/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright 2008 Sun Microsystems, Inc. All rights reserved.
* Use is subject to license terms.
*/
#pragma ident "%Z%%M% %I% %E% SMI"
#pragma dictionary "SUN4U"
#define AGENT_ID_MASK 0x1f
#define AGENT_ID_SHIFT 24
#define HB_FIT 1000
#define HBUS_FIT 1000
#define PCI_BUS_FIT 500
#define PCI_DEV_FIT 1000
#define CPU_FIT 500
#define PCI_HB_DEV_PATH hostbridge/pcibus/pcidev[32]/pcifn[0]
fru hostbridge;
asru hostbridge;
event fault.io.schizo@hostbridge,
FITrate=HB_FIT, FRU=hostbridge, ASRU=hostbridge;
event error.io.sch.saf.dstat@hostbridge;
event error.io.sch.saf.to@hostbridge;
event error.io.sch.saf.bus@hostbridge;
event error.io.sch.ecc.thresh@hostbridge;
event error.io.pci.device-ta@hostbridge/pcibus/pcidev/pcifn;
event ereport.io.sch.saf.to@hostbridge{within(5s)};
event ereport.io.sch.saf.bus@hostbridge{within(5s)};
event ereport.io.sch.saf.bca@hostbridge{within(5s)};
event ereport.io.sch.saf.bcb@hostbridge{within(5s)};
event ereport.io.sch.saf.ciq-to@hostbridge{within(5s)};
event ereport.io.sch.saf.lpq-to@hostbridge{within(5s)};
event ereport.io.sch.saf.sfpq-to@hostbridge{within(5s)};
event ereport.io.sch.saf.ufpq-to@hostbridge{within(5s)};
event ereport.io.sch.saf.ape@hostbridge{within(5s)};
event ereport.io.sch.ecc.pce@hostbridge{within(5s)};
event ereport.io.sch.ecc.pue@hostbridge{within(5s)};
event ereport.io.sch.ecc.s-pce@hostbridge{within(5s)};
event ereport.io.sch.ecc.s-pue@hostbridge{within(5s)};
event ereport.io.sch.ecc.thresh@hostbridge{within(5s)};
event ereport.io.sch.saf.dstat@hostbridge{within(5s)};
/*
* A faulty Schizo hostbridge may cause:
*
* - bca: bad safari command from PCI block A.
* - bcb: bad safari command from PCI block B.
* - ciq-to: coherent input queue timeout.
* - lpq-to: local PIO queue timeout.
* - sfpq-to: safari foreign PIO queue timeout.
* - ufpq-to: UPA foreign PIO queue timeout.
* - ape: address parity error.
* - pue: PIO uncorrectable error, bad reader.
* - s-pue: secondary PIO UE, bad reader.
* - ecc: multiple PIO CEs.
* - to: safari bus timeout.
* - bus: safari bus error.
* - dstat: errant dstat on incoming data.
*
* The to, bus and dstat errors can cause a target abort to be sent onto the
* pci bus in response to a dma request. We represent this using a device-ta
* error to propagate into the generic pci.esc rules.
*/
prop fault.io.schizo@hostbridge (0)->
ereport.io.sch.saf.bca@hostbridge,
ereport.io.sch.saf.bcb@hostbridge,
ereport.io.sch.saf.ciq-to@hostbridge,
ereport.io.sch.saf.lpq-to@hostbridge,
ereport.io.sch.saf.sfpq-to@hostbridge,
ereport.io.sch.saf.ufpq-to@hostbridge,
ereport.io.sch.saf.ape@hostbridge,
ereport.io.sch.ecc.pue@hostbridge,
ereport.io.sch.ecc.s-pue@hostbridge,
error.io.sch.ecc.thresh@hostbridge,
error.io.sch.saf.to@hostbridge,
error.io.sch.saf.bus@hostbridge,
error.io.sch.saf.dstat@hostbridge;
prop error.io.sch.ecc.thresh@hostbridge (2)->
ereport.io.sch.ecc.thresh@hostbridge,
ereport.io.sch.ecc.pce@hostbridge;
prop error.io.sch.saf.to@hostbridge (2)->
ereport.io.sch.saf.to@hostbridge,
error.io.pci.device-ta@PCI_HB_DEV_PATH;
prop error.io.sch.saf.bus@hostbridge (2)->
ereport.io.sch.saf.bus@hostbridge,
error.io.pci.device-ta@PCI_HB_DEV_PATH;
prop error.io.sch.saf.dstat@hostbridge (1)->
ereport.io.sch.saf.dstat@hostbridge;
prop error.io.sch.saf.dstat@hostbridge (0)->
error.io.pci.device-ta@PCI_HB_DEV_PATH;
engine serd.io.schizo.ecc@hostbridge,
N=3, T=1day, method=persistent,
trip=ereport.io.sch.ecc.thresh@hostbridge;
event upset.io.schizo@hostbridge,
engine=serd.io.schizo.ecc@hostbridge;
/*
* An upset schizo may cause:
*
* - pce: PIO correctable error.
*/
prop upset.io.schizo@hostbridge (0)->
ereport.io.sch.ecc.pce@hostbridge;
fru cpu;
event fault.io.datapath@cpu, FITrate=CPU_FIT, FRU=cpu, retire=0;
event error.io.cpu.ecc.thresh@cpu;
event ereport.io.sch.saf.ssm-dis@hostbridge{within(5s)};
event ereport.io.sch.saf.cpu0-par@hostbridge{within(5s)};
event ereport.io.sch.saf.cpu0-bidi@hostbridge{within(5s)};
event ereport.io.sch.saf.cpu1-par@hostbridge{within(5s)};
event ereport.io.sch.saf.cpu1-bidi@hostbridge{within(5s)};
/*
* A faulty xcal CPU[0] may cause:
*
* - cpu0-par: parity error on the unidirectional signals.
* - cpu0-bidi: parity error on the bi-directional signals.
*/
prop fault.io.datapath@cpu[0] (0)->
ereport.io.sch.saf.cpu0-par@hostbridge,
ereport.io.sch.saf.cpu0-bidi@hostbridge;
/*
* A faulty xcal CPU[1] may cause:
*
* - cpu1-par: parity error on the unidirectional signals.
* - cpu1-bidi: parity error on the bidirectional signals.
*/
prop fault.io.datapath@cpu[1] (0)->
ereport.io.sch.saf.cpu1-par@hostbridge,
ereport.io.sch.saf.cpu1-bidi@hostbridge;
/*
* A faulty CPU may cause:
*
* - to: safari bus timeout.
* - bus: safari bus error.
* - dstat: incorrect dstat sent to hostbridge.
* - ssm-dis: ssm command sent to hostbridge when not enabled.
* - ape: safari address parity error.
* - pue: PIO uncorrectable error.
* - ecc: multiple PIO CEs.
*/
prop fault.io.datapath@cpu (0)->
error.io.sch.saf.to@hostbridge,
error.io.sch.saf.bus@hostbridge,
error.io.sch.saf.dstat@hostbridge,
ereport.io.sch.saf.ssm-dis@hostbridge,
ereport.io.sch.saf.ape@hostbridge;
prop fault.io.datapath@cpu[cpuid] (0)->
ereport.io.sch.ecc.pue@hostbridge
{((payloadprop("ecc-afsr") >> AGENT_ID_SHIFT) & AGENT_ID_MASK) == cpuid};
prop fault.io.datapath@cpu (0)->
error.io.cpu.ecc.thresh@cpu;
prop error.io.cpu.ecc.thresh@cpu (1)->
ereport.io.sch.ecc.thresh@hostbridge<>;
prop error.io.cpu.ecc.thresh@cpu[cpuid] (1)->
ereport.io.sch.ecc.pce@hostbridge<>
{((payloadprop("ecc-afsr") >> AGENT_ID_SHIFT) & AGENT_ID_MASK) == cpuid};
event fault.io.hbus@hostbridge,
FITrate=HBUS_FIT, FRU=hostbridge, ASRU=hostbridge;
/*
* A faulty host bus may cause:
*
* - ape: address parity error.
* - cpu0-par: parity error on the unidirectional signals.
* - cpu0-bidi: parity error on the bidirectional signals.
* - cpu1-par: parity error on the unidirectional signals.
* - cpu1-bidi: parity error on the bidirectional signals.
* - pue: PIO uncorrectable error.
* - s-pue: secondary PIO UE.
* - ecc: multiple PIO CEs.
*/
prop fault.io.hbus@hostbridge (0)->
ereport.io.sch.saf.ape@hostbridge,
ereport.io.sch.saf.cpu0-par@hostbridge,
ereport.io.sch.saf.cpu0-bidi@hostbridge,
ereport.io.sch.saf.cpu1-par@hostbridge,
ereport.io.sch.saf.cpu1-bidi@hostbridge,
ereport.io.sch.ecc.pue@hostbridge,
ereport.io.sch.ecc.s-pue@hostbridge,
error.io.sch.ecc.thresh@hostbridge;
/*
* A bad request from a downstream device/driver may cause
*
* - um: safari unmapped address error.
* - mmu: a iommu translation error.
*/
event error.io.pci.badreq-pw-u@hostbridge/pcibus/pcidev/pcifn;
event error.io.pci.badreq-drw-u@hostbridge/pcibus/pcidev/pcifn;
event ereport.io.pci.rserr@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.mmu@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.saf.um@hostbridge{within(5s)};
prop error.io.pci.badreq-pw-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.saf.um@hostbridge;
prop error.io.pci.badreq-pw-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.mmu@PCI_HB_DEV_PATH;
prop error.io.pci.badreq-drw-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.saf.um@hostbridge;
prop error.io.pci.badreq-drw-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.mmu@PCI_HB_DEV_PATH;
fru pcibus;
asru pcibus;
event fault.io.pci.bus@hostbridge/pcibus,
FITrate=PCI_BUS_FIT, FRU=pcibus, ASRU=pcibus;
event ereport.io.sch.bu@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.s-bu@hostbridge/pcibus/pcidev/pcifn{within(5s)};
/*
* A faulty PCI bus may cause:
*
* - bu: PCI bus unusable error.
* - s-bu: secondary PCI bus unusable error.
*/
prop fault.io.pci.bus@hostbridge/pcibus (0)->
ereport.io.sch.bu@PCI_HB_DEV_PATH,
ereport.io.sch.s-bu@PCI_HB_DEV_PATH;
fru pcibus/pcidev;
asru pcibus/pcidev/pcifn;
event fault.io.pci.device-interr@hostbridge/pcibus/pcidev/pcifn,
FITrate=PCI_DEV_FIT, FRU=pcibus/pcidev, ASRU=pcibus/pcidev/pcifn;
event fault.io.pci.device-interr@pcibus/pcidev/pcifn,
FITrate=PCI_DEV_FIT, FRU=pcibus/pcidev, ASRU=pcibus/pcidev/pcifn;
event error.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn;
event error.io.sch.pbm.rl@pcibus/pcidev/pcifn;
event error.io.sch.pbm.rl@pcibus/pcidev/pcifn/pcibus/pcidev/pcifn;
event error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn;
event error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn/pcibus/pcidev/pcifn;
event error.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn;
event error.io.sch.pbm.target-tto@hostbridge/pcibus/pcidev/pcifn;
event error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn;
event error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn/pcibus/pcidev/pcifn;
event error.sch.cpu.berr@cpu;
event error.io.pci.ma-u@hostbridge/pcibus/pcidev/pcifn;
event error.io.pci.perr-pw-u@hostbridge/pcibus/pcidev/pcifn;
event error.io.pci.perr-dw-u@hostbridge/pcibus/pcidev/pcifn;
event error.io.pci.dpdata-dr-u@hostbridge/pcibus/pcidev/pcifn;
event error.io.pci.ta-u@hostbridge/pcibus/pcidev/pcifn;
event error.io.pci.serr-u@hostbridge/pcibus/pcidev/pcifn;
event error.io.pci.retry-to-d@hostbridge/pcibus/pcidev/pcifn;
event ereport.io.sch.sbh@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.pbm.s-rl@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.pbm.s-tto@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.pbm.s-ma@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.pbm.s-rta@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.pbm.s-mdpe@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.pbm.target-rl@pcibus/pcidev/pcifn{within(5s)};
event ereport.io.sch.pbm.target-tto@pcibus/pcidev/pcifn{within(5s)};
event ereport.io.pci.sserr@hostbridge/pcibus/pcidev/pcifn{within(5s)};
event ereport.cpu.ultraSPARC-III.berr@cpu{within(5s)};
event ereport.cpu.ultraSPARC-IIIplus.berr@cpu{within(5s)};
event ereport.cpu.ultraSPARC-IV.berr@cpu{within(5s)};
event ereport.cpu.ultraSPARC-IVplus.berr@cpu{within(5s)};
/*
* A faulty PCI device may cause:
*
* - sbh: a streaming byte hole error.
* - rl: it to exceed the number retriesfor a transaction.
* - tto: it to not assert trdy# within the alloted timeout.
*
* For rl and tto, there may be a target- ereport on a child device. For rl,
* there may also be an associated dto - the retry-to-d error propagates into
* the pci.esc rules to handle this.
*/
prop fault.io.pci.device-interr@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.sbh@PCI_HB_DEV_PATH;
prop fault.io.pci.device-interr@pcibus/pcidev[fromdev]/pcifn (0)->
error.io.sch.pbm.rl@pcibus/pcidev<todev>/pcifn {
fromdev == todev && fromdev != 32 },
error.io.sch.pbm.target-rl@pcibus/pcidev<todev>/pcifn {
fromdev == todev && fromdev != 32 };
prop error.io.sch.pbm.rl@pcibus/pcidev/pcifn/pcibus/pcidev/pcifn (1)->
error.io.sch.pbm.rl@pcibus/pcidev/pcifn;
prop error.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn (1)->
ereport.io.sch.pbm.rl@PCI_HB_DEV_PATH,
ereport.io.sch.pbm.s-rl@PCI_HB_DEV_PATH;
prop error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn (1)->
error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn/pcibus<>/pcidev<>/pcifn<>;
prop error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn (0)->
ereport.io.sch.pbm.target-rl@pcibus/pcidev/pcifn;
prop error.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn (0)->
error.io.pci.retry-to-d@hostbridge/pcibus/pcidev/pcifn;
prop error.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn (0)->
error.sch.cpu.berr@cpu;
prop fault.io.pci.device-interr@hostbridge/pcibus/pcidev[fromdev]/pcifn (0)->
error.io.sch.pbm.tto@hostbridge/pcibus/pcidev<todev>/pcifn {
fromdev == todev && fromdev != 32 };
prop error.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn (1)->
ereport.io.sch.pbm.tto@PCI_HB_DEV_PATH,
ereport.io.sch.pbm.s-tto@PCI_HB_DEV_PATH;
prop error.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn (1)->
error.io.sch.pbm.target-tto@hostbridge/pcibus/pcidev/pcifn;
prop error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn (0)->
ereport.io.sch.pbm.target-tto@pcibus/pcidev/pcifn;
prop error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn (1)->
error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn/pcibus<>/pcidev<>/pcifn<>;
prop error.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn (1)->
ereport.io.sch.bu@PCI_HB_DEV_PATH;
/*
* Need to add the following schizo specific propagations to complete the PCI
* fault tree. These are to allow propagations to secondary errors and cpu
* bus errors, and to represent the way the chip can raise both rserr and sserr
* on detection of SERR#
*/
prop error.io.pci.ma-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.pbm.s-ma@PCI_HB_DEV_PATH;
prop error.io.pci.ta-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.pbm.s-rta@PCI_HB_DEV_PATH;
prop error.io.pci.perr-pw-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.pbm.s-mdpe@PCI_HB_DEV_PATH;
prop error.io.pci.perr-dw-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.pbm.s-mdpe@PCI_HB_DEV_PATH;
prop error.io.pci.dpdata-dr-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.sch.pbm.s-mdpe@PCI_HB_DEV_PATH;
prop error.io.pci.ta-u@hostbridge/pcibus/pcidev/pcifn (0)->
error.sch.cpu.berr@cpu;
prop error.io.pci.dpdata-dr-u@hostbridge/pcibus/pcidev/pcifn (0)->
error.sch.cpu.berr@cpu;
prop error.io.pci.ma-u@hostbridge/pcibus/pcidev/pcifn (0)->
error.sch.cpu.berr@cpu;
prop error.io.pci.serr-u@hostbridge/pcibus/pcidev/pcifn (1)->
ereport.io.pci.rserr@PCI_HB_DEV_PATH;
prop error.io.pci.serr-u@hostbridge/pcibus/pcidev/pcifn (0)->
ereport.io.pci.sserr@PCI_HB_DEV_PATH;
prop error.sch.cpu.berr@cpu (1)->
ereport.cpu.ultraSPARC-III.berr@cpu,
ereport.cpu.ultraSPARC-IIIplus.berr@cpu,
ereport.cpu.ultraSPARC-IV.berr@cpu,
ereport.cpu.ultraSPARC-IVplus.berr@cpu;
event error.io.sch.ecc.drue@hostbridge;
event ereport.io.sch.ecc.drue@hostbridge{within(5s)};
event ereport.io.sch.nodiag@hostbridge;
/*
* Upset used to hide ereports that can not be currently diagnosed.
*
* The drue error can cause a target abort to be sent onto the
* pci bus in response to a dma request. We represent this using a device-ta
* error to propagate into the generic pci.esc rules.
*/
engine serd.io.sch.nodiag@hostbridge,
N=1000, T=1hour, method=persistent,
trip=ereport.io.sch.nodiag@hostbridge;
event upset.io.sch.nodiag@hostbridge,
engine=serd.io.sch.nodiag@hostbridge;
prop upset.io.sch.nodiag@hostbridge (0)->
ereport.io.sch.ecc.s-pce@hostbridge,
error.io.sch.ecc.drue@hostbridge,
ereport.io.sch.nodiag@hostbridge;
prop error.io.sch.ecc.drue@hostbridge (1)->
ereport.io.sch.ecc.drue@hostbridge;
prop error.io.sch.ecc.drue@hostbridge (0)->
error.io.pci.device-ta@PCI_HB_DEV_PATH;