| Message ID | 20200204153704.15934-6-liuwe@microsoft.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | More Hyper-V infrastructures | expand |
On Tue, Feb 04, 2020 at 03:36:59PM +0000, Wei Liu wrote: > Hyper-V uses a technique called overlay page for its hypercall page. It > will insert a backing page to the guest when the hypercall functionality > is enabled. That means we can use a page that is not backed by real > memory for hypercall page. > > To avoid shattering L0 superpages and treading on any MMIO areas > residing in low addresses, use the top-most addressable page for that > purpose. Adjust e820 map accordingly. > > We also need to register Xen's guest OS ID to Hyper-V. Use 0x3 as the > vendor ID. Fix the comment in hyperv-tlfs.h while at it. > > Signed-off-by: Wei Liu <liuwe@microsoft.com> Reviewed-by: Roger Pau Monné <roger.pau@citrix.com> > --- > v7: > 1. Fix a style issue > 2. Initialise ID to 0 > 3. Update commit message > > v6: > 1. Use hv_guest_os_id > 2. Use new e820_fixup hook > 3. Add a BUILD_BUG_ON > > v5: > 1. use hypervisor_reserve_top_pages > 2. add a macro for hypercall page mfn > 3. address other misc comments > > v4: > 1. Use fixmap > 2. Follow routines listed in TLFS > --- > xen/arch/x86/guest/hyperv/hyperv.c | 69 +++++++++++++++++++++++-- > xen/include/asm-x86/guest/hyperv-tlfs.h | 5 +- > xen/include/asm-x86/guest/hyperv.h | 3 ++ > 3 files changed, 72 insertions(+), 5 deletions(-) > > diff --git a/xen/arch/x86/guest/hyperv/hyperv.c b/xen/arch/x86/guest/hyperv/hyperv.c > index 8d38313d7a..2e20a96f30 100644 > --- a/xen/arch/x86/guest/hyperv/hyperv.c > +++ b/xen/arch/x86/guest/hyperv/hyperv.c > @@ -19,15 +19,27 @@ > * Copyright (c) 2019 Microsoft. > */ > #include <xen/init.h> > +#include <xen/version.h> > > +#include <asm/fixmap.h> > #include <asm/guest.h> > #include <asm/guest/hyperv-tlfs.h> > +#include <asm/processor.h> > > struct ms_hyperv_info __read_mostly ms_hyperv; > > -static const struct hypervisor_ops ops = { > - .name = "Hyper-V", > -}; > +static uint64_t generate_guest_id(void) > +{ > + union hv_guest_os_id id = {}; > + > + id.vendor = HV_XEN_VENDOR_ID; > + id.major = xen_major_version(); > + id.minor = xen_minor_version(); > + > + return id.raw; > +} > + > +static const struct hypervisor_ops ops; > > const struct hypervisor_ops *__init hyperv_probe(void) > { > @@ -72,6 +84,57 @@ const struct hypervisor_ops *__init hyperv_probe(void) > return &ops; > } > > +static void __init setup_hypercall_page(void) > +{ > + union hv_x64_msr_hypercall_contents hypercall_msr; > + union hv_guest_os_id guest_id; > + unsigned long mfn; > + > + BUILD_BUG_ON(HV_HYP_PAGE_SHIFT != PAGE_SHIFT); > + > + rdmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); > + if ( !guest_id.raw ) > + { > + guest_id.raw = generate_guest_id(); > + wrmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); > + } > + > + rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); > + if ( !hypercall_msr.enable ) > + { > + mfn = HV_HCALL_MFN; > + hypercall_msr.enable = 1; > + hypercall_msr.guest_physical_address = mfn; > + wrmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); > + } > + else > + mfn = hypercall_msr.guest_physical_address; Is it expected that the guest ID or the hypercal page is already setup? Ie: would virtual firmware setup any of this? (and not clean it up afterwards) Thanks, Roger.
On Wed, Feb 05, 2020 at 03:04:16PM +0100, Roger Pau Monné wrote: [...] > > +static void __init setup_hypercall_page(void) > > +{ > > + union hv_x64_msr_hypercall_contents hypercall_msr; > > + union hv_guest_os_id guest_id; > > + unsigned long mfn; > > + > > + BUILD_BUG_ON(HV_HYP_PAGE_SHIFT != PAGE_SHIFT); > > + > > + rdmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); > > + if ( !guest_id.raw ) > > + { > > + guest_id.raw = generate_guest_id(); > > + wrmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); > > + } > > + > > + rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); > > + if ( !hypercall_msr.enable ) > > + { > > + mfn = HV_HCALL_MFN; > > + hypercall_msr.enable = 1; > > + hypercall_msr.guest_physical_address = mfn; > > + wrmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); > > + } > > + else > > + mfn = hypercall_msr.guest_physical_address; > > Is it expected that the guest ID or the hypercal page is already > setup? > > Ie: would virtual firmware setup any of this? (and not clean it up > afterwards) There is no definitive answer to why/when this happens in TLFS, but it does require guest to respect that is already in the MSR. Wei. > > Thanks, Roger.
On Wed, Feb 05, 2020 at 03:00:24PM +0000, Wei Liu wrote: > On Wed, Feb 05, 2020 at 03:04:16PM +0100, Roger Pau Monné wrote: > [...] > > > +static void __init setup_hypercall_page(void) > > > +{ > > > + union hv_x64_msr_hypercall_contents hypercall_msr; > > > + union hv_guest_os_id guest_id; > > > + unsigned long mfn; > > > + > > > + BUILD_BUG_ON(HV_HYP_PAGE_SHIFT != PAGE_SHIFT); > > > + > > > + rdmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); > > > + if ( !guest_id.raw ) > > > + { > > > + guest_id.raw = generate_guest_id(); > > > + wrmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); > > > + } > > > + > > > + rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); > > > + if ( !hypercall_msr.enable ) > > > + { > > > + mfn = HV_HCALL_MFN; > > > + hypercall_msr.enable = 1; > > > + hypercall_msr.guest_physical_address = mfn; > > > + wrmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); > > > + } > > > + else > > > + mfn = hypercall_msr.guest_physical_address; > > > > Is it expected that the guest ID or the hypercal page is already > > setup? > > > > Ie: would virtual firmware setup any of this? (and not clean it up > > afterwards) > > There is no definitive answer to why/when this happens in TLFS, but it > does require guest to respect that is already in the MSR. My guess would be after migration Hyper-V sets up the same MSR/page FWIW. Wei.
> -----Original Message----- > From: Wei Liu <wei.liu.xen@gmail.com> On Behalf Of Wei Liu > Sent: 04 February 2020 15:37 > To: Xen Development List <xen-devel@lists.xenproject.org> > Cc: Durrant, Paul <pdurrant@amazon.co.uk>; Michael Kelley > <mikelley@microsoft.com>; Wei Liu <liuwe@microsoft.com>; Wei Liu > <wl@xen.org>; Jan Beulich <jbeulich@suse.com>; Andrew Cooper > <andrew.cooper3@citrix.com>; Roger Pau Monné <roger.pau@citrix.com> > Subject: [PATCH v7 05/10] x86/hyperv: setup hypercall page > > Hyper-V uses a technique called overlay page for its hypercall page. It > will insert a backing page to the guest when the hypercall functionality > is enabled. That means we can use a page that is not backed by real > memory for hypercall page. > > To avoid shattering L0 superpages and treading on any MMIO areas > residing in low addresses, use the top-most addressable page for that > purpose. Adjust e820 map accordingly. > > We also need to register Xen's guest OS ID to Hyper-V. Use 0x3 as the > vendor ID. Fix the comment in hyperv-tlfs.h while at it. > > Signed-off-by: Wei Liu <liuwe@microsoft.com> Acked-by: Paul Durrant <pdurrant@amazon.com> > --- > v7: > 1. Fix a style issue > 2. Initialise ID to 0 > 3. Update commit message > > v6: > 1. Use hv_guest_os_id > 2. Use new e820_fixup hook > 3. Add a BUILD_BUG_ON > > v5: > 1. use hypervisor_reserve_top_pages > 2. add a macro for hypercall page mfn > 3. address other misc comments > > v4: > 1. Use fixmap > 2. Follow routines listed in TLFS > --- > xen/arch/x86/guest/hyperv/hyperv.c | 69 +++++++++++++++++++++++-- > xen/include/asm-x86/guest/hyperv-tlfs.h | 5 +- > xen/include/asm-x86/guest/hyperv.h | 3 ++ > 3 files changed, 72 insertions(+), 5 deletions(-) > > diff --git a/xen/arch/x86/guest/hyperv/hyperv.c > b/xen/arch/x86/guest/hyperv/hyperv.c > index 8d38313d7a..2e20a96f30 100644 > --- a/xen/arch/x86/guest/hyperv/hyperv.c > +++ b/xen/arch/x86/guest/hyperv/hyperv.c > @@ -19,15 +19,27 @@ > * Copyright (c) 2019 Microsoft. > */ > #include <xen/init.h> > +#include <xen/version.h> > > +#include <asm/fixmap.h> > #include <asm/guest.h> > #include <asm/guest/hyperv-tlfs.h> > +#include <asm/processor.h> > > struct ms_hyperv_info __read_mostly ms_hyperv; > > -static const struct hypervisor_ops ops = { > - .name = "Hyper-V", > -}; > +static uint64_t generate_guest_id(void) > +{ > + union hv_guest_os_id id = {}; > + > + id.vendor = HV_XEN_VENDOR_ID; > + id.major = xen_major_version(); > + id.minor = xen_minor_version(); > + > + return id.raw; > +} > + > +static const struct hypervisor_ops ops; > > const struct hypervisor_ops *__init hyperv_probe(void) > { > @@ -72,6 +84,57 @@ const struct hypervisor_ops *__init hyperv_probe(void) > return &ops; > } > > +static void __init setup_hypercall_page(void) > +{ > + union hv_x64_msr_hypercall_contents hypercall_msr; > + union hv_guest_os_id guest_id; > + unsigned long mfn; > + > + BUILD_BUG_ON(HV_HYP_PAGE_SHIFT != PAGE_SHIFT); > + > + rdmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); > + if ( !guest_id.raw ) > + { > + guest_id.raw = generate_guest_id(); > + wrmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); > + } > + > + rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); > + if ( !hypercall_msr.enable ) > + { > + mfn = HV_HCALL_MFN; > + hypercall_msr.enable = 1; > + hypercall_msr.guest_physical_address = mfn; > + wrmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); > + } > + else > + mfn = hypercall_msr.guest_physical_address; > + > + rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); > + BUG_ON(!hypercall_msr.enable); > + > + set_fixmap_x(FIX_X_HYPERV_HCALL, mfn << PAGE_SHIFT); > +} > + > +static void __init setup(void) > +{ > + setup_hypercall_page(); > +} > + > +static void __init e820_fixup(struct e820map *e820) > +{ > + uint64_t s = HV_HCALL_MFN << PAGE_SHIFT; > + > + if ( !e820_add_range(e820, s, s + PAGE_SIZE, E820_RESERVED) ) > + panic("Unable to reserve Hyper-V hypercall range\n"); > +} > + > +static const struct hypervisor_ops ops = { > + .name = "Hyper-V", > + .setup = setup, > + .e820_fixup = e820_fixup, > +}; > + > /* > * Local variables: > * mode: C > diff --git a/xen/include/asm-x86/guest/hyperv-tlfs.h b/xen/include/asm- > x86/guest/hyperv-tlfs.h > index 091e25cdd1..0a0f3398c1 100644 > --- a/xen/include/asm-x86/guest/hyperv-tlfs.h > +++ b/xen/include/asm-x86/guest/hyperv-tlfs.h > @@ -318,15 +318,16 @@ struct ms_hyperv_tsc_page { > * > * Bit(s) > * 63 - Indicates if the OS is Open Source or not; 1 is Open Source > - * 62:56 - Os Type; Linux is 0x100 > + * 62:56 - Os Type; Linux 0x1, FreeBSD 0x2, Xen 0x3 > * 55:48 - Distro specific identification > - * 47:16 - Linux kernel version number > + * 47:16 - Guest OS version number > * 15:0 - Distro specific identification > * > * > */ > > #define HV_LINUX_VENDOR_ID 0x8100 > +#define HV_XEN_VENDOR_ID 0x8300 > union hv_guest_os_id > { > uint64_t raw; > diff --git a/xen/include/asm-x86/guest/hyperv.h b/xen/include/asm- > x86/guest/hyperv.h > index c7a7f32bd5..1a1b47831c 100644 > --- a/xen/include/asm-x86/guest/hyperv.h > +++ b/xen/include/asm-x86/guest/hyperv.h > @@ -21,6 +21,9 @@ > > #include <xen/types.h> > > +/* Use top-most MFN for hypercall page */ > +#define HV_HCALL_MFN (((1ull << paddr_bits) - 1) >> HV_HYP_PAGE_SHIFT) > + > /* > * The specification says: "The partition reference time is computed > * by the following formula: > -- > 2.20.1
diff --git a/xen/arch/x86/guest/hyperv/hyperv.c b/xen/arch/x86/guest/hyperv/hyperv.c index 8d38313d7a..2e20a96f30 100644 --- a/xen/arch/x86/guest/hyperv/hyperv.c +++ b/xen/arch/x86/guest/hyperv/hyperv.c @@ -19,15 +19,27 @@ * Copyright (c) 2019 Microsoft. */ #include <xen/init.h> +#include <xen/version.h> +#include <asm/fixmap.h> #include <asm/guest.h> #include <asm/guest/hyperv-tlfs.h> +#include <asm/processor.h> struct ms_hyperv_info __read_mostly ms_hyperv; -static const struct hypervisor_ops ops = { - .name = "Hyper-V", -}; +static uint64_t generate_guest_id(void) +{ + union hv_guest_os_id id = {}; + + id.vendor = HV_XEN_VENDOR_ID; + id.major = xen_major_version(); + id.minor = xen_minor_version(); + + return id.raw; +} + +static const struct hypervisor_ops ops; const struct hypervisor_ops *__init hyperv_probe(void) { @@ -72,6 +84,57 @@ const struct hypervisor_ops *__init hyperv_probe(void) return &ops; } +static void __init setup_hypercall_page(void) +{ + union hv_x64_msr_hypercall_contents hypercall_msr; + union hv_guest_os_id guest_id; + unsigned long mfn; + + BUILD_BUG_ON(HV_HYP_PAGE_SHIFT != PAGE_SHIFT); + + rdmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); + if ( !guest_id.raw ) + { + guest_id.raw = generate_guest_id(); + wrmsrl(HV_X64_MSR_GUEST_OS_ID, guest_id.raw); + } + + rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); + if ( !hypercall_msr.enable ) + { + mfn = HV_HCALL_MFN; + hypercall_msr.enable = 1; + hypercall_msr.guest_physical_address = mfn; + wrmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); + } + else + mfn = hypercall_msr.guest_physical_address; + + rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); + BUG_ON(!hypercall_msr.enable); + + set_fixmap_x(FIX_X_HYPERV_HCALL, mfn << PAGE_SHIFT); +} + +static void __init setup(void) +{ + setup_hypercall_page(); +} + +static void __init e820_fixup(struct e820map *e820) +{ + uint64_t s = HV_HCALL_MFN << PAGE_SHIFT; + + if ( !e820_add_range(e820, s, s + PAGE_SIZE, E820_RESERVED) ) + panic("Unable to reserve Hyper-V hypercall range\n"); +} + +static const struct hypervisor_ops ops = { + .name = "Hyper-V", + .setup = setup, + .e820_fixup = e820_fixup, +}; + /* * Local variables: * mode: C diff --git a/xen/include/asm-x86/guest/hyperv-tlfs.h b/xen/include/asm-x86/guest/hyperv-tlfs.h index 091e25cdd1..0a0f3398c1 100644 --- a/xen/include/asm-x86/guest/hyperv-tlfs.h +++ b/xen/include/asm-x86/guest/hyperv-tlfs.h @@ -318,15 +318,16 @@ struct ms_hyperv_tsc_page { * * Bit(s) * 63 - Indicates if the OS is Open Source or not; 1 is Open Source - * 62:56 - Os Type; Linux is 0x100 + * 62:56 - Os Type; Linux 0x1, FreeBSD 0x2, Xen 0x3 * 55:48 - Distro specific identification - * 47:16 - Linux kernel version number + * 47:16 - Guest OS version number * 15:0 - Distro specific identification * * */ #define HV_LINUX_VENDOR_ID 0x8100 +#define HV_XEN_VENDOR_ID 0x8300 union hv_guest_os_id { uint64_t raw; diff --git a/xen/include/asm-x86/guest/hyperv.h b/xen/include/asm-x86/guest/hyperv.h index c7a7f32bd5..1a1b47831c 100644 --- a/xen/include/asm-x86/guest/hyperv.h +++ b/xen/include/asm-x86/guest/hyperv.h @@ -21,6 +21,9 @@ #include <xen/types.h> +/* Use top-most MFN for hypercall page */ +#define HV_HCALL_MFN (((1ull << paddr_bits) - 1) >> HV_HYP_PAGE_SHIFT) + /* * The specification says: "The partition reference time is computed * by the following formula:
Hyper-V uses a technique called overlay page for its hypercall page. It will insert a backing page to the guest when the hypercall functionality is enabled. That means we can use a page that is not backed by real memory for hypercall page. To avoid shattering L0 superpages and treading on any MMIO areas residing in low addresses, use the top-most addressable page for that purpose. Adjust e820 map accordingly. We also need to register Xen's guest OS ID to Hyper-V. Use 0x3 as the vendor ID. Fix the comment in hyperv-tlfs.h while at it. Signed-off-by: Wei Liu <liuwe@microsoft.com> --- v7: 1. Fix a style issue 2. Initialise ID to 0 3. Update commit message v6: 1. Use hv_guest_os_id 2. Use new e820_fixup hook 3. Add a BUILD_BUG_ON v5: 1. use hypervisor_reserve_top_pages 2. add a macro for hypercall page mfn 3. address other misc comments v4: 1. Use fixmap 2. Follow routines listed in TLFS --- xen/arch/x86/guest/hyperv/hyperv.c | 69 +++++++++++++++++++++++-- xen/include/asm-x86/guest/hyperv-tlfs.h | 5 +- xen/include/asm-x86/guest/hyperv.h | 3 ++ 3 files changed, 72 insertions(+), 5 deletions(-)