[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PULL 35/49] target-i386: hvf: remove MPX support
From: |
Paolo Bonzini |
Subject: |
[Qemu-devel] [PULL 35/49] target-i386: hvf: remove MPX support |
Date: |
Mon, 21 Jan 2019 19:06:06 +0100 |
MPX support is being phased out by Intel and actually I am not sure that
OS X has ever enabled it in XCR0. Drop it from the Hypervisor.framework
acceleration.
Signed-off-by: Paolo Bonzini <address@hidden>
---
target/i386/hvf/x86_cpuid.c | 18 +-----------------
1 file changed, 1 insertion(+), 17 deletions(-)
diff --git a/target/i386/hvf/x86_cpuid.c b/target/i386/hvf/x86_cpuid.c
index 9874a46..4d957fe 100644
--- a/target/i386/hvf/x86_cpuid.c
+++ b/target/i386/hvf/x86_cpuid.c
@@ -38,16 +38,6 @@ static uint64_t xgetbv(uint32_t xcr)
return (((uint64_t)edx) << 32) | eax;
}
-static bool vmx_mpx_supported()
-{
- uint64_t cap_exit, cap_entry;
-
- hv_vmx_read_capability(HV_VMX_CAP_ENTRY, &cap_entry);
- hv_vmx_read_capability(HV_VMX_CAP_EXIT, &cap_exit);
-
- return ((cap_exit & (1 << 23)) && (cap_entry & (1 << 16)));
-}
-
uint32_t hvf_get_supported_cpuid(uint32_t func, uint32_t idx,
int reg)
{
@@ -92,11 +82,8 @@ uint32_t hvf_get_supported_cpuid(uint32_t func, uint32_t idx,
CPUID_7_0_EBX_CLFLUSHOPT | CPUID_7_0_EBX_CLWB |
CPUID_7_0_EBX_AVX512DQ | CPUID_7_0_EBX_SHA_NI |
CPUID_7_0_EBX_AVX512BW | CPUID_7_0_EBX_AVX512VL |
- CPUID_7_0_EBX_INVPCID | CPUID_7_0_EBX_MPX;
+ CPUID_7_0_EBX_INVPCID;
- if (!vmx_mpx_supported()) {
- ebx &= ~CPUID_7_0_EBX_MPX;
- }
hv_vmx_read_capability(HV_VMX_CAP_PROCBASED2, &cap);
if (!(cap & CPU_BASED2_INVPCID)) {
ebx &= ~CPUID_7_0_EBX_INVPCID;
@@ -119,9 +106,6 @@ uint32_t hvf_get_supported_cpuid(uint32_t func, uint32_t
idx,
XSTATE_BNDCSR_MASK | XSTATE_OPMASK_MASK |
XSTATE_ZMM_Hi256_MASK |
XSTATE_Hi16_ZMM_MASK);
eax &= supp_xcr0;
- if (!vmx_mpx_supported()) {
- eax &= ~(XSTATE_BNDREGS_MASK | XSTATE_BNDCSR_MASK);
- }
} else if (idx == 1) {
hv_vmx_read_capability(HV_VMX_CAP_PROCBASED2, &cap);
eax &= CPUID_XSAVE_XSAVEOPT | CPUID_XSAVE_XGETBV1;
--
1.8.3.1
- [Qemu-devel] [PULL 31/49] optionrom/pvh: load initrd from fw_cfg, (continued)
- [Qemu-devel] [PULL 31/49] optionrom/pvh: load initrd from fw_cfg, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 32/49] i386: remove the new CPUID 'PCONFIG' from Icelake-Server CPU model, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 36/49] contrib/elf2dmp: fix elf.h including, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 30/49] hw/i386/pc: use PVH option rom, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 39/49] contrib/elf2dmp: fix structures definitions, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 26/49] pvh: Boot uncompressed kernel using direct boot ABI, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 27/49] pvh: load initrd and expose it through fw_cfg, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 38/49] contrib/elf2dmp: use GLib in PDB processing, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 41/49] configure: enable elf2dmp build for Windows hosts, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 34/49] Revert "i386: Add CPUID bit for PCONFIG", Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 35/49] target-i386: hvf: remove MPX support,
Paolo Bonzini <=
- [Qemu-devel] [PULL 37/49] contrib/elf2dmp: use GLib in ELF processing, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 02/49] include: update Linux headers to 4.21-rc1/5.0-rc1, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 29/49] linuxboot_dma: move common functions in a new header, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 33/49] i386: remove the 'INTEL_PT' CPUID bit from named CPU models, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 42/49] monitor: do not use QTAILQ_FOREACH_SAFE across critical sections, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 43/49] crypto: finish removing TABs, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 44/49] ui: vnc: finish removing TABs, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 47/49] i386: import & use bootparam.h, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 48/49] i386: allow to load initrd below 4 GB for recent linux, Paolo Bonzini, 2019/01/21
- [Qemu-devel] [PULL 46/49] hw/core/loader.c: Read as long as possible in load_image_size(), Paolo Bonzini, 2019/01/21