IB/qib: Add new qib driver for QLogic PCIe InfiniBand adapters

Add a low-level IB driver for QLogic PCIe adapters.

Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

1 files changed, 4618 insertions, 0 deletions

diff --git a/drivers/infiniband/hw/qib/qib_iba7220.c b/drivers/infiniband/hw/qib/qib_iba7220.c
new file mode 100644
index 00000000000..6fd8d74e739
--- /dev/null
+++ b/drivers/infiniband/hw/qib/qib_iba7220.c
@@ -0,0 +1,4618 @@
+/*
+ * Copyright (c) 2006, 2007, 2008, 2009, 2010 QLogic Corporation.
+ * All rights reserved.
+ * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+/*
+ * This file contains all of the code that is specific to the
+ * QLogic_IB 7220 chip (except that specific to the SerDes)
+ */
+
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <rdma/ib_verbs.h>
+
+#include "qib.h"
+#include "qib_7220.h"
+
+static void qib_setup_7220_setextled(struct qib_pportdata *, u32);
+static void qib_7220_handle_hwerrors(struct qib_devdata *, char *, size_t);
+static void sendctrl_7220_mod(struct qib_pportdata *ppd, u32 op);
+static u32 qib_7220_iblink_state(u64);
+static u8 qib_7220_phys_portstate(u64);
+static void qib_sdma_update_7220_tail(struct qib_pportdata *, u16);
+static void qib_set_ib_7220_lstate(struct qib_pportdata *, u16, u16);
+
+/*
+ * This file contains almost all the chip-specific register information and
+ * access functions for the QLogic QLogic_IB 7220 PCI-Express chip, with the
+ * exception of SerDes support, which in in qib_sd7220.c.
+ */
+
+/* Below uses machine-generated qib_chipnum_regs.h file */
+#define KREG_IDX(regname) (QIB_7220_##regname##_OFFS / sizeof(u64))
+
+/* Use defines to tie machine-generated names to lower-case names */
+#define kr_control KREG_IDX(Control)
+#define kr_counterregbase KREG_IDX(CntrRegBase)
+#define kr_errclear KREG_IDX(ErrClear)
+#define kr_errmask KREG_IDX(ErrMask)
+#define kr_errstatus KREG_IDX(ErrStatus)
+#define kr_extctrl KREG_IDX(EXTCtrl)
+#define kr_extstatus KREG_IDX(EXTStatus)
+#define kr_gpio_clear KREG_IDX(GPIOClear)
+#define kr_gpio_mask KREG_IDX(GPIOMask)
+#define kr_gpio_out KREG_IDX(GPIOOut)
+#define kr_gpio_status KREG_IDX(GPIOStatus)
+#define kr_hrtbt_guid KREG_IDX(HRTBT_GUID)
+#define kr_hwdiagctrl KREG_IDX(HwDiagCtrl)
+#define kr_hwerrclear KREG_IDX(HwErrClear)
+#define kr_hwerrmask KREG_IDX(HwErrMask)
+#define kr_hwerrstatus KREG_IDX(HwErrStatus)
+#define kr_ibcctrl KREG_IDX(IBCCtrl)
+#define kr_ibcddrctrl KREG_IDX(IBCDDRCtrl)
+#define kr_ibcddrstatus KREG_IDX(IBCDDRStatus)
+#define kr_ibcstatus KREG_IDX(IBCStatus)
+#define kr_ibserdesctrl KREG_IDX(IBSerDesCtrl)
+#define kr_intclear KREG_IDX(IntClear)
+#define kr_intmask KREG_IDX(IntMask)
+#define kr_intstatus KREG_IDX(IntStatus)
+#define kr_ncmodectrl KREG_IDX(IBNCModeCtrl)
+#define kr_palign KREG_IDX(PageAlign)
+#define kr_partitionkey KREG_IDX(RcvPartitionKey)
+#define kr_portcnt KREG_IDX(PortCnt)
+#define kr_rcvbthqp KREG_IDX(RcvBTHQP)
+#define kr_rcvctrl KREG_IDX(RcvCtrl)
+#define kr_rcvegrbase KREG_IDX(RcvEgrBase)
+#define kr_rcvegrcnt KREG_IDX(RcvEgrCnt)
+#define kr_rcvhdrcnt KREG_IDX(RcvHdrCnt)
+#define kr_rcvhdrentsize KREG_IDX(RcvHdrEntSize)
+#define kr_rcvhdrsize KREG_IDX(RcvHdrSize)
+#define kr_rcvpktledcnt KREG_IDX(RcvPktLEDCnt)
+#define kr_rcvtidbase KREG_IDX(RcvTIDBase)
+#define kr_rcvtidcnt KREG_IDX(RcvTIDCnt)
+#define kr_revision KREG_IDX(Revision)
+#define kr_scratch KREG_IDX(Scratch)
+#define kr_sendbuffererror KREG_IDX(SendBufErr0)
+#define kr_sendctrl KREG_IDX(SendCtrl)
+#define kr_senddmabase KREG_IDX(SendDmaBase)
+#define kr_senddmabufmask0 KREG_IDX(SendDmaBufMask0)
+#define kr_senddmabufmask1 (KREG_IDX(SendDmaBufMask0) + 1)
+#define kr_senddmabufmask2 (KREG_IDX(SendDmaBufMask0) + 2)
+#define kr_senddmahead KREG_IDX(SendDmaHead)
+#define kr_senddmaheadaddr KREG_IDX(SendDmaHeadAddr)
+#define kr_senddmalengen KREG_IDX(SendDmaLenGen)
+#define kr_senddmastatus KREG_IDX(SendDmaStatus)
+#define kr_senddmatail KREG_IDX(SendDmaTail)
+#define kr_sendpioavailaddr KREG_IDX(SendBufAvailAddr)
+#define kr_sendpiobufbase KREG_IDX(SendBufBase)
+#define kr_sendpiobufcnt KREG_IDX(SendBufCnt)
+#define kr_sendpiosize KREG_IDX(SendBufSize)
+#define kr_sendregbase KREG_IDX(SendRegBase)
+#define kr_userregbase KREG_IDX(UserRegBase)
+#define kr_xgxs_cfg KREG_IDX(XGXSCfg)
+
+/* These must only be written via qib_write_kreg_ctxt() */
+#define kr_rcvhdraddr KREG_IDX(RcvHdrAddr0)
+#define kr_rcvhdrtailaddr KREG_IDX(RcvHdrTailAddr0)
+
+
+#define CREG_IDX(regname) ((QIB_7220_##regname##_OFFS - \
+			QIB_7220_LBIntCnt_OFFS) / sizeof(u64))
+
+#define cr_badformat CREG_IDX(RxVersionErrCnt)
+#define cr_erricrc CREG_IDX(RxICRCErrCnt)
+#define cr_errlink CREG_IDX(RxLinkMalformCnt)
+#define cr_errlpcrc CREG_IDX(RxLPCRCErrCnt)
+#define cr_errpkey CREG_IDX(RxPKeyMismatchCnt)
+#define cr_rcvflowctrl_err CREG_IDX(RxFlowCtrlViolCnt)
+#define cr_err_rlen CREG_IDX(RxLenErrCnt)
+#define cr_errslen CREG_IDX(TxLenErrCnt)
+#define cr_errtidfull CREG_IDX(RxTIDFullErrCnt)
+#define cr_errtidvalid CREG_IDX(RxTIDValidErrCnt)
+#define cr_errvcrc CREG_IDX(RxVCRCErrCnt)
+#define cr_ibstatuschange CREG_IDX(IBStatusChangeCnt)
+#define cr_lbint CREG_IDX(LBIntCnt)
+#define cr_invalidrlen CREG_IDX(RxMaxMinLenErrCnt)
+#define cr_invalidslen CREG_IDX(TxMaxMinLenErrCnt)
+#define cr_lbflowstall CREG_IDX(LBFlowStallCnt)
+#define cr_pktrcv CREG_IDX(RxDataPktCnt)
+#define cr_pktrcvflowctrl CREG_IDX(RxFlowPktCnt)
+#define cr_pktsend CREG_IDX(TxDataPktCnt)
+#define cr_pktsendflow CREG_IDX(TxFlowPktCnt)
+#define cr_portovfl CREG_IDX(RxP0HdrEgrOvflCnt)
+#define cr_rcvebp CREG_IDX(RxEBPCnt)
+#define cr_rcvovfl CREG_IDX(RxBufOvflCnt)
+#define cr_senddropped CREG_IDX(TxDroppedPktCnt)
+#define cr_sendstall CREG_IDX(TxFlowStallCnt)
+#define cr_sendunderrun CREG_IDX(TxUnderrunCnt)
+#define cr_wordrcv CREG_IDX(RxDwordCnt)
+#define cr_wordsend CREG_IDX(TxDwordCnt)
+#define cr_txunsupvl CREG_IDX(TxUnsupVLErrCnt)
+#define cr_rxdroppkt CREG_IDX(RxDroppedPktCnt)
+#define cr_iblinkerrrecov CREG_IDX(IBLinkErrRecoveryCnt)
+#define cr_iblinkdown CREG_IDX(IBLinkDownedCnt)
+#define cr_ibsymbolerr CREG_IDX(IBSymbolErrCnt)
+#define cr_vl15droppedpkt CREG_IDX(RxVL15DroppedPktCnt)
+#define cr_rxotherlocalphyerr CREG_IDX(RxOtherLocalPhyErrCnt)
+#define cr_excessbufferovfl CREG_IDX(ExcessBufferOvflCnt)
+#define cr_locallinkintegrityerr CREG_IDX(LocalLinkIntegrityErrCnt)
+#define cr_rxvlerr CREG_IDX(RxVlErrCnt)
+#define cr_rxdlidfltr CREG_IDX(RxDlidFltrCnt)
+#define cr_psstat CREG_IDX(PSStat)
+#define cr_psstart CREG_IDX(PSStart)
+#define cr_psinterval CREG_IDX(PSInterval)
+#define cr_psrcvdatacount CREG_IDX(PSRcvDataCount)
+#define cr_psrcvpktscount CREG_IDX(PSRcvPktsCount)
+#define cr_psxmitdatacount CREG_IDX(PSXmitDataCount)
+#define cr_psxmitpktscount CREG_IDX(PSXmitPktsCount)
+#define cr_psxmitwaitcount CREG_IDX(PSXmitWaitCount)
+#define cr_txsdmadesc CREG_IDX(TxSDmaDescCnt)
+#define cr_pcieretrydiag CREG_IDX(PcieRetryBufDiagQwordCnt)
+
+#define SYM_RMASK(regname, fldname) ((u64)              \
+	QIB_7220_##regname##_##fldname##_RMASK)
+#define SYM_MASK(regname, fldname) ((u64)               \
+	QIB_7220_##regname##_##fldname##_RMASK <<       \
+	 QIB_7220_##regname##_##fldname##_LSB)
+#define SYM_LSB(regname, fldname) (QIB_7220_##regname##_##fldname##_LSB)
+#define SYM_FIELD(value, regname, fldname) ((u64) \
+	(((value) >> SYM_LSB(regname, fldname)) & \
+	 SYM_RMASK(regname, fldname)))
+#define ERR_MASK(fldname) SYM_MASK(ErrMask, fldname##Mask)
+#define HWE_MASK(fldname) SYM_MASK(HwErrMask, fldname##Mask)
+
+/* ibcctrl bits */
+#define QLOGIC_IB_IBCC_LINKINITCMD_DISABLE 1
+/* cycle through TS1/TS2 till OK */
+#define QLOGIC_IB_IBCC_LINKINITCMD_POLL 2
+/* wait for TS1, then go on */
+#define QLOGIC_IB_IBCC_LINKINITCMD_SLEEP 3
+#define QLOGIC_IB_IBCC_LINKINITCMD_SHIFT 16
+
+#define QLOGIC_IB_IBCC_LINKCMD_DOWN 1           /* move to 0x11 */
+#define QLOGIC_IB_IBCC_LINKCMD_ARMED 2          /* move to 0x21 */
+#define QLOGIC_IB_IBCC_LINKCMD_ACTIVE 3 /* move to 0x31 */
+
+#define BLOB_7220_IBCHG 0x81
+
+/*
+ * We could have a single register get/put routine, that takes a group type,
+ * but this is somewhat clearer and cleaner.  It also gives us some error
+ * checking.  64 bit register reads should always work, but are inefficient
+ * on opteron (the northbridge always generates 2 separate HT 32 bit reads),
+ * so we use kreg32 wherever possible.  User register and counter register
+ * reads are always 32 bit reads, so only one form of those routines.
+ */
+
+/**
+ * qib_read_ureg32 - read 32-bit virtualized per-context register
+ * @dd: device
+ * @regno: register number
+ * @ctxt: context number
+ *
+ * Return the contents of a register that is virtualized to be per context.
+ * Returns -1 on errors (not distinguishable from valid contents at
+ * runtime; we may add a separate error variable at some point).
+ */
+static inline u32 qib_read_ureg32(const struct qib_devdata *dd,
+				  enum qib_ureg regno, int ctxt)
+{
+	if (!dd->kregbase || !(dd->flags & QIB_PRESENT))
+		return 0;
+
+	if (dd->userbase)
+		return readl(regno + (u64 __iomem *)
+			     ((char __iomem *)dd->userbase +
+			      dd->ureg_align * ctxt));
+	else
+		return readl(regno + (u64 __iomem *)
+			     (dd->uregbase +
+			      (char __iomem *)dd->kregbase +
+			      dd->ureg_align * ctxt));
+}
+
+/**
+ * qib_write_ureg - write 32-bit virtualized per-context register
+ * @dd: device
+ * @regno: register number
+ * @value: value
+ * @ctxt: context
+ *
+ * Write the contents of a register that is virtualized to be per context.
+ */
+static inline void qib_write_ureg(const struct qib_devdata *dd,
+				  enum qib_ureg regno, u64 value, int ctxt)
+{
+	u64 __iomem *ubase;
+
+	if (dd->userbase)
+		ubase = (u64 __iomem *)
+			((char __iomem *) dd->userbase +
+			 dd->ureg_align * ctxt);
+	else
+		ubase = (u64 __iomem *)
+			(dd->uregbase +
+			 (char __iomem *) dd->kregbase +
+			 dd->ureg_align * ctxt);
+
+	if (dd->kregbase && (dd->flags & QIB_PRESENT))
+		writeq(value, &ubase[regno]);
+}
+
+/**
+ * qib_write_kreg_ctxt - write a device's per-ctxt 64-bit kernel register
+ * @dd: the qlogic_ib device
+ * @regno: the register number to write
+ * @ctxt: the context containing the register
+ * @value: the value to write
+ */
+static inline void qib_write_kreg_ctxt(const struct qib_devdata *dd,
+				       const u16 regno, unsigned ctxt,
+				       u64 value)
+{
+	qib_write_kreg(dd, regno + ctxt, value);
+}
+
+static inline void write_7220_creg(const struct qib_devdata *dd,
+				   u16 regno, u64 value)
+{
+	if (dd->cspec->cregbase && (dd->flags & QIB_PRESENT))
+		writeq(value, &dd->cspec->cregbase[regno]);
+}
+
+static inline u64 read_7220_creg(const struct qib_devdata *dd, u16 regno)
+{
+	if (!dd->cspec->cregbase || !(dd->flags & QIB_PRESENT))
+		return 0;
+	return readq(&dd->cspec->cregbase[regno]);
+}
+
+static inline u32 read_7220_creg32(const struct qib_devdata *dd, u16 regno)
+{
+	if (!dd->cspec->cregbase || !(dd->flags & QIB_PRESENT))
+		return 0;
+	return readl(&dd->cspec->cregbase[regno]);
+}
+
+/* kr_revision bits */
+#define QLOGIC_IB_R_EMULATORREV_MASK ((1ULL << 22) - 1)
+#define QLOGIC_IB_R_EMULATORREV_SHIFT 40
+
+/* kr_control bits */
+#define QLOGIC_IB_C_RESET (1U << 7)
+
+/* kr_intstatus, kr_intclear, kr_intmask bits */
+#define QLOGIC_IB_I_RCVURG_MASK ((1ULL << 17) - 1)
+#define QLOGIC_IB_I_RCVURG_SHIFT 32
+#define QLOGIC_IB_I_RCVAVAIL_MASK ((1ULL << 17) - 1)
+#define QLOGIC_IB_I_RCVAVAIL_SHIFT 0
+#define QLOGIC_IB_I_SERDESTRIMDONE (1ULL << 27)
+
+#define QLOGIC_IB_C_FREEZEMODE 0x00000002
+#define QLOGIC_IB_C_LINKENABLE 0x00000004
+
+#define QLOGIC_IB_I_SDMAINT             0x8000000000000000ULL
+#define QLOGIC_IB_I_SDMADISABLED        0x4000000000000000ULL
+#define QLOGIC_IB_I_ERROR               0x0000000080000000ULL
+#define QLOGIC_IB_I_SPIOSENT            0x0000000040000000ULL
+#define QLOGIC_IB_I_SPIOBUFAVAIL        0x0000000020000000ULL
+#define QLOGIC_IB_I_GPIO                0x0000000010000000ULL
+
+/* variables for sanity checking interrupt and errors */
+#define QLOGIC_IB_I_BITSEXTANT \
+		(QLOGIC_IB_I_SDMAINT | QLOGIC_IB_I_SDMADISABLED | \
+		(QLOGIC_IB_I_RCVURG_MASK << QLOGIC_IB_I_RCVURG_SHIFT) | \
+		(QLOGIC_IB_I_RCVAVAIL_MASK << \
+		 QLOGIC_IB_I_RCVAVAIL_SHIFT) | \
+		QLOGIC_IB_I_ERROR | QLOGIC_IB_I_SPIOSENT | \
+		QLOGIC_IB_I_SPIOBUFAVAIL | QLOGIC_IB_I_GPIO | \
+		QLOGIC_IB_I_SERDESTRIMDONE)
+
+#define IB_HWE_BITSEXTANT \
+	       (HWE_MASK(RXEMemParityErr) | \
+		HWE_MASK(TXEMemParityErr) | \
+		(QLOGIC_IB_HWE_PCIEMEMPARITYERR_MASK <<	 \
+		 QLOGIC_IB_HWE_PCIEMEMPARITYERR_SHIFT) | \
+		QLOGIC_IB_HWE_PCIE1PLLFAILED | \
+		QLOGIC_IB_HWE_PCIE0PLLFAILED | \
+		QLOGIC_IB_HWE_PCIEPOISONEDTLP | \
+		QLOGIC_IB_HWE_PCIECPLTIMEOUT | \
+		QLOGIC_IB_HWE_PCIEBUSPARITYXTLH | \
+		QLOGIC_IB_HWE_PCIEBUSPARITYXADM | \
+		QLOGIC_IB_HWE_PCIEBUSPARITYRADM | \
+		HWE_MASK(PowerOnBISTFailed) |	  \
+		QLOGIC_IB_HWE_COREPLL_FBSLIP | \
+		QLOGIC_IB_HWE_COREPLL_RFSLIP | \
+		QLOGIC_IB_HWE_SERDESPLLFAILED | \
+		HWE_MASK(IBCBusToSPCParityErr) | \
+		HWE_MASK(IBCBusFromSPCParityErr) | \
+		QLOGIC_IB_HWE_PCIECPLDATAQUEUEERR | \
+		QLOGIC_IB_HWE_PCIECPLHDRQUEUEERR | \
+		QLOGIC_IB_HWE_SDMAMEMREADERR | \
+		QLOGIC_IB_HWE_CLK_UC_PLLNOTLOCKED | \
+		QLOGIC_IB_HWE_PCIESERDESQ0PCLKNOTDETECT | \
+		QLOGIC_IB_HWE_PCIESERDESQ1PCLKNOTDETECT | \
+		QLOGIC_IB_HWE_PCIESERDESQ2PCLKNOTDETECT | \
+		QLOGIC_IB_HWE_PCIESERDESQ3PCLKNOTDETECT | \
+		QLOGIC_IB_HWE_DDSRXEQMEMORYPARITYERR | \
+		QLOGIC_IB_HWE_IB_UC_MEMORYPARITYERR | \
+		QLOGIC_IB_HWE_PCIE_UC_OCT0MEMORYPARITYERR | \
+		QLOGIC_IB_HWE_PCIE_UC_OCT1MEMORYPARITYERR)
+
+#define IB_E_BITSEXTANT							\
+	(ERR_MASK(RcvFormatErr) | ERR_MASK(RcvVCRCErr) |		\
+	 ERR_MASK(RcvICRCErr) | ERR_MASK(RcvMinPktLenErr) |		\
+	 ERR_MASK(RcvMaxPktLenErr) | ERR_MASK(RcvLongPktLenErr) |	\
+	 ERR_MASK(RcvShortPktLenErr) | ERR_MASK(RcvUnexpectedCharErr) | \
+	 ERR_MASK(RcvUnsupportedVLErr) | ERR_MASK(RcvEBPErr) |		\
+	 ERR_MASK(RcvIBFlowErr) | ERR_MASK(RcvBadVersionErr) |		\
+	 ERR_MASK(RcvEgrFullErr) | ERR_MASK(RcvHdrFullErr) |		\
+	 ERR_MASK(RcvBadTidErr) | ERR_MASK(RcvHdrLenErr) |		\
+	 ERR_MASK(RcvHdrErr) | ERR_MASK(RcvIBLostLinkErr) |		\
+	 ERR_MASK(SendSpecialTriggerErr) |				\
+	 ERR_MASK(SDmaDisabledErr) | ERR_MASK(SendMinPktLenErr) |	\
+	 ERR_MASK(SendMaxPktLenErr) | ERR_MASK(SendUnderRunErr) |	\
+	 ERR_MASK(SendPktLenErr) | ERR_MASK(SendDroppedSmpPktErr) |	\
+	 ERR_MASK(SendDroppedDataPktErr) |				\
+	 ERR_MASK(SendPioArmLaunchErr) |				\
+	 ERR_MASK(SendUnexpectedPktNumErr) |				\
+	 ERR_MASK(SendUnsupportedVLErr) | ERR_MASK(SendBufMisuseErr) |	\
+	 ERR_MASK(SDmaGenMismatchErr) | ERR_MASK(SDmaOutOfBoundErr) |	\
+	 ERR_MASK(SDmaTailOutOfBoundErr) | ERR_MASK(SDmaBaseErr) |	\
+	 ERR_MASK(SDma1stDescErr) | ERR_MASK(SDmaRpyTagErr) |		\
+	 ERR_MASK(SDmaDwEnErr) | ERR_MASK(SDmaMissingDwErr) |		\
+	 ERR_MASK(SDmaUnexpDataErr) |					\
+	 ERR_MASK(IBStatusChanged) | ERR_MASK(InvalidAddrErr) |		\
+	 ERR_MASK(ResetNegated) | ERR_MASK(HardwareErr) |		\
+	 ERR_MASK(SDmaDescAddrMisalignErr) |				\
+	 ERR_MASK(InvalidEEPCmd))
+
+/* kr_hwerrclear, kr_hwerrmask, kr_hwerrstatus, bits */
+#define QLOGIC_IB_HWE_PCIEMEMPARITYERR_MASK  0x00000000000000ffULL
+#define QLOGIC_IB_HWE_PCIEMEMPARITYERR_SHIFT 0
+#define QLOGIC_IB_HWE_PCIEPOISONEDTLP      0x0000000010000000ULL
+#define QLOGIC_IB_HWE_PCIECPLTIMEOUT       0x0000000020000000ULL
+#define QLOGIC_IB_HWE_PCIEBUSPARITYXTLH    0x0000000040000000ULL
+#define QLOGIC_IB_HWE_PCIEBUSPARITYXADM    0x0000000080000000ULL
+#define QLOGIC_IB_HWE_PCIEBUSPARITYRADM    0x0000000100000000ULL
+#define QLOGIC_IB_HWE_COREPLL_FBSLIP       0x0080000000000000ULL
+#define QLOGIC_IB_HWE_COREPLL_RFSLIP       0x0100000000000000ULL
+#define QLOGIC_IB_HWE_PCIE1PLLFAILED       0x0400000000000000ULL
+#define QLOGIC_IB_HWE_PCIE0PLLFAILED       0x0800000000000000ULL
+#define QLOGIC_IB_HWE_SERDESPLLFAILED      0x1000000000000000ULL
+/* specific to this chip */
+#define QLOGIC_IB_HWE_PCIECPLDATAQUEUEERR         0x0000000000000040ULL
+#define QLOGIC_IB_HWE_PCIECPLHDRQUEUEERR          0x0000000000000080ULL
+#define QLOGIC_IB_HWE_SDMAMEMREADERR              0x0000000010000000ULL
+#define QLOGIC_IB_HWE_CLK_UC_PLLNOTLOCKED          0x2000000000000000ULL
+#define QLOGIC_IB_HWE_PCIESERDESQ0PCLKNOTDETECT   0x0100000000000000ULL
+#define QLOGIC_IB_HWE_PCIESERDESQ1PCLKNOTDETECT   0x0200000000000000ULL
+#define QLOGIC_IB_HWE_PCIESERDESQ2PCLKNOTDETECT   0x0400000000000000ULL
+#define QLOGIC_IB_HWE_PCIESERDESQ3PCLKNOTDETECT   0x0800000000000000ULL
+#define QLOGIC_IB_HWE_DDSRXEQMEMORYPARITYERR       0x0000008000000000ULL
+#define QLOGIC_IB_HWE_IB_UC_MEMORYPARITYERR        0x0000004000000000ULL
+#define QLOGIC_IB_HWE_PCIE_UC_OCT0MEMORYPARITYERR 0x0000001000000000ULL
+#define QLOGIC_IB_HWE_PCIE_UC_OCT1MEMORYPARITYERR 0x0000002000000000ULL
+
+#define IBA7220_IBCC_LINKCMD_SHIFT 19
+
+/* kr_ibcddrctrl bits */
+#define IBA7220_IBC_DLIDLMC_MASK        0xFFFFFFFFUL
+#define IBA7220_IBC_DLIDLMC_SHIFT       32
+
+#define IBA7220_IBC_HRTBT_MASK  (SYM_RMASK(IBCDDRCtrl, HRTBT_AUTO) | \
+				 SYM_RMASK(IBCDDRCtrl, HRTBT_ENB))
+#define IBA7220_IBC_HRTBT_SHIFT SYM_LSB(IBCDDRCtrl, HRTBT_ENB)
+
+#define IBA7220_IBC_LANE_REV_SUPPORTED (1<<8)
+#define IBA7220_IBC_LREV_MASK   1
+#define IBA7220_IBC_LREV_SHIFT  8
+#define IBA7220_IBC_RXPOL_MASK  1
+#define IBA7220_IBC_RXPOL_SHIFT 7
+#define IBA7220_IBC_WIDTH_SHIFT 5
+#define IBA7220_IBC_WIDTH_MASK  0x3
+#define IBA7220_IBC_WIDTH_1X_ONLY       (0 << IBA7220_IBC_WIDTH_SHIFT)
+#define IBA7220_IBC_WIDTH_4X_ONLY       (1 << IBA7220_IBC_WIDTH_SHIFT)
+#define IBA7220_IBC_WIDTH_AUTONEG       (2 << IBA7220_IBC_WIDTH_SHIFT)
+#define IBA7220_IBC_SPEED_AUTONEG       (1 << 1)
+#define IBA7220_IBC_SPEED_SDR           (1 << 2)
+#define IBA7220_IBC_SPEED_DDR           (1 << 3)
+#define IBA7220_IBC_SPEED_AUTONEG_MASK  (0x7 << 1)
+#define IBA7220_IBC_IBTA_1_2_MASK       (1)
+
+/* kr_ibcddrstatus */
+/* link latency shift is 0, don't bother defining */
+#define IBA7220_DDRSTAT_LINKLAT_MASK    0x3ffffff
+
+/* kr_extstatus bits */
+#define QLOGIC_IB_EXTS_FREQSEL 0x2
+#define QLOGIC_IB_EXTS_SERDESSEL 0x4
+#define QLOGIC_IB_EXTS_MEMBIST_ENDTEST     0x0000000000004000
+#define QLOGIC_IB_EXTS_MEMBIST_DISABLED    0x0000000000008000
+
+/* kr_xgxsconfig bits */
+#define QLOGIC_IB_XGXS_RESET          0x5ULL
+#define QLOGIC_IB_XGXS_FC_SAFE        (1ULL << 63)
+
+/* kr_rcvpktledcnt */
+#define IBA7220_LEDBLINK_ON_SHIFT 32 /* 4ns period on after packet */
+#define IBA7220_LEDBLINK_OFF_SHIFT 0 /* 4ns period off before next on */
+
+#define _QIB_GPIO_SDA_NUM 1
+#define _QIB_GPIO_SCL_NUM 0
+#define QIB_TWSI_EEPROM_DEV 0xA2 /* All Production 7220 cards. */
+#define QIB_TWSI_TEMP_DEV 0x98
+
+/* HW counter clock is at 4nsec */
+#define QIB_7220_PSXMITWAIT_CHECK_RATE 4000
+
+#define IBA7220_R_INTRAVAIL_SHIFT 17
+#define IBA7220_R_PKEY_DIS_SHIFT 34
+#define IBA7220_R_TAILUPD_SHIFT 35
+#define IBA7220_R_CTXTCFG_SHIFT 36
+
+#define IBA7220_HDRHEAD_PKTINT_SHIFT 32 /* interrupt cnt in upper 32 bits */
+
+/*
+ * the size bits give us 2^N, in KB units.  0 marks as invalid,
+ * and 7 is reserved.  We currently use only 2KB and 4KB
+ */
+#define IBA7220_TID_SZ_SHIFT 37 /* shift to 3bit size selector */
+#define IBA7220_TID_SZ_2K (1UL << IBA7220_TID_SZ_SHIFT) /* 2KB */
+#define IBA7220_TID_SZ_4K (2UL << IBA7220_TID_SZ_SHIFT) /* 4KB */
+#define IBA7220_TID_PA_SHIFT 11U /* TID addr in chip stored w/o low bits */
+#define PBC_7220_VL15_SEND (1ULL << 63) /* pbc; VL15, no credit check */
+#define PBC_7220_VL15_SEND_CTRL (1ULL << 31) /* control version of same */
+
+#define AUTONEG_TRIES 5 /* sequential retries to negotiate DDR */
+
+/* packet rate matching delay multiplier */
+static u8 rate_to_delay[2][2] = {
+	/* 1x, 4x */
+	{   8, 2 }, /* SDR */
+	{   4, 1 }  /* DDR */
+};
+
+static u8 ib_rate_to_delay[IB_RATE_120_GBPS + 1] = {
+	[IB_RATE_2_5_GBPS] = 8,
+	[IB_RATE_5_GBPS] = 4,
+	[IB_RATE_10_GBPS] = 2,
+	[IB_RATE_20_GBPS] = 1
+};
+
+#define IBA7220_LINKSPEED_SHIFT SYM_LSB(IBCStatus, LinkSpeedActive)
+#define IBA7220_LINKWIDTH_SHIFT SYM_LSB(IBCStatus, LinkWidthActive)
+
+/* link training states, from IBC */
+#define IB_7220_LT_STATE_DISABLED        0x00
+#define IB_7220_LT_STATE_LINKUP          0x01
+#define IB_7220_LT_STATE_POLLACTIVE      0x02
+#define IB_7220_LT_STATE_POLLQUIET       0x03
+#define IB_7220_LT_STATE_SLEEPDELAY      0x04
+#define IB_7220_LT_STATE_SLEEPQUIET      0x05
+#define IB_7220_LT_STATE_CFGDEBOUNCE     0x08
+#define IB_7220_LT_STATE_CFGRCVFCFG      0x09
+#define IB_7220_LT_STATE_CFGWAITRMT      0x0a
+#define IB_7220_LT_STATE_CFGIDLE 0x0b
+#define IB_7220_LT_STATE_RECOVERRETRAIN  0x0c
+#define IB_7220_LT_STATE_RECOVERWAITRMT  0x0e
+#define IB_7220_LT_STATE_RECOVERIDLE     0x0f
+
+/* link state machine states from IBC */
+#define IB_7220_L_STATE_DOWN             0x0
+#define IB_7220_L_STATE_INIT             0x1
+#define IB_7220_L_STATE_ARM              0x2
+#define IB_7220_L_STATE_ACTIVE           0x3
+#define IB_7220_L_STATE_ACT_DEFER        0x4
+
+static const u8 qib_7220_physportstate[0x20] = {
+	[IB_7220_LT_STATE_DISABLED] = IB_PHYSPORTSTATE_DISABLED,
+	[IB_7220_LT_STATE_LINKUP] = IB_PHYSPORTSTATE_LINKUP,
+	[IB_7220_LT_STATE_POLLACTIVE] = IB_PHYSPORTSTATE_POLL,
+	[IB_7220_LT_STATE_POLLQUIET] = IB_PHYSPORTSTATE_POLL,
+	[IB_7220_LT_STATE_SLEEPDELAY] = IB_PHYSPORTSTATE_SLEEP,
+	[IB_7220_LT_STATE_SLEEPQUIET] = IB_PHYSPORTSTATE_SLEEP,
+	[IB_7220_LT_STATE_CFGDEBOUNCE] =
+		IB_PHYSPORTSTATE_CFG_TRAIN,
+	[IB_7220_LT_STATE_CFGRCVFCFG] =
+		IB_PHYSPORTSTATE_CFG_TRAIN,
+	[IB_7220_LT_STATE_CFGWAITRMT] =
+		IB_PHYSPORTSTATE_CFG_TRAIN,
+	[IB_7220_LT_STATE_CFGIDLE] = IB_PHYSPORTSTATE_CFG_TRAIN,
+	[IB_7220_LT_STATE_RECOVERRETRAIN] =
+		IB_PHYSPORTSTATE_LINK_ERR_RECOVER,
+	[IB_7220_LT_STATE_RECOVERWAITRMT] =
+		IB_PHYSPORTSTATE_LINK_ERR_RECOVER,
+	[IB_7220_LT_STATE_RECOVERIDLE] =
+		IB_PHYSPORTSTATE_LINK_ERR_RECOVER,
+	[0x10] = IB_PHYSPORTSTATE_CFG_TRAIN,
+	[0x11] = IB_PHYSPORTSTATE_CFG_TRAIN,
+	[0x12] = IB_PHYSPORTSTATE_CFG_TRAIN,
+	[0x13] = IB_PHYSPORTSTATE_CFG_TRAIN,
+	[0x14] = IB_PHYSPORTSTATE_CFG_TRAIN,
+	[0x15] = IB_PHYSPORTSTATE_CFG_TRAIN,
+	[0x16] = IB_PHYSPORTSTATE_CFG_TRAIN,
+	[0x17] = IB_PHYSPORTSTATE_CFG_TRAIN
+};
+
+int qib_special_trigger;
+module_param_named(special_trigger, qib_special_trigger, int, S_IRUGO);
+MODULE_PARM_DESC(special_trigger, "Enable SpecialTrigger arm/launch");
+
+#define IBCBUSFRSPCPARITYERR HWE_MASK(IBCBusFromSPCParityErr)
+#define IBCBUSTOSPCPARITYERR HWE_MASK(IBCBusToSPCParityErr)
+
+#define SYM_MASK_BIT(regname, fldname, bit) ((u64) \
+	(1ULL << (SYM_LSB(regname, fldname) + (bit))))
+
+#define TXEMEMPARITYERR_PIOBUF \
+	SYM_MASK_BIT(HwErrMask, TXEMemParityErrMask, 0)
+#define TXEMEMPARITYERR_PIOPBC \
+	SYM_MASK_BIT(HwErrMask, TXEMemParityErrMask, 1)
+#define TXEMEMPARITYERR_PIOLAUNCHFIFO \
+	SYM_MASK_BIT(HwErrMask, TXEMemParityErrMask, 2)
+
+#define RXEMEMPARITYERR_RCVBUF \
+	SYM_MASK_BIT(HwErrMask, RXEMemParityErrMask, 0)
+#define RXEMEMPARITYERR_LOOKUPQ \
+	SYM_MASK_BIT(HwErrMask, RXEMemParityErrMask, 1)
+#define RXEMEMPARITYERR_EXPTID \
+	SYM_MASK_BIT(HwErrMask, RXEMemParityErrMask, 2)
+#define RXEMEMPARITYERR_EAGERTID \
+	SYM_MASK_BIT(HwErrMask, RXEMemParityErrMask, 3)
+#define RXEMEMPARITYERR_FLAGBUF \
+	SYM_MASK_BIT(HwErrMask, RXEMemParityErrMask, 4)
+#define RXEMEMPARITYERR_DATAINFO \
+	SYM_MASK_BIT(HwErrMask, RXEMemParityErrMask, 5)
+#define RXEMEMPARITYERR_HDRINFO \
+	SYM_MASK_BIT(HwErrMask, RXEMemParityErrMask, 6)
+
+/* 7220 specific hardware errors... */
+static const struct qib_hwerror_msgs qib_7220_hwerror_msgs[] = {
+	/* generic hardware errors */
+	QLOGIC_IB_HWE_MSG(IBCBUSFRSPCPARITYERR, "QIB2IB Parity"),
+	QLOGIC_IB_HWE_MSG(IBCBUSTOSPCPARITYERR, "IB2QIB Parity"),
+
+	QLOGIC_IB_HWE_MSG(TXEMEMPARITYERR_PIOBUF,
+			  "TXE PIOBUF Memory Parity"),
+	QLOGIC_IB_HWE_MSG(TXEMEMPARITYERR_PIOPBC,
+			  "TXE PIOPBC Memory Parity"),
+	QLOGIC_IB_HWE_MSG(TXEMEMPARITYERR_PIOLAUNCHFIFO,
+			  "TXE PIOLAUNCHFIFO Memory Parity"),
+
+	QLOGIC_IB_HWE_MSG(RXEMEMPARITYERR_RCVBUF,
+			  "RXE RCVBUF Memory Parity"),
+	QLOGIC_IB_HWE_MSG(RXEMEMPARITYERR_LOOKUPQ,
+			  "RXE LOOKUPQ Memory Parity"),
+	QLOGIC_IB_HWE_MSG(RXEMEMPARITYERR_EAGERTID,
+			  "RXE EAGERTID Memory Parity"),
+	QLOGIC_IB_HWE_MSG(RXEMEMPARITYERR_EXPTID,
+			  "RXE EXPTID Memory Parity"),
+	QLOGIC_IB_HWE_MSG(RXEMEMPARITYERR_FLAGBUF,
+			  "RXE FLAGBUF Memory Parity"),
+	QLOGIC_IB_HWE_MSG(RXEMEMPARITYERR_DATAINFO,
+			  "RXE DATAINFO Memory Parity"),
+	QLOGIC_IB_HWE_MSG(RXEMEMPARITYERR_HDRINFO,
+			  "RXE HDRINFO Memory Parity"),
+
+	/* chip-specific hardware errors */
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIEPOISONEDTLP,
+			  "PCIe Poisoned TLP"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIECPLTIMEOUT,
+			  "PCIe completion timeout"),
+	/*
+	 * In practice, it's unlikely wthat we'll see PCIe PLL, or bus
+	 * parity or memory parity error failures, because most likely we
+	 * won't be able to talk to the core of the chip.  Nonetheless, we
+	 * might see them, if they are in parts of the PCIe core that aren't
+	 * essential.
+	 */
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIE1PLLFAILED,
+			  "PCIePLL1"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIE0PLLFAILED,
+			  "PCIePLL0"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIEBUSPARITYXTLH,
+			  "PCIe XTLH core parity"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIEBUSPARITYXADM,
+			  "PCIe ADM TX core parity"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIEBUSPARITYRADM,
+			  "PCIe ADM RX core parity"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_SERDESPLLFAILED,
+			  "SerDes PLL"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIECPLDATAQUEUEERR,
+			  "PCIe cpl header queue"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIECPLHDRQUEUEERR,
+			  "PCIe cpl data queue"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_SDMAMEMREADERR,
+			  "Send DMA memory read"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_CLK_UC_PLLNOTLOCKED,
+			  "uC PLL clock not locked"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIESERDESQ0PCLKNOTDETECT,
+			  "PCIe serdes Q0 no clock"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIESERDESQ1PCLKNOTDETECT,
+			  "PCIe serdes Q1 no clock"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIESERDESQ2PCLKNOTDETECT,
+			  "PCIe serdes Q2 no clock"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIESERDESQ3PCLKNOTDETECT,
+			  "PCIe serdes Q3 no clock"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_DDSRXEQMEMORYPARITYERR,
+			  "DDS RXEQ memory parity"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_IB_UC_MEMORYPARITYERR,
+			  "IB uC memory parity"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIE_UC_OCT0MEMORYPARITYERR,
+			  "PCIe uC oct0 memory parity"),
+	QLOGIC_IB_HWE_MSG(QLOGIC_IB_HWE_PCIE_UC_OCT1MEMORYPARITYERR,
+			  "PCIe uC oct1 memory parity"),
+};
+
+#define RXE_PARITY (RXEMEMPARITYERR_EAGERTID|RXEMEMPARITYERR_EXPTID)
+
+#define QLOGIC_IB_E_PKTERRS (\
+		ERR_MASK(SendPktLenErr) |				\
+		ERR_MASK(SendDroppedDataPktErr) |			\
+		ERR_MASK(RcvVCRCErr) |					\
+		ERR_MASK(RcvICRCErr) |					\
+		ERR_MASK(RcvShortPktLenErr) |				\
+		ERR_MASK(RcvEBPErr))
+
+/* Convenience for decoding Send DMA errors */
+#define QLOGIC_IB_E_SDMAERRS ( \
+		ERR_MASK(SDmaGenMismatchErr) |				\
+		ERR_MASK(SDmaOutOfBoundErr) |				\
+		ERR_MASK(SDmaTailOutOfBoundErr) | ERR_MASK(SDmaBaseErr) | \
+		ERR_MASK(SDma1stDescErr) | ERR_MASK(SDmaRpyTagErr) |	\
+		ERR_MASK(SDmaDwEnErr) | ERR_MASK(SDmaMissingDwErr) |	\
+		ERR_MASK(SDmaUnexpDataErr) |				\
+		ERR_MASK(SDmaDescAddrMisalignErr) |			\
+		ERR_MASK(SDmaDisabledErr) |				\
+		ERR_MASK(SendBufMisuseErr))
+
+/* These are all rcv-related errors which we want to count for stats */
+#define E_SUM_PKTERRS \
+	(ERR_MASK(RcvHdrLenErr) | ERR_MASK(RcvBadTidErr) |		\
+	 ERR_MASK(RcvBadVersionErr) | ERR_MASK(RcvHdrErr) |		\
+	 ERR_MASK(RcvLongPktLenErr) | ERR_MASK(RcvShortPktLenErr) |	\
+	 ERR_MASK(RcvMaxPktLenErr) | ERR_MASK(RcvMinPktLenErr) |	\
+	 ERR_MASK(RcvFormatErr) | ERR_MASK(RcvUnsupportedVLErr) |	\
+	 ERR_MASK(RcvUnexpectedCharErr) | ERR_MASK(RcvEBPErr))
+
+/* These are all send-related errors which we want to count for stats */
+#define E_SUM_ERRS \
+	(ERR_MASK(SendPioArmLaunchErr) | ERR_MASK(SendUnexpectedPktNumErr) | \
+	 ERR_MASK(SendDroppedDataPktErr) | ERR_MASK(SendDroppedSmpPktErr) | \
+	 ERR_MASK(SendMaxPktLenErr) | ERR_MASK(SendUnsupportedVLErr) |	\
+	 ERR_MASK(SendMinPktLenErr) | ERR_MASK(SendPktLenErr) |		\
+	 ERR_MASK(InvalidAddrErr))
+
+/*
+ * this is similar to E_SUM_ERRS, but can't ignore armlaunch, don't ignore
+ * errors not related to freeze and cancelling buffers.  Can't ignore
+ * armlaunch because could get more while still cleaning up, and need
+ * to cancel those as they happen.
+ */
+#define E_SPKT_ERRS_IGNORE \
+	(ERR_MASK(SendDroppedDataPktErr) | ERR_MASK(SendDroppedSmpPktErr) | \
+	 ERR_MASK(SendMaxPktLenErr) | ERR_MASK(SendMinPktLenErr) |	\
+	 ERR_MASK(SendPktLenErr))
+
+/*
+ * these are errors that can occur when the link changes state while
+ * a packet is being sent or received.  This doesn't cover things
+ * like EBP or VCRC that can be the result of a sending having the
+ * link change state, so we receive a "known bad" packet.
+ */
+#define E_SUM_LINK_PKTERRS \
+	(ERR_MASK(SendDroppedDataPktErr) | ERR_MASK(SendDroppedSmpPktErr) | \
+	 ERR_MASK(SendMinPktLenErr) | ERR_MASK(SendPktLenErr) |		\
+	 ERR_MASK(RcvShortPktLenErr) | ERR_MASK(RcvMinPktLenErr) |	\
+	 ERR_MASK(RcvUnexpectedCharErr))
+
+static void autoneg_7220_work(struct work_struct *);
+static u32 __iomem *qib_7220_getsendbuf(struct qib_pportdata *, u64, u32 *);
+
+/*
+ * Called when we might have an error that is specific to a particular
+ * PIO buffer, and may need to cancel that buffer, so it can be re-used.
+ * because we don't need to force the update of pioavail.
+ */
+static void qib_disarm_7220_senderrbufs(struct qib_pportdata *ppd)
+{
+	unsigned long sbuf[3];
+	struct qib_devdata *dd = ppd->dd;
+
+	/*
+	 * It's possible that sendbuffererror could have bits set; might
+	 * have already done this as a result of hardware error handling.
+	 */
+	/* read these before writing errorclear */
+	sbuf[0] = qib_read_kreg64(dd, kr_sendbuffererror);
+	sbuf[1] = qib_read_kreg64(dd, kr_sendbuffererror + 1);
+	sbuf[2] = qib_read_kreg64(dd, kr_sendbuffererror + 2);
+
+	if (sbuf[0] || sbuf[1] || sbuf[2])
+		qib_disarm_piobufs_set(dd, sbuf,
+				       dd->piobcnt2k + dd->piobcnt4k);
+}
+
+static void qib_7220_txe_recover(struct qib_devdata *dd)
+{
+	qib_devinfo(dd->pcidev, "Recovering from TXE PIO parity error\n");
+	qib_disarm_7220_senderrbufs(dd->pport);
+}
+
+/*
+ * This is called with interrupts disabled and sdma_lock held.
+ */
+static void qib_7220_sdma_sendctrl(struct qib_pportdata *ppd, unsigned op)
+{
+	struct qib_devdata *dd = ppd->dd;
+	u64 set_sendctrl = 0;
+	u64 clr_sendctrl = 0;
+
+	if (op & QIB_SDMA_SENDCTRL_OP_ENABLE)
+		set_sendctrl |= SYM_MASK(SendCtrl, SDmaEnable);
+	else
+		clr_sendctrl |= SYM_MASK(SendCtrl, SDmaEnable);
+
+	if (op & QIB_SDMA_SENDCTRL_OP_INTENABLE)
+		set_sendctrl |= SYM_MASK(SendCtrl, SDmaIntEnable);
+	else
+		clr_sendctrl |= SYM_MASK(SendCtrl, SDmaIntEnable);
+
+	if (op & QIB_SDMA_SENDCTRL_OP_HALT)
+		set_sendctrl |= SYM_MASK(SendCtrl, SDmaHalt);
+	else
+		clr_sendctrl |= SYM_MASK(SendCtrl, SDmaHalt);
+
+	spin_lock(&dd->sendctrl_lock);
+
+	dd->sendctrl |= set_sendctrl;
+	dd->sendctrl &= ~clr_sendctrl;
+
+	qib_write_kreg(dd, kr_sendctrl, dd->sendctrl);
+	qib_write_kreg(dd, kr_scratch, 0);
+
+	spin_unlock(&dd->sendctrl_lock);
+}
+
+static void qib_decode_7220_sdma_errs(struct qib_pportdata *ppd,
+				      u64 err, char *buf, size_t blen)
+{
+	static const struct {
+		u64 err;
+		const char *msg;
+	} errs[] = {
+		{ ERR_MASK(SDmaGenMismatchErr),
+		  "SDmaGenMismatch" },
+		{ ERR_MASK(SDmaOutOfBoundErr),
+		  "SDmaOutOfBound" },
+		{ ERR_MASK(SDmaTailOutOfBoundErr),
+		  "SDmaTailOutOfBound" },
+		{ ERR_MASK(SDmaBaseErr),
+		  "SDmaBase" },
+		{ ERR_MASK(SDma1stDescErr),
+		  "SDma1stDesc" },
+		{ ERR_MASK(SDmaRpyTagErr),
+		  "SDmaRpyTag" },
+		{ ERR_MASK(SDmaDwEnErr),
+		  "SDmaDwEn" },
+		{ ERR_MASK(SDmaMissingDwErr),
+		  "SDmaMissingDw" },
+		{ ERR_MASK(SDmaUnexpDataErr),
+		  "SDmaUnexpData" },
+		{ ERR_MASK(SDmaDescAddrMisalignErr),
+		  "SDmaDescAddrMisalign" },
+		{ ERR_MASK(SendBufMisuseErr),
+		  "SendBufMisuse" },
+		{ ERR_MASK(SDmaDisabledErr),
+		  "SDmaDisabled" },
+	};
+	int i;
+	size_t bidx = 0;
+
+	for (i = 0; i < ARRAY_SIZE(errs); i++) {
+		if (err & errs[i].err)
+			bidx += scnprintf(buf + bidx, blen - bidx,
+					 "%s ", errs[i].msg);
+	}
+}
+
+/*
+ * This is called as part of link down clean up so disarm and flush
+ * all send buffers so that SMP packets can be sent.
+ */
+static void qib_7220_sdma_hw_clean_up(struct qib_pportdata *ppd)
+{
+	/* This will trigger the Abort interrupt */
+	sendctrl_7220_mod(ppd, QIB_SENDCTRL_DISARM_ALL | QIB_SENDCTRL_FLUSH |
+			  QIB_SENDCTRL_AVAIL_BLIP);
+	ppd->dd->upd_pio_shadow  = 1; /* update our idea of what's busy */
+}
+
+static void qib_sdma_7220_setlengen(struct qib_pportdata *ppd)
+{
+	/*
+	 * Set SendDmaLenGen and clear and set
+	 * the MSB of the generation count to enable generation checking
+	 * and load the internal generation counter.
+	 */
+	qib_write_kreg(ppd->dd, kr_senddmalengen, ppd->sdma_descq_cnt);
+	qib_write_kreg(ppd->dd, kr_senddmalengen,
+		       ppd->sdma_descq_cnt |
+		       (1ULL << QIB_7220_SendDmaLenGen_Generation_MSB));
+}
+
+static void qib_7220_sdma_hw_start_up(struct qib_pportdata *ppd)
+{
+	qib_sdma_7220_setlengen(ppd);
+	qib_sdma_update_7220_tail(ppd, 0); /* Set SendDmaTail */
+	ppd->sdma_head_dma[0] = 0;
+}
+
+#define DISABLES_SDMA (							\
+		ERR_MASK(SDmaDisabledErr) |				\
+		ERR_MASK(SDmaBaseErr) |					\
+		ERR_MASK(SDmaTailOutOfBoundErr) |			\
+		ERR_MASK(SDmaOutOfBoundErr) |				\
+		ERR_MASK(SDma1stDescErr) |				\
+		ERR_MASK(SDmaRpyTagErr) |				\
+		ERR_MASK(SDmaGenMismatchErr) |				\
+		ERR_MASK(SDmaDescAddrMisalignErr) |			\
+		ERR_MASK(SDmaMissingDwErr) |				\
+		ERR_MASK(SDmaDwEnErr))
+
+static void sdma_7220_errors(struct qib_pportdata *ppd, u64 errs)
+{
+	unsigned long flags;
+	struct qib_devdata *dd = ppd->dd;
+	char *msg;
+
+	errs &= QLOGIC_IB_E_SDMAERRS;
+
+	msg = dd->cspec->sdmamsgbuf;
+	qib_decode_7220_sdma_errs(ppd, errs, msg, sizeof dd->cspec->sdmamsgbuf);
+	spin_lock_irqsave(&ppd->sdma_lock, flags);
+
+	if (errs & ERR_MASK(SendBufMisuseErr)) {
+		unsigned long sbuf[3];
+
+		sbuf[0] = qib_read_kreg64(dd, kr_sendbuffererror);
+		sbuf[1] = qib_read_kreg64(dd, kr_sendbuffererror + 1);
+		sbuf[2] = qib_read_kreg64(dd, kr_sendbuffererror + 2);
+
+		qib_dev_err(ppd->dd,
+			    "IB%u:%u SendBufMisuse: %04lx %016lx %016lx\n",
+			    ppd->dd->unit, ppd->port, sbuf[2], sbuf[1],
+			    sbuf[0]);
+	}
+
+	if (errs & ERR_MASK(SDmaUnexpDataErr))
+		qib_dev_err(dd, "IB%u:%u SDmaUnexpData\n", ppd->dd->unit,
+			    ppd->port);
+
+	switch (ppd->sdma_state.current_state) {
+	case qib_sdma_state_s00_hw_down:
+		/* not expecting any interrupts */
+		break;
+
+	case qib_sdma_state_s10_hw_start_up_wait:
+		/* handled in intr path */
+		break;
+
+	case qib_sdma_state_s20_idle:
+		/* not expecting any interrupts */
+		break;
+
+	case qib_sdma_state_s30_sw_clean_up_wait:
+		/* not expecting any interrupts */
+		break;
+
+	case qib_sdma_state_s40_hw_clean_up_wait:
+		if (errs & ERR_MASK(SDmaDisabledErr))
+			__qib_sdma_process_event(ppd,
+				qib_sdma_event_e50_hw_cleaned);
+		break;
+
+	case qib_sdma_state_s50_hw_halt_wait:
+		/* handled in intr path */
+		break;
+
+	case qib_sdma_state_s99_running:
+		if (errs & DISABLES_SDMA)
+			__qib_sdma_process_event(ppd,
+				qib_sdma_event_e7220_err_halted);
+		break;
+	}
+
+	spin_unlock_irqrestore(&ppd->sdma_lock, flags);
+}
+
+/*
+ * Decode the error status into strings, deciding whether to always
+ * print * it or not depending on "normal packet errors" vs everything
+ * else.   Return 1 if "real" errors, otherwise 0 if only packet
+ * errors, so caller can decide what to print with the string.
+ */
+static int qib_decode_7220_err(struct qib_devdata *dd, char *buf, size_t blen,
+			       u64 err)
+{
+	int iserr = 1;
+
+	*buf = '\0';
+	if (err & QLOGIC_IB_E_PKTERRS) {
+		if (!(err & ~QLOGIC_IB_E_PKTERRS))
+			iserr = 0;
+		if ((err & ERR_MASK(RcvICRCErr)) &&
+		    !(err & (ERR_MASK(RcvVCRCErr) | ERR_MASK(RcvEBPErr))))
+			strlcat(buf, "CRC ", blen);
+		if (!iserr)
+			goto done;
+	}
+	if (err & ERR_MASK(RcvHdrLenErr))
+		strlcat(buf, "rhdrlen ", blen);
+	if (err & ERR_MASK(RcvBadTidErr))
+		strlcat(buf, "rbadtid ", blen);
+	if (err & ERR_MASK(RcvBadVersionErr))
+		strlcat(buf, "rbadversion ", blen);
+	if (err & ERR_MASK(RcvHdrErr))
+		strlcat(buf, "rhdr ", blen);
+	if (err & ERR_MASK(SendSpecialTriggerErr))
+		strlcat(buf, "sendspecialtrigger ", blen);
+	if (err & ERR_MASK(RcvLongPktLenErr))
+		strlcat(buf, "rlongpktlen ", blen);
+	if (err & ERR_MASK(RcvMaxPktLenErr))
+		strlcat(buf, "rmaxpktlen ", blen);
+	if (err & ERR_MASK(RcvMinPktLenErr))
+		strlcat(buf, "rminpktlen ", blen);
+	if (err & ERR_MASK(SendMinPktLenErr))
+		strlcat(buf, "sminpktlen ", blen);
+	if (err & ERR_MASK(RcvFormatErr))
+		strlcat(buf, "rformaterr ", blen);
+	if (err & ERR_MASK(RcvUnsupportedVLErr))
+		strlcat(buf, "runsupvl ", blen);
+	if (err & ERR_MASK(RcvUnexpectedCharErr))
+		strlcat(buf, "runexpchar ", blen);
+	if (err & ERR_MASK(RcvIBFlowErr))
+		strlcat(buf, "ribflow ", blen);
+	if (err & ERR_MASK(SendUnderRunErr))
+		strlcat(buf, "sunderrun ", blen);
+	if (err & ERR_MASK(SendPioArmLaunchErr))
+		strlcat(buf, "spioarmlaunch ", blen);
+	if (err & ERR_MASK(SendUnexpectedPktNumErr))
+		strlcat(buf, "sunexperrpktnum ", blen);
+	if (err & ERR_MASK(SendDroppedSmpPktErr))
+		strlcat(buf, "sdroppedsmppkt ", blen);
+	if (err & ERR_MASK(SendMaxPktLenErr))
+		strlcat(buf, "smaxpktlen ", blen);
+	if (err & ERR_MASK(SendUnsupportedVLErr))
+		strlcat(buf, "sunsupVL ", blen);
+	if (err & ERR_MASK(InvalidAddrErr))
+		strlcat(buf, "invalidaddr ", blen);
+	if (err & ERR_MASK(RcvEgrFullErr))
+		strlcat(buf, "rcvegrfull ", blen);
+	if (err & ERR_MASK(RcvHdrFullErr))
+		strlcat(buf, "rcvhdrfull ", blen);
+	if (err & ERR_MASK(IBStatusChanged))
+		strlcat(buf, "ibcstatuschg ", blen);
+	if (err & ERR_MASK(RcvIBLostLinkErr))
+		strlcat(buf, "riblostlink ", blen);
+	if (err & ERR_MASK(HardwareErr))
+		strlcat(buf, "hardware ", blen);
+	if (err & ERR_MASK(ResetNegated))
+		strlcat(buf, "reset ", blen);
+	if (err & QLOGIC_IB_E_SDMAERRS)
+		qib_decode_7220_sdma_errs(dd->pport, err, buf, blen);
+	if (err & ERR_MASK(InvalidEEPCmd))
+		strlcat(buf, "invalideepromcmd ", blen);
+done:
+	return iserr;
+}
+
+static void reenable_7220_chase(unsigned long opaque)
+{
+	struct qib_pportdata *ppd = (struct qib_pportdata *)opaque;
+	ppd->cpspec->chase_timer.expires = 0;
+	qib_set_ib_7220_lstate(ppd, QLOGIC_IB_IBCC_LINKCMD_DOWN,
+		QLOGIC_IB_IBCC_LINKINITCMD_POLL);
+}
+
+static void handle_7220_chase(struct qib_pportdata *ppd, u64 ibcst)
+{
+	u8 ibclt;
+	u64 tnow;
+
+	ibclt = (u8)SYM_FIELD(ibcst, IBCStatus, LinkTrainingState);
+
+	/*
+	 * Detect and handle the state chase issue, where we can
+	 * get stuck if we are unlucky on timing on both sides of
+	 * the link.   If we are, we disable, set a timer, and
+	 * then re-enable.
+	 */
+	switch (ibclt) {
+	case IB_7220_LT_STATE_CFGRCVFCFG:
+	case IB_7220_LT_STATE_CFGWAITRMT:
+	case IB_7220_LT_STATE_TXREVLANES:
+	case IB_7220_LT_STATE_CFGENH:
+		tnow = get_jiffies_64();
+		if (ppd->cpspec->chase_end &&
+		    time_after64(tnow, ppd->cpspec->chase_end)) {
+			ppd->cpspec->chase_end = 0;
+			qib_set_ib_7220_lstate(ppd,
+				QLOGIC_IB_IBCC_LINKCMD_DOWN,
+				QLOGIC_IB_IBCC_LINKINITCMD_DISABLE);
+			ppd->cpspec->chase_timer.expires = jiffies +
+				QIB_CHASE_DIS_TIME;
+			add_timer(&ppd->cpspec->chase_timer);
+		} else if (!ppd->cpspec->chase_end)
+			ppd->cpspec->chase_end = tnow + QIB_CHASE_TIME;
+		break;
+
+	default:
+		ppd->cpspec->chase_end = 0;
+		break;
+	}
+}
+
+static void handle_7220_errors(struct qib_devdata *dd, u64 errs)
+{
+	char *msg;
+	u64 ignore_this_time = 0;
+	u64 iserr = 0;
+	int log_idx;
+	struct qib_pportdata *ppd = dd->pport;
+	u64 mask;
+
+	/* don't report errors that are masked */
+	errs &= dd->cspec->errormask;
+	msg = dd->cspec->emsgbuf;
+
+	/* do these first, they are most important */
+	if (errs & ERR_MASK(HardwareErr))
+		qib_7220_handle_hwerrors(dd, msg, sizeof dd->cspec->emsgbuf);
+	else
+		for (log_idx = 0; log_idx < QIB_EEP_LOG_CNT; ++log_idx)
+			if (errs & dd->eep_st_masks[log_idx].errs_to_log)
+				qib_inc_eeprom_err(dd, log_idx, 1);
+
+	if (errs & QLOGIC_IB_E_SDMAERRS)
+		sdma_7220_errors(ppd, errs);
+
+	if (errs & ~IB_E_BITSEXTANT)
+		qib_dev_err(dd, "error interrupt with unknown errors "
+			    "%llx set\n", (unsigned long long)
+			    (errs & ~IB_E_BITSEXTANT));
+
+	if (errs & E_SUM_ERRS) {
+		qib_disarm_7220_senderrbufs(ppd);
+		if ((errs & E_SUM_LINK_PKTERRS) &&
+		    !(ppd->lflags & QIBL_LINKACTIVE)) {
+			/*
+			 * This can happen when trying to bring the link
+			 * up, but the IB link changes state at the "wrong"
+			 * time. The IB logic then complains that the packet
+			 * isn't valid.  We don't want to confuse people, so
+			 * we just don't print them, except at debug
+			 */
+			ignore_this_time = errs & E_SUM_LINK_PKTERRS;
+		}
+	} else if ((errs & E_SUM_LINK_PKTERRS) &&
+		   !(ppd->lflags & QIBL_LINKACTIVE)) {
+		/*
+		 * This can happen when SMA is trying to bring the link
+		 * up, but the IB link changes state at the "wrong" time.
+		 * The IB logic then complains that the packet isn't
+		 * valid.  We don't want to confuse people, so we just
+		 * don't print them, except at debug
+		 */
+		ignore_this_time = errs & E_SUM_LINK_PKTERRS;
+	}
+
+	qib_write_kreg(dd, kr_errclear, errs);
+
+	errs &= ~ignore_this_time;
+	if (!errs)
+		goto done;
+
+	/*
+	 * The ones we mask off are handled specially below
+	 * or above.  Also mask SDMADISABLED by default as it
+	 * is too chatty.
+	 */
+	mask = ERR_MASK(IBStatusChanged) |
+		ERR_MASK(RcvEgrFullErr) | ERR_MASK(RcvHdrFullErr) |
+		ERR_MASK(HardwareErr) | ERR_MASK(SDmaDisabledErr);
+
+	qib_decode_7220_err(dd, msg, sizeof dd->cspec->emsgbuf, errs & ~mask);
+
+	if (errs & E_SUM_PKTERRS)
+		qib_stats.sps_rcverrs++;
+	if (errs & E_SUM_ERRS)
+		qib_stats.sps_txerrs++;
+	iserr = errs & ~(E_SUM_PKTERRS | QLOGIC_IB_E_PKTERRS |
+			 ERR_MASK(SDmaDisabledErr));
+
+	if (errs & ERR_MASK(IBStatusChanged)) {
+		u64 ibcs;
+
+		ibcs = qib_read_kreg64(dd, kr_ibcstatus);
+		if (!(ppd->lflags & QIBL_IB_AUTONEG_INPROG))
+			handle_7220_chase(ppd, ibcs);
+
+		/* Update our picture of width and speed from chip */
+		ppd->link_width_active =
+			((ibcs >> IBA7220_LINKWIDTH_SHIFT) & 1) ?
+			    IB_WIDTH_4X : IB_WIDTH_1X;
+		ppd->link_speed_active =
+			((ibcs >> IBA7220_LINKSPEED_SHIFT) & 1) ?
+			    QIB_IB_DDR : QIB_IB_SDR;
+
+		/*
+		 * Since going into a recovery state causes the link state
+		 * to go down and since recovery is transitory, it is better
+		 * if we "miss" ever seeing the link training state go into
+		 * recovery (i.e., ignore this transition for link state
+		 * special handling purposes) without updating lastibcstat.
+		 */
+		if (qib_7220_phys_portstate(ibcs) !=
+					    IB_PHYSPORTSTATE_LINK_ERR_RECOVER)
+			qib_handle_e_ibstatuschanged(ppd, ibcs);
+	}
+
+	if (errs & ERR_MASK(ResetNegated)) {
+		qib_dev_err(dd, "Got reset, requires re-init "
+			    "(unload and reload driver)\n");
+		dd->flags &= ~QIB_INITTED;  /* needs re-init */
+		/* mark as having had error */
+		*dd->devstatusp |= QIB_STATUS_HWERROR;
+		*dd->pport->statusp &= ~QIB_STATUS_IB_CONF;
+	}
+
+	if (*msg && iserr)
+		qib_dev_porterr(dd, ppd->port, "%s error\n", msg);
+
+	if (ppd->state_wanted & ppd->lflags)
+		wake_up_interruptible(&ppd->state_wait);
+
+	/*
+	 * If there were hdrq or egrfull errors, wake up any processes
+	 * waiting in poll.  We used to try to check which contexts had
+	 * the overflow, but given the cost of that and the chip reads
+	 * to support it, it's better to just wake everybody up if we
+	 * get an overflow; waiters can poll again if it's not them.
+	 */
+	if (errs & (ERR_MASK(RcvEgrFullErr) | ERR_MASK(RcvHdrFullErr))) {
+		qib_handle_urcv(dd, ~0U);
+		if (errs & ERR_MASK(RcvEgrFullErr))
+			qib_stats.sps_buffull++;
+		else
+			qib_stats.sps_hdrfull++;
+	}
+done:
+	return;
+}
+
+/* enable/disable chip from delivering interrupts */
+static void qib_7220_set_intr_state(struct qib_devdata *dd, u32 enable)
+{
+	if (enable) {
+		if (dd->flags & QIB_BADINTR)
+			return;
+		qib_write_kreg(dd, kr_intmask, ~0ULL);
+		/* force re-interrupt of any pending interrupts. */
+		qib_write_kreg(dd, kr_intclear, 0ULL);
+	} else
+		qib_write_kreg(dd, kr_intmask, 0ULL);
+}
+
+/*
+ * Try to cleanup as much as possible for anything that might have gone
+ * wrong while in freeze mode, such as pio buffers being written by user
+ * processes (causing armlaunch), send errors due to going into freeze mode,
+ * etc., and try to avoid causing extra interrupts while doing so.
+ * Forcibly update the in-memory pioavail register copies after cleanup
+ * because the chip won't do it while in freeze mode (the register values
+ * themselves are kept correct).
+ * Make sure that we don't lose any important interrupts by using the chip
+ * feature that says that writing 0 to a bit in *clear that is set in
+ * *status will cause an interrupt to be generated again (if allowed by
+ * the *mask value).
+ * This is in chip-specific code because of all of the register accesses,
+ * even though the details are similar on most chips.
+ */
+static void qib_7220_clear_freeze(struct qib_devdata *dd)
+{
+	/* disable error interrupts, to avoid confusion */
+	qib_write_kreg(dd, kr_errmask, 0ULL);
+
+	/* also disable interrupts; errormask is sometimes overwriten */
+	qib_7220_set_intr_state(dd, 0);
+
+	qib_cancel_sends(dd->pport);
+
+	/* clear the freeze, and be sure chip saw it */
+	qib_write_kreg(dd, kr_control, dd->control);
+	qib_read_kreg32(dd, kr_scratch);
+
+	/* force in-memory update now we are out of freeze */
+	qib_force_pio_avail_update(dd);
+
+	/*
+	 * force new interrupt if any hwerr, error or interrupt bits are
+	 * still set, and clear "safe" send packet errors related to freeze
+	 * and cancelling sends.  Re-enable error interrupts before possible
+	 * force of re-interrupt on pending interrupts.
+	 */
+	qib_write_kreg(dd, kr_hwerrclear, 0ULL);
+	qib_write_kreg(dd, kr_errclear, E_SPKT_ERRS_IGNORE);
+	qib_write_kreg(dd, kr_errmask, dd->cspec->errormask);
+	qib_7220_set_intr_state(dd, 1);
+}
+
+/**
+ * qib_7220_handle_hwerrors - display hardware errors.
+ * @dd: the qlogic_ib device
+ * @msg: the output buffer
+ * @msgl: the size of the output buffer
+ *
+ * Use same msg buffer as regular errors to avoid excessive stack
+ * use.  Most hardware errors are catastrophic, but for right now,
+ * we'll print them and continue.  We reuse the same message buffer as
+ * handle_7220_errors() to avoid excessive stack usage.
+ */
+static void qib_7220_handle_hwerrors(struct qib_devdata *dd, char *msg,
+				     size_t msgl)
+{
+	u64 hwerrs;
+	u32 bits, ctrl;
+	int isfatal = 0;
+	char *bitsmsg;
+	int log_idx;
+
+	hwerrs = qib_read_kreg64(dd, kr_hwerrstatus);
+	if (!hwerrs)
+		goto bail;
+	if (hwerrs == ~0ULL) {
+		qib_dev_err(dd, "Read of hardware error status failed "
+			    "(all bits set); ignoring\n");
+		goto bail;
+	}
+	qib_stats.sps_hwerrs++;
+
+	/*
+	 * Always clear the error status register, except MEMBISTFAIL,
+	 * regardless of whether we continue or stop using the chip.
+	 * We want that set so we know it failed, even across driver reload.
+	 * We'll still ignore it in the hwerrmask.  We do this partly for
+	 * diagnostics, but also for support.
+	 */
+	qib_write_kreg(dd, kr_hwerrclear,
+		       hwerrs & ~HWE_MASK(PowerOnBISTFailed));
+
+	hwerrs &= dd->cspec->hwerrmask;
+
+	/* We log some errors to EEPROM, check if we have any of those. */
+	for (log_idx = 0; log_idx < QIB_EEP_LOG_CNT; ++log_idx)
+		if (hwerrs & dd->eep_st_masks[log_idx].hwerrs_to_log)
+			qib_inc_eeprom_err(dd, log_idx, 1);
+	if (hwerrs & ~(TXEMEMPARITYERR_PIOBUF | TXEMEMPARITYERR_PIOPBC |
+		       RXE_PARITY))
+		qib_devinfo(dd->pcidev, "Hardware error: hwerr=0x%llx "
+			 "(cleared)\n", (unsigned long long) hwerrs);
+
+	if (hwerrs & ~IB_HWE_BITSEXTANT)
+		qib_dev_err(dd, "hwerror interrupt with unknown errors "
+			    "%llx set\n", (unsigned long long)
+			    (hwerrs & ~IB_HWE_BITSEXTANT));
+
+	if (hwerrs & QLOGIC_IB_HWE_IB_UC_MEMORYPARITYERR)
+		qib_sd7220_clr_ibpar(dd);
+
+	ctrl = qib_read_kreg32(dd, kr_control);
+	if ((ctrl & QLOGIC_IB_C_FREEZEMODE) && !dd->diag_client) {
+		/*
+		 * Parity errors in send memory are recoverable by h/w
+		 * just do housekeeping, exit freeze mode and continue.
+		 */
+		if (hwerrs & (TXEMEMPARITYERR_PIOBUF |
+			      TXEMEMPARITYERR_PIOPBC)) {
+			qib_7220_txe_recover(dd);
+			hwerrs &= ~(TXEMEMPARITYERR_PIOBUF |
+				    TXEMEMPARITYERR_PIOPBC);
+		}
+		if (hwerrs)
+			isfatal = 1;
+		else
+			qib_7220_clear_freeze(dd);
+	}
+
+	*msg = '\0';
+
+	if (hwerrs & HWE_MASK(PowerOnBISTFailed)) {
+		isfatal = 1;
+		strlcat(msg, "[Memory BIST test failed, "
+			"InfiniPath hardware unusable]", msgl);
+		/* ignore from now on, so disable until driver reloaded */
+		dd->cspec->hwerrmask &= ~HWE_MASK(PowerOnBISTFailed);
+		qib_write_kreg(dd, kr_hwerrmask, dd->cspec->hwerrmask);
+	}
+
+	qib_format_hwerrors(hwerrs, qib_7220_hwerror_msgs,
+			    ARRAY_SIZE(qib_7220_hwerror_msgs), msg, msgl);
+
+	bitsmsg = dd->cspec->bitsmsgbuf;
+	if (hwerrs & (QLOGIC_IB_HWE_PCIEMEMPARITYERR_MASK <<
+		      QLOGIC_IB_HWE_PCIEMEMPARITYERR_SHIFT)) {
+		bits = (u32) ((hwerrs >>
+			       QLOGIC_IB_HWE_PCIEMEMPARITYERR_SHIFT) &
+			      QLOGIC_IB_HWE_PCIEMEMPARITYERR_MASK);
+		snprintf(bitsmsg, sizeof dd->cspec->bitsmsgbuf,
+			 "[PCIe Mem Parity Errs %x] ", bits);
+		strlcat(msg, bitsmsg, msgl);
+	}
+
+#define _QIB_PLL_FAIL (QLOGIC_IB_HWE_COREPLL_FBSLIP |   \
+			 QLOGIC_IB_HWE_COREPLL_RFSLIP)
+
+	if (hwerrs & _QIB_PLL_FAIL) {
+		isfatal = 1;
+		snprintf(bitsmsg, sizeof dd->cspec->bitsmsgbuf,
+			 "[PLL failed (%llx), InfiniPath hardware unusable]",
+			 (unsigned long long) hwerrs & _QIB_PLL_FAIL);
+		strlcat(msg, bitsmsg, msgl);
+		/* ignore from now on, so disable until driver reloaded */
+		dd->cspec->hwerrmask &= ~(hwerrs & _QIB_PLL_FAIL);
+		qib_write_kreg(dd, kr_hwerrmask, dd->cspec->hwerrmask);
+	}
+
+	if (hwerrs & QLOGIC_IB_HWE_SERDESPLLFAILED) {
+		/*
+		 * If it occurs, it is left masked since the eternal
+		 * interface is unused.
+		 */
+		dd->cspec->hwerrmask &= ~QLOGIC_IB_HWE_SERDESPLLFAILED;
+		qib_write_kreg(dd, kr_hwerrmask, dd->cspec->hwerrmask);
+	}
+
+	qib_dev_err(dd, "%s hardware error\n", msg);
+
+	if (isfatal && !dd->diag_client) {
+		qib_dev_err(dd, "Fatal Hardware Error, no longer"
+			    " usable, SN %.16s\n", dd->serial);
+		/*
+		 * For /sys status file and user programs to print; if no
+		 * trailing brace is copied, we'll know it was truncated.
+		 */
+		if (dd->freezemsg)
+			snprintf(dd->freezemsg, dd->freezelen,
+				 "{%s}", msg);
+		qib_disable_after_error(dd);
+	}
+bail:;
+}
+
+/**
+ * qib_7220_init_hwerrors - enable hardware errors
+ * @dd: the qlogic_ib device
+ *
+ * now that we have finished initializing everything that might reasonably
+ * cause a hardware error, and cleared those errors bits as they occur,
+ * we can enable hardware errors in the mask (potentially enabling
+ * freeze mode), and enable hardware errors as errors (along with
+ * everything else) in errormask
+ */
+static void qib_7220_init_hwerrors(struct qib_devdata *dd)
+{
+	u64 val;
+	u64 extsval;
+
+	extsval = qib_read_kreg64(dd, kr_extstatus);
+
+	if (!(extsval & (QLOGIC_IB_EXTS_MEMBIST_ENDTEST |
+			 QLOGIC_IB_EXTS_MEMBIST_DISABLED)))
+		qib_dev_err(dd, "MemBIST did not complete!\n");
+	if (extsval & QLOGIC_IB_EXTS_MEMBIST_DISABLED)
+		qib_devinfo(dd->pcidev, "MemBIST is disabled.\n");
+
+	val = ~0ULL;    /* default to all hwerrors become interrupts, */
+
+	val &= ~QLOGIC_IB_HWE_IB_UC_MEMORYPARITYERR;
+	dd->cspec->hwerrmask = val;
+
+	qib_write_kreg(dd, kr_hwerrclear, ~HWE_MASK(PowerOnBISTFailed));
+	qib_write_kreg(dd, kr_hwerrmask, dd->cspec->hwerrmask);
+
+	/* clear all */
+	qib_write_kreg(dd, kr_errclear, ~0ULL);
+	/* enable errors that are masked, at least this first time. */
+	qib_write_kreg(dd, kr_errmask, ~0ULL);
+	dd->cspec->errormask = qib_read_kreg64(dd, kr_errmask);
+	/* clear any interrupts up to this point (ints still not enabled) */
+	qib_write_kreg(dd, kr_intclear, ~0ULL);
+}
+
+/*
+ * Disable and enable the armlaunch error.  Used for PIO bandwidth testing
+ * on chips that are count-based, rather than trigger-based.  There is no
+ * reference counting, but that's also fine, given the intended use.
+ * Only chip-specific because it's all register accesses
+ */
+static void qib_set_7220_armlaunch(struct qib_devdata *dd, u32 enable)
+{
+	if (enable) {
+		qib_write_kreg(dd, kr_errclear, ERR_MASK(SendPioArmLaunchErr));
+		dd->cspec->errormask |= ERR_MASK(SendPioArmLaunchErr);
+	} else
+		dd->cspec->errormask &= ~ERR_MASK(SendPioArmLaunchErr);
+	qib_write_kreg(dd, kr_errmask, dd->cspec->errormask);
+}
+
+/*
+ * Formerly took parameter <which> in pre-shifted,
+ * pre-merged form with LinkCmd and LinkInitCmd
+ * together, and assuming the zero was NOP.
+ */
+static void qib_set_ib_7220_lstate(struct qib_pportdata *ppd, u16 linkcmd,
+				   u16 linitcmd)
+{
+	u64 mod_wd;
+	struct qib_devdata *dd = ppd->dd;
+	unsigned long flags;
+
+	if (linitcmd == QLOGIC_IB_IBCC_LINKINITCMD_DISABLE) {
+		/*
+		 * If we are told to disable, note that so link-recovery
+		 * code does not attempt to bring us back up.
+		 */
+		spin_lock_irqsave(&ppd->lflags_lock, flags);
+		ppd->lflags |= QIBL_IB_LINK_DISABLED;
+		spin_unlock_irqrestore(&ppd->lflags_lock, flags);
+	} else if (linitcmd || linkcmd == QLOGIC_IB_IBCC_LINKCMD_DOWN) {
+		/*
+		 * Any other linkinitcmd will lead to LINKDOWN and then
+		 * to INIT (if all is well), so clear flag to let
+		 * link-recovery code attempt to bring us back up.
+		 */
+		spin_lock_irqsave(&ppd->lflags_lock, flags);
+		ppd->lflags &= ~QIBL_IB_LINK_DISABLED;
+		spin_unlock_irqrestore(&ppd->lflags_lock, flags);
+	}
+
+	mod_wd = (linkcmd << IBA7220_IBCC_LINKCMD_SHIFT) |
+		(linitcmd << QLOGIC_IB_IBCC_LINKINITCMD_SHIFT);
+
+	qib_write_kreg(dd, kr_ibcctrl, ppd->cpspec->ibcctrl | mod_wd);
+	/* write to chip to prevent back-to-back writes of ibc reg */
+	qib_write_kreg(dd, kr_scratch, 0);
+}
+
+/*
+ * All detailed interaction with the SerDes has been moved to qib_sd7220.c
+ *
+ * The portion of IBA7220-specific bringup_serdes() that actually deals with
+ * registers and memory within the SerDes itself is qib_sd7220_init().
+ */
+
+/**
+ * qib_7220_bringup_serdes - bring up the serdes
+ * @ppd: physical port on the qlogic_ib device
+ */
+static int qib_7220_bringup_serdes(struct qib_pportdata *ppd)
+{
+	struct qib_devdata *dd = ppd->dd;
+	u64 val, prev_val, guid, ibc;
+	int ret = 0;
+
+	/* Put IBC in reset, sends disabled */
+	dd->control &= ~QLOGIC_IB_C_LINKENABLE;
+	qib_write_kreg(dd, kr_control, 0ULL);
+
+	if (qib_compat_ddr_negotiate) {
+		ppd->cpspec->ibdeltainprog = 1;
+		ppd->cpspec->ibsymsnap = read_7220_creg32(dd, cr_ibsymbolerr);
+		ppd->cpspec->iblnkerrsnap =
+			read_7220_creg32(dd, cr_iblinkerrrecov);
+	}
+
+	/* flowcontrolwatermark is in units of KBytes */
+	ibc = 0x5ULL << SYM_LSB(IBCCtrl, FlowCtrlWaterMark);
+	/*
+	 * How often flowctrl sent.  More or less in usecs; balance against
+	 * watermark value, so that in theory senders always get a flow
+	 * control update in time to not let the IB link go idle.
+	 */
+	ibc |= 0x3ULL << SYM_LSB(IBCCtrl, FlowCtrlPeriod);
+	/* max error tolerance */
+	ibc |= 0xfULL << SYM_LSB(IBCCtrl, PhyerrThreshold);
+	/* use "real" buffer space for */
+	ibc |= 4ULL << SYM_LSB(IBCCtrl, CreditScale);
+	/* IB credit flow control. */
+	ibc |= 0xfULL << SYM_LSB(IBCCtrl, OverrunThreshold);
+	/*
+	 * set initial max size pkt IBC will send, including ICRC; it's the
+	 * PIO buffer size in dwords, less 1; also see qib_set_mtu()
+	 */
+	ibc |= ((u64)(ppd->ibmaxlen >> 2) + 1) << SYM_LSB(IBCCtrl, MaxPktLen);
+	ppd->cpspec->ibcctrl = ibc; /* without linkcmd or linkinitcmd! */
+
+	/* initially come up waiting for TS1, without sending anything. */
+	val = ppd->cpspec->ibcctrl | (QLOGIC_IB_IBCC_LINKINITCMD_DISABLE <<
+		QLOGIC_IB_IBCC_LINKINITCMD_SHIFT);
+	qib_write_kreg(dd, kr_ibcctrl, val);
+
+	if (!ppd->cpspec->ibcddrctrl) {
+		/* not on re-init after reset */
+		ppd->cpspec->ibcddrctrl = qib_read_kreg64(dd, kr_ibcddrctrl);
+
+		if (ppd->link_speed_enabled == (QIB_IB_SDR | QIB_IB_DDR))
+			ppd->cpspec->ibcddrctrl |=
+				IBA7220_IBC_SPEED_AUTONEG_MASK |
+				IBA7220_IBC_IBTA_1_2_MASK;
+		else
+			ppd->cpspec->ibcddrctrl |=
+				ppd->link_speed_enabled == QIB_IB_DDR ?
+				IBA7220_IBC_SPEED_DDR : IBA7220_IBC_SPEED_SDR;
+		if ((ppd->link_width_enabled & (IB_WIDTH_1X | IB_WIDTH_4X)) ==
+		    (IB_WIDTH_1X | IB_WIDTH_4X))
+			ppd->cpspec->ibcddrctrl |= IBA7220_IBC_WIDTH_AUTONEG;
+		else
+			ppd->cpspec->ibcddrctrl |=
+				ppd->link_width_enabled == IB_WIDTH_4X ?
+				IBA7220_IBC_WIDTH_4X_ONLY :
+				IBA7220_IBC_WIDTH_1X_ONLY;
+
+		/* always enable these on driver reload, not sticky */
+		ppd->cpspec->ibcddrctrl |=
+			IBA7220_IBC_RXPOL_MASK << IBA7220_IBC_RXPOL_SHIFT;
+		ppd->cpspec->ibcddrctrl |=
+			IBA7220_IBC_HRTBT_MASK << IBA7220_IBC_HRTBT_SHIFT;
+
+		/* enable automatic lane reversal detection for receive */
+		ppd->cpspec->ibcddrctrl |= IBA7220_IBC_LANE_REV_SUPPORTED;
+	} else
+		/* write to chip to prevent back-to-back writes of ibc reg */
+		qib_write_kreg(dd, kr_scratch, 0);
+
+	qib_write_kreg(dd, kr_ibcddrctrl, ppd->cpspec->ibcddrctrl);
+	qib_write_kreg(dd, kr_scratch, 0);
+
+	qib_write_kreg(dd, kr_ncmodectrl, 0Ull);
+	qib_write_kreg(dd, kr_scratch, 0);
+
+	ret = qib_sd7220_init(dd);
+
+	val = qib_read_kreg64(dd, kr_xgxs_cfg);
+	prev_val = val;
+	val |= QLOGIC_IB_XGXS_FC_SAFE;
+	if (val != prev_val) {
+		qib_write_kreg(dd, kr_xgxs_cfg, val);
+		qib_read_kreg32(dd, kr_scratch);
+	}
+	if (val & QLOGIC_IB_XGXS_RESET)
+		val &= ~QLOGIC_IB_XGXS_RESET;
+	if (val != prev_val)
+		qib_write_kreg(dd, kr_xgxs_cfg, val);
+
+	/* first time through, set port guid */
+	if (!ppd->guid)
+		ppd->guid = dd->base_guid;
+	guid = be64_to_cpu(ppd->guid);
+
+	qib_write_kreg(dd, kr_hrtbt_guid, guid);
+	if (!ret) {
+		dd->control |= QLOGIC_IB_C_LINKENABLE;
+		qib_write_kreg(dd, kr_control, dd->control);
+	} else
+		/* write to chip to prevent back-to-back writes of ibc reg */
+		qib_write_kreg(dd, kr_scratch, 0);
+	return ret;
+}
+
+/**
+ * qib_7220_quiet_serdes - set serdes to txidle
+ * @ppd: physical port of the qlogic_ib device
+ * Called when driver is being unloaded
+ */
+static void qib_7220_quiet_serdes(struct qib_pportdata *ppd)
+{
+	u64 val;
+	struct qib_devdata *dd = ppd->dd;
+	unsigned long flags;
+
+	/* disable IBC */
+	dd->control &= ~QLOGIC_IB_C_LINKENABLE;
+	qib_write_kreg(dd, kr_control,
+		       dd->control | QLOGIC_IB_C_FREEZEMODE);
+
+	ppd->cpspec->chase_end = 0;
+	if (ppd->cpspec->chase_timer.data) /* if initted */
+		del_timer_sync(&ppd->cpspec->chase_timer);
+
+	if (ppd->cpspec->ibsymdelta || ppd->cpspec->iblnkerrdelta ||
+	    ppd->cpspec->ibdeltainprog) {
+		u64 diagc;
+
+		/* enable counter writes */
+		diagc = qib_read_kreg64(dd, kr_hwdiagctrl);
+		qib_write_kreg(dd, kr_hwdiagctrl,
+			       diagc | SYM_MASK(HwDiagCtrl, CounterWrEnable));
+
+		if (ppd->cpspec->ibsymdelta || ppd->cpspec->ibdeltainprog) {
+			val = read_7220_creg32(dd, cr_ibsymbolerr);
+			if (ppd->cpspec->ibdeltainprog)
+				val -= val - ppd->cpspec->ibsymsnap;
+			val -= ppd->cpspec->ibsymdelta;
+			write_7220_creg(dd, cr_ibsymbolerr, val);
+		}
+		if (ppd->cpspec->iblnkerrdelta || ppd->cpspec->ibdeltainprog) {
+			val = read_7220_creg32(dd, cr_iblinkerrrecov);
+			if (ppd->cpspec->ibdeltainprog)
+				val -= val - ppd->cpspec->iblnkerrsnap;
+			val -= ppd->cpspec->iblnkerrdelta;
+			write_7220_creg(dd, cr_iblinkerrrecov, val);
+		}
+
+		/* and disable counter writes */
+		qib_write_kreg(dd, kr_hwdiagctrl, diagc);
+	}
+	qib_set_ib_7220_lstate(ppd, 0, QLOGIC_IB_IBCC_LINKINITCMD_DISABLE);
+
+	spin_lock_irqsave(&ppd->lflags_lock, flags);
+	ppd->lflags &= ~QIBL_IB_AUTONEG_INPROG;
+	spin_unlock_irqrestore(&ppd->lflags_lock, flags);
+	wake_up(&ppd->cpspec->autoneg_wait);
+	cancel_delayed_work(&ppd->cpspec->autoneg_work);
+	flush_scheduled_work();
+
+	shutdown_7220_relock_poll(ppd->dd);
+	val = qib_read_kreg64(ppd->dd, kr_xgxs_cfg);
+	val |= QLOGIC_IB_XGXS_RESET;
+	qib_write_kreg(ppd->dd, kr_xgxs_cfg, val);
+}
+
+/**
+ * qib_setup_7220_setextled - set the state of the two external LEDs
+ * @dd: the qlogic_ib device
+ * @on: whether the link is up or not
+ *
+ * The exact combo of LEDs if on is true is determined by looking
+ * at the ibcstatus.
+ *
+ * These LEDs indicate the physical and logical state of IB link.
+ * For this chip (at least with recommended board pinouts), LED1
+ * is Yellow (logical state) and LED2 is Green (physical state),
+ *
+ * Note:  We try to match the Mellanox HCA LED behavior as best
+ * we can.  Green indicates physical link state is OK (something is
+ * plugged in, and we can train).
+ * Amber indicates the link is logically up (ACTIVE).
+ * Mellanox further blinks the amber LED to indicate data packet
+ * activity, but we have no hardware support for that, so it would
+ * require waking up every 10-20 msecs and checking the counters
+ * on the chip, and then turning the LED off if appropriate.  That's
+ * visible overhead, so not something we will do.
+ *
+ */
+static void qib_setup_7220_setextled(struct qib_pportdata *ppd, u32 on)
+{
+	struct qib_devdata *dd = ppd->dd;
+	u64 extctl, ledblink = 0, val, lst, ltst;
+	unsigned long flags;
+
+	/*
+	 * The diags use the LED to indicate diag info, so we leave
+	 * the external LED alone when the diags are running.
+	 */
+	if (dd->diag_client)
+		return;
+
+	if (ppd->led_override) {
+		ltst = (ppd->led_override & QIB_LED_PHYS) ?
+			IB_PHYSPORTSTATE_LINKUP : IB_PHYSPORTSTATE_DISABLED,
+		lst = (ppd->led_override & QIB_LED_LOG) ?
+			IB_PORT_ACTIVE : IB_PORT_DOWN;
+	} else if (on) {
+		val = qib_read_kreg64(dd, kr_ibcstatus);
+		ltst = qib_7220_phys_portstate(val);
+		lst = qib_7220_iblink_state(val);
+	} else {
+		ltst = 0;
+		lst = 0;
+	}
+
+	spin_lock_irqsave(&dd->cspec->gpio_lock, flags);
+	extctl = dd->cspec->extctrl & ~(SYM_MASK(EXTCtrl, LEDPriPortGreenOn) |
+				 SYM_MASK(EXTCtrl, LEDPriPortYellowOn));
+	if (ltst == IB_PHYSPORTSTATE_LINKUP) {
+		extctl |= SYM_MASK(EXTCtrl, LEDPriPortGreenOn);
+		/*
+		 * counts are in chip clock (4ns) periods.
+		 * This is 1/16 sec (66.6ms) on,
+		 * 3/16 sec (187.5 ms) off, with packets rcvd
+		 */
+		ledblink = ((66600 * 1000UL / 4) << IBA7220_LEDBLINK_ON_SHIFT)
+			| ((187500 * 1000UL / 4) << IBA7220_LEDBLINK_OFF_SHIFT);
+	}
+	if (lst == IB_PORT_ACTIVE)
+		extctl |= SYM_MASK(EXTCtrl, LEDPriPortYellowOn);
+	dd->cspec->extctrl = extctl;
+	qib_write_kreg(dd, kr_extctrl, extctl);
+	spin_unlock_irqrestore(&dd->cspec->gpio_lock, flags);
+
+	if (ledblink) /* blink the LED on packet receive */
+		qib_write_kreg(dd, kr_rcvpktledcnt, ledblink);
+}
+
+static void qib_7220_free_irq(struct qib_devdata *dd)
+{
+	if (dd->cspec->irq) {
+		free_irq(dd->cspec->irq, dd);
+		dd->cspec->irq = 0;
+	}
+	qib_nomsi(dd);
+}
+
+/*
+ * qib_setup_7220_cleanup - clean up any per-chip chip-specific stuff
+ * @dd: the qlogic_ib device
+ *
+ * This is called during driver unload.
+ *
+ */
+static void qib_setup_7220_cleanup(struct qib_devdata *dd)
+{
+	qib_7220_free_irq(dd);
+	kfree(dd->cspec->cntrs);
+	kfree(dd->cspec->portcntrs);
+}
+
+/*
+ * This is only called for SDmaInt.
+ * SDmaDisabled is handled on the error path.
+ */
+static void sdma_7220_intr(struct qib_pportdata *ppd, u64 istat)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&ppd->sdma_lock, flags);
+
+	switch (ppd->sdma_state.current_state) {
+	case qib_sdma_state_s00_hw_down:
+		break;
+
+	case qib_sdma_state_s10_hw_start_up_wait:
+		__qib_sdma_process_event(ppd, qib_sdma_event_e20_hw_started);
+		break;
+
+	case qib_sdma_state_s20_idle:
+		break;
+
+	case qib_sdma_state_s30_sw_clean_up_wait:
+		break;
+
+	case qib_sdma_state_s40_hw_clean_up_wait:
+		break;
+
+	case qib_sdma_state_s50_hw_halt_wait:
+		__qib_sdma_process_event(ppd, qib_sdma_event_e60_hw_halted);
+		break;
+
+	case qib_sdma_state_s99_running:
+		/* too chatty to print here */
+		__qib_sdma_intr(ppd);
+		break;
+	}
+	spin_unlock_irqrestore(&ppd->sdma_lock, flags);
+}
+
+static void qib_wantpiobuf_7220_intr(struct qib_devdata *dd, u32 needint)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&dd->sendctrl_lock, flags);
+	if (needint) {
+		if (!(dd->sendctrl & SYM_MASK(SendCtrl, SendBufAvailUpd)))
+			goto done;
+		/*
+		 * blip the availupd off, next write will be on, so
+		 * we ensure an avail update, regardless of threshold or
+		 * buffers becoming free, whenever we want an interrupt
+		 */
+		qib_write_kreg(dd, kr_sendctrl, dd->sendctrl &
+			~SYM_MASK(SendCtrl, SendBufAvailUpd));
+		qib_write_kreg(dd, kr_scratch, 0ULL);
+		dd->sendctrl |= SYM_MASK(SendCtrl, SendIntBufAvail);
+	} else
+		dd->sendctrl &= ~SYM_MASK(SendCtrl, SendIntBufAvail);
+	qib_write_kreg(dd, kr_sendctrl, dd->sendctrl);
+	qib_write_kreg(dd, kr_scratch, 0ULL);
+done:
+	spin_unlock_irqrestore(&dd->sendctrl_lock, flags);
+}
+
+/*
+ * Handle errors and unusual events first, separate function
+ * to improve cache hits for fast path interrupt handling.
+ */
+static noinline void unlikely_7220_intr(struct qib_devdata *dd, u64 istat)
+{
+	if (unlikely(istat & ~QLOGIC_IB_I_BITSEXTANT))
+		qib_dev_err(dd,
+			    "interrupt with unknown interrupts %Lx set\n",
+			    istat & ~QLOGIC_IB_I_BITSEXTANT);
+
+	if (istat & QLOGIC_IB_I_GPIO) {
+		u32 gpiostatus;
+
+		/*
+		 * Boards for this chip currently don't use GPIO interrupts,
+		 * so clear by writing GPIOstatus to GPIOclear, and complain
+		 * to alert developer. To avoid endless repeats, clear
+		 * the bits in the mask, since there is some kind of
+		 * programming error or chip problem.
+		 */
+		gpiostatus = qib_read_kreg32(dd, kr_gpio_status);
+		/*
+		 * In theory, writing GPIOstatus to GPIOclear could
+		 * have a bad side-effect on some diagnostic that wanted
+		 * to poll for a status-change, but the various shadows
+		 * make that problematic at best. Diags will just suppress
+		 * all GPIO interrupts during such tests.
+		 */
+		qib_write_kreg(dd, kr_gpio_clear, gpiostatus);
+
+		if (gpiostatus) {
+			const u32 mask = qib_read_kreg32(dd, kr_gpio_mask);
+			u32 gpio_irq = mask & gpiostatus;
+
+			/*
+			 * A bit set in status and (chip) Mask register
+			 * would cause an interrupt. Since we are not
+			 * expecting any, report it. Also check that the
+			 * chip reflects our shadow, report issues,
+			 * and refresh from the shadow.
+			 */
+			/*
+			 * Clear any troublemakers, and update chip
+			 * from shadow
+			 */
+			dd->cspec->gpio_mask &= ~gpio_irq;
+			qib_write_kreg(dd, kr_gpio_mask, dd->cspec->gpio_mask);
+		}
+	}
+
+	if (istat & QLOGIC_IB_I_ERROR) {
+		u64 estat;
+
+		qib_stats.sps_errints++;
+		estat = qib_read_kreg64(dd, kr_errstatus);
+		if (!estat)
+			qib_devinfo(dd->pcidev, "error interrupt (%Lx), "
+				 "but no error bits set!\n", istat);
+		else
+			handle_7220_errors(dd, estat);
+	}
+}
+
+static irqreturn_t qib_7220intr(int irq, void *data)
+{
+	struct qib_devdata *dd = data;
+	irqreturn_t ret;
+	u64 istat;
+	u64 ctxtrbits;
+	u64 rmask;
+	unsigned i;
+
+	if ((dd->flags & (QIB_PRESENT | QIB_BADINTR)) != QIB_PRESENT) {
+		/*
+		 * This return value is not great, but we do not want the
+		 * interrupt core code to remove our interrupt handler
+		 * because we don't appear to be handling an interrupt
+		 * during a chip reset.
+		 */
+		ret = IRQ_HANDLED;
+		goto bail;
+	}
+
+	istat = qib_read_kreg64(dd, kr_intstatus);
+
+	if (unlikely(!istat)) {
+		ret = IRQ_NONE; /* not our interrupt, or already handled */
+		goto bail;
+	}
+	if (unlikely(istat == -1)) {
+		qib_bad_intrstatus(dd);
+		/* don't know if it was our interrupt or not */
+		ret = IRQ_NONE;
+		goto bail;
+	}
+
+	qib_stats.sps_ints++;
+	if (dd->int_counter != (u32) -1)
+		dd->int_counter++;
+
+	if (unlikely(istat & (~QLOGIC_IB_I_BITSEXTANT |
+			      QLOGIC_IB_I_GPIO | QLOGIC_IB_I_ERROR)))
+		unlikely_7220_intr(dd, istat);
+
+	/*
+	 * Clear the interrupt bits we found set, relatively early, so we
+	 * "know" know the chip will have seen this by the time we process
+	 * the queue, and will re-interrupt if necessary.  The processor
+	 * itself won't take the interrupt again until we return.
+	 */
+	qib_write_kreg(dd, kr_intclear, istat);
+
+	/*
+	 * Handle kernel receive queues before checking for pio buffers
+	 * available since receives can overflow; piobuf waiters can afford
+	 * a few extra cycles, since they were waiting anyway.
+	 */
+	ctxtrbits = istat &
+		((QLOGIC_IB_I_RCVAVAIL_MASK << QLOGIC_IB_I_RCVAVAIL_SHIFT) |
+		 (QLOGIC_IB_I_RCVURG_MASK << QLOGIC_IB_I_RCVURG_SHIFT));
+	if (ctxtrbits) {
+		rmask = (1ULL << QLOGIC_IB_I_RCVAVAIL_SHIFT) |
+			(1ULL << QLOGIC_IB_I_RCVURG_SHIFT);
+		for (i = 0; i < dd->first_user_ctxt; i++) {
+			if (ctxtrbits & rmask) {
+				ctxtrbits &= ~rmask;
+				qib_kreceive(dd->rcd[i], NULL, NULL);
+			}
+			rmask <<= 1;
+		}
+		if (ctxtrbits) {
+			ctxtrbits =
+				(ctxtrbits >> QLOGIC_IB_I_RCVAVAIL_SHIFT) |
+				(ctxtrbits >> QLOGIC_IB_I_RCVURG_SHIFT);
+			qib_handle_urcv(dd, ctxtrbits);
+		}
+	}
+
+	/* only call for SDmaInt */
+	if (istat & QLOGIC_IB_I_SDMAINT)
+		sdma_7220_intr(dd->pport, istat);
+
+	if ((istat & QLOGIC_IB_I_SPIOBUFAVAIL) && (dd->flags & QIB_INITTED))
+		qib_ib_piobufavail(dd);
+
+	ret = IRQ_HANDLED;
+bail:
+	return ret;
+}
+
+/*
+ * Set up our chip-specific interrupt handler.
+ * The interrupt type has already been setup, so
+ * we just need to do the registration and error checking.
+ * If we are using MSI interrupts, we may fall back to
+ * INTx later, if the interrupt handler doesn't get called
+ * within 1/2 second (see verify_interrupt()).
+ */
+static void qib_setup_7220_interrupt(struct qib_devdata *dd)
+{
+	if (!dd->cspec->irq)
+		qib_dev_err(dd, "irq is 0, BIOS error?  Interrupts won't "
+			    "work\n");
+	else {
+		int ret = request_irq(dd->cspec->irq, qib_7220intr,
+			dd->msi_lo ? 0 : IRQF_SHARED,
+			QIB_DRV_NAME, dd);
+
+		if (ret)
+			qib_dev_err(dd, "Couldn't setup %s interrupt "
+				    "(irq=%d): %d\n", dd->msi_lo ?
+				    "MSI" : "INTx", dd->cspec->irq, ret);
+	}
+}
+
+/**
+ * qib_7220_boardname - fill in the board name
+ * @dd: the qlogic_ib device
+ *
+ * info is based on the board revision register
+ */
+static void qib_7220_boardname(struct qib_devdata *dd)
+{
+	char *n;
+	u32 boardid, namelen;
+
+	boardid = SYM_FIELD(dd->revision, Revision,
+			    BoardID);
+
+	switch (boardid) {
+	case 1:
+		n = "InfiniPath_QLE7240";
+		break;
+	case 2:
+		n = "InfiniPath_QLE7280";
+		break;
+	default:
+		qib_dev_err(dd, "Unknown 7220 board with ID %u\n", boardid);
+		n = "Unknown_InfiniPath_7220";
+		break;
+	}
+
+	namelen = strlen(n) + 1;
+	dd->boardname = kmalloc(namelen, GFP_KERNEL);
+	if (!dd->boardname)
+		qib_dev_err(dd, "Failed allocation for board name: %s\n", n);
+	else
+		snprintf(dd->boardname, namelen, "%s", n);
+
+	if (dd->majrev != 5 || !dd->minrev || dd->minrev > 2)
+		qib_dev_err(dd, "Unsupported InfiniPath hardware "
+			    "revision %u.%u!\n",
+			    dd->majrev, dd->minrev);
+
+	snprintf(dd->boardversion, sizeof(dd->boardversion),
+		 "ChipABI %u.%u, %s, InfiniPath%u %u.%u, SW Compat %u\n",
+		 QIB_CHIP_VERS_MAJ, QIB_CHIP_VERS_MIN, dd->boardname,
+		 (unsigned)SYM_FIELD(dd->revision, Revision_R, Arch),
+		 dd->majrev, dd->minrev,
+		 (unsigned)SYM_FIELD(dd->revision, Revision_R, SW));
+}
+
+/*
+ * This routine sleeps, so it can only be called from user context, not
+ * from interrupt context.
+ */
+static int qib_setup_7220_reset(struct qib_devdata *dd)
+{
+	u64 val;
+	int i;
+	int ret;
+	u16 cmdval;
+	u8 int_line, clinesz;
+	unsigned long flags;
+
+	qib_pcie_getcmd(dd, &cmdval, &int_line, &clinesz);
+
+	/* Use dev_err so it shows up in logs, etc. */
+	qib_dev_err(dd, "Resetting InfiniPath unit %u\n", dd->unit);
+
+	/* no interrupts till re-initted */
+	qib_7220_set_intr_state(dd, 0);
+
+	dd->pport->cpspec->ibdeltainprog = 0;
+	dd->pport->cpspec->ibsymdelta = 0;
+	dd->pport->cpspec->iblnkerrdelta = 0;
+
+	/*
+	 * Keep chip from being accessed until we are ready.  Use
+	 * writeq() directly, to allow the write even though QIB_PRESENT
+	 * isnt' set.
+	 */
+	dd->flags &= ~(QIB_INITTED | QIB_PRESENT);
+	dd->int_counter = 0; /* so we check interrupts work again */
+	val = dd->control | QLOGIC_IB_C_RESET;
+	writeq(val, &dd->kregbase[kr_control]);
+	mb(); /* prevent compiler reordering around actual reset */
+
+	for (i = 1; i <= 5; i++) {
+		/*
+		 * Allow MBIST, etc. to complete; longer on each retry.
+		 * We sometimes get machine checks from bus timeout if no
+		 * response, so for now, make it *really* long.
+		 */
+		msleep(1000 + (1 + i) * 2000);
+
+		qib_pcie_reenable(dd, cmdval, int_line, clinesz);
+
+		/*
+		 * Use readq directly, so we don't need to mark it as PRESENT
+		 * until we get a successful indication that all is well.
+		 */
+		val = readq(&dd->kregbase[kr_revision]);
+		if (val == dd->revision) {
+			dd->flags |= QIB_PRESENT; /* it's back */
+			ret = qib_reinit_intr(dd);
+			goto bail;
+		}
+	}
+	ret = 0; /* failed */
+
+bail:
+	if (ret) {
+		if (qib_pcie_params(dd, dd->lbus_width, NULL, NULL))
+			qib_dev_err(dd, "Reset failed to setup PCIe or "
+				    "interrupts; continuing anyway\n");
+
+		/* hold IBC in reset, no sends, etc till later */
+		qib_write_kreg(dd, kr_control, 0ULL);
+
+		/* clear the reset error, init error/hwerror mask */
+		qib_7220_init_hwerrors(dd);
+
+		/* do setup similar to speed or link-width changes */
+		if (dd->pport->cpspec->ibcddrctrl & IBA7220_IBC_IBTA_1_2_MASK)
+			dd->cspec->presets_needed = 1;
+		spin_lock_irqsave(&dd->pport->lflags_lock, flags);
+		dd->pport->lflags |= QIBL_IB_FORCE_NOTIFY;
+		dd->pport->lflags &= ~QIBL_IB_AUTONEG_FAILED;
+		spin_unlock_irqrestore(&dd->pport->lflags_lock, flags);
+	}
+
+	return ret;
+}
+
+/**
+ * qib_7220_put_tid - write a TID to the chip
+ * @dd: the qlogic_ib device
+ * @tidptr: pointer to the expected TID (in chip) to update
+ * @tidtype: 0 for eager, 1 for expected
+ * @pa: physical address of in memory buffer; tidinvalid if freeing
+ */
+static void qib_7220_put_tid(struct qib_devdata *dd, u64 __iomem *tidptr,
+			     u32 type, unsigned long pa)
+{
+	if (pa != dd->tidinvalid) {
+		u64 chippa = pa >> IBA7220_TID_PA_SHIFT;
+
+		/* paranoia checks */
+		if (pa != (chippa << IBA7220_TID_PA_SHIFT)) {
+			qib_dev_err(dd, "Physaddr %lx not 2KB aligned!\n",
+				    pa);
+			return;
+		}
+		if (chippa >= (1UL << IBA7220_TID_SZ_SHIFT)) {
+			qib_dev_err(dd, "Physical page address 0x%lx "
+				"larger than supported\n", pa);
+			return;
+		}
+
+		if (type == RCVHQ_RCV_TYPE_EAGER)
+			chippa |= dd->tidtemplate;
+		else /* for now, always full 4KB page */
+			chippa |= IBA7220_TID_SZ_4K;
+		pa = chippa;
+	}
+	writeq(pa, tidptr);
+	mmiowb();
+}
+
+/**
+ * qib_7220_clear_tids - clear all TID entries for a ctxt, expected and eager
+ * @dd: the qlogic_ib device
+ * @ctxt: the ctxt
+ *
+ * clear all TID entries for a ctxt, expected and eager.
+ * Used from qib_close().  On this chip, TIDs are only 32 bits,
+ * not 64, but they are still on 64 bit boundaries, so tidbase
+ * is declared as u64 * for the pointer math, even though we write 32 bits
+ */
+static void qib_7220_clear_tids(struct qib_devdata *dd,
+				struct qib_ctxtdata *rcd)
+{
+	u64 __iomem *tidbase;
+	unsigned long tidinv;
+	u32 ctxt;
+	int i;
+
+	if (!dd->kregbase || !rcd)
+		return;
+
+	ctxt = rcd->ctxt;
+
+	tidinv = dd->tidinvalid;
+	tidbase = (u64 __iomem *)
+		((char __iomem *)(dd->kregbase) +
+		 dd->rcvtidbase +
+		 ctxt * dd->rcvtidcnt * sizeof(*tidbase));
+
+	for (i = 0; i < dd->rcvtidcnt; i++)
+		qib_7220_put_tid(dd, &tidbase[i], RCVHQ_RCV_TYPE_EXPECTED,
+				 tidinv);
+
+	tidbase = (u64 __iomem *)
+		((char __iomem *)(dd->kregbase) +
+		 dd->rcvegrbase +
+		 rcd->rcvegr_tid_base * sizeof(*tidbase));
+
+	for (i = 0; i < rcd->rcvegrcnt; i++)
+		qib_7220_put_tid(dd, &tidbase[i], RCVHQ_RCV_TYPE_EAGER,
+				 tidinv);
+}
+
+/**
+ * qib_7220_tidtemplate - setup constants for TID updates
+ * @dd: the qlogic_ib device
+ *
+ * We setup stuff that we use a lot, to avoid calculating each time
+ */
+static void qib_7220_tidtemplate(struct qib_devdata *dd)
+{
+	if (dd->rcvegrbufsize == 2048)
+		dd->tidtemplate = IBA7220_TID_SZ_2K;
+	else if (dd->rcvegrbufsize == 4096)
+		dd->tidtemplate = IBA7220_TID_SZ_4K;
+	dd->tidinvalid = 0;
+}
+
+/**
+ * qib_init_7220_get_base_info - set chip-specific flags for user code
+ * @rcd: the qlogic_ib ctxt
+ * @kbase: qib_base_info pointer
+ *
+ * We set the PCIE flag because the lower bandwidth on PCIe vs
+ * HyperTransport can affect some user packet algorithims.
+ */
+static int qib_7220_get_base_info(struct qib_ctxtdata *rcd,
+				  struct qib_base_info *kinfo)
+{
+	kinfo->spi_runtime_flags |= QIB_RUNTIME_PCIE |
+		QIB_RUNTIME_NODMA_RTAIL | QIB_RUNTIME_SDMA;
+
+	if (rcd->dd->flags & QIB_USE_SPCL_TRIG)
+		kinfo->spi_runtime_flags |= QIB_RUNTIME_SPECIAL_TRIGGER;
+
+	return 0;
+}
+
+static struct qib_message_header *
+qib_7220_get_msgheader(struct qib_devdata *dd, __le32 *rhf_addr)
+{
+	u32 offset = qib_hdrget_offset(rhf_addr);
+
+	return (struct qib_message_header *)
+		(rhf_addr - dd->rhf_offset + offset);
+}
+
+static void qib_7220_config_ctxts(struct qib_devdata *dd)
+{
+	unsigned long flags;
+	u32 nchipctxts;
+
+	nchipctxts = qib_read_kreg32(dd, kr_portcnt);
+	dd->cspec->numctxts = nchipctxts;
+	if (qib_n_krcv_queues > 1) {
+		dd->qpn_mask = 0x3f;
+		dd->first_user_ctxt = qib_n_krcv_queues * dd->num_pports;
+		if (dd->first_user_ctxt > nchipctxts)
+			dd->first_user_ctxt = nchipctxts;
+	} else
+		dd->first_user_ctxt = dd->num_pports;
+	dd->n_krcv_queues = dd->first_user_ctxt;
+
+	if (!qib_cfgctxts) {
+		int nctxts = dd->first_user_ctxt + num_online_cpus();
+
+		if (nctxts <= 5)
+			dd->ctxtcnt = 5;
+		else if (nctxts <= 9)
+			dd->ctxtcnt = 9;
+		else if (nctxts <= nchipctxts)
+			dd->ctxtcnt = nchipctxts;
+	} else if (qib_cfgctxts <= nchipctxts)
+		dd->ctxtcnt = qib_cfgctxts;
+	if (!dd->ctxtcnt) /* none of the above, set to max */
+		dd->ctxtcnt = nchipctxts;
+
+	/*
+	 * Chip can be configured for 5, 9, or 17 ctxts, and choice
+	 * affects number of eager TIDs per ctxt (1K, 2K, 4K).
+	 * Lock to be paranoid about later motion, etc.
+	 */
+	spin_lock_irqsave(&dd->cspec->rcvmod_lock, flags);
+	if (dd->ctxtcnt > 9)
+		dd->rcvctrl |= 2ULL << IBA7220_R_CTXTCFG_SHIFT;
+	else if (dd->ctxtcnt > 5)
+		dd->rcvctrl |= 1ULL << IBA7220_R_CTXTCFG_SHIFT;
+	/* else configure for default 5 receive ctxts */
+	if (dd->qpn_mask)
+		dd->rcvctrl |= 1ULL << QIB_7220_RcvCtrl_RcvQPMapEnable_LSB;
+	qib_write_kreg(dd, kr_rcvctrl, dd->rcvctrl);
+	spin_unlock_irqrestore(&dd->cspec->rcvmod_lock, flags);
+
+	/* kr_rcvegrcnt changes based on the number of contexts enabled */
+	dd->cspec->rcvegrcnt = qib_read_kreg32(dd, kr_rcvegrcnt);
+	dd->rcvhdrcnt = max(dd->cspec->rcvegrcnt, IBA7220_KRCVEGRCNT);
+}
+
+static int qib_7220_get_ib_cfg(struct qib_pportdata *ppd, int which)
+{
+	int lsb, ret = 0;
+	u64 maskr; /* right-justified mask */
+
+	switch (which) {
+	case QIB_IB_CFG_LWID_ENB: /* Get allowed Link-width */
+		ret = ppd->link_width_enabled;
+		goto done;
+
+	case QIB_IB_CFG_LWID: /* Get currently active Link-width */
+		ret = ppd->link_width_active;
+		goto done;
+
+	case QIB_IB_CFG_SPD_ENB: /* Get allowed Link speeds */
+		ret = ppd->link_speed_enabled;
+		goto done;
+
+	case QIB_IB_CFG_SPD: /* Get current Link spd */
+		ret = ppd->link_speed_active;
+		goto done;
+
+	case QIB_IB_CFG_RXPOL_ENB: /* Get Auto-RX-polarity enable */
+		lsb = IBA7220_IBC_RXPOL_SHIFT;
+		maskr = IBA7220_IBC_RXPOL_MASK;
+		break;
+
+	case QIB_IB_CFG_LREV_ENB: /* Get Auto-Lane-reversal enable */
+		lsb = IBA7220_IBC_LREV_SHIFT;
+		maskr = IBA7220_IBC_LREV_MASK;
+		break;
+
+	case QIB_IB_CFG_LINKLATENCY:
+		ret = qib_read_kreg64(ppd->dd, kr_ibcddrstatus)
+			& IBA7220_DDRSTAT_LINKLAT_MASK;
+		goto done;
+
+	case QIB_IB_CFG_OP_VLS:
+		ret = ppd->vls_operational;
+		goto done;
+
+	case QIB_IB_CFG_VL_HIGH_CAP:
+		ret = 0;
+		goto done;
+
+	case QIB_IB_CFG_VL_LOW_CAP:
+		ret = 0;
+		goto done;
+
+	case QIB_IB_CFG_OVERRUN_THRESH: /* IB overrun threshold */
+		ret = SYM_FIELD(ppd->cpspec->ibcctrl, IBCCtrl,
+				OverrunThreshold);
+		goto done;
+
+	case QIB_IB_CFG_PHYERR_THRESH: /* IB PHY error threshold */
+		ret = SYM_FIELD(ppd->cpspec->ibcctrl, IBCCtrl,
+				PhyerrThreshold);
+		goto done;
+
+	case QIB_IB_CFG_LINKDEFAULT: /* IB link default (sleep/poll) */
+		/* will only take effect when the link state changes */
+		ret = (ppd->cpspec->ibcctrl &
+		       SYM_MASK(IBCCtrl, LinkDownDefaultState)) ?
+			IB_LINKINITCMD_SLEEP : IB_LINKINITCMD_POLL;
+		goto done;
+
+	case QIB_IB_CFG_HRTBT: /* Get Heartbeat off/enable/auto */
+		lsb = IBA7220_IBC_HRTBT_SHIFT;
+		maskr = IBA7220_IBC_HRTBT_MASK;
+		break;
+
+	case QIB_IB_CFG_PMA_TICKS:
+		/*
+		 * 0x00 = 10x link transfer rate or 4 nsec. for 2.5Gbs
+		 * Since the clock is always 250MHz, the value is 1 or 0.
+		 */
+		ret = (ppd->link_speed_active == QIB_IB_DDR);
+		goto done;
+
+	default:
+		ret = -EINVAL;
+		goto done;
+	}
+	ret = (int)((ppd->cpspec->ibcddrctrl >> lsb) & maskr);
+done:
+	return ret;
+}
+
+static int qib_7220_set_ib_cfg(struct qib_pportdata *ppd, int which, u32 val)
+{
+	struct qib_devdata *dd = ppd->dd;
+	u64 maskr; /* right-justified mask */
+	int lsb, ret = 0, setforce = 0;
+	u16 lcmd, licmd;
+	unsigned long flags;
+
+	switch (which) {
+	case QIB_IB_CFG_LIDLMC:
+		/*
+		 * Set LID and LMC. Combined to avoid possible hazard
+		 * caller puts LMC in 16MSbits, DLID in 16LSbits of val
+		 */
+		lsb = IBA7220_IBC_DLIDLMC_SHIFT;
+		maskr = IBA7220_IBC_DLIDLMC_MASK;
+		break;
+
+	case QIB_IB_CFG_LWID_ENB: /* set allowed Link-width */
+		/*
+		 * As with speed, only write the actual register if
+		 * the link is currently down, otherwise takes effect
+		 * on next link change.
+		 */
+		ppd->link_width_enabled = val;
+		if (!(ppd->lflags & QIBL_LINKDOWN))
+			goto bail;
+		/*
+		 * We set the QIBL_IB_FORCE_NOTIFY bit so updown
+		 * will get called because we want update
+		 * link_width_active, and the change may not take
+		 * effect for some time (if we are in POLL), so this
+		 * flag will force the updown routine to be called
+		 * on the next ibstatuschange down interrupt, even
+		 * if it's not an down->up transition.
+		 */
+		val--; /* convert from IB to chip */
+		maskr = IBA7220_IBC_WIDTH_MASK;
+		lsb = IBA7220_IBC_WIDTH_SHIFT;
+		setforce = 1;
+		spin_lock_irqsave(&ppd->lflags_lock, flags);
+		ppd->lflags |= QIBL_IB_FORCE_NOTIFY;
+		spin_unlock_irqrestore(&ppd->lflags_lock, flags);
+		break;
+
+	case QIB_IB_CFG_SPD_ENB: /* set allowed Link speeds */
+		/*
+		 * If we turn off IB1.2, need to preset SerDes defaults,
+		 * but not right now. Set a flag for the next time
+		 * we command the link down.  As with width, only write the
+		 * actual register if the link is currently down, otherwise
+		 * takes effect on next link change.  Since setting is being
+		 * explictly requested (via MAD or sysfs), clear autoneg
+		 * failure status if speed autoneg is enabled.
+		 */
+		ppd->link_speed_enabled = val;
+		if ((ppd->cpspec->ibcddrctrl & IBA7220_IBC_IBTA_1_2_MASK) &&
+		    !(val & (val - 1)))
+			dd->cspec->presets_needed = 1;
+		if (!(ppd->lflags & QIBL_LINKDOWN))
+			goto bail;
+		/*
+		 * We set the QIBL_IB_FORCE_NOTIFY bit so updown
+		 * will get called because we want update
+		 * link_speed_active, and the change may not take
+		 * effect for some time (if we are in POLL), so this
+		 * flag will force the updown routine to be called
+		 * on the next ibstatuschange down interrupt, even
+		 * if it's not an down->up transition.
+		 */
+		if (val == (QIB_IB_SDR | QIB_IB_DDR)) {
+			val = IBA7220_IBC_SPEED_AUTONEG_MASK |
+				IBA7220_IBC_IBTA_1_2_MASK;
+			spin_lock_irqsave(&ppd->lflags_lock, flags);
+			ppd->lflags &= ~QIBL_IB_AUTONEG_FAILED;
+			spin_unlock_irqrestore(&ppd->lflags_lock, flags);
+		} else
+			val = val == QIB_IB_DDR ?
+				IBA7220_IBC_SPEED_DDR : IBA7220_IBC_SPEED_SDR;
+		maskr = IBA7220_IBC_SPEED_AUTONEG_MASK |
+			IBA7220_IBC_IBTA_1_2_MASK;
+		/* IBTA 1.2 mode + speed bits are contiguous */
+		lsb = SYM_LSB(IBCDDRCtrl, IB_ENHANCED_MODE);
+		setforce = 1;
+		break;
+
+	case QIB_IB_CFG_RXPOL_ENB: /* set Auto-RX-polarity enable */
+		lsb = IBA7220_IBC_RXPOL_SHIFT;
+		maskr = IBA7220_IBC_RXPOL_MASK;
+		break;
+
+	case QIB_IB_CFG_LREV_ENB: /* set Auto-Lane-reversal enable */
+		lsb = IBA7220_IBC_LREV_SHIFT;
+		maskr = IBA7220_IBC_LREV_MASK;
+		break;
+
+	case QIB_IB_CFG_OVERRUN_THRESH: /* IB overrun threshold */
+		maskr = SYM_FIELD(ppd->cpspec->ibcctrl, IBCCtrl,
+				  OverrunThreshold);
+		if (maskr != val) {
+			ppd->cpspec->ibcctrl &=
+				~SYM_MASK(IBCCtrl, OverrunThreshold);
+			ppd->cpspec->ibcctrl |= (u64) val <<
+				SYM_LSB(IBCCtrl, OverrunThreshold);
+			qib_write_kreg(dd, kr_ibcctrl, ppd->cpspec->ibcctrl);
+			qib_write_kreg(dd, kr_scratch, 0);
+		}
+		goto bail;
+
+	case QIB_IB_CFG_PHYERR_THRESH: /* IB PHY error threshold */
+		maskr = SYM_FIELD(ppd->cpspec->ibcctrl, IBCCtrl,
+				  PhyerrThreshold);
+		if (maskr != val) {
+			ppd->cpspec->ibcctrl &=
+				~SYM_MASK(IBCCtrl, PhyerrThreshold);
+			ppd->cpspec->ibcctrl |= (u64) val <<
+				SYM_LSB(IBCCtrl, PhyerrThreshold);
+			qib_write_kreg(dd, kr_ibcctrl, ppd->cpspec->ibcctrl);
+			qib_write_kreg(dd, kr_scratch, 0);
+		}
+		goto bail;
+
+	case QIB_IB_CFG_PKEYS: /* update pkeys */
+		maskr = (u64) ppd->pkeys[0] | ((u64) ppd->pkeys[1] << 16) |
+			((u64) ppd->pkeys[2] << 32) |
+			((u64) ppd->pkeys[3] << 48);
+		qib_write_kreg(dd, kr_partitionkey, maskr);
+		goto bail;
+
+	case QIB_IB_CFG_LINKDEFAULT: /* IB link default (sleep/poll) */
+		/* will only take effect when the link state changes */
+		if (val == IB_LINKINITCMD_POLL)
+			ppd->cpspec->ibcctrl &=
+				~SYM_MASK(IBCCtrl, LinkDownDefaultState);
+		else /* SLEEP */
+			ppd->cpspec->ibcctrl |=
+				SYM_MASK(IBCCtrl, LinkDownDefaultState);
+		qib_write_kreg(dd, kr_ibcctrl, ppd->cpspec->ibcctrl);
+		qib_write_kreg(dd, kr_scratch, 0);
+		goto bail;
+
+	case QIB_IB_CFG_MTU: /* update the MTU in IBC */
+		/*
+		 * Update our housekeeping variables, and set IBC max
+		 * size, same as init code; max IBC is max we allow in
+		 * buffer, less the qword pbc, plus 1 for ICRC, in dwords
+		 * Set even if it's unchanged, print debug message only
+		 * on changes.
+		 */
+		val = (ppd->ibmaxlen >> 2) + 1;
+		ppd->cpspec->ibcctrl &= ~SYM_MASK(IBCCtrl, MaxPktLen);
+		ppd->cpspec->ibcctrl |= (u64)val << SYM_LSB(IBCCtrl, MaxPktLen);
+		qib_write_kreg(dd, kr_ibcctrl, ppd->cpspec->ibcctrl);
+		qib_write_kreg(dd, kr_scratch, 0);
+		goto bail;
+
+	case QIB_IB_CFG_LSTATE: /* set the IB link state */
+		switch (val & 0xffff0000) {
+		case IB_LINKCMD_DOWN:
+			lcmd = QLOGIC_IB_IBCC_LINKCMD_DOWN;
+			if (!ppd->cpspec->ibdeltainprog &&
+			    qib_compat_ddr_negotiate) {
+				ppd->cpspec->ibdeltainprog = 1;
+				ppd->cpspec->ibsymsnap =
+					read_7220_creg32(dd, cr_ibsymbolerr);
+				ppd->cpspec->iblnkerrsnap =
+					read_7220_creg32(dd, cr_iblinkerrrecov);
+			}
+			break;
+
+		case IB_LINKCMD_ARMED:
+			lcmd = QLOGIC_IB_IBCC_LINKCMD_ARMED;
+			break;
+
+		case IB_LINKCMD_ACTIVE:
+			lcmd = QLOGIC_IB_IBCC_LINKCMD_ACTIVE;
+			break;
+
+		default:
+			ret = -EINVAL;
+			qib_dev_err(dd, "bad linkcmd req 0x%x\n", val >> 16);
+			goto bail;
+		}
+		switch (val & 0xffff) {
+		case IB_LINKINITCMD_NOP:
+			licmd = 0;
+			break;
+
+		case IB_LINKINITCMD_POLL:
+			licmd = QLOGIC_IB_IBCC_LINKINITCMD_POLL;
+			break;
+
+		case IB_LINKINITCMD_SLEEP:
+			licmd = QLOGIC_IB_IBCC_LINKINITCMD_SLEEP;
+			break;
+
+		case IB_LINKINITCMD_DISABLE:
+			licmd = QLOGIC_IB_IBCC_LINKINITCMD_DISABLE;
+			ppd->cpspec->chase_end = 0;
+			/*
+			 * stop state chase counter and timer, if running.
+			 * wait forpending timer, but don't clear .data (ppd)!
+			 */
+			if (ppd->cpspec->chase_timer.expires) {
+				del_timer_sync(&ppd->cpspec->chase_timer);
+				ppd->cpspec->chase_timer.expires = 0;
+			}
+			break;
+
+		default:
+			ret = -EINVAL;
+			qib_dev_err(dd, "bad linkinitcmd req 0x%x\n",
+				    val & 0xffff);
+			goto bail;
+		}
+		qib_set_ib_7220_lstate(ppd, lcmd, licmd);
+		goto bail;
+
+	case QIB_IB_CFG_HRTBT: /* set Heartbeat off/enable/auto */
+		if (val > IBA7220_IBC_HRTBT_MASK) {
+			ret = -EINVAL;
+			goto bail;
+		}
+		lsb = IBA7220_IBC_HRTBT_SHIFT;
+		maskr = IBA7220_IBC_HRTBT_MASK;
+		break;
+
+	default:
+		ret = -EINVAL;
+		goto bail;
+	}
+	ppd->cpspec->ibcddrctrl &= ~(maskr << lsb);
+	ppd->cpspec->ibcddrctrl |= (((u64) val & maskr) << lsb);
+	qib_write_kreg(dd, kr_ibcddrctrl, ppd->cpspec->ibcddrctrl);
+	qib_write_kreg(dd, kr_scratch, 0);
+	if (setforce) {
+		spin_lock_irqsave(&ppd->lflags_lock, flags);
+		ppd->lflags |= QIBL_IB_FORCE_NOTIFY;
+		spin_unlock_irqrestore(&ppd->lflags_lock, flags);
+	}
+bail:
+	return ret;
+}
+
+static int qib_7220_set_loopback(struct qib_pportdata *ppd, const char *what)
+{
+	int ret = 0;
+	u64 val, ddr;
+
+	if (!strncmp(what, "ibc", 3)) {
+		ppd->cpspec->ibcctrl |= SYM_MASK(IBCCtrl, Loopback);
+		val = 0; /* disable heart beat, so link will come up */
+		qib_devinfo(ppd->dd->pcidev, "Enabling IB%u:%u IBC loopback\n",
+			 ppd->dd->unit, ppd->port);
+	} else if (!strncmp(what, "off", 3)) {
+		ppd->cpspec->ibcctrl &= ~SYM_MASK(IBCCtrl, Loopback);
+		/* enable heart beat again */
+		val = IBA7220_IBC_HRTBT_MASK << IBA7220_IBC_HRTBT_SHIFT;
+		qib_devinfo(ppd->dd->pcidev, "Disabling IB%u:%u IBC loopback "
+			    "(normal)\n", ppd->dd->unit, ppd->port);
+	} else
+		ret = -EINVAL;
+	if (!ret) {
+		qib_write_kreg(ppd->dd, kr_ibcctrl, ppd->cpspec->ibcctrl);
+		ddr = ppd->cpspec->ibcddrctrl & ~(IBA7220_IBC_HRTBT_MASK
+					     << IBA7220_IBC_HRTBT_SHIFT);
+		ppd->cpspec->ibcddrctrl = ddr | val;
+		qib_write_kreg(ppd->dd, kr_ibcddrctrl,
+			       ppd->cpspec->ibcddrctrl);
+		qib_write_kreg(ppd->dd, kr_scratch, 0);
+	}
+	return ret;
+}
+
+static void qib_update_7220_usrhead(struct qib_ctxtdata *rcd, u64 hd,
+				    u32 updegr, u32 egrhd)
+{
+	qib_write_ureg(rcd->dd, ur_rcvhdrhead, hd, rcd->ctxt);
+	if (updegr)
+		qib_write_ureg(rcd->dd, ur_rcvegrindexhead, egrhd, rcd->ctxt);
+}
+
+static u32 qib_7220_hdrqempty(struct qib_ctxtdata *rcd)
+{
+	u32 head, tail;
+
+	head = qib_read_ureg32(rcd->dd, ur_rcvhdrhead, rcd->ctxt);
+	if (rcd->rcvhdrtail_kvaddr)
+		tail = qib_get_rcvhdrtail(rcd);
+	else
+		tail = qib_read_ureg32(rcd->dd, ur_rcvhdrtail, rcd->ctxt);
+	return head == tail;
+}
+
+/*
+ * Modify the RCVCTRL register in chip-specific way. This
+ * is a function because bit positions and (future) register
+ * location is chip-specifc, but the needed operations are
+ * generic. <op> is a bit-mask because we often want to
+ * do multiple modifications.
+ */
+static void rcvctrl_7220_mod(struct qib_pportdata *ppd, unsigned int op,
+			     int ctxt)
+{
+	struct qib_devdata *dd = ppd->dd;
+	u64 mask, val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&dd->cspec->rcvmod_lock, flags);
+	if (op & QIB_RCVCTRL_TAILUPD_ENB)
+		dd->rcvctrl |= (1ULL << IBA7220_R_TAILUPD_SHIFT);
+	if (op & QIB_RCVCTRL_TAILUPD_DIS)
+		dd->rcvctrl &= ~(1ULL << IBA7220_R_TAILUPD_SHIFT);
+	if (op & QIB_RCVCTRL_PKEY_ENB)
+		dd->rcvctrl &= ~(1ULL << IBA7220_R_PKEY_DIS_SHIFT);
+	if (op & QIB_RCVCTRL_PKEY_DIS)
+		dd->rcvctrl |= (1ULL << IBA7220_R_PKEY_DIS_SHIFT);
+	if (ctxt < 0)
+		mask = (1ULL << dd->ctxtcnt) - 1;
+	else
+		mask = (1ULL << ctxt);
+	if (op & QIB_RCVCTRL_CTXT_ENB) {
+		/* always done for specific ctxt */
+		dd->rcvctrl |= (mask << SYM_LSB(RcvCtrl, PortEnable));
+		if (!(dd->flags & QIB_NODMA_RTAIL))
+			dd->rcvctrl |= 1ULL << IBA7220_R_TAILUPD_SHIFT;
+		/* Write these registers before the context is enabled. */
+		qib_write_kreg_ctxt(dd, kr_rcvhdrtailaddr, ctxt,
+			dd->rcd[ctxt]->rcvhdrqtailaddr_phys);
+		qib_write_kreg_ctxt(dd, kr_rcvhdraddr, ctxt,
+			dd->rcd[ctxt]->rcvhdrq_phys);
+		dd->rcd[ctxt]->seq_cnt = 1;
+	}
+	if (op & QIB_RCVCTRL_CTXT_DIS)
+		dd->rcvctrl &= ~(mask << SYM_LSB(RcvCtrl, PortEnable));
+	if (op & QIB_RCVCTRL_INTRAVAIL_ENB)
+		dd->rcvctrl |= (mask << IBA7220_R_INTRAVAIL_SHIFT);
+	if (op & QIB_RCVCTRL_INTRAVAIL_DIS)
+		dd->rcvctrl &= ~(mask << IBA7220_R_INTRAVAIL_SHIFT);
+	qib_write_kreg(dd, kr_rcvctrl, dd->rcvctrl);
+	if ((op & QIB_RCVCTRL_INTRAVAIL_ENB) && dd->rhdrhead_intr_off) {
+		/* arm rcv interrupt */
+		val = qib_read_ureg32(dd, ur_rcvhdrhead, ctxt) |
+			dd->rhdrhead_intr_off;
+		qib_write_ureg(dd, ur_rcvhdrhead, val, ctxt);
+	}
+	if (op & QIB_RCVCTRL_CTXT_ENB) {
+		/*
+		 * Init the context registers also; if we were
+		 * disabled, tail and head should both be zero
+		 * already from the enable, but since we don't
+		 * know, we have to do it explictly.
+		 */
+		val = qib_read_ureg32(dd, ur_rcvegrindextail, ctxt);
+		qib_write_ureg(dd, ur_rcvegrindexhead, val, ctxt);
+
+		val = qib_read_ureg32(dd, ur_rcvhdrtail, ctxt);
+		dd->rcd[ctxt]->head = val;
+		/* If kctxt, interrupt on next receive. */
+		if (ctxt < dd->first_user_ctxt)
+			val |= dd->rhdrhead_intr_off;
+		qib_write_ureg(dd, ur_rcvhdrhead, val, ctxt);
+	}
+	if (op & QIB_RCVCTRL_CTXT_DIS) {
+		if (ctxt >= 0) {
+			qib_write_kreg_ctxt(dd, kr_rcvhdrtailaddr, ctxt, 0);
+			qib_write_kreg_ctxt(dd, kr_rcvhdraddr, ctxt, 0);
+		} else {
+			unsigned i;
+
+			for (i = 0; i < dd->cfgctxts; i++) {
+				qib_write_kreg_ctxt(dd, kr_rcvhdrtailaddr,
+						    i, 0);
+				qib_write_kreg_ctxt(dd, kr_rcvhdraddr, i, 0);
+			}
+		}
+	}
+	spin_unlock_irqrestore(&dd->cspec->rcvmod_lock, flags);
+}
+
+/*
+ * Modify the SENDCTRL register in chip-specific way. This
+ * is a function there may be multiple such registers with
+ * slightly different layouts. To start, we assume the
+ * "canonical" register layout of the first chips.
+ * Chip requires no back-back sendctrl writes, so write
+ * scratch register after writing sendctrl
+ */
+static void sendctrl_7220_mod(struct qib_pportdata *ppd, u32 op)
+{
+	struct qib_devdata *dd = ppd->dd;
+	u64 tmp_dd_sendctrl;
+	unsigned long flags;
+
+	spin_lock_irqsave(&dd->sendctrl_lock, flags);
+
+	/* First the ones that are "sticky", saved in shadow */
+	if (op & QIB_SENDCTRL_CLEAR)
+		dd->sendctrl = 0;
+	if (op & QIB_SENDCTRL_SEND_DIS)
+		dd->sendctrl &= ~SYM_MASK(SendCtrl, SPioEnable);
+	else if (op & QIB_SENDCTRL_SEND_ENB) {
+		dd->sendctrl |= SYM_MASK(SendCtrl, SPioEnable);
+		if (dd->flags & QIB_USE_SPCL_TRIG)
+			dd->sendctrl |= SYM_MASK(SendCtrl,
+						 SSpecialTriggerEn);
+	}
+	if (op & QIB_SENDCTRL_AVAIL_DIS)
+		dd->sendctrl &= ~SYM_MASK(SendCtrl, SendBufAvailUpd);
+	else if (op & QIB_SENDCTRL_AVAIL_ENB)
+		dd->sendctrl |= SYM_MASK(SendCtrl, SendBufAvailUpd);
+
+	if (op & QIB_SENDCTRL_DISARM_ALL) {
+		u32 i, last;
+
+		tmp_dd_sendctrl = dd->sendctrl;
+		/*
+		 * disarm any that are not yet launched, disabling sends
+		 * and updates until done.
+		 */
+		last = dd->piobcnt2k + dd->piobcnt4k;
+		tmp_dd_sendctrl &=
+			~(SYM_MASK(SendCtrl, SPioEnable) |
+			  SYM_MASK(SendCtrl, SendBufAvailUpd));
+		for (i = 0; i < last; i++) {
+			qib_write_kreg(dd, kr_sendctrl,
+				       tmp_dd_sendctrl |
+				       SYM_MASK(SendCtrl, Disarm) | i);
+			qib_write_kreg(dd, kr_scratch, 0);
+		}
+	}
+
+	tmp_dd_sendctrl = dd->sendctrl;
+
+	if (op & QIB_SENDCTRL_FLUSH)
+		tmp_dd_sendctrl |= SYM_MASK(SendCtrl, Abort);
+	if (op & QIB_SENDCTRL_DISARM)
+		tmp_dd_sendctrl |= SYM_MASK(SendCtrl, Disarm) |
+			((op & QIB_7220_SendCtrl_DisarmPIOBuf_RMASK) <<
+			 SYM_LSB(SendCtrl, DisarmPIOBuf));
+	if ((op & QIB_SENDCTRL_AVAIL_BLIP) &&
+	    (dd->sendctrl & SYM_MASK(SendCtrl, SendBufAvailUpd)))
+		tmp_dd_sendctrl &= ~SYM_MASK(SendCtrl, SendBufAvailUpd);
+
+	qib_write_kreg(dd, kr_sendctrl, tmp_dd_sendctrl);
+	qib_write_kreg(dd, kr_scratch, 0);
+
+	if (op & QIB_SENDCTRL_AVAIL_BLIP) {
+		qib_write_kreg(dd, kr_sendctrl, dd->sendctrl);
+		qib_write_kreg(dd, kr_scratch, 0);
+	}
+
+	spin_unlock_irqrestore(&dd->sendctrl_lock, flags);
+
+	if (op & QIB_SENDCTRL_FLUSH) {
+		u32 v;
+		/*
+		 * ensure writes have hit chip, then do a few
+		 * more reads, to allow DMA of pioavail registers
+		 * to occur, so in-memory copy is in sync with
+		 * the chip.  Not always safe to sleep.
+		 */
+		v = qib_read_kreg32(dd, kr_scratch);
+		qib_write_kreg(dd, kr_scratch, v);
+		v = qib_read_kreg32(dd, kr_scratch);
+		qib_write_kreg(dd, kr_scratch, v);
+		qib_read_kreg32(dd, kr_scratch);
+	}
+}
+
+/**
+ * qib_portcntr_7220 - read a per-port counter
+ * @dd: the qlogic_ib device
+ * @creg: the counter to snapshot
+ */
+static u64 qib_portcntr_7220(struct qib_pportdata *ppd, u32 reg)
+{
+	u64 ret = 0ULL;
+	struct qib_devdata *dd = ppd->dd;
+	u16 creg;
+	/* 0xffff for unimplemented or synthesized counters */
+	static const u16 xlator[] = {
+		[QIBPORTCNTR_PKTSEND] = cr_pktsend,
+		[QIBPORTCNTR_WORDSEND] = cr_wordsend,
+		[QIBPORTCNTR_PSXMITDATA] = cr_psxmitdatacount,
+		[QIBPORTCNTR_PSXMITPKTS] = cr_psxmitpktscount,
+		[QIBPORTCNTR_PSXMITWAIT] = cr_psxmitwaitcount,
+		[QIBPORTCNTR_SENDSTALL] = cr_sendstall,
+		[QIBPORTCNTR_PKTRCV] = cr_pktrcv,
+		[QIBPORTCNTR_PSRCVDATA] = cr_psrcvdatacount,
+		[QIBPORTCNTR_PSRCVPKTS] = cr_psrcvpktscount,
+		[QIBPORTCNTR_RCVEBP] = cr_rcvebp,
+		[QIBPORTCNTR_RCVOVFL] = cr_rcvovfl,
+		[QIBPORTCNTR_WORDRCV] = cr_wordrcv,
+		[QIBPORTCNTR_RXDROPPKT] = cr_rxdroppkt,
+		[QIBPORTCNTR_RXLOCALPHYERR] = cr_rxotherlocalphyerr,
+		[QIBPORTCNTR_RXVLERR] = cr_rxvlerr,
+		[QIBPORTCNTR_ERRICRC] = cr_erricrc,
+		[QIBPORTCNTR_ERRVCRC] = cr_errvcrc,
+		[QIBPORTCNTR_ERRLPCRC] = cr_errlpcrc,
+		[QIBPORTCNTR_BADFORMAT] = cr_badformat,
+		[QIBPORTCNTR_ERR_RLEN] = cr_err_rlen,
+		[QIBPORTCNTR_IBSYMBOLERR] = cr_ibsymbolerr,
+		[QIBPORTCNTR_INVALIDRLEN] = cr_invalidrlen,
+		[QIBPORTCNTR_UNSUPVL] = cr_txunsupvl,
+		[QIBPORTCNTR_EXCESSBUFOVFL] = cr_excessbufferovfl,
+		[QIBPORTCNTR_ERRLINK] = cr_errlink,
+		[QIBPORTCNTR_IBLINKDOWN] = cr_iblinkdown,
+		[QIBPORTCNTR_IBLINKERRRECOV] = cr_iblinkerrrecov,
+		[QIBPORTCNTR_LLI] = cr_locallinkintegrityerr,
+		[QIBPORTCNTR_PSINTERVAL] = cr_psinterval,
+		[QIBPORTCNTR_PSSTART] = cr_psstart,
+		[QIBPORTCNTR_PSSTAT] = cr_psstat,
+		[QIBPORTCNTR_VL15PKTDROP] = cr_vl15droppedpkt,
+		[QIBPORTCNTR_ERRPKEY] = cr_errpkey,
+		[QIBPORTCNTR_KHDROVFL] = 0xffff,
+	};
+
+	if (reg >= ARRAY_SIZE(xlator)) {
+		qib_devinfo(ppd->dd->pcidev,
+			 "Unimplemented portcounter %u\n", reg);
+		goto done;
+	}
+	creg = xlator[reg];
+
+	if (reg == QIBPORTCNTR_KHDROVFL) {
+		int i;
+
+		/* sum over all kernel contexts */
+		for (i = 0; i < dd->first_user_ctxt; i++)
+			ret += read_7220_creg32(dd, cr_portovfl + i);
+	}
+	if (creg == 0xffff)
+		goto done;
+
+	/*
+	 * only fast incrementing counters are 64bit; use 32 bit reads to
+	 * avoid two independent reads when on opteron
+	 */
+	if ((creg == cr_wordsend || creg == cr_wordrcv ||
+	     creg == cr_pktsend || creg == cr_pktrcv))
+		ret = read_7220_creg(dd, creg);
+	else
+		ret = read_7220_creg32(dd, creg);
+	if (creg == cr_ibsymbolerr) {
+		if (dd->pport->cpspec->ibdeltainprog)
+			ret -= ret - ppd->cpspec->ibsymsnap;
+		ret -= dd->pport->cpspec->ibsymdelta;
+	} else if (creg == cr_iblinkerrrecov) {
+		if (dd->pport->cpspec->ibdeltainprog)
+			ret -= ret - ppd->cpspec->iblnkerrsnap;
+		ret -= dd->pport->cpspec->iblnkerrdelta;
+	}
+done:
+	return ret;
+}
+
+/*
+ * Device counter names (not port-specific), one line per stat,
+ * single string.  Used by utilities like ipathstats to print the stats
+ * in a way which works for different versions of drivers, without changing
+ * the utility.  Names need to be 12 chars or less (w/o newline), for proper
+ * display by utility.
+ * Non-error counters are first.
+ * Start of "error" conters is indicated by a leading "E " on the first
+ * "error" counter, and doesn't count in label length.
+ * The EgrOvfl list needs to be last so we truncate them at the configured
+ * context count for the device.
+ * cntr7220indices contains the corresponding register indices.
+ */
+static const char cntr7220names[] =
+	"Interrupts\n"
+	"HostBusStall\n"
+	"E RxTIDFull\n"
+	"RxTIDInvalid\n"
+	"Ctxt0EgrOvfl\n"
+	"Ctxt1EgrOvfl\n"
+	"Ctxt2EgrOvfl\n"
+	"Ctxt3EgrOvfl\n"
+	"Ctxt4EgrOvfl\n"
+	"Ctxt5EgrOvfl\n"
+	"Ctxt6EgrOvfl\n"
+	"Ctxt7EgrOvfl\n"
+	"Ctxt8EgrOvfl\n"
+	"Ctxt9EgrOvfl\n"
+	"Ctx10EgrOvfl\n"
+	"Ctx11EgrOvfl\n"
+	"Ctx12EgrOvfl\n"
+	"Ctx13EgrOvfl\n"
+	"Ctx14EgrOvfl\n"
+	"Ctx15EgrOvfl\n"
+	"Ctx16EgrOvfl\n";
+
+static const size_t cntr7220indices[] = {
+	cr_lbint,
+	cr_lbflowstall,
+	cr_errtidfull,
+	cr_errtidvalid,
+	cr_portovfl + 0,
+	cr_portovfl + 1,
+	cr_portovfl + 2,
+	cr_portovfl + 3,
+	cr_portovfl + 4,
+	cr_portovfl + 5,
+	cr_portovfl + 6,
+	cr_portovfl + 7,
+	cr_portovfl + 8,
+	cr_portovfl + 9,
+	cr_portovfl + 10,
+	cr_portovfl + 11,
+	cr_portovfl + 12,
+	cr_portovfl + 13,
+	cr_portovfl + 14,
+	cr_portovfl + 15,
+	cr_portovfl + 16,
+};
+
+/*
+ * same as cntr7220names and cntr7220indices, but for port-specific counters.
+ * portcntr7220indices is somewhat complicated by some registers needing
+ * adjustments of various kinds, and those are ORed with _PORT_VIRT_FLAG
+ */
+static const char portcntr7220names[] =
+	"TxPkt\n"
+	"TxFlowPkt\n"
+	"TxWords\n"
+	"RxPkt\n"
+	"RxFlowPkt\n"
+	"RxWords\n"
+	"TxFlowStall\n"
+	"TxDmaDesc\n"  /* 7220 and 7322-only */
+	"E RxDlidFltr\n"  /* 7220 and 7322-only */
+	"IBStatusChng\n"
+	"IBLinkDown\n"
+	"IBLnkRecov\n"
+	"IBRxLinkErr\n"
+	"IBSymbolErr\n"
+	"RxLLIErr\n"
+	"RxBadFormat\n"
+	"RxBadLen\n"
+	"RxBufOvrfl\n"
+	"RxEBP\n"
+	"RxFlowCtlErr\n"
+	"RxICRCerr\n"
+	"RxLPCRCerr\n"
+	"RxVCRCerr\n"
+	"RxInvalLen\n"
+	"RxInvalPKey\n"
+	"RxPktDropped\n"
+	"TxBadLength\n"
+	"TxDropped\n"
+	"TxInvalLen\n"
+	"TxUnderrun\n"
+	"TxUnsupVL\n"
+	"RxLclPhyErr\n" /* 7220 and 7322-only */
+	"RxVL15Drop\n" /* 7220 and 7322-only */
+	"RxVlErr\n" /* 7220 and 7322-only */
+	"XcessBufOvfl\n" /* 7220 and 7322-only */
+	;
+
+#define _PORT_VIRT_FLAG 0x8000 /* "virtual", need adjustments */
+static const size_t portcntr7220indices[] = {
+	QIBPORTCNTR_PKTSEND | _PORT_VIRT_FLAG,
+	cr_pktsendflow,
+	QIBPORTCNTR_WORDSEND | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_PKTRCV | _PORT_VIRT_FLAG,
+	cr_pktrcvflowctrl,
+	QIBPORTCNTR_WORDRCV | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_SENDSTALL | _PORT_VIRT_FLAG,
+	cr_txsdmadesc,
+	cr_rxdlidfltr,
+	cr_ibstatuschange,
+	QIBPORTCNTR_IBLINKDOWN | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_IBLINKERRRECOV | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_ERRLINK | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_IBSYMBOLERR | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_LLI | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_BADFORMAT | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_ERR_RLEN | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_RCVOVFL | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_RCVEBP | _PORT_VIRT_FLAG,
+	cr_rcvflowctrl_err,
+	QIBPORTCNTR_ERRICRC | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_ERRLPCRC | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_ERRVCRC | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_INVALIDRLEN | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_ERRPKEY | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_RXDROPPKT | _PORT_VIRT_FLAG,
+	cr_invalidslen,
+	cr_senddropped,
+	cr_errslen,
+	cr_sendunderrun,
+	cr_txunsupvl,
+	QIBPORTCNTR_RXLOCALPHYERR | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_VL15PKTDROP | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_RXVLERR | _PORT_VIRT_FLAG,
+	QIBPORTCNTR_EXCESSBUFOVFL | _PORT_VIRT_FLAG,
+};
+
+/* do all the setup to make the counter reads efficient later */
+static void init_7220_cntrnames(struct qib_devdata *dd)
+{
+	int i, j = 0;
+	char *s;
+
+	for (i = 0, s = (char *)cntr7220names; s && j <= dd->cfgctxts;
+	     i++) {
+		/* we always have at least one counter before the egrovfl */
+		if (!j && !strncmp("Ctxt0EgrOvfl", s + 1, 12))
+			j = 1;
+		s = strchr(s + 1, '\n');
+		if (s && j)
+			j++;
+	}
+	dd->cspec->ncntrs = i;
+	if (!s)
+		/* full list; size is without terminating null */
+		dd->cspec->cntrnamelen = sizeof(cntr7220names) - 1;
+	else
+		dd->cspec->cntrnamelen = 1 + s - cntr7220names;
+	dd->cspec->cntrs = kmalloc(dd->cspec->ncntrs
+		* sizeof(u64), GFP_KERNEL);
+	if (!dd->cspec->cntrs)
+		qib_dev_err(dd, "Failed allocation for counters\n");
+
+	for (i = 0, s = (char *)portcntr7220names; s; i++)
+		s = strchr(s + 1, '\n');
+	dd->cspec->nportcntrs = i - 1;
+	dd->cspec->portcntrnamelen = sizeof(portcntr7220names) - 1;
+	dd->cspec->portcntrs = kmalloc(dd->cspec->nportcntrs
+		* sizeof(u64), GFP_KERNEL);
+	if (!dd->cspec->portcntrs)
+		qib_dev_err(dd, "Failed allocation for portcounters\n");
+}
+
+static u32 qib_read_7220cntrs(struct qib_devdata *dd, loff_t pos, char **namep,
+			      u64 **cntrp)
+{
+	u32 ret;
+
+	if (!dd->cspec->cntrs) {
+		ret = 0;
+		goto done;
+	}
+
+	if (namep) {
+		*namep = (char *)cntr7220names;
+		ret = dd->cspec->cntrnamelen;
+		if (pos >= ret)
+			ret = 0; /* final read after getting everything */
+	} else {
+		u64 *cntr = dd->cspec->cntrs;
+		int i;
+
+		ret = dd->cspec->ncntrs * sizeof(u64);
+		if (!cntr || pos >= ret) {
+			/* everything read, or couldn't get memory */
+			ret = 0;
+			goto done;
+		}
+
+		*cntrp = cntr;
+		for (i = 0; i < dd->cspec->ncntrs; i++)
+			*cntr++ = read_7220_creg32(dd, cntr7220indices[i]);
+	}
+done:
+	return ret;
+}
+
+static u32 qib_read_7220portcntrs(struct qib_devdata *dd, loff_t pos, u32 port,
+				  char **namep, u64 **cntrp)
+{
+	u32 ret;
+
+	if (!dd->cspec->portcntrs) {
+		ret = 0;
+		goto done;
+	}
+	if (namep) {
+		*namep = (char *)portcntr7220names;
+		ret = dd->cspec->portcntrnamelen;
+		if (pos >= ret)
+			ret = 0; /* final read after getting everything */
+	} else {
+		u64 *cntr = dd->cspec->portcntrs;
+		struct qib_pportdata *ppd = &dd->pport[port];
+		int i;
+
+		ret = dd->cspec->nportcntrs * sizeof(u64);
+		if (!cntr || pos >= ret) {
+			/* everything read, or couldn't get memory */
+			ret = 0;
+			goto done;
+		}
+		*cntrp = cntr;
+		for (i = 0; i < dd->cspec->nportcntrs; i++) {
+			if (portcntr7220indices[i] & _PORT_VIRT_FLAG)
+				*cntr++ = qib_portcntr_7220(ppd,
+					portcntr7220indices[i] &
+					~_PORT_VIRT_FLAG);
+			else
+				*cntr++ = read_7220_creg32(dd,
+					   portcntr7220indices[i]);
+		}
+	}
+done:
+	return ret;
+}
+
+/**
+ * qib_get_7220_faststats - get word counters from chip before they overflow
+ * @opaque - contains a pointer to the qlogic_ib device qib_devdata
+ *
+ * This needs more work; in particular, decision on whether we really
+ * need traffic_wds done the way it is
+ * called from add_timer
+ */
+static void qib_get_7220_faststats(unsigned long opaque)
+{
+	struct qib_devdata *dd = (struct qib_devdata *) opaque;
+	struct qib_pportdata *ppd = dd->pport;
+	unsigned long flags;
+	u64 traffic_wds;
+
+	/*
+	 * don't access the chip while running diags, or memory diags can
+	 * fail
+	 */
+	if (!(dd->flags & QIB_INITTED) || dd->diag_client)
+		/* but re-arm the timer, for diags case; won't hurt other */
+		goto done;
+
+	/*
+	 * We now try to maintain an activity timer, based on traffic
+	 * exceeding a threshold, so we need to check the word-counts
+	 * even if they are 64-bit.
+	 */
+	traffic_wds = qib_portcntr_7220(ppd, cr_wordsend) +
+		qib_portcntr_7220(ppd, cr_wordrcv);
+	spin_lock_irqsave(&dd->eep_st_lock, flags);
+	traffic_wds -= dd->traffic_wds;
+	dd->traffic_wds += traffic_wds;
+	if (traffic_wds  >= QIB_TRAFFIC_ACTIVE_THRESHOLD)
+		atomic_add(5, &dd->active_time); /* S/B #define */
+	spin_unlock_irqrestore(&dd->eep_st_lock, flags);
+done:
+	mod_timer(&dd->stats_timer, jiffies + HZ * ACTIVITY_TIMER);
+}
+
+/*
+ * If we are using MSI, try to fallback to INTx.
+ */
+static int qib_7220_intr_fallback(struct qib_devdata *dd)
+{
+	if (!dd->msi_lo)
+		return 0;
+
+	qib_devinfo(dd->pcidev, "MSI interrupt not detected,"
+		 " trying INTx interrupts\n");
+	qib_7220_free_irq(dd);
+	qib_enable_intx(dd->pcidev);
+	/*
+	 * Some newer kernels require free_irq before disable_msi,
+	 * and irq can be changed during disable and INTx enable
+	 * and we need to therefore use the pcidev->irq value,
+	 * not our saved MSI value.
+	 */
+	dd->cspec->irq = dd->pcidev->irq;
+	qib_setup_7220_interrupt(dd);
+	return 1;
+}
+
+/*
+ * Reset the XGXS (between serdes and IBC).  Slightly less intrusive
+ * than resetting the IBC or external link state, and useful in some
+ * cases to cause some retraining.  To do this right, we reset IBC
+ * as well.
+ */
+static void qib_7220_xgxs_reset(struct qib_pportdata *ppd)
+{
+	u64 val, prev_val;
+	struct qib_devdata *dd = ppd->dd;
+
+	prev_val = qib_read_kreg64(dd, kr_xgxs_cfg);
+	val = prev_val | QLOGIC_IB_XGXS_RESET;
+	prev_val &= ~QLOGIC_IB_XGXS_RESET; /* be sure */
+	qib_write_kreg(dd, kr_control,
+		       dd->control & ~QLOGIC_IB_C_LINKENABLE);
+	qib_write_kreg(dd, kr_xgxs_cfg, val);
+	qib_read_kreg32(dd, kr_scratch);
+	qib_write_kreg(dd, kr_xgxs_cfg, prev_val);
+	qib_write_kreg(dd, kr_control, dd->control);
+}
+
+/*
+ * For this chip, we want to use the same buffer every time
+ * when we are trying to bring the link up (they are always VL15
+ * packets).  At that link state the packet should always go out immediately
+ * (or at least be discarded at the tx interface if the link is down).
+ * If it doesn't, and the buffer isn't available, that means some other
+ * sender has gotten ahead of us, and is preventing our packet from going
+ * out.  In that case, we flush all packets, and try again.  If that still
+ * fails, we fail the request, and hope things work the next time around.
+ *
+ * We don't need very complicated heuristics on whether the packet had
+ * time to go out or not, since even at SDR 1X, it goes out in very short
+ * time periods, covered by the chip reads done here and as part of the
+ * flush.
+ */
+static u32 __iomem *get_7220_link_buf(struct qib_pportdata *ppd, u32 *bnum)
+{
+	u32 __iomem *buf;
+	u32 lbuf = ppd->dd->cspec->lastbuf_for_pio;
+	int do_cleanup;
+	unsigned long flags;
+
+	/*
+	 * always blip to get avail list updated, since it's almost
+	 * always needed, and is fairly cheap.
+	 */
+	sendctrl_7220_mod(ppd->dd->pport, QIB_SENDCTRL_AVAIL_BLIP);
+	qib_read_kreg64(ppd->dd, kr_scratch); /* extra chip flush */
+	buf = qib_getsendbuf_range(ppd->dd, bnum, lbuf, lbuf);
+	if (buf)
+		goto done;
+
+	spin_lock_irqsave(&ppd->sdma_lock, flags);
+	if (ppd->sdma_state.current_state == qib_sdma_state_s20_idle &&
+	    ppd->sdma_state.current_state != qib_sdma_state_s00_hw_down) {
+		__qib_sdma_process_event(ppd, qib_sdma_event_e00_go_hw_down);
+		do_cleanup = 0;
+	} else {
+		do_cleanup = 1;
+		qib_7220_sdma_hw_clean_up(ppd);
+	}
+	spin_unlock_irqrestore(&ppd->sdma_lock, flags);
+
+	if (do_cleanup) {
+		qib_read_kreg64(ppd->dd, kr_scratch); /* extra chip flush */
+		buf = qib_getsendbuf_range(ppd->dd, bnum, lbuf, lbuf);
+	}
+done:
+	return buf;
+}
+
+/*
+ * This code for non-IBTA-compliant IB speed negotiation is only known to
+ * work for the SDR to DDR transition, and only between an HCA and a switch
+ * with recent firmware.  It is based on observed heuristics, rather than
+ * actual knowledge of the non-compliant speed negotiation.
+ * It has a number of hard-coded fields, since the hope is to rewrite this
+ * when a spec is available on how the negoation is intended to work.
+ */
+static void autoneg_7220_sendpkt(struct qib_pportdata *ppd, u32 *hdr,
+				 u32 dcnt, u32 *data)
+{
+	int i;
+	u64 pbc;
+	u32 __iomem *piobuf;
+	u32 pnum;
+	struct qib_devdata *dd = ppd->dd;
+
+	i = 0;
+	pbc = 7 + dcnt + 1; /* 7 dword header, dword data, icrc */
+	pbc |= PBC_7220_VL15_SEND;
+	while (!(piobuf = get_7220_link_buf(ppd, &pnum))) {
+		if (i++ > 5)
+			return;
+		udelay(2);
+	}
+	sendctrl_7220_mod(dd->pport, QIB_SENDCTRL_DISARM_BUF(pnum));
+	writeq(pbc, piobuf);
+	qib_flush_wc();
+	qib_pio_copy(piobuf + 2, hdr, 7);
+	qib_pio_copy(piobuf + 9, data, dcnt);
+	if (dd->flags & QIB_USE_SPCL_TRIG) {
+		u32 spcl_off = (pnum >= dd->piobcnt2k) ? 2047 : 1023;
+
+		qib_flush_wc();
+		__raw_writel(0xaebecede, piobuf + spcl_off);
+	}
+	qib_flush_wc();
+	qib_sendbuf_done(dd, pnum);
+}
+
+/*
+ * _start packet gets sent twice at start, _done gets sent twice at end
+ */
+static void autoneg_7220_send(struct qib_pportdata *ppd, int which)
+{
+	struct qib_devdata *dd = ppd->dd;
+	static u32 swapped;
+	u32 dw, i, hcnt, dcnt, *data;
+	static u32 hdr[7] = { 0xf002ffff, 0x48ffff, 0x6400abba };
+	static u32 madpayload_start[0x40] = {
+		0x1810103, 0x1, 0x0, 0x0, 0x2c90000, 0x2c9, 0x0, 0x0,
+		0xffffffff, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0,
+		0x1, 0x1388, 0x15e, 0x1, /* rest 0's */
+		};
+	static u32 madpayload_done[0x40] = {
+		0x1810103, 0x1, 0x0, 0x0, 0x2c90000, 0x2c9, 0x0, 0x0,
+		0xffffffff, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0,
+		0x40000001, 0x1388, 0x15e, /* rest 0's */
+		};
+
+	dcnt = ARRAY_SIZE(madpayload_start);
+	hcnt = ARRAY_SIZE(hdr);
+	if (!swapped) {
+		/* for maintainability, do it at runtime */
+		for (i = 0; i < hcnt; i++) {
+			dw = (__force u32) cpu_to_be32(hdr[i]);
+			hdr[i] = dw;
+		}
+		for (i = 0; i < dcnt; i++) {
+			dw = (__force u32) cpu_to_be32(madpayload_start[i]);
+			madpayload_start[i] = dw;
+			dw = (__force u32) cpu_to_be32(madpayload_done[i]);
+			madpayload_done[i] = dw;
+		}
+		swapped = 1;
+	}
+
+	data = which ? madpayload_done : madpayload_start;
+
+	autoneg_7220_sendpkt(ppd, hdr, dcnt, data);
+	qib_read_kreg64(dd, kr_scratch);
+	udelay(2);
+	autoneg_7220_sendpkt(ppd, hdr, dcnt, data);
+	qib_read_kreg64(dd, kr_scratch);
+	udelay(2);
+}
+
+/*
+ * Do the absolute minimum to cause an IB speed change, and make it
+ * ready, but don't actually trigger the change.   The caller will
+ * do that when ready (if link is in Polling training state, it will
+ * happen immediately, otherwise when link next goes down)
+ *
+ * This routine should only be used as part of the DDR autonegotation
+ * code for devices that are not compliant with IB 1.2 (or code that
+ * fixes things up for same).
+ *
+ * When link has gone down, and autoneg enabled, or autoneg has
+ * failed and we give up until next time we set both speeds, and
+ * then we want IBTA enabled as well as "use max enabled speed.
+ */
+static void set_7220_ibspeed_fast(struct qib_pportdata *ppd, u32 speed)
+{
+	ppd->cpspec->ibcddrctrl &= ~(IBA7220_IBC_SPEED_AUTONEG_MASK |
+		IBA7220_IBC_IBTA_1_2_MASK);
+
+	if (speed == (QIB_IB_SDR | QIB_IB_DDR))
+		ppd->cpspec->ibcddrctrl |= IBA7220_IBC_SPEED_AUTONEG_MASK |
+			IBA7220_IBC_IBTA_1_2_MASK;
+	else
+		ppd->cpspec->ibcddrctrl |= speed == QIB_IB_DDR ?
+			IBA7220_IBC_SPEED_DDR : IBA7220_IBC_SPEED_SDR;
+
+	qib_write_kreg(ppd->dd, kr_ibcddrctrl, ppd->cpspec->ibcddrctrl);
+	qib_write_kreg(ppd->dd, kr_scratch, 0);
+}
+
+/*
+ * This routine is only used when we are not talking to another
+ * IB 1.2-compliant device that we think can do DDR.
+ * (This includes all existing switch chips as of Oct 2007.)
+ * 1.2-compliant devices go directly to DDR prior to reaching INIT
+ */
+static void try_7220_autoneg(struct qib_pportdata *ppd)
+{
+	unsigned long flags;
+
+	/*
+	 * Required for older non-IB1.2 DDR switches.  Newer
+	 * non-IB-compliant switches don't need it, but so far,
+	 * aren't bothered by it either.  "Magic constant"
+	 */
+	qib_write_kreg(ppd->dd, kr_ncmodectrl, 0x3b9dc07);
+
+	spin_lock_irqsave(&ppd->lflags_lock, flags);
+	ppd->lflags |= QIBL_IB_AUTONEG_INPROG;
+	spin_unlock_irqrestore(&ppd->lflags_lock, flags);
+	autoneg_7220_send(ppd, 0);
+	set_7220_ibspeed_fast(ppd, QIB_IB_DDR);
+
+	toggle_7220_rclkrls(ppd->dd);
+	/* 2 msec is minimum length of a poll cycle */
+	schedule_delayed_work(&ppd->cpspec->autoneg_work,
+			      msecs_to_jiffies(2));
+}
+
+/*
+ * Handle the empirically determined mechanism for auto-negotiation
+ * of DDR speed with switches.
+ */
+static void autoneg_7220_work(struct work_struct *work)
+{
+	struct qib_pportdata *ppd;
+	struct qib_devdata *dd;
+	u64 startms;
+	u32 i;
+	unsigned long flags;
+
+	ppd = &container_of(work, struct qib_chippport_specific,
+			    autoneg_work.work)->pportdata;
+	dd = ppd->dd;
+
+	startms = jiffies_to_msecs(jiffies);
+
+	/*
+	 * Busy wait for this first part, it should be at most a
+	 * few hundred usec, since we scheduled ourselves for 2msec.
+	 */
+	for (i = 0; i < 25; i++) {
+		if (SYM_FIELD(ppd->lastibcstat, IBCStatus, LinkTrainingState)
+		     == IB_7220_LT_STATE_POLLQUIET) {
+			qib_set_linkstate(ppd, QIB_IB_LINKDOWN_DISABLE);
+			break;
+		}
+		udelay(100);
+	}
+
+	if (!(ppd->lflags & QIBL_IB_AUTONEG_INPROG))
+		goto done; /* we got there early or told to stop */
+
+	/* we expect this to timeout */
+	if (wait_event_timeout(ppd->cpspec->autoneg_wait,
+			       !(ppd->lflags & QIBL_IB_AUTONEG_INPROG),
+			       msecs_to_jiffies(90)))
+		goto done;
+
+	toggle_7220_rclkrls(dd);
+
+	/* we expect this to timeout */
+	if (wait_event_timeout(ppd->cpspec->autoneg_wait,
+			       !(ppd->lflags & QIBL_IB_AUTONEG_INPROG),
+			       msecs_to_jiffies(1700)))
+		goto done;
+
+	set_7220_ibspeed_fast(ppd, QIB_IB_SDR);
+	toggle_7220_rclkrls(dd);
+
+	/*
+	 * Wait up to 250 msec for link to train and get to INIT at DDR;
+	 * this should terminate early.
+	 */
+	wait_event_timeout(ppd->cpspec->autoneg_wait,
+		!(ppd->lflags & QIBL_IB_AUTONEG_INPROG),
+		msecs_to_jiffies(250));
+done:
+	if (ppd->lflags & QIBL_IB_AUTONEG_INPROG) {
+		spin_lock_irqsave(&ppd->lflags_lock, flags);
+		ppd->lflags &= ~QIBL_IB_AUTONEG_INPROG;
+		if (dd->cspec->autoneg_tries == AUTONEG_TRIES) {
+			ppd->lflags |= QIBL_IB_AUTONEG_FAILED;
+			dd->cspec->autoneg_tries = 0;
+		}
+		spin_unlock_irqrestore(&ppd->lflags_lock, flags);
+		set_7220_ibspeed_fast(ppd, ppd->link_speed_enabled);
+	}
+}
+
+static u32 qib_7220_iblink_state(u64 ibcs)
+{
+	u32 state = (u32)SYM_FIELD(ibcs, IBCStatus, LinkState);
+
+	switch (state) {
+	case IB_7220_L_STATE_INIT:
+		state = IB_PORT_INIT;
+		break;
+	case IB_7220_L_STATE_ARM:
+		state = IB_PORT_ARMED;
+		break;
+	case IB_7220_L_STATE_ACTIVE:
+		/* fall through */
+	case IB_7220_L_STATE_ACT_DEFER:
+		state = IB_PORT_ACTIVE;
+		break;
+	default: /* fall through */
+	case IB_7220_L_STATE_DOWN:
+		state = IB_PORT_DOWN;
+		break;
+	}
+	return state;
+}
+
+/* returns the IBTA port state, rather than the IBC link training state */
+static u8 qib_7220_phys_portstate(u64 ibcs)
+{
+	u8 state = (u8)SYM_FIELD(ibcs, IBCStatus, LinkTrainingState);
+	return qib_7220_physportstate[state];
+}
+
+static int qib_7220_ib_updown(struct qib_pportdata *ppd, int ibup, u64 ibcs)
+{
+	int ret = 0, symadj = 0;
+	struct qib_devdata *dd = ppd->dd;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ppd->lflags_lock, flags);
+	ppd->lflags &= ~QIBL_IB_FORCE_NOTIFY;
+	spin_unlock_irqrestore(&ppd->lflags_lock, flags);
+
+	if (!ibup) {
+		/*
+		 * When the link goes down we don't want AEQ running, so it
+		 * won't interfere with IBC training, etc., and we need
+		 * to go back to the static SerDes preset values.
+		 */
+		if (!(ppd->lflags & (QIBL_IB_AUTONEG_FAILED |
+				     QIBL_IB_AUTONEG_INPROG)))
+			set_7220_ibspeed_fast(ppd, ppd->link_speed_enabled);
+		if (!(ppd->lflags & QIBL_IB_AUTONEG_INPROG)) {
+			qib_sd7220_presets(dd);
+			qib_cancel_sends(ppd); /* initial disarm, etc. */
+			spin_lock_irqsave(&ppd->sdma_lock, flags);
+			if (__qib_sdma_running(ppd))
+				__qib_sdma_process_event(ppd,
+					qib_sdma_event_e70_go_idle);
+			spin_unlock_irqrestore(&ppd->sdma_lock, flags);
+		}
+		/* this might better in qib_sd7220_presets() */
+		set_7220_relock_poll(dd, ibup);
+	} else {
+		if (qib_compat_ddr_negotiate &&
+		    !(ppd->lflags & (QIBL_IB_AUTONEG_FAILED |
+				     QIBL_IB_AUTONEG_INPROG)) &&
+		    ppd->link_speed_active == QIB_IB_SDR &&
+		    (ppd->link_speed_enabled & (QIB_IB_DDR | QIB_IB_SDR)) ==
+		    (QIB_IB_DDR | QIB_IB_SDR) &&
+		    dd->cspec->autoneg_tries < AUTONEG_TRIES) {
+			/* we are SDR, and DDR auto-negotiation enabled */
+			++dd->cspec->autoneg_tries;
+			if (!ppd->cpspec->ibdeltainprog) {
+				ppd->cpspec->ibdeltainprog = 1;
+				ppd->cpspec->ibsymsnap = read_7220_creg32(dd,
+					cr_ibsymbolerr);
+				ppd->cpspec->iblnkerrsnap = read_7220_creg32(dd,
+					cr_iblinkerrrecov);
+			}
+			try_7220_autoneg(ppd);
+			ret = 1; /* no other IB status change processing */
+		} else if ((ppd->lflags & QIBL_IB_AUTONEG_INPROG) &&
+			   ppd->link_speed_active == QIB_IB_SDR) {
+			autoneg_7220_send(ppd, 1);
+			set_7220_ibspeed_fast(ppd, QIB_IB_DDR);
+			udelay(2);
+			toggle_7220_rclkrls(dd);
+			ret = 1; /* no other IB status change processing */
+		} else {
+			if ((ppd->lflags & QIBL_IB_AUTONEG_INPROG) &&
+			    (ppd->link_speed_active & QIB_IB_DDR)) {
+				spin_lock_irqsave(&ppd->lflags_lock, flags);
+				ppd->lflags &= ~(QIBL_IB_AUTONEG_INPROG |
+						 QIBL_IB_AUTONEG_FAILED);
+				spin_unlock_irqrestore(&ppd->lflags_lock,
+						       flags);
+				dd->cspec->autoneg_tries = 0;
+				/* re-enable SDR, for next link down */
+				set_7220_ibspeed_fast(ppd,
+						      ppd->link_speed_enabled);
+				wake_up(&ppd->cpspec->autoneg_wait);
+				symadj = 1;
+			} else if (ppd->lflags & QIBL_IB_AUTONEG_FAILED) {
+				/*
+				 * Clear autoneg failure flag, and do setup
+				 * so we'll try next time link goes down and
+				 * back to INIT (possibly connected to a
+				 * different device).
+				 */
+				spin_lock_irqsave(&ppd->lflags_lock, flags);
+				ppd->lflags &= ~QIBL_IB_AUTONEG_FAILED;
+				spin_unlock_irqrestore(&ppd->lflags_lock,
+						       flags);
+				ppd->cpspec->ibcddrctrl |=
+					IBA7220_IBC_IBTA_1_2_MASK;
+				qib_write_kreg(dd, kr_ncmodectrl, 0);
+				symadj = 1;
+			}
+		}
+
+		if (!(ppd->lflags & QIBL_IB_AUTONEG_INPROG))
+			symadj = 1;
+
+		if (!ret) {
+			ppd->delay_mult = rate_to_delay
+			    [(ibcs >> IBA7220_LINKSPEED_SHIFT) & 1]
+			    [(ibcs >> IBA7220_LINKWIDTH_SHIFT) & 1];
+
+			set_7220_relock_poll(dd, ibup);
+			spin_lock_irqsave(&ppd->sdma_lock, flags);
+			/*
+			 * Unlike 7322, the 7220 needs this, due to lack of
+			 * interrupt in some cases when we have sdma active
+			 * when the link goes down.
+			 */
+			if (ppd->sdma_state.current_state !=
+			    qib_sdma_state_s20_idle)
+				__qib_sdma_process_event(ppd,
+					qib_sdma_event_e00_go_hw_down);
+			spin_unlock_irqrestore(&ppd->sdma_lock, flags);
+		}
+	}
+
+	if (symadj) {
+		if (ppd->cpspec->ibdeltainprog) {
+			ppd->cpspec->ibdeltainprog = 0;
+			ppd->cpspec->ibsymdelta += read_7220_creg32(ppd->dd,
+				cr_ibsymbolerr) - ppd->cpspec->ibsymsnap;
+			ppd->cpspec->iblnkerrdelta += read_7220_creg32(ppd->dd,
+				cr_iblinkerrrecov) - ppd->cpspec->iblnkerrsnap;
+		}
+	} else if (!ibup && qib_compat_ddr_negotiate &&
+		   !ppd->cpspec->ibdeltainprog &&
+			!(ppd->lflags & QIBL_IB_AUTONEG_INPROG)) {
+		ppd->cpspec->ibdeltainprog = 1;
+		ppd->cpspec->ibsymsnap = read_7220_creg32(ppd->dd,
+							  cr_ibsymbolerr);
+		ppd->cpspec->iblnkerrsnap = read_7220_creg32(ppd->dd,
+						     cr_iblinkerrrecov);
+	}
+
+	if (!ret)
+		qib_setup_7220_setextled(ppd, ibup);
+	return ret;
+}
+
+/*
+ * Does read/modify/write to appropriate registers to
+ * set output and direction bits selected by mask.
+ * these are in their canonical postions (e.g. lsb of
+ * dir will end up in D48 of extctrl on existing chips).
+ * returns contents of GP Inputs.
+ */
+static int gpio_7220_mod(struct qib_devdata *dd, u32 out, u32 dir, u32 mask)
+{
+	u64 read_val, new_out;
+	unsigned long flags;
+
+	if (mask) {
+		/* some bits being written, lock access to GPIO */
+		dir &= mask;
+		out &= mask;
+		spin_lock_irqsave(&dd->cspec->gpio_lock, flags);
+		dd->cspec->extctrl &= ~((u64)mask << SYM_LSB(EXTCtrl, GPIOOe));
+		dd->cspec->extctrl |= ((u64) dir << SYM_LSB(EXTCtrl, GPIOOe));
+		new_out = (dd->cspec->gpio_out & ~mask) | out;
+
+		qib_write_kreg(dd, kr_extctrl, dd->cspec->extctrl);
+		qib_write_kreg(dd, kr_gpio_out, new_out);
+		dd->cspec->gpio_out = new_out;
+		spin_unlock_irqrestore(&dd->cspec->gpio_lock, flags);
+	}
+	/*
+	 * It is unlikely that a read at this time would get valid
+	 * data on a pin whose direction line was set in the same
+	 * call to this function. We include the read here because
+	 * that allows us to potentially combine a change on one pin with
+	 * a read on another, and because the old code did something like
+	 * this.
+	 */
+	read_val = qib_read_kreg64(dd, kr_extstatus);
+	return SYM_FIELD(read_val, EXTStatus, GPIOIn);
+}
+
+/*
+ * Read fundamental info we need to use the chip.  These are
+ * the registers that describe chip capabilities, and are
+ * saved in shadow registers.
+ */
+static void get_7220_chip_params(struct qib_devdata *dd)
+{
+	u64 val;
+	u32 piobufs;
+	int mtu;
+
+	dd->uregbase = qib_read_kreg32(dd, kr_userregbase);
+
+	dd->rcvtidcnt = qib_read_kreg32(dd, kr_rcvtidcnt);
+	dd->rcvtidbase = qib_read_kreg32(dd, kr_rcvtidbase);
+	dd->rcvegrbase = qib_read_kreg32(dd, kr_rcvegrbase);
+	dd->palign = qib_read_kreg32(dd, kr_palign);
+	dd->piobufbase = qib_read_kreg64(dd, kr_sendpiobufbase);
+	dd->pio2k_bufbase = dd->piobufbase & 0xffffffff;
+
+	val = qib_read_kreg64(dd, kr_sendpiosize);
+	dd->piosize2k = val & ~0U;
+	dd->piosize4k = val >> 32;
+
+	mtu = ib_mtu_enum_to_int(qib_ibmtu);
+	if (mtu == -1)
+		mtu = QIB_DEFAULT_MTU;
+	dd->pport->ibmtu = (u32)mtu;
+
+	val = qib_read_kreg64(dd, kr_sendpiobufcnt);
+	dd->piobcnt2k = val & ~0U;
+	dd->piobcnt4k = val >> 32;
+	/* these may be adjusted in init_chip_wc_pat() */
+	dd->pio2kbase = (u32 __iomem *)
+		((char __iomem *) dd->kregbase + dd->pio2k_bufbase);
+	if (dd->piobcnt4k) {
+		dd->pio4kbase = (u32 __iomem *)
+			((char __iomem *) dd->kregbase +
+			 (dd->piobufbase >> 32));
+		/*
+		 * 4K buffers take 2 pages; we use roundup just to be
+		 * paranoid; we calculate it once here, rather than on
+		 * ever buf allocate
+		 */
+		dd->align4k = ALIGN(dd->piosize4k, dd->palign);
+	}
+
+	piobufs = dd->piobcnt4k + dd->piobcnt2k;
+
+	dd->pioavregs = ALIGN(piobufs, sizeof(u64) * BITS_PER_BYTE / 2) /
+		(sizeof(u64) * BITS_PER_BYTE / 2);
+}
+
+/*
+ * The chip base addresses in cspec and cpspec have to be set
+ * after possible init_chip_wc_pat(), rather than in
+ * qib_get_7220_chip_params(), so split out as separate function
+ */
+static void set_7220_baseaddrs(struct qib_devdata *dd)
+{
+	u32 cregbase;
+	/* init after possible re-map in init_chip_wc_pat() */
+	cregbase = qib_read_kreg32(dd, kr_counterregbase);
+	dd->cspec->cregbase = (u64 __iomem *)
+		((char __iomem *) dd->kregbase + cregbase);
+
+	dd->egrtidbase = (u64 __iomem *)
+		((char __iomem *) dd->kregbase + dd->rcvegrbase);
+}
+
+
+#define SENDCTRL_SHADOWED (SYM_MASK(SendCtrl, SendIntBufAvail) |	\
+			   SYM_MASK(SendCtrl, SPioEnable) |		\
+			   SYM_MASK(SendCtrl, SSpecialTriggerEn) |	\
+			   SYM_MASK(SendCtrl, SendBufAvailUpd) |	\
+			   SYM_MASK(SendCtrl, AvailUpdThld) |		\
+			   SYM_MASK(SendCtrl, SDmaEnable) |		\
+			   SYM_MASK(SendCtrl, SDmaIntEnable) |		\
+			   SYM_MASK(SendCtrl, SDmaHalt) |		\
+			   SYM_MASK(SendCtrl, SDmaSingleDescriptor))
+
+static int sendctrl_hook(struct qib_devdata *dd,
+			 const struct diag_observer *op,
+			 u32 offs, u64 *data, u64 mask, int only_32)
+{
+	unsigned long flags;
+	unsigned idx = offs / sizeof(u64);
+	u64 local_data, all_bits;
+
+	if (idx != kr_sendctrl) {
+		qib_dev_err(dd, "SendCtrl Hook called with offs %X, %s-bit\n",
+			    offs, only_32 ? "32" : "64");
+		return 0;
+	}
+
+	all_bits = ~0ULL;
+	if (only_32)
+		all_bits >>= 32;
+	spin_lock_irqsave(&dd->sendctrl_lock, flags);
+	if ((mask & all_bits) != all_bits) {
+		/*
+		 * At least some mask bits are zero, so we need
+		 * to read. The judgement call is whether from
+		 * reg or shadow. First-cut: read reg, and complain
+		 * if any bits which should be shadowed are different
+		 * from their shadowed value.
+		 */
+		if (only_32)
+			local_data = (u64)qib_read_kreg32(dd, idx);
+		else
+			local_data = qib_read_kreg64(dd, idx);
+		qib_dev_err(dd, "Sendctrl -> %X, Shad -> %X\n",
+			    (u32)local_data, (u32)dd->sendctrl);
+		if ((local_data & SENDCTRL_SHADOWED) !=
+		    (dd->sendctrl & SENDCTRL_SHADOWED))
+			qib_dev_err(dd, "Sendctrl read: %X shadow is %X\n",
+				(u32)local_data, (u32) dd->sendctrl);
+		*data = (local_data & ~mask) | (*data & mask);
+	}
+	if (mask) {
+		/*
+		 * At least some mask bits are one, so we need
+		 * to write, but only shadow some bits.
+		 */
+		u64 sval, tval; /* Shadowed, transient */
+
+		/*
+		 * New shadow val is bits we don't want to touch,
+		 * ORed with bits we do, that are intended for shadow.
+		 */
+		sval = (dd->sendctrl & ~mask);
+		sval |= *data & SENDCTRL_SHADOWED & mask;
+		dd->sendctrl = sval;
+		tval = sval | (*data & ~SENDCTRL_SHADOWED & mask);
+		qib_dev_err(dd, "Sendctrl <- %X, Shad <- %X\n",
+			    (u32)tval, (u32)sval);
+		qib_write_kreg(dd, kr_sendctrl, tval);
+		qib_write_kreg(dd, kr_scratch, 0Ull);
+	}
+	spin_unlock_irqrestore(&dd->sendctrl_lock, flags);
+
+	return only_32 ? 4 : 8;
+}
+
+static const struct diag_observer sendctrl_observer = {
+	sendctrl_hook, kr_sendctrl * sizeof(u64),
+	kr_sendctrl * sizeof(u64)
+};
+
+/*
+ * write the final few registers that depend on some of the
+ * init setup.  Done late in init, just before bringing up
+ * the serdes.
+ */
+static int qib_late_7220_initreg(struct qib_devdata *dd)
+{
+	int ret = 0;
+	u64 val;
+
+	qib_write_kreg(dd, kr_rcvhdrentsize, dd->rcvhdrentsize);
+	qib_write_kreg(dd, kr_rcvhdrsize, dd->rcvhdrsize);
+	qib_write_kreg(dd, kr_rcvhdrcnt, dd->rcvhdrcnt);
+	qib_write_kreg(dd, kr_sendpioavailaddr, dd->pioavailregs_phys);
+	val = qib_read_kreg64(dd, kr_sendpioavailaddr);
+	if (val != dd->pioavailregs_phys) {
+		qib_dev_err(dd, "Catastrophic software error, "
+			    "SendPIOAvailAddr written as %lx, "
+			    "read back as %llx\n",
+			    (unsigned long) dd->pioavailregs_phys,
+			    (unsigned long long) val);
+		ret = -EINVAL;
+	}
+	qib_register_observer(dd, &sendctrl_observer);
+	return ret;
+}
+
+static int qib_init_7220_variables(struct qib_devdata *dd)
+{
+	struct qib_chippport_specific *cpspec;
+	struct qib_pportdata *ppd;
+	int ret = 0;
+	u32 sbufs, updthresh;
+
+	cpspec = (struct qib_chippport_specific *)(dd + 1);
+	ppd = &cpspec->pportdata;
+	dd->pport = ppd;
+	dd->num_pports = 1;
+
+	dd->cspec = (struct qib_chip_specific *)(cpspec + dd->num_pports);
+	ppd->cpspec = cpspec;
+
+	spin_lock_init(&dd->cspec->sdepb_lock);
+	spin_lock_init(&dd->cspec->rcvmod_lock);
+	spin_lock_init(&dd->cspec->gpio_lock);
+
+	/* we haven't yet set QIB_PRESENT, so use read directly */
+	dd->revision = readq(&dd->kregbase[kr_revision]);
+
+	if ((dd->revision & 0xffffffffU) == 0xffffffffU) {
+		qib_dev_err(dd, "Revision register read failure, "
+			    "giving up initialization\n");
+		ret = -ENODEV;
+		goto bail;
+	}
+	dd->flags |= QIB_PRESENT;  /* now register routines work */
+
+	dd->majrev = (u8) SYM_FIELD(dd->revision, Revision_R,
+				    ChipRevMajor);
+	dd->minrev = (u8) SYM_FIELD(dd->revision, Revision_R,
+				    ChipRevMinor);
+
+	get_7220_chip_params(dd);
+	qib_7220_boardname(dd);
+
+	/*
+	 * GPIO bits for TWSI data and clock,
+	 * used for serial EEPROM.
+	 */
+	dd->gpio_sda_num = _QIB_GPIO_SDA_NUM;
+	dd->gpio_scl_num = _QIB_GPIO_SCL_NUM;
+	dd->twsi_eeprom_dev = QIB_TWSI_EEPROM_DEV;
+
+	dd->flags |= QIB_HAS_INTX | QIB_HAS_LINK_LATENCY |
+		QIB_NODMA_RTAIL | QIB_HAS_THRESH_UPDATE;
+	dd->flags |= qib_special_trigger ?
+		QIB_USE_SPCL_TRIG : QIB_HAS_SEND_DMA;
+
+	/*
+	 * EEPROM error log 0 is TXE Parity errors. 1 is RXE Parity.
+	 * 2 is Some Misc, 3 is reserved for future.
+	 */
+	dd->eep_st_masks[0].hwerrs_to_log = HWE_MASK(TXEMemParityErr);
+
+	dd->eep_st_masks[1].hwerrs_to_log = HWE_MASK(RXEMemParityErr);
+
+	dd->eep_st_masks[2].errs_to_log = ERR_MASK(ResetNegated);
+
+	init_waitqueue_head(&cpspec->autoneg_wait);
+	INIT_DELAYED_WORK(&cpspec->autoneg_work, autoneg_7220_work);
+
+	qib_init_pportdata(ppd, dd, 0, 1);
+	ppd->link_width_supported = IB_WIDTH_1X | IB_WIDTH_4X;
+	ppd->link_speed_supported = QIB_IB_SDR | QIB_IB_DDR;
+
+	ppd->link_width_enabled = ppd->link_width_supported;
+	ppd->link_speed_enabled = ppd->link_speed_supported;
+	/*
+	 * Set the initial values to reasonable default, will be set
+	 * for real when link is up.
+	 */
+	ppd->link_width_active = IB_WIDTH_4X;
+	ppd->link_speed_active = QIB_IB_SDR;
+	ppd->delay_mult = rate_to_delay[0][1];
+	ppd->vls_supported = IB_VL_VL0;
+	ppd->vls_operational = ppd->vls_supported;
+
+	if (!qib_mini_init)
+		qib_write_kreg(dd, kr_rcvbthqp, QIB_KD_QP);
+
+	init_timer(&ppd->cpspec->chase_timer);
+	ppd->cpspec->chase_timer.function = reenable_7220_chase;
+	ppd->cpspec->chase_timer.data = (unsigned long)ppd;
+
+	qib_num_cfg_vls = 1; /* if any 7220's, only one VL */
+
+	dd->rcvhdrentsize = QIB_RCVHDR_ENTSIZE;
+	dd->rcvhdrsize = QIB_DFLT_RCVHDRSIZE;
+	dd->rhf_offset =
+		dd->rcvhdrentsize - sizeof(u64) / sizeof(u32);
+
+	/* we always allocate at least 2048 bytes for eager buffers */
+	ret = ib_mtu_enum_to_int(qib_ibmtu);
+	dd->rcvegrbufsize = ret != -1 ? max(ret, 2048) : QIB_DEFAULT_MTU;
+
+	qib_7220_tidtemplate(dd);
+
+	/*
+	 * We can request a receive interrupt for 1 or
+	 * more packets from current offset.  For now, we set this
+	 * up for a single packet.
+	 */
+	dd->rhdrhead_intr_off = 1ULL << 32;
+
+	/* setup the stats timer; the add_timer is done at end of init */
+	init_timer(&dd->stats_timer);
+	dd->stats_timer.function = qib_get_7220_faststats;
+	dd->stats_timer.data = (unsigned long) dd;
+	dd->stats_timer.expires = jiffies + ACTIVITY_TIMER * HZ;
+
+	/*
+	 * Control[4] has been added to change the arbitration within
+	 * the SDMA engine between favoring data fetches over descriptor
+	 * fetches.  qib_sdma_fetch_arb==0 gives data fetches priority.
+	 */
+	if (qib_sdma_fetch_arb)
+		dd->control |= 1 << 4;
+
+	dd->ureg_align = 0x10000;  /* 64KB alignment */
+
+	dd->piosize2kmax_dwords = (dd->piosize2k >> 2)-1;
+	qib_7220_config_ctxts(dd);
+	qib_set_ctxtcnt(dd);  /* needed for PAT setup */
+
+	if (qib_wc_pat) {
+		ret = init_chip_wc_pat(dd, 0);
+		if (ret)
+			goto bail;
+	}
+	set_7220_baseaddrs(dd); /* set chip access pointers now */
+
+	ret = 0;
+	if (qib_mini_init)
+		goto bail;
+
+	ret = qib_create_ctxts(dd);
+	init_7220_cntrnames(dd);
+
+	/* use all of 4KB buffers for the kernel SDMA, zero if !SDMA.
+	 * reserve the update threshold amount for other kernel use, such
+	 * as sending SMI, MAD, and ACKs, or 3, whichever is greater,
+	 * unless we aren't enabling SDMA, in which case we want to use
+	 * all the 4k bufs for the kernel.
+	 * if this was less than the update threshold, we could wait
+	 * a long time for an update.  Coded this way because we
+	 * sometimes change the update threshold for various reasons,
+	 * and we want this to remain robust.
+	 */
+	updthresh = 8U; /* update threshold */
+	if (dd->flags & QIB_HAS_SEND_DMA) {
+		dd->cspec->sdmabufcnt =  dd->piobcnt4k;
+		sbufs = updthresh > 3 ? updthresh : 3;
+	} else {
+		dd->cspec->sdmabufcnt = 0;
+		sbufs = dd->piobcnt4k;
+	}
+
+	dd->cspec->lastbuf_for_pio = dd->piobcnt2k + dd->piobcnt4k -
+		dd->cspec->sdmabufcnt;
+	dd->lastctxt_piobuf = dd->cspec->lastbuf_for_pio - sbufs;
+	dd->cspec->lastbuf_for_pio--; /* range is <= , not < */
+	dd->pbufsctxt = dd->lastctxt_piobuf /
+		(dd->cfgctxts - dd->first_user_ctxt);
+
+	/*
+	 * if we are at 16 user contexts, we will have one 7 sbufs
+	 * per context, so drop the update threshold to match.  We
+	 * want to update before we actually run out, at low pbufs/ctxt
+	 * so give ourselves some margin
+	 */
+	if ((dd->pbufsctxt - 2) < updthresh)
+		updthresh = dd->pbufsctxt - 2;
+
+	dd->cspec->updthresh_dflt = updthresh;
+	dd->cspec->updthresh = updthresh;
+
+	/* before full enable, no interrupts, no locking needed */
+	dd->sendctrl |= (updthresh & SYM_RMASK(SendCtrl, AvailUpdThld))
+			     << SYM_LSB(SendCtrl, AvailUpdThld);
+
+	dd->psxmitwait_supported = 1;
+	dd->psxmitwait_check_rate = QIB_7220_PSXMITWAIT_CHECK_RATE;
+bail:
+	return ret;
+}
+
+static u32 __iomem *qib_7220_getsendbuf(struct qib_pportdata *ppd, u64 pbc,
+					u32 *pbufnum)
+{
+	u32 first, last, plen = pbc & QIB_PBC_LENGTH_MASK;
+	struct qib_devdata *dd = ppd->dd;
+	u32 __iomem *buf;
+
+	if (((pbc >> 32) & PBC_7220_VL15_SEND_CTRL) &&
+		!(ppd->lflags & (QIBL_IB_AUTONEG_INPROG | QIBL_LINKACTIVE)))
+		buf = get_7220_link_buf(ppd, pbufnum);
+	else {
+		if ((plen + 1) > dd->piosize2kmax_dwords)
+			first = dd->piobcnt2k;
+		else
+			first = 0;
+		/* try 4k if all 2k busy, so same last for both sizes */
+		last = dd->cspec->lastbuf_for_pio;
+		buf = qib_getsendbuf_range(dd, pbufnum, first, last);
+	}
+	return buf;
+}
+
+/* these 2 "counters" are really control registers, and are always RW */
+static void qib_set_cntr_7220_sample(struct qib_pportdata *ppd, u32 intv,
+				     u32 start)
+{
+	write_7220_creg(ppd->dd, cr_psinterval, intv);
+	write_7220_creg(ppd->dd, cr_psstart, start);
+}
+
+/*
+ * NOTE: no real attempt is made to generalize the SDMA stuff.
+ * At some point "soon" we will have a new more generalized
+ * set of sdma interface, and then we'll clean this up.
+ */
+
+/* Must be called with sdma_lock held, or before init finished */
+static void qib_sdma_update_7220_tail(struct qib_pportdata *ppd, u16 tail)
+{
+	/* Commit writes to memory and advance the tail on the chip */
+	wmb();
+	ppd->sdma_descq_tail = tail;
+	qib_write_kreg(ppd->dd, kr_senddmatail, tail);
+}
+
+static void qib_sdma_set_7220_desc_cnt(struct qib_pportdata *ppd, unsigned cnt)
+{
+}
+
+static struct sdma_set_state_action sdma_7220_action_table[] = {
+	[qib_sdma_state_s00_hw_down] = {
+		.op_enable = 0,
+		.op_intenable = 0,
+		.op_halt = 0,
+		.go_s99_running_tofalse = 1,
+	},
+	[qib_sdma_state_s10_hw_start_up_wait] = {
+		.op_enable = 1,
+		.op_intenable = 1,
+		.op_halt = 1,
+	},
+	[qib_sdma_state_s20_idle] = {
+		.op_enable = 1,
+		.op_intenable = 1,
+		.op_halt = 1,
+	},
+	[qib_sdma_state_s30_sw_clean_up_wait] = {
+		.op_enable = 0,
+		.op_intenable = 1,
+		.op_halt = 0,
+	},
+	[qib_sdma_state_s40_hw_clean_up_wait] = {
+		.op_enable = 1,
+		.op_intenable = 1,
+		.op_halt = 1,
+	},
+	[qib_sdma_state_s50_hw_halt_wait] = {
+		.op_enable = 1,
+		.op_intenable = 1,
+		.op_halt = 1,
+	},
+	[qib_sdma_state_s99_running] = {
+		.op_enable = 1,
+		.op_intenable = 1,
+		.op_halt = 0,
+		.go_s99_running_totrue = 1,
+	},
+};
+
+static void qib_7220_sdma_init_early(struct qib_pportdata *ppd)
+{
+	ppd->sdma_state.set_state_action = sdma_7220_action_table;
+}
+
+static int init_sdma_7220_regs(struct qib_pportdata *ppd)
+{
+	struct qib_devdata *dd = ppd->dd;
+	unsigned i, n;
+	u64 senddmabufmask[3] = { 0 };
+
+	/* Set SendDmaBase */
+	qib_write_kreg(dd, kr_senddmabase, ppd->sdma_descq_phys);
+	qib_sdma_7220_setlengen(ppd);
+	qib_sdma_update_7220_tail(ppd, 0); /* Set SendDmaTail */
+	/* Set SendDmaHeadAddr */
+	qib_write_kreg(dd, kr_senddmaheadaddr, ppd->sdma_head_phys);
+
+	/*
+	 * Reserve all the former "kernel" piobufs, using high number range
+	 * so we get as many 4K buffers as possible
+	 */
+	n = dd->piobcnt2k + dd->piobcnt4k;
+	i = n - dd->cspec->sdmabufcnt;
+
+	for (; i < n; ++i) {
+		unsigned word = i / 64;
+		unsigned bit = i & 63;
+
+		BUG_ON(word >= 3);
+		senddmabufmask[word] |= 1ULL << bit;
+	}
+	qib_write_kreg(dd, kr_senddmabufmask0, senddmabufmask[0]);
+	qib_write_kreg(dd, kr_senddmabufmask1, senddmabufmask[1]);
+	qib_write_kreg(dd, kr_senddmabufmask2, senddmabufmask[2]);
+
+	ppd->sdma_state.first_sendbuf = i;
+	ppd->sdma_state.last_sendbuf = n;
+
+	return 0;
+}
+
+/* sdma_lock must be held */
+static u16 qib_sdma_7220_gethead(struct qib_pportdata *ppd)
+{
+	struct qib_devdata *dd = ppd->dd;
+	int sane;
+	int use_dmahead;
+	u16 swhead;
+	u16 swtail;
+	u16 cnt;
+	u16 hwhead;
+
+	use_dmahead = __qib_sdma_running(ppd) &&
+		(dd->flags & QIB_HAS_SDMA_TIMEOUT);
+retry:
+	hwhead = use_dmahead ?
+		(u16)le64_to_cpu(*ppd->sdma_head_dma) :
+		(u16)qib_read_kreg32(dd, kr_senddmahead);
+
+	swhead = ppd->sdma_descq_head;
+	swtail = ppd->sdma_descq_tail;
+	cnt = ppd->sdma_descq_cnt;
+
+	if (swhead < swtail) {
+		/* not wrapped */
+		sane = (hwhead >= swhead) & (hwhead <= swtail);
+	} else if (swhead > swtail) {
+		/* wrapped around */
+		sane = ((hwhead >= swhead) && (hwhead < cnt)) ||
+			(hwhead <= swtail);
+	} else {
+		/* empty */
+		sane = (hwhead == swhead);
+	}
+
+	if (unlikely(!sane)) {
+		if (use_dmahead) {
+			/* try one more time, directly from the register */
+			use_dmahead = 0;
+			goto retry;
+		}
+		/* assume no progress */
+		hwhead = swhead;
+	}
+
+	return hwhead;
+}
+
+static int qib_sdma_7220_busy(struct qib_pportdata *ppd)
+{
+	u64 hwstatus = qib_read_kreg64(ppd->dd, kr_senddmastatus);
+
+	return (hwstatus & SYM_MASK(SendDmaStatus, ScoreBoardDrainInProg)) ||
+	       (hwstatus & SYM_MASK(SendDmaStatus, AbortInProg)) ||
+	       (hwstatus & SYM_MASK(SendDmaStatus, InternalSDmaEnable)) ||
+	       !(hwstatus & SYM_MASK(SendDmaStatus, ScbEmpty));
+}
+
+/*
+ * Compute the amount of delay before sending the next packet if the
+ * port's send rate differs from the static rate set for the QP.
+ * Since the delay affects this packet but the amount of the delay is
+ * based on the length of the previous packet, use the last delay computed
+ * and save the delay count for this packet to be used next time
+ * we get here.
+ */
+static u32 qib_7220_setpbc_control(struct qib_pportdata *ppd, u32 plen,
+				   u8 srate, u8 vl)
+{
+	u8 snd_mult = ppd->delay_mult;
+	u8 rcv_mult = ib_rate_to_delay[srate];
+	u32 ret = ppd->cpspec->last_delay_mult;
+
+	ppd->cpspec->last_delay_mult = (rcv_mult > snd_mult) ?
+		(plen * (rcv_mult - snd_mult) + 1) >> 1 : 0;
+
+	/* Indicate VL15, if necessary */
+	if (vl == 15)
+		ret |= PBC_7220_VL15_SEND_CTRL;
+	return ret;
+}
+
+static void qib_7220_initvl15_bufs(struct qib_devdata *dd)
+{
+}
+
+static void qib_7220_init_ctxt(struct qib_ctxtdata *rcd)
+{
+	if (!rcd->ctxt) {
+		rcd->rcvegrcnt = IBA7220_KRCVEGRCNT;
+		rcd->rcvegr_tid_base = 0;
+	} else {
+		rcd->rcvegrcnt = rcd->dd->cspec->rcvegrcnt;
+		rcd->rcvegr_tid_base = IBA7220_KRCVEGRCNT +
+			(rcd->ctxt - 1) * rcd->rcvegrcnt;
+	}
+}
+
+static void qib_7220_txchk_change(struct qib_devdata *dd, u32 start,
+				  u32 len, u32 which, struct qib_ctxtdata *rcd)
+{
+	int i;
+	unsigned long flags;
+
+	switch (which) {
+	case TXCHK_CHG_TYPE_KERN:
+		/* see if we need to raise avail update threshold */
+		spin_lock_irqsave(&dd->uctxt_lock, flags);
+		for (i = dd->first_user_ctxt;
+		     dd->cspec->updthresh != dd->cspec->updthresh_dflt
+		     && i < dd->cfgctxts; i++)
+			if (dd->rcd[i] && dd->rcd[i]->subctxt_cnt &&
+			   ((dd->rcd[i]->piocnt / dd->rcd[i]->subctxt_cnt) - 1)
+			   < dd->cspec->updthresh_dflt)
+				break;
+		spin_unlock_irqrestore(&dd->uctxt_lock, flags);
+		if (i == dd->cfgctxts) {
+			spin_lock_irqsave(&dd->sendctrl_lock, flags);
+			dd->cspec->updthresh = dd->cspec->updthresh_dflt;
+			dd->sendctrl &= ~SYM_MASK(SendCtrl, AvailUpdThld);
+			dd->sendctrl |= (dd->cspec->updthresh &
+					 SYM_RMASK(SendCtrl, AvailUpdThld)) <<
+					   SYM_LSB(SendCtrl, AvailUpdThld);
+			spin_unlock_irqrestore(&dd->sendctrl_lock, flags);
+			sendctrl_7220_mod(dd->pport, QIB_SENDCTRL_AVAIL_BLIP);
+		}
+		break;
+	case TXCHK_CHG_TYPE_USER:
+		spin_lock_irqsave(&dd->sendctrl_lock, flags);
+		if (rcd && rcd->subctxt_cnt && ((rcd->piocnt
+			/ rcd->subctxt_cnt) - 1) < dd->cspec->updthresh) {
+			dd->cspec->updthresh = (rcd->piocnt /
+						rcd->subctxt_cnt) - 1;
+			dd->sendctrl &= ~SYM_MASK(SendCtrl, AvailUpdThld);
+			dd->sendctrl |= (dd->cspec->updthresh &
+					SYM_RMASK(SendCtrl, AvailUpdThld))
+					<< SYM_LSB(SendCtrl, AvailUpdThld);
+			spin_unlock_irqrestore(&dd->sendctrl_lock, flags);
+			sendctrl_7220_mod(dd->pport, QIB_SENDCTRL_AVAIL_BLIP);
+		} else
+			spin_unlock_irqrestore(&dd->sendctrl_lock, flags);
+		break;
+	}
+}
+
+static void writescratch(struct qib_devdata *dd, u32 val)
+{
+	qib_write_kreg(dd, kr_scratch, val);
+}
+
+#define VALID_TS_RD_REG_MASK 0xBF
+/**
+ * qib_7220_tempsense_read - read register of temp sensor via TWSI
+ * @dd: the qlogic_ib device
+ * @regnum: register to read from
+ *
+ * returns reg contents (0..255) or < 0 for error
+ */
+static int qib_7220_tempsense_rd(struct qib_devdata *dd, int regnum)
+{
+	int ret;
+	u8 rdata;
+
+	if (regnum > 7) {
+		ret = -EINVAL;
+		goto bail;
+	}
+
+	/* return a bogus value for (the one) register we do not have */
+	if (!((1 << regnum) & VALID_TS_RD_REG_MASK)) {
+		ret = 0;
+		goto bail;
+	}
+
+	ret = mutex_lock_interruptible(&dd->eep_lock);
+	if (ret)
+		goto bail;
+
+	ret = qib_twsi_blk_rd(dd, QIB_TWSI_TEMP_DEV, regnum, &rdata, 1);
+	if (!ret)
+		ret = rdata;
+
+	mutex_unlock(&dd->eep_lock);
+
+	/*
+	 * There are three possibilities here:
+	 * ret is actual value (0..255)
+	 * ret is -ENXIO or -EINVAL from twsi code or this file
+	 * ret is -EINTR from mutex_lock_interruptible.
+	 */
+bail:
+	return ret;
+}
+
+/* Dummy function, as 7220 boards never disable EEPROM Write */
+static int qib_7220_eeprom_wen(struct qib_devdata *dd, int wen)
+{
+	return 1;
+}
+
+/**
+ * qib_init_iba7220_funcs - set up the chip-specific function pointers
+ * @dev: the pci_dev for qlogic_ib device
+ * @ent: pci_device_id struct for this dev
+ *
+ * This is global, and is called directly at init to set up the
+ * chip-specific function pointers for later use.
+ */
+struct qib_devdata *qib_init_iba7220_funcs(struct pci_dev *pdev,
+					   const struct pci_device_id *ent)
+{
+	struct qib_devdata *dd;
+	int ret;
+	u32 boardid, minwidth;
+
+	dd = qib_alloc_devdata(pdev, sizeof(struct qib_chip_specific) +
+		sizeof(struct qib_chippport_specific));
+	if (IS_ERR(dd))
+		goto bail;
+
+	dd->f_bringup_serdes    = qib_7220_bringup_serdes;
+	dd->f_cleanup           = qib_setup_7220_cleanup;
+	dd->f_clear_tids        = qib_7220_clear_tids;
+	dd->f_free_irq          = qib_7220_free_irq;
+	dd->f_get_base_info     = qib_7220_get_base_info;
+	dd->f_get_msgheader     = qib_7220_get_msgheader;
+	dd->f_getsendbuf        = qib_7220_getsendbuf;
+	dd->f_gpio_mod          = gpio_7220_mod;
+	dd->f_eeprom_wen        = qib_7220_eeprom_wen;
+	dd->f_hdrqempty         = qib_7220_hdrqempty;
+	dd->f_ib_updown         = qib_7220_ib_updown;
+	dd->f_init_ctxt         = qib_7220_init_ctxt;
+	dd->f_initvl15_bufs     = qib_7220_initvl15_bufs;
+	dd->f_intr_fallback     = qib_7220_intr_fallback;
+	dd->f_late_initreg      = qib_late_7220_initreg;
+	dd->f_setpbc_control    = qib_7220_setpbc_control;
+	dd->f_portcntr          = qib_portcntr_7220;
+	dd->f_put_tid           = qib_7220_put_tid;
+	dd->f_quiet_serdes      = qib_7220_quiet_serdes;
+	dd->f_rcvctrl           = rcvctrl_7220_mod;
+	dd->f_read_cntrs        = qib_read_7220cntrs;
+	dd->f_read_portcntrs    = qib_read_7220portcntrs;
+	dd->f_reset             = qib_setup_7220_reset;
+	dd->f_init_sdma_regs    = init_sdma_7220_regs;
+	dd->f_sdma_busy         = qib_sdma_7220_busy;
+	dd->f_sdma_gethead      = qib_sdma_7220_gethead;
+	dd->f_sdma_sendctrl     = qib_7220_sdma_sendctrl;
+	dd->f_sdma_set_desc_cnt = qib_sdma_set_7220_desc_cnt;
+	dd->f_sdma_update_tail  = qib_sdma_update_7220_tail;
+	dd->f_sdma_hw_clean_up  = qib_7220_sdma_hw_clean_up;
+	dd->f_sdma_hw_start_up  = qib_7220_sdma_hw_start_up;
+	dd->f_sdma_init_early   = qib_7220_sdma_init_early;
+	dd->f_sendctrl          = sendctrl_7220_mod;
+	dd->f_set_armlaunch     = qib_set_7220_armlaunch;
+	dd->f_set_cntr_sample   = qib_set_cntr_7220_sample;
+	dd->f_iblink_state      = qib_7220_iblink_state;
+	dd->f_ibphys_portstate  = qib_7220_phys_portstate;
+	dd->f_get_ib_cfg        = qib_7220_get_ib_cfg;
+	dd->f_set_ib_cfg        = qib_7220_set_ib_cfg;
+	dd->f_set_ib_loopback   = qib_7220_set_loopback;
+	dd->f_set_intr_state    = qib_7220_set_intr_state;
+	dd->f_setextled         = qib_setup_7220_setextled;
+	dd->f_txchk_change      = qib_7220_txchk_change;
+	dd->f_update_usrhead    = qib_update_7220_usrhead;
+	dd->f_wantpiobuf_intr   = qib_wantpiobuf_7220_intr;
+	dd->f_xgxs_reset        = qib_7220_xgxs_reset;
+	dd->f_writescratch      = writescratch;
+	dd->f_tempsense_rd	= qib_7220_tempsense_rd;
+	/*
+	 * Do remaining pcie setup and save pcie values in dd.
+	 * Any error printing is already done by the init code.
+	 * On return, we have the chip mapped, but chip registers
+	 * are not set up until start of qib_init_7220_variables.
+	 */
+	ret = qib_pcie_ddinit(dd, pdev, ent);
+	if (ret < 0)
+		goto bail_free;
+
+	/* initialize chip-specific variables */
+	ret = qib_init_7220_variables(dd);
+	if (ret)
+		goto bail_cleanup;
+
+	if (qib_mini_init)
+		goto bail;
+
+	boardid = SYM_FIELD(dd->revision, Revision,
+			    BoardID);
+	switch (boardid) {
+	case 0:
+	case 2:
+	case 10:
+	case 12:
+		minwidth = 16; /* x16 capable boards */
+		break;
+	default:
+		minwidth = 8; /* x8 capable boards */
+		break;
+	}
+	if (qib_pcie_params(dd, minwidth, NULL, NULL))
+		qib_dev_err(dd, "Failed to setup PCIe or interrupts; "
+			    "continuing anyway\n");
+
+	/* save IRQ for possible later use */
+	dd->cspec->irq = pdev->irq;
+
+	if (qib_read_kreg64(dd, kr_hwerrstatus) &
+	    QLOGIC_IB_HWE_SERDESPLLFAILED)
+		qib_write_kreg(dd, kr_hwerrclear,
+			       QLOGIC_IB_HWE_SERDESPLLFAILED);
+
+	/* setup interrupt handler (interrupt type handled above) */
+	qib_setup_7220_interrupt(dd);
+	qib_7220_init_hwerrors(dd);
+
+	/* clear diagctrl register, in case diags were running and crashed */
+	qib_write_kreg(dd, kr_hwdiagctrl, 0);
+
+	goto bail;
+
+bail_cleanup:
+	qib_pcie_ddcleanup(dd);
+bail_free:
+	qib_free_devdata(dd);
+	dd = ERR_PTR(ret);
+bail:
+	return dd;
+}