Hello, Patch in the bottom extends pmov insns. Bootstrapped. AVX-512* tests on top of patch-set all pass under simulator.
Is it ok for trunk? gcc/ * config/i386/sse.md (define_insn "avx2_<code>v16qiv16hi2<mask_name>"): Add masking. (define_insn "avx512bw_<code>v32qiv32hi2<mask_name>"): New. (define_insn "sse4_1_<code>v8qiv8hi2<mask_name>"): Add masking. (define_insn "avx2_<code>v8qiv8si2<mask_name>"): Ditto. (define_insn "sse4_1_<code>v4qiv4si2<mask_name>"): Ditto. (define_insn "avx2_<code>v8hiv8si2<mask_name>"): Ditto. (define_insn "sse4_1_<code>v4hiv4si2<mask_name>"): Ditto. (define_insn "avx2_<code>v4qiv4di2<mask_name>"): Ditto. (define_insn "sse4_1_<code>v2qiv2di2<mask_name>"): Ditto. (define_insn "avx2_<code>v4hiv4di2<mask_name>"): Ditto. (define_insn "sse4_1_<code>v2hiv2di2<mask_name>"): Ditto. (define_insn "avx2_<code>v4siv4di2<mask_name>"): Ditto. (define_insn "sse4_1_<code>v2siv2di2<mask_name>"): Ditto. -- Thanks, diff --git a/gcc/config/i386/sse.md b/gcc/config/i386/sse.md index 9e0c0e8..a7cc5ad 100644 --- a/gcc/config/i386/sse.md +++ b/gcc/config/i386/sse.md @@ -13330,28 +13330,39 @@ (set_attr "prefix" "maybe_vex") (set_attr "mode" "TI")]) -(define_insn "avx2_<code>v16qiv16hi2" - [(set (match_operand:V16HI 0 "register_operand" "=x") +(define_insn "avx2_<code>v16qiv16hi2<mask_name>" + [(set (match_operand:V16HI 0 "register_operand" "=v") (any_extend:V16HI - (match_operand:V16QI 1 "nonimmediate_operand" "xm")))] - "TARGET_AVX2" - "vpmov<extsuffix>bw\t{%1, %0|%0, %1}" + (match_operand:V16QI 1 "nonimmediate_operand" "vm")))] + "TARGET_AVX2 && <mask_avx512bw_condition> && <mask_avx512vl_condition>" + "vpmov<extsuffix>bw\t{%1, %0<mask_operand2>|%0<mask_operand2>, %1}" [(set_attr "type" "ssemov") (set_attr "prefix_extra" "1") - (set_attr "prefix" "vex") + (set_attr "prefix" "maybe_evex") (set_attr "mode" "OI")]) -(define_insn "sse4_1_<code>v8qiv8hi2" - [(set (match_operand:V8HI 0 "register_operand" "=x") +(define_insn "avx512bw_<code>v32qiv32hi2<mask_name>" + [(set (match_operand:V32HI 0 "register_operand" "=v") + (any_extend:V32HI + (match_operand:V32QI 1 "nonimmediate_operand" "vm")))] + "TARGET_AVX512BW" + "vpmov<extsuffix>bw\t{%1, %0<mask_operand2>|%0<mask_operand2>, %1}" + [(set_attr "type" "ssemov") + (set_attr "prefix_extra" "1") + (set_attr "prefix" "evex") + (set_attr "mode" "XI")]) + +(define_insn "sse4_1_<code>v8qiv8hi2<mask_name>" + [(set (match_operand:V8HI 0 "register_operand" "=v") (any_extend:V8HI (vec_select:V8QI - (match_operand:V16QI 1 "nonimmediate_operand" "xm") + (match_operand:V16QI 1 "nonimmediate_operand" "vm") (parallel [(const_int 0) (const_int 1) (const_int 2) (const_int 3) (const_int 4) (const_int 5) (const_int 6) (const_int 7)]))))] - "TARGET_SSE4_1" - "%vpmov<extsuffix>bw\t{%1, %0|%0, %q1}" + "TARGET_SSE4_1 && <mask_avx512bw_condition> && <mask_avx512vl_condition>" + "%vpmov<extsuffix>bw\t{%1, %0<mask_operand2>|%0<mask_operand2>, %q1}" [(set_attr "type" "ssemov") (set_attr "ssememalign" "64") (set_attr "prefix_extra" "1") @@ -13368,31 +13379,31 @@ (set_attr "prefix" "evex") (set_attr "mode" "XI")]) -(define_insn "avx2_<code>v8qiv8si2" - [(set (match_operand:V8SI 0 "register_operand" "=x") +(define_insn "avx2_<code>v8qiv8si2<mask_name>" + [(set (match_operand:V8SI 0 "register_operand" "=v") (any_extend:V8SI (vec_select:V8QI - (match_operand:V16QI 1 "nonimmediate_operand" "xm") + (match_operand:V16QI 1 "nonimmediate_operand" "vm") (parallel [(const_int 0) (const_int 1) (const_int 2) (const_int 3) (const_int 4) (const_int 5) (const_int 6) (const_int 7)]))))] - "TARGET_AVX2" - "vpmov<extsuffix>bd\t{%1, %0|%0, %q1}" + "TARGET_AVX2 && <mask_avx512vl_condition>" + "vpmov<extsuffix>bd\t{%1, %0<mask_operand2>|%0<mask_operand2>, %q1}" [(set_attr "type" "ssemov") (set_attr "prefix_extra" "1") - (set_attr "prefix" "vex") + (set_attr "prefix" "maybe_evex") (set_attr "mode" "OI")]) -(define_insn "sse4_1_<code>v4qiv4si2" - [(set (match_operand:V4SI 0 "register_operand" "=x") +(define_insn "sse4_1_<code>v4qiv4si2<mask_name>" + [(set (match_operand:V4SI 0 "register_operand" "=v") (any_extend:V4SI (vec_select:V4QI - (match_operand:V16QI 1 "nonimmediate_operand" "xm") + (match_operand:V16QI 1 "nonimmediate_operand" "vm") (parallel [(const_int 0) (const_int 1) (const_int 2) (const_int 3)]))))] - "TARGET_SSE4_1" - "%vpmov<extsuffix>bd\t{%1, %0|%0, %k1}" + "TARGET_SSE4_1 && <mask_avx512vl_condition>" + "%vpmov<extsuffix>bd\t{%1, %0<mask_operand2>|%0<mask_operand2>, %k1}" [(set_attr "type" "ssemov") (set_attr "ssememalign" "32") (set_attr "prefix_extra" "1") @@ -13409,26 +13420,26 @@ (set_attr "prefix" "evex") (set_attr "mode" "XI")]) -(define_insn "avx2_<code>v8hiv8si2" - [(set (match_operand:V8SI 0 "register_operand" "=x") +(define_insn "avx2_<code>v8hiv8si2<mask_name>" + [(set (match_operand:V8SI 0 "register_operand" "=v") (any_extend:V8SI - (match_operand:V8HI 1 "nonimmediate_operand" "xm")))] - "TARGET_AVX2" - "vpmov<extsuffix>wd\t{%1, %0|%0, %1}" + (match_operand:V8HI 1 "nonimmediate_operand" "vm")))] + "TARGET_AVX2 && <mask_avx512vl_condition>" + "vpmov<extsuffix>wd\t{%1, %0<mask_operand2>|%0<mask_operand2>, %1}" [(set_attr "type" "ssemov") (set_attr "prefix_extra" "1") - (set_attr "prefix" "vex") + (set_attr "prefix" "maybe_evex") (set_attr "mode" "OI")]) -(define_insn "sse4_1_<code>v4hiv4si2" - [(set (match_operand:V4SI 0 "register_operand" "=x") +(define_insn "sse4_1_<code>v4hiv4si2<mask_name>" + [(set (match_operand:V4SI 0 "register_operand" "=v") (any_extend:V4SI (vec_select:V4HI - (match_operand:V8HI 1 "nonimmediate_operand" "xm") + (match_operand:V8HI 1 "nonimmediate_operand" "vm") (parallel [(const_int 0) (const_int 1) (const_int 2) (const_int 3)]))))] - "TARGET_SSE4_1" - "%vpmov<extsuffix>wd\t{%1, %0|%0, %q1}" + "TARGET_SSE4_1 && <mask_avx512vl_condition>" + "%vpmov<extsuffix>wd\t{%1, %0<mask_operand2>|%0<mask_operand2>, %q1}" [(set_attr "type" "ssemov") (set_attr "ssememalign" "64") (set_attr "prefix_extra" "1") @@ -13450,28 +13461,28 @@ (set_attr "prefix" "evex") (set_attr "mode" "XI")]) -(define_insn "avx2_<code>v4qiv4di2" - [(set (match_operand:V4DI 0 "register_operand" "=x") +(define_insn "avx2_<code>v4qiv4di2<mask_name>" + [(set (match_operand:V4DI 0 "register_operand" "=v") (any_extend:V4DI (vec_select:V4QI - (match_operand:V16QI 1 "nonimmediate_operand" "xm") + (match_operand:V16QI 1 "nonimmediate_operand" "vm") (parallel [(const_int 0) (const_int 1) (const_int 2) (const_int 3)]))))] - "TARGET_AVX2" - "vpmov<extsuffix>bq\t{%1, %0|%0, %k1}" + "TARGET_AVX2 && <mask_avx512vl_condition>" + "vpmov<extsuffix>bq\t{%1, %0<mask_operand2>|%0<mask_operand2>, %k1}" [(set_attr "type" "ssemov") (set_attr "prefix_extra" "1") - (set_attr "prefix" "vex") + (set_attr "prefix" "maybe_evex") (set_attr "mode" "OI")]) -(define_insn "sse4_1_<code>v2qiv2di2" - [(set (match_operand:V2DI 0 "register_operand" "=x") +(define_insn "sse4_1_<code>v2qiv2di2<mask_name>" + [(set (match_operand:V2DI 0 "register_operand" "=v") (any_extend:V2DI (vec_select:V2QI - (match_operand:V16QI 1 "nonimmediate_operand" "xm") + (match_operand:V16QI 1 "nonimmediate_operand" "vm") (parallel [(const_int 0) (const_int 1)]))))] - "TARGET_SSE4_1" - "%vpmov<extsuffix>bq\t{%1, %0|%0, %w1}" + "TARGET_SSE4_1 && <mask_avx512vl_condition>" + "%vpmov<extsuffix>bq\t{%1, %0<mask_operand2>|%0<mask_operand2>, %w1}" [(set_attr "type" "ssemov") (set_attr "ssememalign" "16") (set_attr "prefix_extra" "1") @@ -13488,28 +13499,28 @@ (set_attr "prefix" "evex") (set_attr "mode" "XI")]) -(define_insn "avx2_<code>v4hiv4di2" - [(set (match_operand:V4DI 0 "register_operand" "=x") +(define_insn "avx2_<code>v4hiv4di2<mask_name>" + [(set (match_operand:V4DI 0 "register_operand" "=v") (any_extend:V4DI (vec_select:V4HI - (match_operand:V8HI 1 "nonimmediate_operand" "xm") + (match_operand:V8HI 1 "nonimmediate_operand" "vm") (parallel [(const_int 0) (const_int 1) (const_int 2) (const_int 3)]))))] - "TARGET_AVX2" - "vpmov<extsuffix>wq\t{%1, %0|%0, %q1}" + "TARGET_AVX2 && <mask_avx512vl_condition>" + "vpmov<extsuffix>wq\t{%1, %0<mask_operand2>|%0<mask_operand2>, %q1}" [(set_attr "type" "ssemov") (set_attr "prefix_extra" "1") - (set_attr "prefix" "vex") + (set_attr "prefix" "maybe_evex") (set_attr "mode" "OI")]) -(define_insn "sse4_1_<code>v2hiv2di2" - [(set (match_operand:V2DI 0 "register_operand" "=x") +(define_insn "sse4_1_<code>v2hiv2di2<mask_name>" + [(set (match_operand:V2DI 0 "register_operand" "=v") (any_extend:V2DI (vec_select:V2HI - (match_operand:V8HI 1 "nonimmediate_operand" "xm") + (match_operand:V8HI 1 "nonimmediate_operand" "vm") (parallel [(const_int 0) (const_int 1)]))))] - "TARGET_SSE4_1" - "%vpmov<extsuffix>wq\t{%1, %0|%0, %k1}" + "TARGET_SSE4_1 && <mask_avx512vl_condition>" + "%vpmov<extsuffix>wq\t{%1, %0<mask_operand2>|%0<mask_operand2>, %k1}" [(set_attr "type" "ssemov") (set_attr "ssememalign" "32") (set_attr "prefix_extra" "1") @@ -13526,24 +13537,25 @@ (set_attr "prefix" "evex") (set_attr "mode" "XI")]) -(define_insn "avx2_<code>v4siv4di2" - [(set (match_operand:V4DI 0 "register_operand" "=x") +(define_insn "avx2_<code>v4siv4di2<mask_name>" + [(set (match_operand:V4DI 0 "register_operand" "=v") (any_extend:V4DI - (match_operand:V4SI 1 "nonimmediate_operand" "xm")))] - "TARGET_AVX2" - "vpmov<extsuffix>dq\t{%1, %0|%0, %1}" + (match_operand:V4SI 1 "nonimmediate_operand" "vm")))] + "TARGET_AVX2 && <mask_avx512vl_condition>" + "vpmov<extsuffix>dq\t{%1, %0<mask_operand2>|%0<mask_operand2>, %1}" [(set_attr "type" "ssemov") + (set_attr "prefix" "maybe_evex") (set_attr "prefix_extra" "1") (set_attr "mode" "OI")]) -(define_insn "sse4_1_<code>v2siv2di2" - [(set (match_operand:V2DI 0 "register_operand" "=x") +(define_insn "sse4_1_<code>v2siv2di2<mask_name>" + [(set (match_operand:V2DI 0 "register_operand" "=v") (any_extend:V2DI (vec_select:V2SI - (match_operand:V4SI 1 "nonimmediate_operand" "xm") + (match_operand:V4SI 1 "nonimmediate_operand" "vm") (parallel [(const_int 0) (const_int 1)]))))] - "TARGET_SSE4_1" - "%vpmov<extsuffix>dq\t{%1, %0|%0, %q1}" + "TARGET_SSE4_1 && <mask_avx512vl_condition>" + "%vpmov<extsuffix>dq\t{%1, %0<mask_operand2>|%0<mask_operand2>, %q1}" [(set_attr "type" "ssemov") (set_attr "ssememalign" "64") (set_attr "prefix_extra" "1")