From 971eda734150ea9cdea47be259486c3a8d087037 Mon Sep 17 00:00:00 2001 From: Przemyslaw Wirkus Date: Wed, 17 Nov 2021 19:21:33 +0000 Subject: aarch64: [SME] Add SME instructions Patch is adding new SME matrix instructions. Please note additional instructions will be added in following patches. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_zada_operand): New parser. * config/tc-aarch64.c (parse_reg_with_qual): New reg parser. * config/tc-aarch64.c (R_ZA): New egister type. (parse_operands): New parser. * testsuite/gas/aarch64/sme-illegal.d: New test. * testsuite/gas/aarch64/sme-illegal.l: New test. * testsuite/gas/aarch64/sme-illegal.s: New test. * testsuite/gas/aarch64/sme.d: New test. * testsuite/gas/aarch64/sme.s: New test. * testsuite/gas/aarch64/sme-f64.d: New test. * testsuite/gas/aarch64/sme-f64.s: New test. * testsuite/gas/aarch64/sme-i64.d: New test. * testsuite/gas/aarch64/sme-i64.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands AARCH64_OPND_SME_ZAda_2b, AARCH64_OPND_SME_ZAda_3b and AARCH64_OPND_SME_Pm. (enum aarch64_insn_class): New instruction class sme_misc. opcodes/ChangeLog: * aarch64-opc.c (aarch64_print_operand): Print OPND_SME_ZAda_2b and OPND_SME_ZAda_3b operands. (verify_constraints): Handle OPND_SME_Pm. * aarch64-opc.h (enum aarch64_field_kind): New bit fields FLD_SME_ZAda_2b, FLD_SME_ZAda_3b and FLD_SME_Pm. * aarch64-tbl.h (OP_SME_ZADA_PN_PM_ZN_S): New qualifier set. (OP_SME_ZADA_PN_PM_ZN_D): New qualifier. (OP_SME_ZADA_PN_PM_ZN_ZM): New qualifier. (OP_SME_ZADA_S_PM_PM_S_S): New qualifier. (OP_SME_ZADA_D_PM_PM_D_D): New qualifier. (OP_SME_ZADA_S_PM_PM_H_H): New qualifier. (OP_SME_ZADA_S_PM_PM_B_B): New qualifier. (OP_SME_ZADA_D_PM_PM_H_H): New qualifier. (SME_INSN): New instruction macro. (SME_F64_INSN): New instruction macro. (SME_I64_INSN): New instruction macro. (SME_INSNC): New instruction macro. (struct aarch64_opcode): New SME instructions. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate. --- opcodes/aarch64-tbl.h | 83 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 83 insertions(+) (limited to 'opcodes/aarch64-tbl.h') diff --git a/opcodes/aarch64-tbl.h b/opcodes/aarch64-tbl.h index d63b081..f725f2f 100644 --- a/opcodes/aarch64-tbl.h +++ b/opcodes/aarch64-tbl.h @@ -2168,6 +2168,42 @@ { \ QLF3(X,X,NIL), \ } +/* e.g. ADDVA .S, /M, /M, .S */ +#define OP_SME_ZADA_PN_PM_ZN_S \ +{ \ + QLF4(S_S,P_M,P_M,S_S), \ +} +/* e.g. ADDVA .D, /M, /M, .D */ +#define OP_SME_ZADA_PN_PM_ZN_D \ +{ \ + QLF4(S_D,P_M,P_M,S_D), \ +} +/* e.g. BFMOPA .S, /M, /M, .H, .H */ +#define OP_SME_ZADA_PN_PM_ZN_ZM \ +{ \ + QLF5(S_S,P_M,P_M,S_H,S_H), \ +} +#define OP_SME_ZADA_S_PM_PM_S_S \ +{ \ + QLF5(S_S,P_M,P_M,S_S,S_S) \ +} +#define OP_SME_ZADA_D_PM_PM_D_D \ +{ \ + QLF5(S_D,P_M,P_M,S_D,S_D) \ +} +#define OP_SME_ZADA_S_PM_PM_H_H \ +{ \ + QLF5(S_S,P_M,P_M,S_H,S_H) \ +} +#define OP_SME_ZADA_S_PM_PM_B_B \ +{ \ + QLF5(S_S,P_M,P_M,S_B,S_B) \ +} +#define OP_SME_ZADA_D_PM_PM_H_H \ +{ \ + QLF5(S_D,P_M,P_M,S_H,S_H) \ +} + /* e.g. UDOT .2S, .8B, .8B. */ #define QL_V3DOT \ { \ @@ -2564,6 +2600,18 @@ static const aarch64_feature_set aarch64_feature_flagm = #define SVE2SM4_INSNC(NAME,OPCODE,MASK,CLASS,OP,OPS,QUALS,FLAGS,CONSTRAINTS,TIED) \ { NAME, OPCODE, MASK, CLASS, OP, SVE2_SM4, OPS, QUALS, \ FLAGS | F_STRICT, CONSTRAINTS, TIED, NULL } +#define SME_INSN(NAME,OPCODE,MASK,CLASS,OP,OPS,QUALS,FLAGS,TIED) \ + { NAME, OPCODE, MASK, CLASS, OP, SME, OPS, QUALS, \ + FLAGS, 0, TIED, NULL } +#define SME_F64_INSN(NAME,OPCODE,MASK,CLASS,OP,OPS,QUALS,FLAGS,TIED) \ + { NAME, OPCODE, MASK, CLASS, OP, SME_F64, OPS, QUALS, \ + FLAGS, 0, TIED, NULL } +#define SME_I64_INSN(NAME,OPCODE,MASK,CLASS,OP,OPS,QUALS,FLAGS,TIED) \ + { NAME, OPCODE, MASK, CLASS, OP, SME_I64, OPS, QUALS, \ + FLAGS, 0, TIED, NULL } +#define SME_INSNC(NAME,OPCODE,MASK,CLASS,OP,OPS,QUALS,FLAGS,CONSTRAINTS,TIED) \ + { NAME, OPCODE, MASK, CLASS, OP, SME, OPS, QUALS, \ + FLAGS, CONSTRAINTS, TIED, NULL } #define SVE2BITPERM_INSN(NAME,OPCODE,MASK,CLASS,OP,OPS,QUALS,FLAGS,TIED) \ { NAME, OPCODE, MASK, CLASS, OP, SVE2_BITPERM, OPS, QUALS, \ FLAGS | F_STRICT, 0, TIED, NULL } @@ -5045,6 +5093,35 @@ const struct aarch64_opcode aarch64_opcode_table[] = SVE2BITPERM_INSN ("bdep", 0x4500b400, 0xff20fc00, sve_size_bhsd, 0, OP3 (SVE_Zd, SVE_Zn, SVE_Zm_16), OP_SVE_VVV_BHSD, 0, 0), SVE2BITPERM_INSN ("bext", 0x4500b000, 0xff20fc00, sve_size_bhsd, 0, OP3 (SVE_Zd, SVE_Zn, SVE_Zm_16), OP_SVE_VVV_BHSD, 0, 0), SVE2BITPERM_INSN ("bgrp", 0x4500b800, 0xff20fc00, sve_size_bhsd, 0, OP3 (SVE_Zd, SVE_Zn, SVE_Zm_16), OP_SVE_VVV_BHSD, 0, 0), + /* SME instructions. */ + SME_INSN ("addha", 0xc0900000, 0xffff001c, sme_misc, 0, OP4 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn), OP_SME_ZADA_PN_PM_ZN_S, 0, 0), + SME_I64_INSN ("addha", 0xc0d00000, 0xffff0018, sme_misc, 0, OP4 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn), OP_SME_ZADA_PN_PM_ZN_D, 0, 0), + SME_INSN ("addva", 0xc0910000, 0xffff001c, sme_misc, 0, OP4 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn), OP_SME_ZADA_PN_PM_ZN_S, 0, 0), + SME_I64_INSN ("addva", 0xc0d10000, 0xffff0018, sme_misc, 0, OP4 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn), OP_SME_ZADA_PN_PM_ZN_D, 0, 0), + SME_INSN ("bfmopa", 0x81800000, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_PN_PM_ZN_ZM, 0, 0), + SME_INSN ("bfmops", 0x81800010, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_PN_PM_ZN_ZM, 0, 0), + SME_INSN ("fmopa", 0x80800000, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_S_S, 0, 0), + SME_F64_INSN ("fmopa", 0x80c00000, 0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_D_D, 0, 0), + SME_INSN ("fmopa", 0x81a00000, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_H_H, 0, 0), + SME_INSN ("fmops", 0x80800010, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_S_S, 0, 0), + SME_F64_INSN ("fmops", 0x80c00010, 0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_D_D, 0, 0), + SME_INSN ("fmops", 0x81a00010, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_H_H, 0, 0), + SME_INSN ("smopa", 0xa0800000, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_B_B, 0, 0), + SME_I64_INSN ("smopa", 0xa0c00000, 0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_H_H, 0, 0), + SME_INSN ("smops", 0xa0800010, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_B_B, 0, 0), + SME_I64_INSN ("smops", 0xa0c00010, 0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_H_H, 0, 0), + SME_INSN ("sumopa", 0xa0a00000, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_B_B, 0, 0), + SME_I64_INSN ("sumopa", 0xa0e00000, 0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_H_H, 0, 0), + SME_INSN ("sumops", 0xa0a00010 ,0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_B_B, 0, 0), + SME_I64_INSN ("sumops", 0xa0e00010 ,0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_H_H, 0, 0), + SME_INSN ("umopa", 0xa1a00000, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_B_B, 0, 0), + SME_I64_INSN ("umopa", 0xa1e00000, 0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_H_H, 0, 0), + SME_INSN ("umops", 0xa1a00010 ,0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_B_B, 0, 0), + SME_I64_INSN ("umops", 0xa1e00010 ,0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_H_H, 0, 0), + SME_INSN ("usmopa", 0xa1800000, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_B_B, 0, 0), + SME_I64_INSN ("usmopa", 0xa1c00000, 0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_H_H, 0, 0), + SME_INSN ("usmops", 0xa1800010, 0xffe0001c, sme_misc, 0, OP5 (SME_ZAda_2b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_S_PM_PM_B_B, 0, 0), + SME_I64_INSN ("usmops", 0xa1c00010, 0xffe00018, sme_misc, 0, OP5 (SME_ZAda_3b, SVE_Pg3, SME_Pm, SVE_Zn, SVE_Zm_16), OP_SME_ZADA_D_PM_PM_H_H, 0, 0), /* SIMD Dot Product (optional in v8.2-A). */ DOT_INSN ("udot", 0x2e009400, 0xbf20fc00, dotproduct, OP3 (Vd, Vn, Vm), QL_V3DOT, F_SIZEQ), @@ -5613,6 +5690,12 @@ const struct aarch64_opcode aarch64_opcode_table[] = "an SVE vector register") \ Y(SVE_REG, sve_reglist, "SVE_ZtxN", 0, F(FLD_SVE_Zt), \ "a list of SVE vector registers") \ + Y(SVE_REG, regno, "SME_ZAda_2b", 0, F(FLD_SME_ZAda_2b), \ + "an SME ZA tile ZA0-ZA3") \ + Y(SVE_REG, regno, "SME_ZAda_3b", 0, F(FLD_SME_ZAda_3b), \ + "an SME ZA tile ZA0-ZA7") \ + Y(PRED_REG, regno, "SME_Pm", 0, F(FLD_SME_Pm), \ + "an SVE predicate register") \ Y(IMMEDIATE, imm, "TME_UIMM16", 0, F(FLD_imm16), \ "a 16-bit unsigned immediate for TME tcancel") \ Y(SIMD_ELEMENT, reglane, "SM3_IMM2", 0, F(FLD_SM3_imm2), \ -- cgit v1.1