ab.Put1(byte(p.From.Offset))

			case Zevex_i_rm_k_r:
				imm, from, kmask, to := unpackOps4(p)
				ab.evex = newEVEXBits(z, &o.op)
				ab.asmevex(ctxt, p, from, nil, to, kmask)
				ab.asmand(ctxt, cursym, p, from, to)
				ab.Put1(byte(imm.Offset))

			case Zevex_i_rm_v_r:
				imm, from, from3, to := unpackOps4(p)
				ab.evex = newEVEXBits(z, &o.op)

		}
		buf.WriteString(s3Select.Output.CSVArgs.RecordDelimiter)

		return nil
	case jsonFormat:
		err := record.WriteJSON(buf)
		if err != nil {
			return err
		}
		// Trim trailing newline from non-simd output
		if buf.Bytes()[buf.Len()-1] == '\n' {
			buf.Truncate(buf.Len() - 1)
		}
		buf.WriteString(s3Select.Output.JSONArgs.RecordDelimiter)

		return nil
	}

github.com/minio/mc v0.0.0-20240612143403-e7c9a733c680/go.mod h1:21/cb+wUd+lLRsdX7ACqyO8DzPNSpXftp1bOkQlIbh8=
github.com/minio/md5-simd v1.1.2 h1:Gdi1DZK69+ZVMoNHRXJyNcxrMA4dSxoYHZSQbirFg34=
github.com/minio/md5-simd v1.1.2/go.mod h1:MzdKDxYpY2BT9XQFocsiZf/NKVtR7nkE4RoEpN+20RM=
github.com/minio/minio-go/v6 v6.0.46/go.mod h1:qD0lajrGW49lKZLtXKtCB4X/qkMf0a5tBvN2PaZg7Gg=

//		Encoding:
//			type = TYPE_REG
//			reg = REG_[US]XT[BHWX] + register + shift amount
//			offset = ((reg&31) << 16) | (exttype << 13) | (amount<<10)
//
//	reg.<T>
//		Register arrangement for ARM64 SIMD register
//		e.g.: V1.S4, V2.S2, V7.D2, V2.H4, V6.B16
//		Encoding:
//			type = TYPE_REG
//			reg = REG_ARNG + register + arrangement
//
//	reg.<T>[index]
//		Register element for ARM64
//		Encoding:

			esize = 32
		case ARNG_2D:
			imax = 127
			esize = 64
		}

		imm := 0
		switch p.As {
		case AVUSHR, AVSRI, AVUSRA:
			imm = esize*2 - shift
			if imm < esize || imm > imax {
				c.ctxt.Diag("shift out of range: %v", p)
			}
		case AVSHL, AVSLI:
			imm = esize + shift
			if imm > imax {
				c.ctxt.Diag("shift out of range: %v", p)
			}
		default:

        op, result_type, op.getX(), indices, updates);
    return success();
  }
};

// Approximates lgamma using Lanczos' approximation from
// "A Precision Approximation of the Gamma Function". SIAM Journal on Numerical
// Analysis series B. Vol. 1:
// lgamma(z + 1) = (log(2) + log(pi)) / 2 + (z + 1/2) * log(t(z)) - t(z) + A(z)
// t(z) = z + kLanczosGamma + 1/2
// A(z) = kBaseLanczosCoeff

//
// TF unsorted segment reduction op peforms the following calculation:
//
// Assume segment ids' shape is [SI0, SI1, ..., SIm] and data's  shape is
// [D0, D1, ..., Dn]. Note that segment ids' shape must be a prefix of data's
// shape, so we can have data's shape represented as [SI0, SI1, ..., SIm,
// Dm+1, ..., Dn]. Then
//   output[segment_ids[SI_i0, SI_i1, ..., SI_im], D_im+1, ..., D_in] =

peeyush18, Peng Yu, Pierre, preciousdp11, qjivy, Raingo, raoqiyu, ribx, Richard
S. Imaoka, Rishabh Patel, Robert Walecki, Rockford Wei, Ryan Kung, Sahil Dua,
Sandip Giri, Sayed Hadi Hashemi, sgt101, Shitian Ni, Shuolongbj, Siim PõDer,
Simon Perkins, sj6077, SOLARIS, Spotlight0xff, Steffen Eberbach, Stephen Fox,
superryanguo, Sven Mayer, Tapan Prakash, Tiago Morais Morgado, Till Hoffmann, Tj

<https://www.gnu.org/licenses/>.

================================================================

github.com/minio/md5-simd
https://github.com/minio/md5-simd
----------------------------------------------------------------

                                 Apache License
                           Version 2.0, January 2004