ID_AA64ISAR1_EL1, AArch64 Instruction Set Attribute Register 1
The ID_AA64ISAR1_EL1 characteristics are:
Purpose
Provides information about the features and instructions implemented in AArch64 state.
For general information about the interpretation of the ID registers, see 'Principles of the ID scheme for fields in ID registers'.
Configuration
If ID_AA64ISAR1_EL1.{API, APA} == {0000, 0000}, then:
- The TCR_EL1.{TBID,TBID0}, TCR_EL2.{TBID0,TBID1}, TCR_EL2.TBID and TCR_EL3.TBID bits are RES0.
- APIAKeyHi_EL1, APIAKeyLo_EL1, APIBKeyHi_EL1, APIBKeyLo_EL1, APDAKeyHi_EL1, APDAKeyLo_EL1, APDBKeyHi_EL1, APDBKeyLo_EL1 are not allocated.
- SCTLR_ELx.EnIA, SCTLR_ELx.EnIB, SCTLR_ELx.EnDA, SCTLR_ELx.EnDB are all RES0.
If ID_AA64ISAR1_EL1.{GPI, GPA, API, APA} == {0000, 0000, 0000, 0000}, then:
Attributes
ID_AA64ISAR1_EL1 is a 64-bit register.
Field descriptions
The ID_AA64ISAR1_EL1 bit assignments are:
63 | 62 | 61 | 60 | 59 | 58 | 57 | 56 | 55 | 54 | 53 | 52 | 51 | 50 | 49 | 48 | 47 | 46 | 45 | 44 | 43 | 42 | 41 | 40 | 39 | 38 | 37 | 36 | 35 | 34 | 33 | 32 |
LS64 | XS | I8MM | DGH | BF16 | SPECRES | SB | FRINTTS | ||||||||||||||||||||||||
GPI | GPA | LRCPC | FCMA | JSCVT | API | APA | DPB | ||||||||||||||||||||||||
31 | 30 | 29 | 28 | 27 | 26 | 25 | 24 | 23 | 22 | 21 | 20 | 19 | 18 | 17 | 16 | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
LS64, bits [63:60]
Indicates support for LD64B and ST64B* instructions, and the ACCDATA_EL1 register. Defined values of this field are:
LS64 | Meaning |
---|---|
0b0000 |
The LD64B and ST64B* instructions, the ACCDATA_EL1 register, and associated traps are not supported. |
0b0001 |
The LD64B and ST64B instructions are supported. |
0b0010 |
The LD64B, ST64B, and ST64BV instructions, and their associated traps are supported. |
0b0011 |
The LD64 and ST64B* instructions, the ACCDATA_EL1 register, and their associated traps are supported. |
All other values are reserved.
FEAT_LS64 implements the functionality identified by 0b0001.
FEAT_LS64_V implements the functionality identified by 0b0010.
FEAT_LS64_ACCDATA implements the functionality identified by 0b0011.
From Armv8.7, the permitted values are 0b0001, 0b0010, and 0b0011.
XS, bits [59:56]
Indicates support for the XS attribute, the TLBI and DSB instructions with the nXS qualifier, and the HCRX_EL2.{FGTnXS, FnXS} fields in AArch64 state. Defined values are:
XS | Meaning |
---|---|
0b0000 |
The XS attribute, the TLBI and DSB instructions with the nXS qualifier, and the HCRX_EL2.{FGTnXS, FnXS} fields are not supported. |
0b0001 |
The XS attribute, the TLBI and DSB instructions with the nXS qualifier, and the HCRX_EL2.{FGTnXS, FnXS} fields are supported. |
All other values are reserved.
FEAT_XS implements the functionality identified by 0b0001.
From Armv8.7, the only permitted value is 0b0001.
I8MM, bits [55:52]
Indicates support for Advanced SIMD and Floating-point Int8 matrix multiplication instructions in AArch64 state. Defined values are:
I8MM | Meaning |
---|---|
0b0000 |
Int8 matrix multiplication instructions are not implemented. |
0b0001 |
SMMLA, SUDOT, UMMLA, USMMLA, and USDOT instructions are implemented. |
All other values are reserved.
FEAT_I8MM implements the functionality identified by 0b0001.
When Advanced SIMD and SVE are both implemented, this field must return the same value as ID_AA64ZFR0_EL1.I8MM.
From Armv8.6, the only permitted value is 0b0001.
DGH, bits [51:48]
Indicates support for the Data Gathering Hint instruction. Defined values are:
DGH | Meaning |
---|---|
0b0000 |
Data Gathering Hint is not implemented. |
0b0001 |
Data Gathering Hint is implemented. |
All other values are reserved.
FEAT_DGH implements the functionality identified by 0b0001.
From ARMv8.0, the permitted values are 0b0000 and 0b0001.
If the DGH instruction has no effect in preventing the merging of memory accesses, the value of this field is 0b0000.
BF16, bits [47:44]
Indicates support for Advanced SIMD and Floating-point BFloat16 instructions in AArch64 state. Defined values are:
BF16 | Meaning |
---|---|
0b0000 |
BFloat16 instructions are not implemented. |
0b0001 |
BFCVT, BFCVTN, BFCVTN2, BFDOT, BFMLALB, BFMLALT, and BFMMLA instructions are implemented. |
All other values are reserved.
FEAT_BF16 implements the functionality identified by 0b0001.
When Advanced SIMD and SVE are both implemented, this field must return the same value as ID_AA64ZFR0_EL1.BF16.
From Armv8.6, the only permitted value is 0b0001.
SPECRES, bits [43:40]
Indicates support for prediction invalidation instructions in AArch64 state. Defined values are:
SPECRES | Meaning |
---|---|
0b0000 |
CFP RCTX, DVP RCTX, and CPP RCTX instructions are not implemented. |
0b0001 |
CFP RCTX, DVP RCTX, and CPP RCTX instructions are implemented. |
All other values are reserved.
FEAT_SPECRES implements the functionality identified by 0b0001.
In Armv8.0, the permitted values are 0b0000 and 0b0001.
From Armv8.5, the only permitted value is 0b0001.
SB, bits [39:36]
Indicates support for SB instruction in AArch64 state. Defined values are:
SB | Meaning |
---|---|
0b0000 |
SB instruction is not implemented. |
0b0001 |
SB instruction is implemented. |
All other values are reserved.
FEAT_SB implements the functionality identified by 0b0001.
In Armv8.0, the permitted values are 0b0000 and 0b0001.
From Armv8.5, the only permitted value is 0b0001.
FRINTTS, bits [35:32]
Indicates support for the FRINT32Z, FRINT32X, FRINT64Z, and FRINT64X instructions are implemented. Defined values are:
FRINTTS | Meaning |
---|---|
0b0000 |
FRINT32Z, FRINT32X, FRINT64Z, and FRINT64X instructions are not implemented. |
0b0001 |
FRINT32Z, FRINT32X, FRINT64Z, and FRINT64X instructions are implemented. |
All other values are reserved.
FEAT_FRINTTS implements the functionality identified by 0b0001.
From Armv8.5, the only permitted value is 0b0001.
GPI, bits [31:28]
Indicates support for an IMPLEMENTATION DEFINED algorithm is implemented in the PE for generic code authentication in AArch64 state. Defined values are:
GPI | Meaning |
---|---|
0b0000 |
Generic Authentication using an IMPLEMENTATION DEFINED algorithm is not implemented. |
0b0001 |
Generic Authentication using an IMPLEMENTATION DEFINED algorithm is implemented. This includes the PACGA instruction. |
All other values are reserved.
From Armv8.3, the permitted values are 0b0000 and 0b0001.
If the value of ID_AA64ISAR1_EL1.GPA is non-zero, this field must have the value 0b0000.
GPA, bits [27:24]
Indicates whether QARMA or Architected algorithm is implemented in the PE for generic code authentication in AArch64 state. Defined values are:
GPA | Meaning |
---|---|
0b0000 |
Generic Authentication using an Architected algorithm is not implemented. |
0b0001 |
Generic Authentication using the QARMA algorithm is implemented. This includes the PACGA instruction. |
All other values are reserved.
From Armv8.3, the permitted values are 0b0000 and 0b0001.
If the value of ID_AA64ISAR1_EL1.GPI is non-zero, this field must have the value 0b0000.
LRCPC, bits [23:20]
Indicates support for weaker release consistency, RCpc, based model. Defined values are:
LRCPC | Meaning |
---|---|
0b0000 |
The LDAPR*, LDAPUR*, and STLUR* instructions are not implemented. |
0b0001 | The LDAPR* instructions are implemented. The LDAPUR*, and STLUR* instructions are not implemented. |
0b0010 |
The LDAPR*, LDAPUR*, and STLUR* instructions are implemented. |
All other values are reserved.
FEAT_LRCPC implements the functionality identified by the value 0b0001.
FEAT_LRCPC2 implements the functionality identified by the value 0b0010.
In Armv8.2, the permitted values are 0b0000, 0b0001, and 0b0010.
In Armv8.3, the permitted values are 0b0001 and 0b0010.
From Armv8.4, the only permitted value is 0b0010.
FCMA, bits [19:16]
Indicates support for complex number addition and multiplication, where numbers are stored in vectors. Defined values are:
FCMA | Meaning |
---|---|
0b0000 |
The FCMLA and FCADD instructions are not implemented. |
0b0001 |
The FCMLA and FCADD instructions are implemented. |
All other values are reserved.
FEAT_FCMA implements the functionality identified by the value 0b0001.
In Armv8.0, Armv8.1, and Armv8.2, the only permitted value is 0b0000.
From Armv8.3, if Advanced SIMD or Floating-point is implemented, the only permitted value is 0b0001.
From Armv8.3, if Advanced SIMD or Floating-point is not implemented, the only permitted value is 0b0000.
JSCVT, bits [15:12]
Indicates support for JavaScript conversion from double precision floating point values to integers in AArch64 state. Defined values are:
JSCVT | Meaning |
---|---|
0b0000 |
The FJCVTZS instruction is not implemented. |
0b0001 |
The FJCVTZS instruction is implemented. |
All other values are reserved.
FEAT_JSCVT implements the functionality identified by 0b0001.
In Armv8.0, Armv8.1, and Armv8.2, the only permitted value is 0b0000.
From Armv8.3, if Advanced SIMD or Floating-point is implemented, the only permitted value is 0b0001.
From Armv8.3, if Advanced SIMD or Floating-point is not implemented, the only permitted value is 0b0000.
API, bits [11:8]
Indicates whether an IMPLEMENTATION DEFINED algorithm is implemented in the PE for address authentication, in AArch64 state. This applies to all Pointer Authentication instructions other than the PACGA instruction. Defined values are:
API | Meaning |
---|---|
0b0000 |
Address Authentication using an IMPLEMENTATION DEFINED algorithm is not implemented. |
0b0001 |
Address Authentication using an IMPLEMENTATION DEFINED algorithm is implemented, with the HaveEnhancedPAC() and HaveEnhancedPAC2() functions returning FALSE. |
0b0010 |
Address Authentication using an IMPLEMENTATION DEFINED algorithm is implemented, with the HaveEnhancedPAC() function returning TRUE, and the HaveEnhancedPAC2() function returning FALSE. |
0b0011 |
Address Authentication using an IMPLEMENTATION DEFINED algorithm is implemented, with the HaveEnhancedPAC2() function returning TRUE, and the HaveEnhancedPAC() function returning FALSE. |
0b0100 |
Address Authentication using an IMPLEMENTATION DEFINED algorithm is implemented, with the HaveEnhancedPAC2() function returning TRUE, the HaveFPAC() function returning TRUE, the HaveFPACCombined() function returning FALSE, and the HaveEnhancedPAC() function returning FALSE. |
0b0101 |
Address Authentication using an IMPLEMENTATION DEFINED algorithm is implemented, with the HaveEnhancedPAC2() function returning TRUE, the HaveFPAC() function returning TRUE, the HaveFPACCombined() function returning TRUE, and the HaveEnhancedPAC() function returning FALSE. |
All other values are reserved.
FEAT_PAuth implements the functionality added by the values 0b0000, 0b0001, and 0b0010.
FEAT_PAuth2 implements the functionality added by the value 0b0011.
FEAT_FPAC implements the functionality added by the values 0b0100 and 0b0101.
From Armv8.6, the permitted values are 0b0011, 0b0100, and 0b0101.
If the value of ID_AA64ISAR1_EL1.APA is non-zero, this field must have the value 0b0000.
APA, bits [7:4]
Indicates whether QARMA or Architected algorithm is implemented in the PE for address authentication, in AArch64 state. This applies to all Pointer Authentication instructions other than the PACGA instruction. Defined values are:
APA | Meaning |
---|---|
0b0000 |
Address Authentication using an Architected algorithm is not implemented. |
0b0001 |
Address Authentication using the QARMA algorithm is implemented, with the HaveEnhancedPAC() and HaveEnhancedPAC2() functions returning FALSE. |
0b0010 |
Address Authentication using the QARMA algorithm is implemented, with the HaveEnhancedPAC() function returning TRUE and the HaveEnhancedPAC2() function returning FALSE. |
0b0011 |
Address Authentication using the QARMA algorithm is implemented, with the HaveEnhancedPAC2() function returning TRUE, the HaveFPAC() function returning FALSE, the HaveFPACCombined() function returning FALSE, and the HaveEnhancedPAC() function returning FALSE. |
0b0100 |
Address Authentication using the QARMA algorithm is implemented, with the HaveEnhancedPAC2() function returning TRUE, the HaveFPAC() function returning TRUE, the HaveFPACCombined() function returning FALSE, and the HaveEnhancedPAC() function returning FALSE. |
0b0101 |
Address Authentication using the QARMA algorithm is implemented, with the HaveEnhancedPAC2() function returning TRUE, the HaveFPAC() function returning TRUE, the HaveFPACCombined() function returning TRUE, and the HaveEnhancedPAC() function returning FALSE. |
All other values are reserved.
FEAT_PAuth implements the functionality added by the values 0b0000, 0b0001, and 0b0010.
FEAT_PAuth2 implements the functionality added by the value 0b0011.
FEAT_FPAC implements the functionality added by the values 0b0100 and 0b0101.
From Armv8.6, the permitted values are 0b0011, 0b0100, and 0b0101.
If the value of the ID_AA64ISAR1_EL1.API is non-zero, this field must have the value 0b0000.
DPB, bits [3:0]
Data Persistence writeback. Indicates support for the DC CVAP and DC CVADP instructions in AArch64 state. Defined values are:
DPB | Meaning |
---|---|
0b0000 |
DC CVAP not supported. |
0b0001 |
DC CVAP supported. |
0b0010 |
All other values are reserved.
FEAT_DPB implements the functionality identified by the value 0b0001.
FEAT_DPB2 implements the functionality identified by the value 0b0010.
In Armv8.2, the permitted values are 0b0001 and 0b0010.
From Armv8.5, the only permitted value is 0b0010.
Accessing the ID_AA64ISAR1_EL1
Accesses to this register use the following encodings:
MRS <Xt>, ID_AA64ISAR1_EL1
op0 | op1 | CRn | CRm | op2 |
---|---|---|---|---|
0b11 | 0b000 | 0b0000 | 0b0110 | 0b001 |
if PSTATE.EL == EL0 then if IsFeatureImplemented(FEAT_IDST) then if EL2Enabled() && HCR_EL2.TGE == '1' then AArch64.SystemAccessTrap(EL2, 0x18); else AArch64.SystemAccessTrap(EL1, 0x18); else UNDEFINED; elsif PSTATE.EL == EL1 then if EL2Enabled() && HCR_EL2.TID3 == '1' then AArch64.SystemAccessTrap(EL2, 0x18); else return ID_AA64ISAR1_EL1; elsif PSTATE.EL == EL2 then return ID_AA64ISAR1_EL1; elsif PSTATE.EL == EL3 then return ID_AA64ISAR1_EL1;