATCH for-next v3 0/9] Port Xen to Hyper-V

[Wei Liu]

Hi all

This is version 3 of the patch series.

This is the very first stage for porting Xen to run on Hyper-V with all the
goodies Hyper-V has to offer.  With this series, Xen can successfully detect
Hyper-V and prints out a message.  I would like to first get the code structure
and kconfig options agreed upon.

There are two major areas to be worked on:
  * Make Dom0 able to use Hyper-V's synthetic devices.
  * Make Xen use of the synthetic timer, reference TSC and enlightenment VMCS
    and other interfaces.

They aren't trivial, and time can be scarce on my side, so I intend to post
patches piecemeal when they are ready.

Questions and comments are welcome.

Thanks,
Wei.

---
Changes in v3: See invididual patches

Changes in v2:
1. Introduce and use a hypervisor framework
2. Keep memmap infra under Xen for now

Cc: Jan Beulich <jbeulich@suse.com>
Cc: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: Wei Liu <wl@xen.org>
Cc: Roger Pau Monné <roger.pau@citrix.com>

Wei Liu (9):
  x86: introduce CONFIG_GUEST and move code
  x86: include asm_defns.h directly in hypercall.h
  x86: drop hypervisor_cpuid_base
  x86: include xen/lib.h in guest/hypercall.h
  x86: introduce hypervisor framework
  x86: rename hypervisor_{alloc,free}_unused_page
  x86: switch xen implementation to use hypervisor framework
  x86: be more verbose when running on a hypervisor
  x86: introduce CONFIG_HYPERV and detection code

 xen/arch/x86/Kconfig                          | 13 +++
 xen/arch/x86/Makefile                         |  2 +-
 xen/arch/x86/guest/Makefile                   |  6 +-
 xen/arch/x86/guest/hyperv/Makefile            |  1 +
 xen/arch/x86/guest/hyperv/hyperv.c            | 54 +++++++++++
 xen/arch/x86/guest/hypervisor.c               | 89 +++++++++++++++++++
 xen/arch/x86/guest/xen/Makefile               |  4 +
 xen/arch/x86/guest/{ => xen}/hypercall_page.S |  0
 xen/arch/x86/guest/{ => xen}/pvh-boot.c       |  2 +-
 xen/arch/x86/guest/{ => xen}/xen.c            | 39 ++++----
 xen/arch/x86/pv/shim.c                        |  6 +-
 xen/arch/x86/setup.c                          |  6 +-
 xen/include/asm-x86/guest.h                   |  2 +
 xen/include/asm-x86/guest/hypercall.h         |  4 +
 xen/include/asm-x86/guest/hyperv.h            | 45 ++++++++++
 xen/include/asm-x86/guest/hypervisor.h        | 64 +++++++++++++
 xen/include/asm-x86/guest/xen.h               | 24 ++---
 17 files changed, 314 insertions(+), 47 deletions(-)
 create mode 100644 xen/arch/x86/guest/hyperv/Makefile
 create mode 100644 xen/arch/x86/guest/hyperv/hyperv.c
 create mode 100644 xen/arch/x86/guest/hypervisor.c
 create mode 100644 xen/arch/x86/guest/xen/Makefile
 rename xen/arch/x86/guest/{ => xen}/hypercall_page.S (100%)
 rename xen/arch/x86/guest/{ => xen}/pvh-boot.c (99%)
 rename xen/arch/x86/guest/{ => xen}/xen.c (93%)
 create mode 100644 xen/include/asm-x86/guest/hyperv.h
 create mode 100644 xen/include/asm-x86/guest/hypervisor.h

-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH for-next v3 1/9] x86: introduce CONFIG_GUEST and move code

[Wei Liu]

Xen is able to run as a guest on Xen. We plan to make it able to run
on Hyper-V as well.

Introduce CONFIG_GUEST which is set to true if either running on Xen
or Hyper-V is desired. Restructure code hierarchy for new code to
come.

No functional change intended.

Signed-off-by: Wei Liu <liuwe@microsoft.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
---
 xen/arch/x86/Kconfig                          | 4 ++++
 xen/arch/x86/Makefile                         | 2 +-
 xen/arch/x86/guest/Makefile                   | 5 +----
 xen/arch/x86/guest/xen/Makefile               | 4 ++++
 xen/arch/x86/guest/{ => xen}/hypercall_page.S | 0
 xen/arch/x86/guest/{ => xen}/pvh-boot.c       | 0
 xen/arch/x86/guest/{ => xen}/xen.c            | 0
 7 files changed, 10 insertions(+), 5 deletions(-)
 create mode 100644 xen/arch/x86/guest/xen/Makefile
 rename xen/arch/x86/guest/{ => xen}/hypercall_page.S (100%)
 rename xen/arch/x86/guest/{ => xen}/pvh-boot.c (100%)
 rename xen/arch/x86/guest/{ => xen}/xen.c (100%)

diff --git a/xen/arch/x86/Kconfig b/xen/arch/x86/Kconfig
index 28b3b4692a..867de857e8 100644
--- a/xen/arch/x86/Kconfig
+++ b/xen/arch/x86/Kconfig
@@ -161,8 +161,12 @@ config XEN_ALIGN_2M
 
 endchoice
 
+config GUEST
+	bool
+
 config XEN_GUEST
 	def_bool n
+	select GUEST
 	prompt "Xen Guest"
 	---help---
 	  Support for Xen detecting when it is running under Xen.
diff --git a/xen/arch/x86/Makefile b/xen/arch/x86/Makefile
index 2443fd2cc5..99a12d0090 100644
--- a/xen/arch/x86/Makefile
+++ b/xen/arch/x86/Makefile
@@ -1,7 +1,7 @@
 subdir-y += acpi
 subdir-y += cpu
 subdir-y += genapic
-subdir-$(CONFIG_XEN_GUEST) += guest
+subdir-$(CONFIG_GUEST) += guest
 subdir-$(CONFIG_HVM) += hvm
 subdir-y += mm
 subdir-$(CONFIG_XENOPROF) += oprofile
diff --git a/xen/arch/x86/guest/Makefile b/xen/arch/x86/guest/Makefile
index 26fb4b1007..6806f04947 100644
--- a/xen/arch/x86/guest/Makefile
+++ b/xen/arch/x86/guest/Makefile
@@ -1,4 +1 @@
-obj-y += hypercall_page.o
-obj-y += xen.o
-
-obj-bin-$(CONFIG_PVH_GUEST) += pvh-boot.init.o
+subdir-$(CONFIG_XEN_GUEST) += xen
diff --git a/xen/arch/x86/guest/xen/Makefile b/xen/arch/x86/guest/xen/Makefile
new file mode 100644
index 0000000000..26fb4b1007
--- /dev/null
+++ b/xen/arch/x86/guest/xen/Makefile
@@ -0,0 +1,4 @@
+obj-y += hypercall_page.o
+obj-y += xen.o
+
+obj-bin-$(CONFIG_PVH_GUEST) += pvh-boot.init.o
diff --git a/xen/arch/x86/guest/hypercall_page.S b/xen/arch/x86/guest/xen/hypercall_page.S
similarity index 100%
rename from xen/arch/x86/guest/hypercall_page.S
rename to xen/arch/x86/guest/xen/hypercall_page.S
diff --git a/xen/arch/x86/guest/pvh-boot.c b/xen/arch/x86/guest/xen/pvh-boot.c
similarity index 100%
rename from xen/arch/x86/guest/pvh-boot.c
rename to xen/arch/x86/guest/xen/pvh-boot.c
diff --git a/xen/arch/x86/guest/xen.c b/xen/arch/x86/guest/xen/xen.c
similarity index 100%
rename from xen/arch/x86/guest/xen.c
rename to xen/arch/x86/guest/xen/xen.c
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH for-next v3 2/9] x86: include asm_defns.h directly in hypercall.h

[Wei Liu]

ASM_CALL_CONSTRAINT is defined there.

No functional change.

Signed-off-by: Wei Liu <liuwe@microsoft.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
---
 xen/include/asm-x86/guest/hypercall.h | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/xen/include/asm-x86/guest/hypercall.h b/xen/include/asm-x86/guest/hypercall.h
index d548816b30..c9deca6ffc 100644
--- a/xen/include/asm-x86/guest/hypercall.h
+++ b/xen/include/asm-x86/guest/hypercall.h
@@ -23,6 +23,8 @@
 
 #include <xen/types.h>
 
+#include <asm/asm_defns.h>
+
 #include <public/xen.h>
 #include <public/sched.h>
 #include <public/hvm/hvm_op.h>
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH for-next v3 3/9] x86: drop hypervisor_cpuid_base

[Wei Liu]

The only user is Xen specific code in PV shim. We can therefore export
the variable directly.

Signed-off-by: Wei Liu <liuwe@microsoft.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
---
 xen/arch/x86/guest/xen/xen.c    | 7 +------
 xen/arch/x86/pv/shim.c          | 2 +-
 xen/include/asm-x86/guest/xen.h | 2 +-
 3 files changed, 3 insertions(+), 8 deletions(-)

diff --git a/xen/arch/x86/guest/xen/xen.c b/xen/arch/x86/guest/xen/xen.c
index 7b7a5badab..78fc603996 100644
--- a/xen/arch/x86/guest/xen/xen.c
+++ b/xen/arch/x86/guest/xen/xen.c
@@ -37,7 +37,7 @@
 
 bool __read_mostly xen_guest;
 
-static __read_mostly uint32_t xen_cpuid_base;
+__read_mostly uint32_t xen_cpuid_base;
 extern char hypercall_page[];
 static struct rangeset *mem;
 
@@ -301,11 +301,6 @@ int hypervisor_free_unused_page(mfn_t mfn)
     return rangeset_remove_range(mem, mfn_x(mfn), mfn_x(mfn));
 }
 
-uint32_t hypervisor_cpuid_base(void)
-{
-    return xen_cpuid_base;
-}
-
 static void ap_resume(void *unused)
 {
     map_vcpuinfo();
diff --git a/xen/arch/x86/pv/shim.c b/xen/arch/x86/pv/shim.c
index 5edbcd9ac5..5441e89de2 100644
--- a/xen/arch/x86/pv/shim.c
+++ b/xen/arch/x86/pv/shim.c
@@ -990,7 +990,7 @@ domid_t get_initial_domain_id(void)
     if ( !pv_shim )
         return 0;
 
-    cpuid(hypervisor_cpuid_base() + 4, &eax, &ebx, &ecx, &edx);
+    cpuid(xen_cpuid_base + 4, &eax, &ebx, &ecx, &edx);
 
     return (eax & XEN_HVM_CPUID_DOMID_PRESENT) ? ecx : 1;
 }
diff --git a/xen/include/asm-x86/guest/xen.h b/xen/include/asm-x86/guest/xen.h
index 7e04e4a7ab..b015ed1883 100644
--- a/xen/include/asm-x86/guest/xen.h
+++ b/xen/include/asm-x86/guest/xen.h
@@ -30,13 +30,13 @@
 
 extern bool xen_guest;
 extern bool pv_console;
+extern uint32_t xen_cpuid_base;
 
 void probe_hypervisor(void);
 void hypervisor_setup(void);
 void hypervisor_ap_setup(void);
 int hypervisor_alloc_unused_page(mfn_t *mfn);
 int hypervisor_free_unused_page(mfn_t mfn);
-uint32_t hypervisor_cpuid_base(void);
 void hypervisor_resume(void);
 
 DECLARE_PER_CPU(unsigned int, vcpu_id);
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH for-next v3 4/9] x86: include xen/lib.h in guest/hypercall.h

[Wei Liu]

We need ASSERT_UNREACHABLE.

Signed-off-by: Wei Liu <liuwe@microsoft.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
---
 xen/include/asm-x86/guest/hypercall.h | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/xen/include/asm-x86/guest/hypercall.h b/xen/include/asm-x86/guest/hypercall.h
index c9deca6ffc..d0d2f5022d 100644
--- a/xen/include/asm-x86/guest/hypercall.h
+++ b/xen/include/asm-x86/guest/hypercall.h
@@ -182,6 +182,8 @@ static inline long xen_hypercall_set_evtchn_upcall_vector(
 
 #else /* CONFIG_XEN_GUEST */
 
+#include <xen/lib.h>
+
 #include <public/sched.h>
 
 static inline void xen_hypercall_console_write(
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH for-next v3 5/9] x86: introduce hypervisor framework

[Wei Liu]

We will soon implement Hyper-V support for Xen. Add a framework for
that.

This requires moving some of the hypervisor_* functions from xen.h to
hypervisor.h.

Signed-off-by: Wei Liu <liuwe@microsoft.com>
---
 xen/arch/x86/guest/Makefile            |  2 +
 xen/arch/x86/guest/hypervisor.c        | 45 +++++++++++++++++++
 xen/include/asm-x86/guest.h            |  1 +
 xen/include/asm-x86/guest/hypervisor.h | 61 ++++++++++++++++++++++++++
 xen/include/asm-x86/guest/xen.h        | 12 -----
 5 files changed, 109 insertions(+), 12 deletions(-)
 create mode 100644 xen/arch/x86/guest/hypervisor.c
 create mode 100644 xen/include/asm-x86/guest/hypervisor.h

diff --git a/xen/arch/x86/guest/Makefile b/xen/arch/x86/guest/Makefile
index 6806f04947..f63d64bbee 100644
--- a/xen/arch/x86/guest/Makefile
+++ b/xen/arch/x86/guest/Makefile
@@ -1 +1,3 @@
+obj-y += hypervisor.o
+
 subdir-$(CONFIG_XEN_GUEST) += xen
diff --git a/xen/arch/x86/guest/hypervisor.c b/xen/arch/x86/guest/hypervisor.c
new file mode 100644
index 0000000000..89b9ae4de0
--- /dev/null
+++ b/xen/arch/x86/guest/hypervisor.c
@@ -0,0 +1,45 @@
+/******************************************************************************
+ * arch/x86/guest/hypervisor.c
+ *
+ * Support for detecting and running under a hypervisor.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; If not, see <http://www.gnu.org/licenses/>.
+ *
+ * Copyright (c) 2019 Microsoft.
+ */
+
+#include <xen/types.h>
+
+#include <asm/cache.h>
+#include <asm/guest/hypervisor.h>
+
+static struct hypervisor_ops *hops __read_mostly;
+
+bool hypervisor_probe(void)
+{
+    if ( hops )
+        return true;
+
+    return false;
+}
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/xen/include/asm-x86/guest.h b/xen/include/asm-x86/guest.h
index a38c6b5b3f..8e167165ae 100644
--- a/xen/include/asm-x86/guest.h
+++ b/xen/include/asm-x86/guest.h
@@ -20,6 +20,7 @@
 #define __X86_GUEST_H__
 
 #include <asm/guest/hypercall.h>
+#include <asm/guest/hypervisor.h>
 #include <asm/guest/pvh-boot.h>
 #include <asm/guest/xen.h>
 #include <asm/pv/shim.h>
diff --git a/xen/include/asm-x86/guest/hypervisor.h b/xen/include/asm-x86/guest/hypervisor.h
new file mode 100644
index 0000000000..38344e2e89
--- /dev/null
+++ b/xen/include/asm-x86/guest/hypervisor.h
@@ -0,0 +1,61 @@
+/******************************************************************************
+ * asm-x86/guest/hypervisor.h
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms and conditions of the GNU General Public
+ * License, version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; If not, see <http://www.gnu.org/licenses/>.
+ *
+ * Copyright (c) 2019 Microsoft.
+ */
+
+#ifndef __X86_HYPERVISOR_H__
+#define __X86_HYPERVISOR_H__
+
+#ifdef CONFIG_GUEST
+
+struct hypervisor_ops {
+    /* Name of the hypervisor */
+    const char *name;
+    /* Main setup routine */
+    void (*setup)(void);
+    /* AP setup */
+    void (*ap_setup)(void);
+    /* Resume from suspension */
+    void (*resume)(void);
+};
+
+bool hypervisor_probe(void);
+void hypervisor_setup(void);
+void hypervisor_ap_setup(void);
+void hypervisor_resume(void);
+
+#else
+
+#include <xen/types.h>
+
+static inline bool hypervisor_probe(void) { return false; }
+static inline void hypervisor_setup(void) {}
+static inline void hypervisor_ap_setup(void) {}
+static inline void hypervisor_resume(void) {}
+
+#endif  /* CONFIG_GUEST */
+
+#endif /* __X86_HYPERVISOR_H__ */
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/xen/include/asm-x86/guest/xen.h b/xen/include/asm-x86/guest/xen.h
index b015ed1883..3145f75361 100644
--- a/xen/include/asm-x86/guest/xen.h
+++ b/xen/include/asm-x86/guest/xen.h
@@ -33,11 +33,8 @@ extern bool pv_console;
 extern uint32_t xen_cpuid_base;
 
 void probe_hypervisor(void);
-void hypervisor_setup(void);
-void hypervisor_ap_setup(void);
 int hypervisor_alloc_unused_page(mfn_t *mfn);
 int hypervisor_free_unused_page(mfn_t mfn);
-void hypervisor_resume(void);
 
 DECLARE_PER_CPU(unsigned int, vcpu_id);
 DECLARE_PER_CPU(struct vcpu_info *, vcpu_info);
@@ -49,15 +46,6 @@ DECLARE_PER_CPU(struct vcpu_info *, vcpu_info);
 
 static inline void probe_hypervisor(void) {}
 
-static inline void hypervisor_setup(void)
-{
-    ASSERT_UNREACHABLE();
-}
-static inline void hypervisor_ap_setup(void)
-{
-    ASSERT_UNREACHABLE();
-}
-
 #endif /* CONFIG_XEN_GUEST */
 #endif /* __X86_GUEST_XEN_H__ */
 
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH for-next v3 6/9] x86: rename hypervisor_{alloc, free}_unused_page

[Wei Liu]

They are used in Xen code only.

No functional change.

Signed-off-by: Wei Liu <liuwe@microsoft.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
---
 xen/arch/x86/guest/xen/xen.c    | 6 +++---
 xen/arch/x86/pv/shim.c          | 4 ++--
 xen/include/asm-x86/guest/xen.h | 4 ++--
 3 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/xen/arch/x86/guest/xen/xen.c b/xen/arch/x86/guest/xen/xen.c
index 78fc603996..9895025d02 100644
--- a/xen/arch/x86/guest/xen/xen.c
+++ b/xen/arch/x86/guest/xen/xen.c
@@ -97,7 +97,7 @@ static void map_shared_info(void)
     unsigned int i;
     unsigned long rc;
 
-    if ( hypervisor_alloc_unused_page(&mfn) )
+    if ( xen_alloc_unused_page(&mfn) )
         panic("unable to reserve shared info memory page\n");
 
     xatp.gpfn = mfn_x(mfn);
@@ -284,7 +284,7 @@ void hypervisor_ap_setup(void)
     init_evtchn();
 }
 
-int hypervisor_alloc_unused_page(mfn_t *mfn)
+int xen_alloc_unused_page(mfn_t *mfn)
 {
     unsigned long m;
     int rc;
@@ -296,7 +296,7 @@ int hypervisor_alloc_unused_page(mfn_t *mfn)
     return rc;
 }
 
-int hypervisor_free_unused_page(mfn_t mfn)
+int xen_free_unused_page(mfn_t mfn)
 {
     return rangeset_remove_range(mem, mfn_x(mfn), mfn_x(mfn));
 }
diff --git a/xen/arch/x86/pv/shim.c b/xen/arch/x86/pv/shim.c
index 5441e89de2..41b4665649 100644
--- a/xen/arch/x86/pv/shim.c
+++ b/xen/arch/x86/pv/shim.c
@@ -739,7 +739,7 @@ static long pv_shim_grant_table_op(unsigned int cmd,
                 };
                 mfn_t mfn;
 
-                rc = hypervisor_alloc_unused_page(&mfn);
+                rc = xen_alloc_unused_page(&mfn);
                 if ( rc )
                 {
                     gprintk(XENLOG_ERR,
@@ -751,7 +751,7 @@ static long pv_shim_grant_table_op(unsigned int cmd,
                 rc = xen_hypercall_memory_op(XENMEM_add_to_physmap, &xatp);
                 if ( rc )
                 {
-                    hypervisor_free_unused_page(mfn);
+                    xen_free_unused_page(mfn);
                     break;
                 }
 
diff --git a/xen/include/asm-x86/guest/xen.h b/xen/include/asm-x86/guest/xen.h
index 3145f75361..8221fc1325 100644
--- a/xen/include/asm-x86/guest/xen.h
+++ b/xen/include/asm-x86/guest/xen.h
@@ -33,8 +33,8 @@ extern bool pv_console;
 extern uint32_t xen_cpuid_base;
 
 void probe_hypervisor(void);
-int hypervisor_alloc_unused_page(mfn_t *mfn);
-int hypervisor_free_unused_page(mfn_t mfn);
+int xen_alloc_unused_page(mfn_t *mfn);
+int xen_free_unused_page(mfn_t mfn);
 
 DECLARE_PER_CPU(unsigned int, vcpu_id);
 DECLARE_PER_CPU(struct vcpu_info *, vcpu_info);
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH for-next v3 7/9] x86: switch xen implementation to use hypervisor framework

[Wei Liu]

Take the chance to change probe_hypervisor to hypervisor_probe.

Signed-off-by: Wei Liu <liuwe@microsoft.com>
---
V3:
1. Address Roger's comments
2. Change xen_hypervisor_ops to xen_ops
---
 xen/arch/x86/guest/hypervisor.c   | 32 ++++++++++++++++++++++++++++++-
 xen/arch/x86/guest/xen/pvh-boot.c |  2 +-
 xen/arch/x86/guest/xen/xen.c      | 26 +++++++++++++------------
 xen/arch/x86/setup.c              |  2 +-
 xen/include/asm-x86/guest/xen.h   |  6 ++++--
 5 files changed, 51 insertions(+), 17 deletions(-)

diff --git a/xen/arch/x86/guest/hypervisor.c b/xen/arch/x86/guest/hypervisor.c
index 89b9ae4de0..33bf1a769d 100644
--- a/xen/arch/x86/guest/hypervisor.c
+++ b/xen/arch/x86/guest/hypervisor.c
@@ -22,7 +22,7 @@
 #include <xen/types.h>
 
 #include <asm/cache.h>
-#include <asm/guest/hypervisor.h>
+#include <asm/guest.h>
 
 static struct hypervisor_ops *hops __read_mostly;
 
@@ -31,9 +31,39 @@ bool hypervisor_probe(void)
     if ( hops )
         return true;
 
+    /* Too early to use cpu_has_hypervisor */
+    if ( !(cpuid_ecx(1) & cpufeat_mask(X86_FEATURE_HYPERVISOR)) )
+        return false;
+
+#ifdef CONFIG_XEN_GUEST
+    if ( xen_probe() )
+    {
+        hops = &xen_ops;
+        return true;
+    }
+#endif
+
     return false;
 }
 
+void hypervisor_setup(void)
+{
+    if ( hops && hops->setup )
+        hops->setup();
+}
+
+void hypervisor_ap_setup(void)
+{
+    if ( hops && hops->ap_setup )
+        hops->ap_setup();
+}
+
+void hypervisor_resume(void)
+{
+    if ( hops && hops->resume )
+        hops->resume();
+}
+
 /*
  * Local variables:
  * mode: C
diff --git a/xen/arch/x86/guest/xen/pvh-boot.c b/xen/arch/x86/guest/xen/pvh-boot.c
index ca8e156f7d..498625eae0 100644
--- a/xen/arch/x86/guest/xen/pvh-boot.c
+++ b/xen/arch/x86/guest/xen/pvh-boot.c
@@ -103,7 +103,7 @@ void __init pvh_init(multiboot_info_t **mbi, module_t **mod)
 {
     convert_pvh_info(mbi, mod);
 
-    probe_hypervisor();
+    hypervisor_probe();
     ASSERT(xen_guest);
 
     get_memory_map();
diff --git a/xen/arch/x86/guest/xen/xen.c b/xen/arch/x86/guest/xen/xen.c
index 9895025d02..655435c1f7 100644
--- a/xen/arch/x86/guest/xen/xen.c
+++ b/xen/arch/x86/guest/xen/xen.c
@@ -67,24 +67,19 @@ static void __init find_xen_leaves(void)
     }
 }
 
-void __init probe_hypervisor(void)
+bool __init xen_probe(void)
 {
-    if ( xen_guest )
-        return;
-
-    /* Too early to use cpu_has_hypervisor */
-    if ( !(cpuid_ecx(1) & cpufeat_mask(X86_FEATURE_HYPERVISOR)) )
-        return;
-
     find_xen_leaves();
 
     if ( !xen_cpuid_base )
-        return;
+        return false;
 
     /* Fill the hypercall page. */
     wrmsrl(cpuid_ebx(xen_cpuid_base + 2), __pa(hypercall_page));
 
     xen_guest = true;
+
+    return true;
 }
 
 static void map_shared_info(void)
@@ -249,7 +244,7 @@ static void init_evtchn(void)
     }
 }
 
-void __init hypervisor_setup(void)
+static void __init xen_setup(void)
 {
     init_memmap();
 
@@ -277,7 +272,7 @@ void __init hypervisor_setup(void)
     init_evtchn();
 }
 
-void hypervisor_ap_setup(void)
+static void xen_ap_setup(void)
 {
     set_vcpu_id();
     map_vcpuinfo();
@@ -307,7 +302,7 @@ static void ap_resume(void *unused)
     init_evtchn();
 }
 
-void hypervisor_resume(void)
+static void xen_resume(void)
 {
     /* Reset shared info page. */
     map_shared_info();
@@ -330,6 +325,13 @@ void hypervisor_resume(void)
         pv_console_init();
 }
 
+struct hypervisor_ops xen_ops = {
+    .name = "Xen",
+    .setup = xen_setup,
+    .ap_setup = xen_ap_setup,
+    .resume = xen_resume,
+};
+
 /*
  * Local variables:
  * mode: C
diff --git a/xen/arch/x86/setup.c b/xen/arch/x86/setup.c
index dec60d0301..0ee11b15a6 100644
--- a/xen/arch/x86/setup.c
+++ b/xen/arch/x86/setup.c
@@ -763,7 +763,7 @@ void __init noreturn __start_xen(unsigned long mbi_p)
      * allocing any xenheap structures wanted in lower memory. */
     kexec_early_calculations();
 
-    probe_hypervisor();
+    hypervisor_probe();
 
     parse_video_info();
 
diff --git a/xen/include/asm-x86/guest/xen.h b/xen/include/asm-x86/guest/xen.h
index 8221fc1325..c3c6bea24f 100644
--- a/xen/include/asm-x86/guest/xen.h
+++ b/xen/include/asm-x86/guest/xen.h
@@ -23,6 +23,7 @@
 
 #include <asm/e820.h>
 #include <asm/fixmap.h>
+#include <asm/guest/hypervisor.h>
 
 #define XEN_shared_info ((struct shared_info *)fix_to_virt(FIX_XEN_SHARED_INFO))
 
@@ -31,8 +32,9 @@
 extern bool xen_guest;
 extern bool pv_console;
 extern uint32_t xen_cpuid_base;
+extern struct hypervisor_ops xen_ops;
 
-void probe_hypervisor(void);
+bool xen_probe(void);
 int xen_alloc_unused_page(mfn_t *mfn);
 int xen_free_unused_page(mfn_t mfn);
 
@@ -44,7 +46,7 @@ DECLARE_PER_CPU(struct vcpu_info *, vcpu_info);
 #define xen_guest 0
 #define pv_console 0
 
-static inline void probe_hypervisor(void) {}
+static inline bool xen_probe(void) { return false; }
 
 #endif /* CONFIG_XEN_GUEST */
 #endif /* __X86_GUEST_XEN_H__ */
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH for-next v3 8/9] x86: be more verbose when running on a hypervisor

[Wei Liu]

Signed-off-by: Wei Liu <liuwe@microsoft.com>
---
V3: Address Roger's comment, add ASSERTs
---
 xen/arch/x86/guest/hypervisor.c        | 6 ++++++
 xen/arch/x86/setup.c                   | 6 +++++-
 xen/include/asm-x86/guest/hypervisor.h | 3 +++
 3 files changed, 14 insertions(+), 1 deletion(-)

diff --git a/xen/arch/x86/guest/hypervisor.c b/xen/arch/x86/guest/hypervisor.c
index 33bf1a769d..a666ad9526 100644
--- a/xen/arch/x86/guest/hypervisor.c
+++ b/xen/arch/x86/guest/hypervisor.c
@@ -46,6 +46,12 @@ bool hypervisor_probe(void)
     return false;
 }
 
+const char *hypervisor_name(void)
+{
+    ASSERT(hops);
+    return hops->name;
+}
+
 void hypervisor_setup(void)
 {
     if ( hops && hops->setup )
diff --git a/xen/arch/x86/setup.c b/xen/arch/x86/setup.c
index 0ee11b15a6..cf5a7b8e1e 100644
--- a/xen/arch/x86/setup.c
+++ b/xen/arch/x86/setup.c
@@ -689,6 +689,7 @@ void __init noreturn __start_xen(unsigned long mbi_p)
     int i, j, e820_warn = 0, bytes = 0;
     bool acpi_boot_table_init_done = false, relocated = false;
     int ret;
+    bool running_on_hypervisor;
     struct ns16550_defaults ns16550 = {
         .data_bits = 8,
         .parity    = 'n',
@@ -763,7 +764,7 @@ void __init noreturn __start_xen(unsigned long mbi_p)
      * allocing any xenheap structures wanted in lower memory. */
     kexec_early_calculations();
 
-    hypervisor_probe();
+    running_on_hypervisor = hypervisor_probe();
 
     parse_video_info();
 
@@ -789,6 +790,9 @@ void __init noreturn __start_xen(unsigned long mbi_p)
 
     printk("Xen image load base address: %#lx\n", xen_phys_start);
 
+    if ( running_on_hypervisor )
+        printk("Running on %s\n", hypervisor_name());
+
 #ifdef CONFIG_VIDEO
     printk("Video information:\n");
 
diff --git a/xen/include/asm-x86/guest/hypervisor.h b/xen/include/asm-x86/guest/hypervisor.h
index 38344e2e89..b583722f5d 100644
--- a/xen/include/asm-x86/guest/hypervisor.h
+++ b/xen/include/asm-x86/guest/hypervisor.h
@@ -36,15 +36,18 @@ bool hypervisor_probe(void);
 void hypervisor_setup(void);
 void hypervisor_ap_setup(void);
 void hypervisor_resume(void);
+const char *hypervisor_name(void);
 
 #else
 
+#include <xen/lib.h>
 #include <xen/types.h>
 
 static inline bool hypervisor_probe(void) { return false; }
 static inline void hypervisor_setup(void) {}
 static inline void hypervisor_ap_setup(void) {}
 static inline void hypervisor_resume(void) {}
+static inline char *hypervisor_name(void) { ASSERT_UNREACHABLE(); return NULL; }
 
 #endif  /* CONFIG_GUEST */
 
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

[Xen-devel] [PATCH for-next v3 9/9] x86: introduce CONFIG_HYPERV and detection code

[Wei Liu]

We use the same code structure as we did for Xen.

As starters, detect Hyper-V in probe routine. More complex
functionalities will be added later.

Signed-off-by: Wei Liu <liuwe@microsoft.com>
---
V3:
1. Remove some unused code
2. Rename structure
3. Also detect HV#1 signature
---
 xen/arch/x86/Kconfig               |  9 +++++
 xen/arch/x86/guest/Makefile        |  1 +
 xen/arch/x86/guest/hyperv/Makefile |  1 +
 xen/arch/x86/guest/hyperv/hyperv.c | 54 ++++++++++++++++++++++++++++++
 xen/arch/x86/guest/hypervisor.c    |  8 +++++
 xen/include/asm-x86/guest.h        |  1 +
 xen/include/asm-x86/guest/hyperv.h | 45 +++++++++++++++++++++++++
 7 files changed, 119 insertions(+)
 create mode 100644 xen/arch/x86/guest/hyperv/Makefile
 create mode 100644 xen/arch/x86/guest/hyperv/hyperv.c
 create mode 100644 xen/include/asm-x86/guest/hyperv.h

diff --git a/xen/arch/x86/Kconfig b/xen/arch/x86/Kconfig
index 867de857e8..56513c771c 100644
--- a/xen/arch/x86/Kconfig
+++ b/xen/arch/x86/Kconfig
@@ -164,6 +164,15 @@ endchoice
 config GUEST
 	bool
 
+config HYPERV_GUEST
+	def_bool n
+	select GUEST
+	prompt "Hyper-V Guest"
+	---help---
+	  Support for Xen detecting when it is running under Hyper-V.
+
+	  If unsure, say N.
+
 config XEN_GUEST
 	def_bool n
 	select GUEST
diff --git a/xen/arch/x86/guest/Makefile b/xen/arch/x86/guest/Makefile
index f63d64bbee..f164196772 100644
--- a/xen/arch/x86/guest/Makefile
+++ b/xen/arch/x86/guest/Makefile
@@ -1,3 +1,4 @@
 obj-y += hypervisor.o
 
+subdir-$(CONFIG_HYPERV_GUEST) += hyperv
 subdir-$(CONFIG_XEN_GUEST) += xen
diff --git a/xen/arch/x86/guest/hyperv/Makefile b/xen/arch/x86/guest/hyperv/Makefile
new file mode 100644
index 0000000000..68170109a9
--- /dev/null
+++ b/xen/arch/x86/guest/hyperv/Makefile
@@ -0,0 +1 @@
+obj-y += hyperv.o
diff --git a/xen/arch/x86/guest/hyperv/hyperv.c b/xen/arch/x86/guest/hyperv/hyperv.c
new file mode 100644
index 0000000000..7ab4b127f3
--- /dev/null
+++ b/xen/arch/x86/guest/hyperv/hyperv.c
@@ -0,0 +1,54 @@
+/******************************************************************************
+ * arch/x86/guest/hyperv/hyperv.c
+ *
+ * Support for detecting and running under Hyper-V.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; If not, see <http://www.gnu.org/licenses/>.
+ *
+ * Copyright (c) 2019 Microsoft.
+ */
+#include <xen/init.h>
+
+#include <asm/guest.h>
+
+bool __init hyperv_probe(void)
+{
+    uint32_t eax, ebx, ecx, edx;
+
+    cpuid(0x40000000, &eax, &ebx, &ecx, &edx);
+    if ( !((ebx == 0x7263694d) &&  /* "Micr" */
+           (ecx == 0x666f736f) &&  /* "osof" */
+           (edx == 0x76482074)) )  /* "t Hv" */
+        return false;
+
+    cpuid(0x40000001, &eax, &ebx, &ecx, &edx);
+    if ( eax != 0x31237648 )    /* Hv#1 */
+        return false;
+
+    return true;
+}
+
+struct hypervisor_ops hyperv_ops = {
+    .name = "Hyper-V",
+};
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/xen/arch/x86/guest/hypervisor.c b/xen/arch/x86/guest/hypervisor.c
index a666ad9526..17392d1ffa 100644
--- a/xen/arch/x86/guest/hypervisor.c
+++ b/xen/arch/x86/guest/hypervisor.c
@@ -43,6 +43,14 @@ bool hypervisor_probe(void)
     }
 #endif
 
+#ifdef CONFIG_HYPERV_GUEST
+    if ( hyperv_probe() )
+    {
+        hops = &hyperv_ops;
+        return true;
+    }
+#endif
+
     return false;
 }
 
diff --git a/xen/include/asm-x86/guest.h b/xen/include/asm-x86/guest.h
index 8e167165ae..94448606d4 100644
--- a/xen/include/asm-x86/guest.h
+++ b/xen/include/asm-x86/guest.h
@@ -20,6 +20,7 @@
 #define __X86_GUEST_H__
 
 #include <asm/guest/hypercall.h>
+#include <asm/guest/hyperv.h>
 #include <asm/guest/hypervisor.h>
 #include <asm/guest/pvh-boot.h>
 #include <asm/guest/xen.h>
diff --git a/xen/include/asm-x86/guest/hyperv.h b/xen/include/asm-x86/guest/hyperv.h
new file mode 100644
index 0000000000..4b9cc5a836
--- /dev/null
+++ b/xen/include/asm-x86/guest/hyperv.h
@@ -0,0 +1,45 @@
+/******************************************************************************
+ * asm-x86/guest/hyperv.h
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms and conditions of the GNU General Public
+ * License, version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; If not, see <http://www.gnu.org/licenses/>.
+ *
+ * Copyright (c) 2019 Microsoft.
+ */
+
+#ifndef __X86_GUEST_HYPERV_H__
+#define __X86_GUEST_HYPERV_H__
+
+#ifdef CONFIG_HYPERV_GUEST
+
+#include <asm/guest/hypervisor.h>
+
+extern struct hypervisor_ops hyperv_ops;
+
+bool hyperv_probe(void);
+
+#else
+
+static inline bool hyperv_probe(void) { return false; }
+
+#endif /* CONFIG_HYPERV_GUEST */
+#endif /* __X86_GUEST_HYPERV_H__ */
+
+/*
+ * Local variables:
+ * mode: C
+ * c-file-style: "BSD"
+ * c-basic-offset: 4
+ * tab-width: 4
+ * indent-tabs-mode: nil
+ * End:
+ */
-- 
2.20.1


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 1/9] x86: introduce CONFIG_GUEST and move code

[Paul Durrant]

On Mon, 21 Oct 2019 at 16:59, Wei Liu <wl@xen.org> wrote:
>
> Xen is able to run as a guest on Xen. We plan to make it able to run
> on Hyper-V as well.
>
> Introduce CONFIG_GUEST which is set to true if either running on Xen
> or Hyper-V is desired. Restructure code hierarchy for new code to
> come.
>
> No functional change intended.
>
> Signed-off-by: Wei Liu <liuwe@microsoft.com>
> Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
> ---
>  xen/arch/x86/Kconfig                          | 4 ++++
>  xen/arch/x86/Makefile                         | 2 +-
>  xen/arch/x86/guest/Makefile                   | 5 +----
>  xen/arch/x86/guest/xen/Makefile               | 4 ++++
>  xen/arch/x86/guest/{ => xen}/hypercall_page.S | 0
>  xen/arch/x86/guest/{ => xen}/pvh-boot.c       | 0
>  xen/arch/x86/guest/{ => xen}/xen.c            | 0
>  7 files changed, 10 insertions(+), 5 deletions(-)
>  create mode 100644 xen/arch/x86/guest/xen/Makefile
>  rename xen/arch/x86/guest/{ => xen}/hypercall_page.S (100%)
>  rename xen/arch/x86/guest/{ => xen}/pvh-boot.c (100%)
>  rename xen/arch/x86/guest/{ => xen}/xen.c (100%)
>
> diff --git a/xen/arch/x86/Kconfig b/xen/arch/x86/Kconfig
> index 28b3b4692a..867de857e8 100644
> --- a/xen/arch/x86/Kconfig
> +++ b/xen/arch/x86/Kconfig
> @@ -161,8 +161,12 @@ config XEN_ALIGN_2M
>
>  endchoice
>
> +config GUEST
> +       bool
> +
>  config XEN_GUEST
>         def_bool n
> +       select GUEST
>         prompt "Xen Guest"
>         ---help---
>           Support for Xen detecting when it is running under Xen.

This text needs to be modified. Perhaps just say 'when it is running
in a virtual machine'. That should be future-proof.

  Paul

> diff --git a/xen/arch/x86/Makefile b/xen/arch/x86/Makefile
> index 2443fd2cc5..99a12d0090 100644
> --- a/xen/arch/x86/Makefile
> +++ b/xen/arch/x86/Makefile
> @@ -1,7 +1,7 @@
>  subdir-y += acpi
>  subdir-y += cpu
>  subdir-y += genapic
> -subdir-$(CONFIG_XEN_GUEST) += guest
> +subdir-$(CONFIG_GUEST) += guest
>  subdir-$(CONFIG_HVM) += hvm
>  subdir-y += mm
>  subdir-$(CONFIG_XENOPROF) += oprofile
> diff --git a/xen/arch/x86/guest/Makefile b/xen/arch/x86/guest/Makefile
> index 26fb4b1007..6806f04947 100644
> --- a/xen/arch/x86/guest/Makefile
> +++ b/xen/arch/x86/guest/Makefile
> @@ -1,4 +1 @@
> -obj-y += hypercall_page.o
> -obj-y += xen.o
> -
> -obj-bin-$(CONFIG_PVH_GUEST) += pvh-boot.init.o
> +subdir-$(CONFIG_XEN_GUEST) += xen
> diff --git a/xen/arch/x86/guest/xen/Makefile b/xen/arch/x86/guest/xen/Makefile
> new file mode 100644
> index 0000000000..26fb4b1007
> --- /dev/null
> +++ b/xen/arch/x86/guest/xen/Makefile
> @@ -0,0 +1,4 @@
> +obj-y += hypercall_page.o
> +obj-y += xen.o
> +
> +obj-bin-$(CONFIG_PVH_GUEST) += pvh-boot.init.o
> diff --git a/xen/arch/x86/guest/hypercall_page.S b/xen/arch/x86/guest/xen/hypercall_page.S
> similarity index 100%
> rename from xen/arch/x86/guest/hypercall_page.S
> rename to xen/arch/x86/guest/xen/hypercall_page.S
> diff --git a/xen/arch/x86/guest/pvh-boot.c b/xen/arch/x86/guest/xen/pvh-boot.c
> similarity index 100%
> rename from xen/arch/x86/guest/pvh-boot.c
> rename to xen/arch/x86/guest/xen/pvh-boot.c
> diff --git a/xen/arch/x86/guest/xen.c b/xen/arch/x86/guest/xen/xen.c
> similarity index 100%
> rename from xen/arch/x86/guest/xen.c
> rename to xen/arch/x86/guest/xen/xen.c
> --
> 2.20.1
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xenproject.org
> https://lists.xenproject.org/mailman/listinfo/xen-devel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 5/9] x86: introduce hypervisor framework

[Paul Durrant]

On Mon, 21 Oct 2019 at 17:01, Wei Liu <wl@xen.org> wrote:
>
> We will soon implement Hyper-V support for Xen. Add a framework for
> that.
>
> This requires moving some of the hypervisor_* functions from xen.h to
> hypervisor.h.
>
> Signed-off-by: Wei Liu <liuwe@microsoft.com>

Reviewed-by: Paul Durrant <paul@xen.org>

> ---
>  xen/arch/x86/guest/Makefile            |  2 +
>  xen/arch/x86/guest/hypervisor.c        | 45 +++++++++++++++++++
>  xen/include/asm-x86/guest.h            |  1 +
>  xen/include/asm-x86/guest/hypervisor.h | 61 ++++++++++++++++++++++++++
>  xen/include/asm-x86/guest/xen.h        | 12 -----
>  5 files changed, 109 insertions(+), 12 deletions(-)
>  create mode 100644 xen/arch/x86/guest/hypervisor.c
>  create mode 100644 xen/include/asm-x86/guest/hypervisor.h
>
> diff --git a/xen/arch/x86/guest/Makefile b/xen/arch/x86/guest/Makefile
> index 6806f04947..f63d64bbee 100644
> --- a/xen/arch/x86/guest/Makefile
> +++ b/xen/arch/x86/guest/Makefile
> @@ -1 +1,3 @@
> +obj-y += hypervisor.o
> +
>  subdir-$(CONFIG_XEN_GUEST) += xen
> diff --git a/xen/arch/x86/guest/hypervisor.c b/xen/arch/x86/guest/hypervisor.c
> new file mode 100644
> index 0000000000..89b9ae4de0
> --- /dev/null
> +++ b/xen/arch/x86/guest/hypervisor.c
> @@ -0,0 +1,45 @@
> +/******************************************************************************
> + * arch/x86/guest/hypervisor.c
> + *
> + * Support for detecting and running under a hypervisor.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; If not, see <http://www.gnu.org/licenses/>.
> + *
> + * Copyright (c) 2019 Microsoft.
> + */
> +
> +#include <xen/types.h>
> +
> +#include <asm/cache.h>
> +#include <asm/guest/hypervisor.h>
> +
> +static struct hypervisor_ops *hops __read_mostly;
> +
> +bool hypervisor_probe(void)
> +{
> +    if ( hops )
> +        return true;
> +
> +    return false;
> +}
> +
> +/*
> + * Local variables:
> + * mode: C
> + * c-file-style: "BSD"
> + * c-basic-offset: 4
> + * tab-width: 4
> + * indent-tabs-mode: nil
> + * End:
> + */
> diff --git a/xen/include/asm-x86/guest.h b/xen/include/asm-x86/guest.h
> index a38c6b5b3f..8e167165ae 100644
> --- a/xen/include/asm-x86/guest.h
> +++ b/xen/include/asm-x86/guest.h
> @@ -20,6 +20,7 @@
>  #define __X86_GUEST_H__
>
>  #include <asm/guest/hypercall.h>
> +#include <asm/guest/hypervisor.h>
>  #include <asm/guest/pvh-boot.h>
>  #include <asm/guest/xen.h>
>  #include <asm/pv/shim.h>
> diff --git a/xen/include/asm-x86/guest/hypervisor.h b/xen/include/asm-x86/guest/hypervisor.h
> new file mode 100644
> index 0000000000..38344e2e89
> --- /dev/null
> +++ b/xen/include/asm-x86/guest/hypervisor.h
> @@ -0,0 +1,61 @@
> +/******************************************************************************
> + * asm-x86/guest/hypervisor.h
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms and conditions of the GNU General Public
> + * License, version 2, as published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> + * General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public
> + * License along with this program; If not, see <http://www.gnu.org/licenses/>.
> + *
> + * Copyright (c) 2019 Microsoft.
> + */
> +
> +#ifndef __X86_HYPERVISOR_H__
> +#define __X86_HYPERVISOR_H__
> +
> +#ifdef CONFIG_GUEST
> +
> +struct hypervisor_ops {
> +    /* Name of the hypervisor */
> +    const char *name;
> +    /* Main setup routine */
> +    void (*setup)(void);
> +    /* AP setup */
> +    void (*ap_setup)(void);
> +    /* Resume from suspension */
> +    void (*resume)(void);
> +};
> +
> +bool hypervisor_probe(void);
> +void hypervisor_setup(void);
> +void hypervisor_ap_setup(void);
> +void hypervisor_resume(void);
> +
> +#else
> +
> +#include <xen/types.h>
> +
> +static inline bool hypervisor_probe(void) { return false; }
> +static inline void hypervisor_setup(void) {}
> +static inline void hypervisor_ap_setup(void) {}
> +static inline void hypervisor_resume(void) {}
> +
> +#endif  /* CONFIG_GUEST */
> +
> +#endif /* __X86_HYPERVISOR_H__ */
> +
> +/*
> + * Local variables:
> + * mode: C
> + * c-file-style: "BSD"
> + * c-basic-offset: 4
> + * tab-width: 4
> + * indent-tabs-mode: nil
> + * End:
> + */
> diff --git a/xen/include/asm-x86/guest/xen.h b/xen/include/asm-x86/guest/xen.h
> index b015ed1883..3145f75361 100644
> --- a/xen/include/asm-x86/guest/xen.h
> +++ b/xen/include/asm-x86/guest/xen.h
> @@ -33,11 +33,8 @@ extern bool pv_console;
>  extern uint32_t xen_cpuid_base;
>
>  void probe_hypervisor(void);
> -void hypervisor_setup(void);
> -void hypervisor_ap_setup(void);
>  int hypervisor_alloc_unused_page(mfn_t *mfn);
>  int hypervisor_free_unused_page(mfn_t mfn);
> -void hypervisor_resume(void);
>
>  DECLARE_PER_CPU(unsigned int, vcpu_id);
>  DECLARE_PER_CPU(struct vcpu_info *, vcpu_info);
> @@ -49,15 +46,6 @@ DECLARE_PER_CPU(struct vcpu_info *, vcpu_info);
>
>  static inline void probe_hypervisor(void) {}
>
> -static inline void hypervisor_setup(void)
> -{
> -    ASSERT_UNREACHABLE();
> -}
> -static inline void hypervisor_ap_setup(void)
> -{
> -    ASSERT_UNREACHABLE();
> -}
> -
>  #endif /* CONFIG_XEN_GUEST */
>  #endif /* __X86_GUEST_XEN_H__ */
>
> --
> 2.20.1
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xenproject.org
> https://lists.xenproject.org/mailman/listinfo/xen-devel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 7/9] x86: switch xen implementation to use hypervisor framework

[Paul Durrant]

On Mon, 21 Oct 2019 at 17:00, Wei Liu <wl@xen.org> wrote:
>
> Take the chance to change probe_hypervisor to hypervisor_probe.
>
> Signed-off-by: Wei Liu <liuwe@microsoft.com>

Reviewed-by: Paul Durrant <paul@xen.org>

> ---
> V3:
> 1. Address Roger's comments
> 2. Change xen_hypervisor_ops to xen_ops
> ---
>  xen/arch/x86/guest/hypervisor.c   | 32 ++++++++++++++++++++++++++++++-
>  xen/arch/x86/guest/xen/pvh-boot.c |  2 +-
>  xen/arch/x86/guest/xen/xen.c      | 26 +++++++++++++------------
>  xen/arch/x86/setup.c              |  2 +-
>  xen/include/asm-x86/guest/xen.h   |  6 ++++--
>  5 files changed, 51 insertions(+), 17 deletions(-)
>
> diff --git a/xen/arch/x86/guest/hypervisor.c b/xen/arch/x86/guest/hypervisor.c
> index 89b9ae4de0..33bf1a769d 100644
> --- a/xen/arch/x86/guest/hypervisor.c
> +++ b/xen/arch/x86/guest/hypervisor.c
> @@ -22,7 +22,7 @@
>  #include <xen/types.h>
>
>  #include <asm/cache.h>
> -#include <asm/guest/hypervisor.h>
> +#include <asm/guest.h>
>
>  static struct hypervisor_ops *hops __read_mostly;
>
> @@ -31,9 +31,39 @@ bool hypervisor_probe(void)
>      if ( hops )
>          return true;
>
> +    /* Too early to use cpu_has_hypervisor */
> +    if ( !(cpuid_ecx(1) & cpufeat_mask(X86_FEATURE_HYPERVISOR)) )
> +        return false;
> +
> +#ifdef CONFIG_XEN_GUEST
> +    if ( xen_probe() )
> +    {
> +        hops = &xen_ops;
> +        return true;
> +    }
> +#endif
> +
>      return false;
>  }
>
> +void hypervisor_setup(void)
> +{
> +    if ( hops && hops->setup )
> +        hops->setup();
> +}
> +
> +void hypervisor_ap_setup(void)
> +{
> +    if ( hops && hops->ap_setup )
> +        hops->ap_setup();
> +}
> +
> +void hypervisor_resume(void)
> +{
> +    if ( hops && hops->resume )
> +        hops->resume();
> +}
> +
>  /*
>   * Local variables:
>   * mode: C
> diff --git a/xen/arch/x86/guest/xen/pvh-boot.c b/xen/arch/x86/guest/xen/pvh-boot.c
> index ca8e156f7d..498625eae0 100644
> --- a/xen/arch/x86/guest/xen/pvh-boot.c
> +++ b/xen/arch/x86/guest/xen/pvh-boot.c
> @@ -103,7 +103,7 @@ void __init pvh_init(multiboot_info_t **mbi, module_t **mod)
>  {
>      convert_pvh_info(mbi, mod);
>
> -    probe_hypervisor();
> +    hypervisor_probe();
>      ASSERT(xen_guest);
>
>      get_memory_map();
> diff --git a/xen/arch/x86/guest/xen/xen.c b/xen/arch/x86/guest/xen/xen.c
> index 9895025d02..655435c1f7 100644
> --- a/xen/arch/x86/guest/xen/xen.c
> +++ b/xen/arch/x86/guest/xen/xen.c
> @@ -67,24 +67,19 @@ static void __init find_xen_leaves(void)
>      }
>  }
>
> -void __init probe_hypervisor(void)
> +bool __init xen_probe(void)
>  {
> -    if ( xen_guest )
> -        return;
> -
> -    /* Too early to use cpu_has_hypervisor */
> -    if ( !(cpuid_ecx(1) & cpufeat_mask(X86_FEATURE_HYPERVISOR)) )
> -        return;
> -
>      find_xen_leaves();
>
>      if ( !xen_cpuid_base )
> -        return;
> +        return false;
>
>      /* Fill the hypercall page. */
>      wrmsrl(cpuid_ebx(xen_cpuid_base + 2), __pa(hypercall_page));
>
>      xen_guest = true;
> +
> +    return true;
>  }
>
>  static void map_shared_info(void)
> @@ -249,7 +244,7 @@ static void init_evtchn(void)
>      }
>  }
>
> -void __init hypervisor_setup(void)
> +static void __init xen_setup(void)
>  {
>      init_memmap();
>
> @@ -277,7 +272,7 @@ void __init hypervisor_setup(void)
>      init_evtchn();
>  }
>
> -void hypervisor_ap_setup(void)
> +static void xen_ap_setup(void)
>  {
>      set_vcpu_id();
>      map_vcpuinfo();
> @@ -307,7 +302,7 @@ static void ap_resume(void *unused)
>      init_evtchn();
>  }
>
> -void hypervisor_resume(void)
> +static void xen_resume(void)
>  {
>      /* Reset shared info page. */
>      map_shared_info();
> @@ -330,6 +325,13 @@ void hypervisor_resume(void)
>          pv_console_init();
>  }
>
> +struct hypervisor_ops xen_ops = {
> +    .name = "Xen",
> +    .setup = xen_setup,
> +    .ap_setup = xen_ap_setup,
> +    .resume = xen_resume,
> +};
> +
>  /*
>   * Local variables:
>   * mode: C
> diff --git a/xen/arch/x86/setup.c b/xen/arch/x86/setup.c
> index dec60d0301..0ee11b15a6 100644
> --- a/xen/arch/x86/setup.c
> +++ b/xen/arch/x86/setup.c
> @@ -763,7 +763,7 @@ void __init noreturn __start_xen(unsigned long mbi_p)
>       * allocing any xenheap structures wanted in lower memory. */
>      kexec_early_calculations();
>
> -    probe_hypervisor();
> +    hypervisor_probe();
>
>      parse_video_info();
>
> diff --git a/xen/include/asm-x86/guest/xen.h b/xen/include/asm-x86/guest/xen.h
> index 8221fc1325..c3c6bea24f 100644
> --- a/xen/include/asm-x86/guest/xen.h
> +++ b/xen/include/asm-x86/guest/xen.h
> @@ -23,6 +23,7 @@
>
>  #include <asm/e820.h>
>  #include <asm/fixmap.h>
> +#include <asm/guest/hypervisor.h>
>
>  #define XEN_shared_info ((struct shared_info *)fix_to_virt(FIX_XEN_SHARED_INFO))
>
> @@ -31,8 +32,9 @@
>  extern bool xen_guest;
>  extern bool pv_console;
>  extern uint32_t xen_cpuid_base;
> +extern struct hypervisor_ops xen_ops;
>
> -void probe_hypervisor(void);
> +bool xen_probe(void);
>  int xen_alloc_unused_page(mfn_t *mfn);
>  int xen_free_unused_page(mfn_t mfn);
>
> @@ -44,7 +46,7 @@ DECLARE_PER_CPU(struct vcpu_info *, vcpu_info);
>  #define xen_guest 0
>  #define pv_console 0
>
> -static inline void probe_hypervisor(void) {}
> +static inline bool xen_probe(void) { return false; }
>
>  #endif /* CONFIG_XEN_GUEST */
>  #endif /* __X86_GUEST_XEN_H__ */
> --
> 2.20.1
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xenproject.org
> https://lists.xenproject.org/mailman/listinfo/xen-devel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 8/9] x86: be more verbose when running on a hypervisor

[Paul Durrant]

On Mon, 21 Oct 2019 at 17:00, Wei Liu <wl@xen.org> wrote:
>
> Signed-off-by: Wei Liu <liuwe@microsoft.com>

Reviewed-by: Paul Durrant <paul@xen.org>

> ---
> V3: Address Roger's comment, add ASSERTs
> ---
>  xen/arch/x86/guest/hypervisor.c        | 6 ++++++
>  xen/arch/x86/setup.c                   | 6 +++++-
>  xen/include/asm-x86/guest/hypervisor.h | 3 +++
>  3 files changed, 14 insertions(+), 1 deletion(-)
>
> diff --git a/xen/arch/x86/guest/hypervisor.c b/xen/arch/x86/guest/hypervisor.c
> index 33bf1a769d..a666ad9526 100644
> --- a/xen/arch/x86/guest/hypervisor.c
> +++ b/xen/arch/x86/guest/hypervisor.c
> @@ -46,6 +46,12 @@ bool hypervisor_probe(void)
>      return false;
>  }
>
> +const char *hypervisor_name(void)
> +{
> +    ASSERT(hops);
> +    return hops->name;
> +}
> +
>  void hypervisor_setup(void)
>  {
>      if ( hops && hops->setup )
> diff --git a/xen/arch/x86/setup.c b/xen/arch/x86/setup.c
> index 0ee11b15a6..cf5a7b8e1e 100644
> --- a/xen/arch/x86/setup.c
> +++ b/xen/arch/x86/setup.c
> @@ -689,6 +689,7 @@ void __init noreturn __start_xen(unsigned long mbi_p)
>      int i, j, e820_warn = 0, bytes = 0;
>      bool acpi_boot_table_init_done = false, relocated = false;
>      int ret;
> +    bool running_on_hypervisor;
>      struct ns16550_defaults ns16550 = {
>          .data_bits = 8,
>          .parity    = 'n',
> @@ -763,7 +764,7 @@ void __init noreturn __start_xen(unsigned long mbi_p)
>       * allocing any xenheap structures wanted in lower memory. */
>      kexec_early_calculations();
>
> -    hypervisor_probe();
> +    running_on_hypervisor = hypervisor_probe();
>
>      parse_video_info();
>
> @@ -789,6 +790,9 @@ void __init noreturn __start_xen(unsigned long mbi_p)
>
>      printk("Xen image load base address: %#lx\n", xen_phys_start);
>
> +    if ( running_on_hypervisor )
> +        printk("Running on %s\n", hypervisor_name());
> +
>  #ifdef CONFIG_VIDEO
>      printk("Video information:\n");
>
> diff --git a/xen/include/asm-x86/guest/hypervisor.h b/xen/include/asm-x86/guest/hypervisor.h
> index 38344e2e89..b583722f5d 100644
> --- a/xen/include/asm-x86/guest/hypervisor.h
> +++ b/xen/include/asm-x86/guest/hypervisor.h
> @@ -36,15 +36,18 @@ bool hypervisor_probe(void);
>  void hypervisor_setup(void);
>  void hypervisor_ap_setup(void);
>  void hypervisor_resume(void);
> +const char *hypervisor_name(void);
>
>  #else
>
> +#include <xen/lib.h>
>  #include <xen/types.h>
>
>  static inline bool hypervisor_probe(void) { return false; }
>  static inline void hypervisor_setup(void) {}
>  static inline void hypervisor_ap_setup(void) {}
>  static inline void hypervisor_resume(void) {}
> +static inline char *hypervisor_name(void) { ASSERT_UNREACHABLE(); return NULL; }
>
>  #endif  /* CONFIG_GUEST */
>
> --
> 2.20.1
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xenproject.org
> https://lists.xenproject.org/mailman/listinfo/xen-devel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 9/9] x86: introduce CONFIG_HYPERV and detection code

[Paul Durrant]

On Mon, 21 Oct 2019 at 17:01, Wei Liu <wl@xen.org> wrote:
>
> We use the same code structure as we did for Xen.
>
> As starters, detect Hyper-V in probe routine. More complex
> functionalities will be added later.
>
> Signed-off-by: Wei Liu <liuwe@microsoft.com>

Reviewed-by: Paul Durrant <paul@xen.org>

...with one suggestion...

> ---
> V3:
> 1. Remove some unused code
> 2. Rename structure
> 3. Also detect HV#1 signature
> ---
>  xen/arch/x86/Kconfig               |  9 +++++
>  xen/arch/x86/guest/Makefile        |  1 +
>  xen/arch/x86/guest/hyperv/Makefile |  1 +
>  xen/arch/x86/guest/hyperv/hyperv.c | 54 ++++++++++++++++++++++++++++++
>  xen/arch/x86/guest/hypervisor.c    |  8 +++++
>  xen/include/asm-x86/guest.h        |  1 +
>  xen/include/asm-x86/guest/hyperv.h | 45 +++++++++++++++++++++++++
>  7 files changed, 119 insertions(+)
>  create mode 100644 xen/arch/x86/guest/hyperv/Makefile
>  create mode 100644 xen/arch/x86/guest/hyperv/hyperv.c
>  create mode 100644 xen/include/asm-x86/guest/hyperv.h
>
> diff --git a/xen/arch/x86/Kconfig b/xen/arch/x86/Kconfig
> index 867de857e8..56513c771c 100644
> --- a/xen/arch/x86/Kconfig
> +++ b/xen/arch/x86/Kconfig
> @@ -164,6 +164,15 @@ endchoice
>  config GUEST
>         bool
>
> +config HYPERV_GUEST
> +       def_bool n
> +       select GUEST
> +       prompt "Hyper-V Guest"
> +       ---help---
> +         Support for Xen detecting when it is running under Hyper-V.
> +
> +         If unsure, say N.
> +
>  config XEN_GUEST
>         def_bool n
>         select GUEST
> diff --git a/xen/arch/x86/guest/Makefile b/xen/arch/x86/guest/Makefile
> index f63d64bbee..f164196772 100644
> --- a/xen/arch/x86/guest/Makefile
> +++ b/xen/arch/x86/guest/Makefile
> @@ -1,3 +1,4 @@
>  obj-y += hypervisor.o
>
> +subdir-$(CONFIG_HYPERV_GUEST) += hyperv
>  subdir-$(CONFIG_XEN_GUEST) += xen
> diff --git a/xen/arch/x86/guest/hyperv/Makefile b/xen/arch/x86/guest/hyperv/Makefile
> new file mode 100644
> index 0000000000..68170109a9
> --- /dev/null
> +++ b/xen/arch/x86/guest/hyperv/Makefile
> @@ -0,0 +1 @@
> +obj-y += hyperv.o
> diff --git a/xen/arch/x86/guest/hyperv/hyperv.c b/xen/arch/x86/guest/hyperv/hyperv.c
> new file mode 100644
> index 0000000000..7ab4b127f3
> --- /dev/null
> +++ b/xen/arch/x86/guest/hyperv/hyperv.c
> @@ -0,0 +1,54 @@
> +/******************************************************************************
> + * arch/x86/guest/hyperv/hyperv.c
> + *
> + * Support for detecting and running under Hyper-V.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; If not, see <http://www.gnu.org/licenses/>.
> + *
> + * Copyright (c) 2019 Microsoft.
> + */
> +#include <xen/init.h>
> +
> +#include <asm/guest.h>
> +
> +bool __init hyperv_probe(void)
> +{
> +    uint32_t eax, ebx, ecx, edx;
> +
> +    cpuid(0x40000000, &eax, &ebx, &ecx, &edx);
> +    if ( !((ebx == 0x7263694d) &&  /* "Micr" */
> +           (ecx == 0x666f736f) &&  /* "osof" */
> +           (edx == 0x76482074)) )  /* "t Hv" */
> +        return false;
> +
> +    cpuid(0x40000001, &eax, &ebx, &ecx, &edx);
> +    if ( eax != 0x31237648 )    /* Hv#1 */
> +        return false;
> +
> +    return true;
> +}
> +
> +struct hypervisor_ops hyperv_ops = {
> +    .name = "Hyper-V",
> +};
> +
> +/*
> + * Local variables:
> + * mode: C
> + * c-file-style: "BSD"
> + * c-basic-offset: 4
> + * tab-width: 4
> + * indent-tabs-mode: nil
> + * End:
> + */
> diff --git a/xen/arch/x86/guest/hypervisor.c b/xen/arch/x86/guest/hypervisor.c
> index a666ad9526..17392d1ffa 100644
> --- a/xen/arch/x86/guest/hypervisor.c
> +++ b/xen/arch/x86/guest/hypervisor.c
> @@ -43,6 +43,14 @@ bool hypervisor_probe(void)
>      }
>  #endif
>

Add a comment here to point out that hyperv_probe() needs to be called
after xen_probe() to avoid a false positive due to viridian support.

> +#ifdef CONFIG_HYPERV_GUEST
> +    if ( hyperv_probe() )
> +    {
> +        hops = &hyperv_ops;
> +        return true;
> +    }
> +#endif
> +
>      return false;
>  }
>
> diff --git a/xen/include/asm-x86/guest.h b/xen/include/asm-x86/guest.h
> index 8e167165ae..94448606d4 100644
> --- a/xen/include/asm-x86/guest.h
> +++ b/xen/include/asm-x86/guest.h
> @@ -20,6 +20,7 @@
>  #define __X86_GUEST_H__
>
>  #include <asm/guest/hypercall.h>
> +#include <asm/guest/hyperv.h>
>  #include <asm/guest/hypervisor.h>
>  #include <asm/guest/pvh-boot.h>
>  #include <asm/guest/xen.h>
> diff --git a/xen/include/asm-x86/guest/hyperv.h b/xen/include/asm-x86/guest/hyperv.h
> new file mode 100644
> index 0000000000..4b9cc5a836
> --- /dev/null
> +++ b/xen/include/asm-x86/guest/hyperv.h
> @@ -0,0 +1,45 @@
> +/******************************************************************************
> + * asm-x86/guest/hyperv.h
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms and conditions of the GNU General Public
> + * License, version 2, as published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> + * General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public
> + * License along with this program; If not, see <http://www.gnu.org/licenses/>.
> + *
> + * Copyright (c) 2019 Microsoft.
> + */
> +
> +#ifndef __X86_GUEST_HYPERV_H__
> +#define __X86_GUEST_HYPERV_H__
> +
> +#ifdef CONFIG_HYPERV_GUEST
> +
> +#include <asm/guest/hypervisor.h>
> +
> +extern struct hypervisor_ops hyperv_ops;
> +
> +bool hyperv_probe(void);
> +
> +#else
> +
> +static inline bool hyperv_probe(void) { return false; }
> +
> +#endif /* CONFIG_HYPERV_GUEST */
> +#endif /* __X86_GUEST_HYPERV_H__ */
> +
> +/*
> + * Local variables:
> + * mode: C
> + * c-file-style: "BSD"
> + * c-basic-offset: 4
> + * tab-width: 4
> + * indent-tabs-mode: nil
> + * End:
> + */
> --
> 2.20.1
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xenproject.org
> https://lists.xenproject.org/mailman/listinfo/xen-devel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 1/9] x86: introduce CONFIG_GUEST and move code

[Wei Liu]

On Wed, Oct 23, 2019 at 08:55:24AM +0100, Paul Durrant wrote:
> On Mon, 21 Oct 2019 at 16:59, Wei Liu <wl@xen.org> wrote:
> >
> > Xen is able to run as a guest on Xen. We plan to make it able to run
> > on Hyper-V as well.
> >
> > Introduce CONFIG_GUEST which is set to true if either running on Xen
> > or Hyper-V is desired. Restructure code hierarchy for new code to
> > come.
> >
> > No functional change intended.
> >
> > Signed-off-by: Wei Liu <liuwe@microsoft.com>
> > Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
> > ---
> >  xen/arch/x86/Kconfig                          | 4 ++++
> >  xen/arch/x86/Makefile                         | 2 +-
> >  xen/arch/x86/guest/Makefile                   | 5 +----
> >  xen/arch/x86/guest/xen/Makefile               | 4 ++++
> >  xen/arch/x86/guest/{ => xen}/hypercall_page.S | 0
> >  xen/arch/x86/guest/{ => xen}/pvh-boot.c       | 0
> >  xen/arch/x86/guest/{ => xen}/xen.c            | 0
> >  7 files changed, 10 insertions(+), 5 deletions(-)
> >  create mode 100644 xen/arch/x86/guest/xen/Makefile
> >  rename xen/arch/x86/guest/{ => xen}/hypercall_page.S (100%)
> >  rename xen/arch/x86/guest/{ => xen}/pvh-boot.c (100%)
> >  rename xen/arch/x86/guest/{ => xen}/xen.c (100%)
> >
> > diff --git a/xen/arch/x86/Kconfig b/xen/arch/x86/Kconfig
> > index 28b3b4692a..867de857e8 100644
> > --- a/xen/arch/x86/Kconfig
> > +++ b/xen/arch/x86/Kconfig
> > @@ -161,8 +161,12 @@ config XEN_ALIGN_2M
> >
> >  endchoice
> >
> > +config GUEST
> > +       bool
> > +
> >  config XEN_GUEST
> >         def_bool n
> > +       select GUEST
> >         prompt "Xen Guest"
> >         ---help---
> >           Support for Xen detecting when it is running under Xen.
> 
> This text needs to be modified. Perhaps just say 'when it is running
> in a virtual machine'. That should be future-proof.
> 

I don't think it should be modified -- XEN_GUEST here really means
specifically running under Xen but not others.

We will later introduced HYPERV_GUEST.

Wei.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 9/9] x86: introduce CONFIG_HYPERV and detection code

[Wei Liu]

On Wed, Oct 23, 2019 at 10:07:24AM +0100, Paul Durrant wrote:
> On Mon, 21 Oct 2019 at 17:01, Wei Liu <wl@xen.org> wrote:
> >
> > We use the same code structure as we did for Xen.
> >
> > As starters, detect Hyper-V in probe routine. More complex
> > functionalities will be added later.
> >
> > Signed-off-by: Wei Liu <liuwe@microsoft.com>
> 
> Reviewed-by: Paul Durrant <paul@xen.org>

Thanks.

[...]
> > diff --git a/xen/arch/x86/guest/hypervisor.c b/xen/arch/x86/guest/hypervisor.c
> > index a666ad9526..17392d1ffa 100644
> > --- a/xen/arch/x86/guest/hypervisor.c
> > +++ b/xen/arch/x86/guest/hypervisor.c
> > @@ -43,6 +43,14 @@ bool hypervisor_probe(void)
> >      }
> >  #endif
> >
> 
> Add a comment here to point out that hyperv_probe() needs to be called
> after xen_probe() to avoid a false positive due to viridian support.
> 

Yes I can do that.

Wei.  

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 1/9] x86: introduce CONFIG_GUEST and move code

[Jan Beulich]

On 21.10.2019 17:57, Wei Liu wrote:
> Xen is able to run as a guest on Xen. We plan to make it able to run
> on Hyper-V as well.
> 
> Introduce CONFIG_GUEST which is set to true if either running on Xen
> or Hyper-V is desired. Restructure code hierarchy for new code to
> come.
> 
> No functional change intended.
> 
> Signed-off-by: Wei Liu <liuwe@microsoft.com>
> Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>

Acked-by: Jan Beulich <jbeulich@suse.com>

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 2/9] x86: include asm_defns.h directly in hypercall.h

[Jan Beulich]

On 21.10.2019 17:57, Wei Liu wrote:
> ASM_CALL_CONSTRAINT is defined there.
> 
> No functional change.
> 
> Signed-off-by: Wei Liu <liuwe@microsoft.com>
> Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>

Acked-by: Jan Beulich <jbeulich@suse.com>

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 3/9] x86: drop hypervisor_cpuid_base

[Jan Beulich]

On 21.10.2019 17:57, Wei Liu wrote:
> The only user is Xen specific code in PV shim. We can therefore export
> the variable directly.
> 
> Signed-off-by: Wei Liu <liuwe@microsoft.com>
> Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>

Acked-by: Jan Beulich <jbeulich@suse.com>
with, at this occasion, ...

> --- a/xen/arch/x86/guest/xen/xen.c
> +++ b/xen/arch/x86/guest/xen/xen.c
> @@ -37,7 +37,7 @@
>  
>  bool __read_mostly xen_guest;
>  
> -static __read_mostly uint32_t xen_cpuid_base;
> +__read_mostly uint32_t xen_cpuid_base;

... the __read_mostly moved to its more standard place (an example
is in context).

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 4/9] x86: include xen/lib.h in guest/hypercall.h

[Jan Beulich]

On 21.10.2019 17:57, Wei Liu wrote:
> We need ASSERT_UNREACHABLE.
> 
> Signed-off-by: Wei Liu <liuwe@microsoft.com>
> Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>

Acked-by: Jan Beulich <jbeulich@suse.com>

Albeit perhaps fold into patch 2?

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 5/9] x86: introduce hypervisor framework

[Jan Beulich]

On 21.10.2019 17:57, Wei Liu wrote:
> --- /dev/null
> +++ b/xen/arch/x86/guest/hypervisor.c
> @@ -0,0 +1,45 @@
> +/******************************************************************************
> + * arch/x86/guest/hypervisor.c
> + *
> + * Support for detecting and running under a hypervisor.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; If not, see <http://www.gnu.org/licenses/>.
> + *
> + * Copyright (c) 2019 Microsoft.
> + */
> +
> +#include <xen/types.h>
> +
> +#include <asm/cache.h>
> +#include <asm/guest/hypervisor.h>
> +
> +static struct hypervisor_ops *hops __read_mostly;

The __read_mostly wants to again go between type and identifier.

> +bool hypervisor_probe(void)
> +{
> +    if ( hops )
> +        return true;
> +
> +    return false;

I assume this isn't simply "return hops" because more is going to be
added here?

> --- /dev/null
> +++ b/xen/include/asm-x86/guest/hypervisor.h
> @@ -0,0 +1,61 @@
> +/******************************************************************************
> + * asm-x86/guest/hypervisor.h
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms and conditions of the GNU General Public
> + * License, version 2, as published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> + * General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public
> + * License along with this program; If not, see <http://www.gnu.org/licenses/>.
> + *
> + * Copyright (c) 2019 Microsoft.
> + */
> +
> +#ifndef __X86_HYPERVISOR_H__
> +#define __X86_HYPERVISOR_H__
> +
> +#ifdef CONFIG_GUEST
> +
> +struct hypervisor_ops {
> +    /* Name of the hypervisor */
> +    const char *name;
> +    /* Main setup routine */
> +    void (*setup)(void);
> +    /* AP setup */
> +    void (*ap_setup)(void);
> +    /* Resume from suspension */
> +    void (*resume)(void);

None of these fields look to be used, despite ...

> +};
> +
> +bool hypervisor_probe(void);
> +void hypervisor_setup(void);
> +void hypervisor_ap_setup(void);
> +void hypervisor_resume(void);

... the latter three of these being the apparent wrappers for
them. Perhaps just a side effect of how exactly this series
was split up, so not strictly a request to change anything.

> @@ -49,15 +46,6 @@ DECLARE_PER_CPU(struct vcpu_info *, vcpu_info);
>  
>  static inline void probe_hypervisor(void) {}
>  
> -static inline void hypervisor_setup(void)
> -{
> -    ASSERT_UNREACHABLE();
> -}
> -static inline void hypervisor_ap_setup(void)
> -{
> -    ASSERT_UNREACHABLE();
> -}

Why did the ASSERT_UNREACHABLE() get lost?

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 6/9] x86: rename hypervisor_{alloc, free}_unused_page

[Jan Beulich]

On 21.10.2019 17:57, Wei Liu wrote:
> --- a/xen/arch/x86/guest/xen/xen.c
> +++ b/xen/arch/x86/guest/xen/xen.c
> @@ -97,7 +97,7 @@ static void map_shared_info(void)
>      unsigned int i;
>      unsigned long rc;
>  
> -    if ( hypervisor_alloc_unused_page(&mfn) )
> +    if ( xen_alloc_unused_page(&mfn) )

But is just "xen_" as a prefix a good name in Xen code? Perhaps
"xen_guest_" or (shorter) "xg_"?

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 7/9] x86: switch xen implementation to use hypervisor framework

[Jan Beulich]

On 21.10.2019 17:57, Wei Liu wrote:
> @@ -31,9 +31,39 @@ bool hypervisor_probe(void)
>      if ( hops )
>          return true;
>  
> +    /* Too early to use cpu_has_hypervisor */
> +    if ( !(cpuid_ecx(1) & cpufeat_mask(X86_FEATURE_HYPERVISOR)) )
> +        return false;
> +
> +#ifdef CONFIG_XEN_GUEST
> +    if ( xen_probe() )
> +    {
> +        hops = &xen_ops;
> +        return true;
> +    }
> +#endif
> +
>      return false;
>  }
>  
> +void hypervisor_setup(void)

__init?

> @@ -330,6 +325,13 @@ void hypervisor_resume(void)
>          pv_console_init();
>  }
>  
> +struct hypervisor_ops xen_ops = {

const (will require a corresponding adjustment in the earlier patch
then too)?

With these taken care of and the naming possibly adjusted according
to whatever the decision on the earlier patch s going to be
Acked-by: Jan Beulich <jbeulich@suse.com>

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 8/9] x86: be more verbose when running on a hypervisor

[Jan Beulich]

On 21.10.2019 17:57, Wei Liu wrote:
> --- a/xen/arch/x86/setup.c
> +++ b/xen/arch/x86/setup.c
> @@ -689,6 +689,7 @@ void __init noreturn __start_xen(unsigned long mbi_p)
>      int i, j, e820_warn = 0, bytes = 0;
>      bool acpi_boot_table_init_done = false, relocated = false;
>      int ret;
> +    bool running_on_hypervisor;
>      struct ns16550_defaults ns16550 = {
>          .data_bits = 8,
>          .parity    = 'n',
> @@ -763,7 +764,7 @@ void __init noreturn __start_xen(unsigned long mbi_p)
>       * allocing any xenheap structures wanted in lower memory. */
>      kexec_early_calculations();
>  
> -    hypervisor_probe();
> +    running_on_hypervisor = hypervisor_probe();
>  
>      parse_video_info();
>  
> @@ -789,6 +790,9 @@ void __init noreturn __start_xen(unsigned long mbi_p)
>  
>      printk("Xen image load base address: %#lx\n", xen_phys_start);
>  
> +    if ( running_on_hypervisor )
> +        printk("Running on %s\n", hypervisor_name());

Instead of a bool, why don't you make hypervisor_probe() return the
name (or NULL), avoiding the separate hypervisor_name() (and making
it moot for me to ask for it to become __init)?

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 9/9] x86: introduce CONFIG_HYPERV and detection code

[Jan Beulich]

On 21.10.2019 17:57, Wei Liu wrote:
> --- a/xen/arch/x86/Kconfig
> +++ b/xen/arch/x86/Kconfig
> @@ -164,6 +164,15 @@ endchoice
>  config GUEST
>  	bool
>  
> +config HYPERV_GUEST
> +	def_bool n
> +	select GUEST
> +	prompt "Hyper-V Guest"

Please can you avoid following the bad example XEN_GUEST gives (and
perhaps even take the opportunity here or in the earlier patch
adding GUEST to change that one as well)? What you want is

config HYPERV_GUEST
	bool "Hyper-V Guest"
	select GUEST

> --- /dev/null
> +++ b/xen/arch/x86/guest/hyperv/hyperv.c
> @@ -0,0 +1,54 @@
> +/******************************************************************************
> + * arch/x86/guest/hyperv/hyperv.c
> + *
> + * Support for detecting and running under Hyper-V.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; If not, see <http://www.gnu.org/licenses/>.
> + *
> + * Copyright (c) 2019 Microsoft.
> + */
> +#include <xen/init.h>
> +
> +#include <asm/guest.h>
> +
> +bool __init hyperv_probe(void)
> +{
> +    uint32_t eax, ebx, ecx, edx;
> +
> +    cpuid(0x40000000, &eax, &ebx, &ecx, &edx);
> +    if ( !((ebx == 0x7263694d) &&  /* "Micr" */
> +           (ecx == 0x666f736f) &&  /* "osof" */
> +           (edx == 0x76482074)) )  /* "t Hv" */
> +        return false;
> +
> +    cpuid(0x40000001, &eax, &ebx, &ecx, &edx);
> +    if ( eax != 0x31237648 )    /* Hv#1 */
> +        return false;
> +
> +    return true;
> +}
> +
> +struct hypervisor_ops hyperv_ops = {

const again.

> --- a/xen/arch/x86/guest/hypervisor.c
> +++ b/xen/arch/x86/guest/hypervisor.c
> @@ -43,6 +43,14 @@ bool hypervisor_probe(void)
>      }
>  #endif
>  
> +#ifdef CONFIG_HYPERV_GUEST
> +    if ( hyperv_probe() )
> +    {
> +        hops = &hyperv_ops;
> +        return true;
> +    }
> +#endif

This recurring #ifdef CONFIG_*_GUEST is going to start looking ugly
the latest when one or two more get added. Perhaps better providing
*_probe() stubs returning false, and (like we do elsewhere) rely on
DCE to get rid of the *_ops reference? (And really you already have
such a stub - all you need to do is put the hyperv_ops declaration
outside the #ifdef (but read on).

Also how about having *_probe() return the address of *_ops, such
that the latter could all become static?

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 9/9] x86: introduce CONFIG_HYPERV and detection code

[Wei Liu]

On Fri, Nov 15, 2019 at 03:07:29PM +0100, Jan Beulich wrote:
> On 21.10.2019 17:57, Wei Liu wrote:
> > --- a/xen/arch/x86/Kconfig
> > +++ b/xen/arch/x86/Kconfig
> > @@ -164,6 +164,15 @@ endchoice
> >  config GUEST
> >  	bool
> >  
> > +config HYPERV_GUEST
> > +	def_bool n
> > +	select GUEST
> > +	prompt "Hyper-V Guest"
> 
> Please can you avoid following the bad example XEN_GUEST gives (and
> perhaps even take the opportunity here or in the earlier patch
> adding GUEST to change that one as well)? What you want is
> 
> config HYPERV_GUEST
> 	bool "Hyper-V Guest"
> 	select GUEST

Ack.

> 
> > --- /dev/null
> > +++ b/xen/arch/x86/guest/hyperv/hyperv.c
> > @@ -0,0 +1,54 @@
> > +/******************************************************************************
> > + * arch/x86/guest/hyperv/hyperv.c
> > + *
> > + * Support for detecting and running under Hyper-V.
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License as published by
> > + * the Free Software Foundation; either version 2 of the License, or
> > + * (at your option) any later version.
> > + *
> > + * This program is distributed in the hope that it will be useful,
> > + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > + * GNU General Public License for more details.
> > + *
> > + * You should have received a copy of the GNU General Public License
> > + * along with this program; If not, see <http://www.gnu.org/licenses/>.
> > + *
> > + * Copyright (c) 2019 Microsoft.
> > + */
> > +#include <xen/init.h>
> > +
> > +#include <asm/guest.h>
> > +
> > +bool __init hyperv_probe(void)
> > +{
> > +    uint32_t eax, ebx, ecx, edx;
> > +
> > +    cpuid(0x40000000, &eax, &ebx, &ecx, &edx);
> > +    if ( !((ebx == 0x7263694d) &&  /* "Micr" */
> > +           (ecx == 0x666f736f) &&  /* "osof" */
> > +           (edx == 0x76482074)) )  /* "t Hv" */
> > +        return false;
> > +
> > +    cpuid(0x40000001, &eax, &ebx, &ecx, &edx);
> > +    if ( eax != 0x31237648 )    /* Hv#1 */
> > +        return false;
> > +
> > +    return true;
> > +}
> > +
> > +struct hypervisor_ops hyperv_ops = {
> 
> const again.
> 
> > --- a/xen/arch/x86/guest/hypervisor.c
> > +++ b/xen/arch/x86/guest/hypervisor.c
> > @@ -43,6 +43,14 @@ bool hypervisor_probe(void)
> >      }
> >  #endif
> >  
> > +#ifdef CONFIG_HYPERV_GUEST
> > +    if ( hyperv_probe() )
> > +    {
> > +        hops = &hyperv_ops;
> > +        return true;
> > +    }
> > +#endif
> 
> This recurring #ifdef CONFIG_*_GUEST is going to start looking ugly
> the latest when one or two more get added. Perhaps better providing
> *_probe() stubs returning false, and (like we do elsewhere) rely on
> DCE to get rid of the *_ops reference? (And really you already have
> such a stub - all you need to do is put the hyperv_ops declaration
> outside the #ifdef (but read on).
> 
> Also how about having *_probe() return the address of *_ops, such
> that the latter could all become static?

Previously you made a suggestion to make probe return the name of the
hypervisor. Here you ask for address of ops. I actually prefer the
method suggested here, but this means I will need to keep
hypervisor_name around.

Wei.

> 
> Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 6/9] x86: rename hypervisor_{alloc, free}_unused_page

[Wei Liu]

On Fri, Nov 15, 2019 at 02:49:57PM +0100, Jan Beulich wrote:
> On 21.10.2019 17:57, Wei Liu wrote:
> > --- a/xen/arch/x86/guest/xen/xen.c
> > +++ b/xen/arch/x86/guest/xen/xen.c
> > @@ -97,7 +97,7 @@ static void map_shared_info(void)
> >      unsigned int i;
> >      unsigned long rc;
> >  
> > -    if ( hypervisor_alloc_unused_page(&mfn) )
> > +    if ( xen_alloc_unused_page(&mfn) )
> 
> But is just "xen_" as a prefix a good name in Xen code? Perhaps
> "xen_guest_" or (shorter) "xg_"?

That's fine by me.

Wei.

> 
> Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 5/9] x86: introduce hypervisor framework

[Wei Liu]

On Fri, Nov 15, 2019 at 02:48:18PM +0100, Jan Beulich wrote:
> On 21.10.2019 17:57, Wei Liu wrote:
> > --- /dev/null
> > +++ b/xen/arch/x86/guest/hypervisor.c
> > @@ -0,0 +1,45 @@
> > +/******************************************************************************
> > + * arch/x86/guest/hypervisor.c
> > + *
> > + * Support for detecting and running under a hypervisor.
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License as published by
> > + * the Free Software Foundation; either version 2 of the License, or
> > + * (at your option) any later version.
> > + *
> > + * This program is distributed in the hope that it will be useful,
> > + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > + * GNU General Public License for more details.
> > + *
> > + * You should have received a copy of the GNU General Public License
> > + * along with this program; If not, see <http://www.gnu.org/licenses/>.
> > + *
> > + * Copyright (c) 2019 Microsoft.
> > + */
> > +
> > +#include <xen/types.h>
> > +
> > +#include <asm/cache.h>
> > +#include <asm/guest/hypervisor.h>
> > +
> > +static struct hypervisor_ops *hops __read_mostly;
> 
> The __read_mostly wants to again go between type and identifier.

Ack.

> 
> > +bool hypervisor_probe(void)
> > +{
> > +    if ( hops )
> > +        return true;
> > +
> > +    return false;
> 
> I assume this isn't simply "return hops" because more is going to be
> added here?
> 

That's right.

[...]
> > @@ -49,15 +46,6 @@ DECLARE_PER_CPU(struct vcpu_info *, vcpu_info);
> >  
> >  static inline void probe_hypervisor(void) {}
> >  
> > -static inline void hypervisor_setup(void)
> > -{
> > -    ASSERT_UNREACHABLE();
> > -}
> > -static inline void hypervisor_ap_setup(void)
> > -{
> > -    ASSERT_UNREACHABLE();
> > -}
> 
> Why did the ASSERT_UNREACHABLE() get lost?

I will add it back to all those stubs.

Wei.

> 
> Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 9/9] x86: introduce CONFIG_HYPERV and detection code

[Jan Beulich]

On 21.11.2019 17:27, Wei Liu wrote:
> On Fri, Nov 15, 2019 at 03:07:29PM +0100, Jan Beulich wrote:
>> On 21.10.2019 17:57, Wei Liu wrote:
>>> --- a/xen/arch/x86/Kconfig
>>> +++ b/xen/arch/x86/Kconfig
>>> @@ -164,6 +164,15 @@ endchoice
>>>  config GUEST
>>>  	bool
>>>  
>>> +config HYPERV_GUEST
>>> +	def_bool n
>>> +	select GUEST
>>> +	prompt "Hyper-V Guest"
>>
>> Please can you avoid following the bad example XEN_GUEST gives (and
>> perhaps even take the opportunity here or in the earlier patch
>> adding GUEST to change that one as well)? What you want is
>>
>> config HYPERV_GUEST
>> 	bool "Hyper-V Guest"
>> 	select GUEST
> 
> Ack.
> 
>>
>>> --- /dev/null
>>> +++ b/xen/arch/x86/guest/hyperv/hyperv.c
>>> @@ -0,0 +1,54 @@
>>> +/******************************************************************************
>>> + * arch/x86/guest/hyperv/hyperv.c
>>> + *
>>> + * Support for detecting and running under Hyper-V.
>>> + *
>>> + * This program is free software; you can redistribute it and/or modify
>>> + * it under the terms of the GNU General Public License as published by
>>> + * the Free Software Foundation; either version 2 of the License, or
>>> + * (at your option) any later version.
>>> + *
>>> + * This program is distributed in the hope that it will be useful,
>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>> + * GNU General Public License for more details.
>>> + *
>>> + * You should have received a copy of the GNU General Public License
>>> + * along with this program; If not, see <http://www.gnu.org/licenses/>.
>>> + *
>>> + * Copyright (c) 2019 Microsoft.
>>> + */
>>> +#include <xen/init.h>
>>> +
>>> +#include <asm/guest.h>
>>> +
>>> +bool __init hyperv_probe(void)
>>> +{
>>> +    uint32_t eax, ebx, ecx, edx;
>>> +
>>> +    cpuid(0x40000000, &eax, &ebx, &ecx, &edx);
>>> +    if ( !((ebx == 0x7263694d) &&  /* "Micr" */
>>> +           (ecx == 0x666f736f) &&  /* "osof" */
>>> +           (edx == 0x76482074)) )  /* "t Hv" */
>>> +        return false;
>>> +
>>> +    cpuid(0x40000001, &eax, &ebx, &ecx, &edx);
>>> +    if ( eax != 0x31237648 )    /* Hv#1 */
>>> +        return false;
>>> +
>>> +    return true;
>>> +}
>>> +
>>> +struct hypervisor_ops hyperv_ops = {
>>
>> const again.
>>
>>> --- a/xen/arch/x86/guest/hypervisor.c
>>> +++ b/xen/arch/x86/guest/hypervisor.c
>>> @@ -43,6 +43,14 @@ bool hypervisor_probe(void)
>>>      }
>>>  #endif
>>>  
>>> +#ifdef CONFIG_HYPERV_GUEST
>>> +    if ( hyperv_probe() )
>>> +    {
>>> +        hops = &hyperv_ops;
>>> +        return true;
>>> +    }
>>> +#endif
>>
>> This recurring #ifdef CONFIG_*_GUEST is going to start looking ugly
>> the latest when one or two more get added. Perhaps better providing
>> *_probe() stubs returning false, and (like we do elsewhere) rely on
>> DCE to get rid of the *_ops reference? (And really you already have
>> such a stub - all you need to do is put the hyperv_ops declaration
>> outside the #ifdef (but read on).
>>
>> Also how about having *_probe() return the address of *_ops, such
>> that the latter could all become static?
> 
> Previously you made a suggestion to make probe return the name of the
> hypervisor. Here you ask for address of ops. I actually prefer the
> method suggested here, but this means I will need to keep
> hypervisor_name around.

Is there actually any user of the name field other than the caller
of probe? If not, surely that caller could access the name field
without a hypervisor_name() wrapper.

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: [Xen-devel] [PATCH for-next v3 9/9] x86: introduce CONFIG_HYPERV and detection code

[Wei Liu]

On Thu, Nov 21, 2019 at 05:59:16PM +0100, Jan Beulich wrote:
> >>
> >> Also how about having *_probe() return the address of *_ops, such
> >> that the latter could all become static?
> > 
> > Previously you made a suggestion to make probe return the name of the
> > hypervisor. Here you ask for address of ops. I actually prefer the
> > method suggested here, but this means I will need to keep
> > hypervisor_name around.
> 
> Is there actually any user of the name field other than the caller
> of probe? If not, surely that caller could access the name field
> without a hypervisor_name() wrapper.

I don't envision more users at this stage.

I'm fine with accessing that field directly. We can always introduce the
function again if it becomes necessary.

Wei.

> 
> Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel