DISPATCH(runtime·call1024, 1024)
	DISPATCH(runtime·call2048, 2048)
	DISPATCH(runtime·call4096, 4096)
	DISPATCH(runtime·call8192, 8192)
	DISPATCH(runtime·call16384, 16384)
	DISPATCH(runtime·call32768, 32768)
	DISPATCH(runtime·call65536, 65536)
	DISPATCH(runtime·call131072, 131072)
	DISPATCH(runtime·call262144, 262144)
	DISPATCH(runtime·call524288, 524288)
	DISPATCH(runtime·call1048576, 1048576)

	DISPATCH(runtime·call1024, 1024)
	DISPATCH(runtime·call2048, 2048)
	DISPATCH(runtime·call4096, 4096)
	DISPATCH(runtime·call8192, 8192)
	DISPATCH(runtime·call16384, 16384)
	DISPATCH(runtime·call32768, 32768)
	DISPATCH(runtime·call65536, 65536)
	DISPATCH(runtime·call131072, 131072)
	DISPATCH(runtime·call262144, 262144)
	DISPATCH(runtime·call524288, 524288)
	DISPATCH(runtime·call1048576, 1048576)

  burst: 10
  contentType: application/vnd.kubernetes.protobuf
  kubeconfig: /var/lib/kube-proxy/kubeconfig.conf
  qps: 5
clusterCIDR: 10.244.0.0/16
configSyncPeriod: 15m0s
conntrack:
  maxPerCore: 32768
  min: 131072
  tcpCloseWaitTimeout: 1h0m0s
  tcpEstablishedTimeout: 24h0m0s
enableProfiling: false
healthzBindAddress: 0.0.0.0:10256
hostnameOverride: ""
iptables:
  masqueradeAll: false

		totalAlloc += m2.TotalAlloc - m1.TotalAlloc
	}
	// 32k is large and we actually allocate substantially less,
	// but prior to the fix for #58975 we allocated ~400k in this case.
	if got, want := totalAlloc/count, uint64(32768); got > want {
		t.Fatalf("ReadMIMEHeader allocated %v bytes, want < %v", got, want)
	}
}

type readResponseTest struct {
	in       string
	inCode   int
	wantCode int
	wantMsg  string
}

	FORMAT_MESSAGE_FROM_HMODULE    = 2048
	FORMAT_MESSAGE_FROM_SYSTEM     = 4096
	FORMAT_MESSAGE_ARGUMENT_ARRAY  = 8192
	FORMAT_MESSAGE_MAX_WIDTH_MASK  = 255

	MAX_PATH      = 260
	MAX_LONG_PATH = 32768

	MAX_COMPUTERNAME_LENGTH = 15

	TIME_ZONE_ID_UNKNOWN  = 0
	TIME_ZONE_ID_STANDARD = 1

	TIME_ZONE_ID_DAYLIGHT = 2
	IGNORE                = 0
	INFINITE              = 0xffffffff

func.func @convert_argmin_i16(%arg0: tensor<2xi16>) -> (tensor<i16>, tensor<i32>) {
  %0 = mhlo.constant dense<false> : tensor<i1>
  %1 = "mhlo.iota"() <{iota_dimension = 0 : i64}> : () -> tensor<2xi32>
  %2 = mhlo.constant dense<32767> : tensor<i16>
  %3 = mhlo.constant dense<0> : tensor<i32>
  %4:2 = "mhlo.reduce"(%arg0, %1, %2, %3) ({
  ^bb0(%arg1: tensor<i16>, %arg2: tensor<i32>, %arg3: tensor<i16>, %arg4: tensor<i32>):

	C_LAUTO       // any other constant up to 64 bits

	C_SEXT1  // 0 to 4095, direct
	C_SEXT2  // 0 to 8190
	C_SEXT4  // 0 to 16380
	C_SEXT8  // 0 to 32760
	C_SEXT16 // 0 to 65520
	C_LEXT

	C_ZOREG     // 0(R)
	C_NSOREG_16 // must mirror C_NSAUTO_16, etc
	C_NSOREG_8
	C_NSOREG_4
	C_NSOREG
	C_NPOREG_16
	C_NPOREG
	C_NQOREG_16
	C_NOREG4K

	MOVD    $·byteswapcons(SB),R3
	LVX	(R3),byteswap
#endif

	CMPU	R5,$256		// length of bytes
	BLT	short

	RLDICR	$0,R5,$56,R6 // chunk to process

	// First step for larger sizes
l1:	MOVD	$32768,R7
	MOVD	R7,R9
	CMP	R6,R7   // compare R6, R7 (MAX SIZE)
	BGT	top	// less than MAX, just do remainder
	MOVD	R6,R7
top:
	SUB	R7,R6,R6

	// mainloop does 128 bytes at a time
	SRD	$7,R7

	DISPATCH(runtime·call1024, 1024)
	DISPATCH(runtime·call2048, 2048)
	DISPATCH(runtime·call4096, 4096)
	DISPATCH(runtime·call8192, 8192)
	DISPATCH(runtime·call16384, 16384)
	DISPATCH(runtime·call32768, 32768)
	DISPATCH(runtime·call65536, 65536)
	DISPATCH(runtime·call131072, 131072)
	DISPATCH(runtime·call262144, 262144)
	DISPATCH(runtime·call524288, 524288)
	DISPATCH(runtime·call1048576, 1048576)

	DISPATCH(runtime·call1024, 1024)
	DISPATCH(runtime·call2048, 2048)
	DISPATCH(runtime·call4096, 4096)
	DISPATCH(runtime·call8192, 8192)
	DISPATCH(runtime·call16384, 16384)
	DISPATCH(runtime·call32768, 32768)
	DISPATCH(runtime·call65536, 65536)
	DISPATCH(runtime·call131072, 131072)
	DISPATCH(runtime·call262144, 262144)
	DISPATCH(runtime·call524288, 524288)
	DISPATCH(runtime·call1048576, 1048576)