From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Cc: "Justin Terry (VM)" <juterry@microsoft.com>
Subject: [Qemu-devel] [PULL 41/60] WHPX: register for unrecognized MSR exits
Date: Thu, 28 Jun 2018 22:04:51 +0200 [thread overview]
Message-ID: <1530216310-52873-42-git-send-email-pbonzini@redhat.com> (raw)
In-Reply-To: <1530216310-52873-1-git-send-email-pbonzini@redhat.com>
From: "Justin Terry (VM)" <juterry@microsoft.com>
Some variations of Linux kernels end up accessing MSR's that the Windows
Hypervisor doesn't implement which causes a GP to be returned to the guest.
This fix registers QEMU for unimplemented MSR access and globally returns 0 on
reads and ignores writes. This behavior is allows the Linux kernel to probe the
MSR with a write/read/check sequence it does often without failing the access.
Signed-off-by: Justin Terry (VM) <juterry@microsoft.com>
Message-Id: <20180605221500.21674-2-juterry@microsoft.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
target/i386/whpx-all.c | 41 ++++++++++++++++++++++++++++++++++++++---
1 file changed, 38 insertions(+), 3 deletions(-)
diff --git a/target/i386/whpx-all.c b/target/i386/whpx-all.c
index 99501ba..57e53e1 100644
--- a/target/i386/whpx-all.c
+++ b/target/i386/whpx-all.c
@@ -932,6 +932,7 @@ static int whpx_vcpu_run(CPUState *cpu)
case WHvRunVpExitReasonX64InterruptWindow:
vcpu->window_registered = 0;
+ ret = 0;
break;
case WHvRunVpExitReasonX64Halt:
@@ -943,6 +944,40 @@ static int whpx_vcpu_run(CPUState *cpu)
ret = 1;
break;
+ case WHvRunVpExitReasonX64MsrAccess: {
+ WHV_REGISTER_VALUE reg_values[3] = {0};
+ WHV_REGISTER_NAME reg_names[3];
+ UINT32 reg_count;
+
+ reg_names[0] = WHvX64RegisterRip;
+ reg_names[1] = WHvX64RegisterRax;
+ reg_names[2] = WHvX64RegisterRdx;
+
+ reg_values[0].Reg64 =
+ vcpu->exit_ctx.VpContext.Rip +
+ vcpu->exit_ctx.VpContext.InstructionLength;
+
+ /*
+ * For all unsupported MSR access we:
+ * ignore writes
+ * return 0 on read.
+ */
+ reg_count = vcpu->exit_ctx.MsrAccess.AccessInfo.IsWrite ?
+ 1 : 3;
+
+ hr = whp_dispatch.WHvSetVirtualProcessorRegisters(
+ whpx->partition,
+ cpu->cpu_index,
+ reg_names, reg_count,
+ reg_values);
+
+ if (FAILED(hr)) {
+ error_report("WHPX: Failed to set MsrAccess state "
+ " registers, hr=%08lx", hr);
+ }
+ ret = 0;
+ break;
+ }
case WHvRunVpExitReasonX64Cpuid: {
WHV_REGISTER_VALUE reg_values[5];
WHV_REGISTER_NAME reg_names[5];
@@ -1010,7 +1045,6 @@ static int whpx_vcpu_run(CPUState *cpu)
case WHvRunVpExitReasonUnrecoverableException:
case WHvRunVpExitReasonInvalidVpRegisterValue:
case WHvRunVpExitReasonUnsupportedFeature:
- case WHvRunVpExitReasonX64MsrAccess:
case WHvRunVpExitReasonException:
default:
error_report("WHPX: Unexpected VP exit code %d",
@@ -1378,6 +1412,7 @@ static int whpx_accel_init(MachineState *ms)
}
memset(&prop, 0, sizeof(WHV_PARTITION_PROPERTY));
+ prop.ExtendedVmExits.X64MsrExit = 1;
prop.ExtendedVmExits.X64CpuidExit = 1;
hr = whp_dispatch.WHvSetPartitionProperty(
whpx->partition,
@@ -1386,8 +1421,8 @@ static int whpx_accel_init(MachineState *ms)
sizeof(WHV_PARTITION_PROPERTY));
if (FAILED(hr)) {
- error_report("WHPX: Failed to enable partition extended X64CpuidExit"
- " hr=%08lx", hr);
+ error_report("WHPX: Failed to enable partition extended X64MsrExit and"
+ " X64CpuidExit hr=%08lx", hr);
ret = -EINVAL;
goto error;
}
--
1.8.3.1
next prev parent reply other threads:[~2018-06-28 20:05 UTC|newest]
Thread overview: 64+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-06-28 20:04 [Qemu-devel] [PULL 00/60] Misc patches for soft freeze Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 01/60] exec: Fix MAP_RAM for cached access Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 02/60] move public invalidate APIs out of translate-all.{c, h}, clean up Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 03/60] chardev: don't splatter terminal settings on exit if not previously set Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 04/60] main-loop: document IOCanReadHandler Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 05/60] hw/char/serial: Only retry if qemu_chr_fe_write returns 0 Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 06/60] target/i386: Fix BLSR and BLSI Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 07/60] whpx: commit missing file Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 08/60] memory-device: turn alignment assert into check Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 09/60] exec: check that alignment is a power of two Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 10/60] kvm: Delete the slot if and only if the KVM_MEM_READONLY flag is changed Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 11/60] tests/atomic_add-bench: add -m option to use mutexes Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 12/60] qemu-thread: introduce qemu-thread-common.h Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 13/60] QemuMutex: support --enable-debug-mutex Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 14/60] configure: enable debug-mutex if debug enabled Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 15/60] Replace '-enable-kvm' with '-accel kvm' in docs and help texts Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 16/60] qemu-options: Add missing newline to -accel help text Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 17/60] pc-dimm: remove leftover "struct pc_dimms_capacity" Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 18/60] pc: rename pc_dimm_(plug|unplug|...)* into pc_memory_(plug|unplug|...)* Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 19/60] pc-dimm: rename pc_dimm_memory_* to pc_dimm_* Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 20/60] pc-dimm: remove pc_dimm_get_free_slot() from header Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 21/60] pc: factor out pc specific dimm checks into pc_memory_pre_plug() Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 22/60] nvdimm: no need to overwrite get_vmstate_memory_region() Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 23/60] hostmem: drop error variable from host_memory_backend_get_memory() Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 24/60] pc-dimm: merge get_(vmstate_)memory_region() Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 25/60] nvdimm: convert "unarmed" into a static property Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 26/60] nvdimm: convert nvdimm_mr into a pointer Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 27/60] nvdimm: make get_memory_region() perform checks and initialization Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 28/60] pc-dimm: get_memory_region() will not fail after realize Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 29/60] numa: report all DIMM/NVDIMMs as plugged memory Paolo Bonzini
2018-06-28 20:12 ` David Hildenbrand
2018-06-28 20:04 ` [Qemu-devel] [PULL 30/60] osdep: work around Coverity parsing errors Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 31/60] Deprecate the -enable-hax option Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 32/60] pr-helper: fix --socket-path default in help Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 33/60] pr-helper: fix assertion failure on failed multipath PERSISTENT RESERVE IN Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 34/60] pr-manager-helper: avoid SIGSEGV when writing to the socket fail Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 35/60] pr-manager: put stubs in .c file Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 36/60] pr-manager: add query-pr-managers QMP command Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 37/60] pr-manager-helper: report event on connection/disconnection Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 38/60] hw/mips/jazz: create ESP device directly via qdev Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 39/60] esp: remove legacy esp_init() function Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 40/60] WHPX workaround bug in OSVW handling Paolo Bonzini
2018-06-28 20:04 ` Paolo Bonzini [this message]
2018-06-28 20:04 ` [Qemu-devel] [PULL 42/60] memory/hmp: Print owners/parents in "info mtree" Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 43/60] target-i386: Add NMI interception to SVM Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 44/60] target-i386: Allow interrupt injection after STGI Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 45/60] target-i386: Mark cpu_vmexit noreturn Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 46/60] doc: another fix to "info pic" Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 47/60] ioapic: support " Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 48/60] ioapic: some proper indents when dump info Paolo Bonzini
2018-06-28 20:04 ` [Qemu-devel] [PULL 49/60] ioapic: support "info irq" Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 50/60] hmp: obsolete "info ioapic" Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 51/60] kvm: support -overcommit cpu-pm=on|off Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 52/60] i386/cpu: make -cpu host support monitor/mwait Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 53/60] dump: add Windows dump format to dump-guest-memory Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 54/60] dump: use system context in Windows dump Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 55/60] dump: add fallback KDBG using " Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 56/60] dump: add Windows live system dump Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 57/60] hw/scsi: cleanups before VPD BL emulation Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 58/60] hw/scsi: centralize SG_IO calls into single function Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 59/60] hw/scsi: add VPD Block Limits emulation Paolo Bonzini
2018-06-28 20:05 ` [Qemu-devel] [PULL 60/60] tests/boot-serial: Do not delete the output file in case of errors Paolo Bonzini
2018-06-29 9:25 ` [Qemu-devel] [PULL 00/60] Misc patches for soft freeze Peter Maydell
2018-06-29 9:44 ` Paolo Bonzini
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1530216310-52873-42-git-send-email-pbonzini@redhat.com \
--to=pbonzini@redhat.com \
--cc=juterry@microsoft.com \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).