1.. _Arm Emulation: 2 3A-profile CPU architecture support 4================================== 5 6QEMU's TCG emulation includes support for the Armv5, Armv6, Armv7 and 7Armv8 versions of the A-profile architecture. It also has support for 8the following architecture extensions: 9 10- FEAT_AA32BF16 (AArch32 BFloat16 instructions) 11- FEAT_AA32HPD (AArch32 hierarchical permission disables) 12- FEAT_AA32I8MM (AArch32 Int8 matrix multiplication instructions) 13- FEAT_AES (AESD and AESE instructions) 14- FEAT_BBM at level 2 (Translation table break-before-make levels) 15- FEAT_BF16 (AArch64 BFloat16 instructions) 16- FEAT_BTI (Branch Target Identification) 17- FEAT_CSV2 (Cache speculation variant 2) 18- FEAT_CSV2_1p1 (Cache speculation variant 2, version 1.1) 19- FEAT_CSV2_1p2 (Cache speculation variant 2, version 1.2) 20- FEAT_CSV2_2 (Cache speculation variant 2, version 2) 21- FEAT_CSV3 (Cache speculation variant 3) 22- FEAT_DGH (Data gathering hint) 23- FEAT_DIT (Data Independent Timing instructions) 24- FEAT_DPB (DC CVAP instruction) 25- FEAT_Debugv8p2 (Debug changes for v8.2) 26- FEAT_Debugv8p4 (Debug changes for v8.4) 27- FEAT_DotProd (Advanced SIMD dot product instructions) 28- FEAT_DoubleFault (Double Fault Extension) 29- FEAT_E0PD (Preventing EL0 access to halves of address maps) 30- FEAT_ETS (Enhanced Translation Synchronization) 31- FEAT_EVT (Enhanced Virtualization Traps) 32- FEAT_FCMA (Floating-point complex number instructions) 33- FEAT_FGT (Fine-Grained Traps) 34- FEAT_FHM (Floating-point half-precision multiplication instructions) 35- FEAT_FP16 (Half-precision floating-point data processing) 36- FEAT_FRINTTS (Floating-point to integer instructions) 37- FEAT_FlagM (Flag manipulation instructions v2) 38- FEAT_FlagM2 (Enhancements to flag manipulation instructions) 39- FEAT_GTG (Guest translation granule size) 40- FEAT_HAFDBS (Hardware management of the access flag and dirty bit state) 41- FEAT_HCX (Support for the HCRX_EL2 register) 42- FEAT_HPDS (Hierarchical permission disables) 43- FEAT_I8MM (AArch64 Int8 matrix multiplication instructions) 44- FEAT_IDST (ID space trap handling) 45- FEAT_IESB (Implicit error synchronization event) 46- FEAT_JSCVT (JavaScript conversion instructions) 47- FEAT_LOR (Limited ordering regions) 48- FEAT_LPA (Large Physical Address space) 49- FEAT_LPA2 (Large Physical and virtual Address space v2) 50- FEAT_LRCPC (Load-acquire RCpc instructions) 51- FEAT_LRCPC2 (Load-acquire RCpc instructions v2) 52- FEAT_LSE (Large System Extensions) 53- FEAT_LSE2 (Large System Extensions v2) 54- FEAT_LVA (Large Virtual Address space) 55- FEAT_MTE (Memory Tagging Extension) 56- FEAT_MTE2 (Memory Tagging Extension) 57- FEAT_MTE3 (MTE Asymmetric Fault Handling) 58- FEAT_PAN (Privileged access never) 59- FEAT_PAN2 (AT S1E1R and AT S1E1W instruction variants affected by PSTATE.PAN) 60- FEAT_PAN3 (Support for SCTLR_ELx.EPAN) 61- FEAT_PAuth (Pointer authentication) 62- FEAT_PMULL (PMULL, PMULL2 instructions) 63- FEAT_PMUv3p1 (PMU Extensions v3.1) 64- FEAT_PMUv3p4 (PMU Extensions v3.4) 65- FEAT_PMUv3p5 (PMU Extensions v3.5) 66- FEAT_RAS (Reliability, availability, and serviceability) 67- FEAT_RASv1p1 (RAS Extension v1.1) 68- FEAT_RDM (Advanced SIMD rounding double multiply accumulate instructions) 69- FEAT_RME (Realm Management Extension) (NB: support status in QEMU is experimental) 70- FEAT_RNG (Random number generator) 71- FEAT_S2FWB (Stage 2 forced Write-Back) 72- FEAT_SB (Speculation Barrier) 73- FEAT_SEL2 (Secure EL2) 74- FEAT_SHA1 (SHA1 instructions) 75- FEAT_SHA256 (SHA256 instructions) 76- FEAT_SHA3 (Advanced SIMD SHA3 instructions) 77- FEAT_SHA512 (Advanced SIMD SHA512 instructions) 78- FEAT_SM3 (Advanced SIMD SM3 instructions) 79- FEAT_SM4 (Advanced SIMD SM4 instructions) 80- FEAT_SME (Scalable Matrix Extension) 81- FEAT_SME_FA64 (Full A64 instruction set in Streaming SVE mode) 82- FEAT_SME_F64F64 (Double-precision floating-point outer product instructions) 83- FEAT_SME_I16I64 (16-bit to 64-bit integer widening outer product instructions) 84- FEAT_SPECRES (Speculation restriction instructions) 85- FEAT_SSBS (Speculative Store Bypass Safe) 86- FEAT_TLBIOS (TLB invalidate instructions in Outer Shareable domain) 87- FEAT_TLBIRANGE (TLB invalidate range instructions) 88- FEAT_TTCNP (Translation table Common not private translations) 89- FEAT_TTL (Translation Table Level) 90- FEAT_TTST (Small translation tables) 91- FEAT_UAO (Unprivileged Access Override control) 92- FEAT_VHE (Virtualization Host Extensions) 93- FEAT_VMID16 (16-bit VMID) 94- FEAT_XNX (Translation table stage 2 Unprivileged Execute-never) 95- SVE (The Scalable Vector Extension) 96- SVE2 (The Scalable Vector Extension v2) 97 98For information on the specifics of these extensions, please refer 99to the `Armv8-A Arm Architecture Reference Manual 100<https://developer.arm.com/documentation/ddi0487/latest>`_. 101 102When a specific named CPU is being emulated, only those features which 103are present in hardware for that CPU are emulated. (If a feature is 104not in the list above then it is not supported, even if the real 105hardware should have it.) The ``max`` CPU enables all features. 106 107R-profile CPU architecture support 108================================== 109 110QEMU's TCG emulation support for R-profile CPUs is currently limited. 111We emulate only the Cortex-R5 and Cortex-R5F CPUs. 112 113M-profile CPU architecture support 114================================== 115 116QEMU's TCG emulation includes support for Armv6-M, Armv7-M, Armv8-M, and 117Armv8.1-M versions of the M-profile architucture. It also has support 118for the following architecture extensions: 119 120- FP (Floating-point Extension) 121- FPCXT (FPCXT access instructions) 122- HP (Half-precision floating-point instructions) 123- LOB (Low Overhead loops and Branch future) 124- M (Main Extension) 125- MPU (Memory Protection Unit Extension) 126- PXN (Privileged Execute Never) 127- RAS (Reliability, Serviceability and Availability): "minimum RAS Extension" only 128- S (Security Extension) 129- ST (System Timer Extension) 130 131For information on the specifics of these extensions, please refer 132to the `Armv8-M Arm Architecture Reference Manual 133<https://developer.arm.com/documentation/ddi0553/latest>`_. 134 135When a specific named CPU is being emulated, only those features which 136are present in hardware for that CPU are emulated. (If a feature is 137not in the list above then it is not supported, even if the real 138hardware should have it.) There is no equivalent of the ``max`` CPU for 139M-profile. 140