4 * The contents of this file are subject to the terms of the
5 * Common Development and Distribution License (the "License").
6 * You may not use this file except in compliance with the License.
8 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
9 * or http://www.opensolaris.org/os/licensing.
10 * See the License for the specific language governing permissions
11 * and limitations under the License.
13 * When distributing Covered Code, include this CDDL HEADER in each
14 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
15 * If applicable, add the following below this CDDL HEADER, with the
16 * fields enclosed by brackets "[]" replaced with your own identifying
17 * information: Portions Copyright [yyyy] [name of copyright owner]
22 * Copyright 2008 Sun Microsystems, Inc. All rights reserved.
23 * Use is subject to license terms.
26 #pragma ident "%Z%%M% %I% %E% SMI"
28 #pragma dictionary "SUN4U"
30 #define AGENT_ID_MASK 0x1f
31 #define AGENT_ID_SHIFT 24
35 #define PCI_BUS_FIT 500
36 #define PCI_DEV_FIT 1000
39 #define PCI_HB_DEV_PATH hostbridge/pcibus/pcidev[32]/pcifn[0]
44 event fault.io.schizo@hostbridge,
45 FITrate=HB_FIT, FRU=hostbridge, ASRU=hostbridge;
47 event error.io.sch.saf.dstat@hostbridge;
48 event error.io.sch.saf.to@hostbridge;
49 event error.io.sch.saf.bus@hostbridge;
50 event error.io.sch.ecc.thresh@hostbridge;
51 event error.io.pci.device-ta@hostbridge/pcibus/pcidev/pcifn;
53 event ereport.io.sch.saf.to@hostbridge{within(5s)};
54 event ereport.io.sch.saf.bus@hostbridge{within(5s)};
55 event ereport.io.sch.saf.bca@hostbridge{within(5s)};
56 event ereport.io.sch.saf.bcb@hostbridge{within(5s)};
57 event ereport.io.sch.saf.ciq-to@hostbridge{within(5s)};
58 event ereport.io.sch.saf.lpq-to@hostbridge{within(5s)};
59 event ereport.io.sch.saf.sfpq-to@hostbridge{within(5s)};
60 event ereport.io.sch.saf.ufpq-to@hostbridge{within(5s)};
61 event ereport.io.sch.saf.ape@hostbridge{within(5s)};
62 event ereport.io.sch.ecc.pce@hostbridge{within(5s)};
63 event ereport.io.sch.ecc.pue@hostbridge{within(5s)};
64 event ereport.io.sch.ecc.s-pce@hostbridge{within(5s)};
65 event ereport.io.sch.ecc.s-pue@hostbridge{within(5s)};
66 event ereport.io.sch.ecc.thresh@hostbridge{within(5s)};
67 event ereport.io.sch.saf.dstat@hostbridge{within(5s)};
70 * A faulty Schizo hostbridge may cause:
72 * - bca: bad safari command from PCI block A.
73 * - bcb: bad safari command from PCI block B.
74 * - ciq-to: coherent input queue timeout.
75 * - lpq-to: local PIO queue timeout.
76 * - sfpq-to: safari foreign PIO queue timeout.
77 * - ufpq-to: UPA foreign PIO queue timeout.
78 * - ape: address parity error.
79 * - pue: PIO uncorrectable error, bad reader.
80 * - s-pue: secondary PIO UE, bad reader.
81 * - ecc: multiple PIO CEs.
82 * - to: safari bus timeout.
83 * - bus: safari bus error.
84 * - dstat: errant dstat on incoming data.
86 * The to, bus and dstat errors can cause a target abort to be sent onto the
87 * pci bus in response to a dma request. We represent this using a device-ta
88 * error to propagate into the generic pci.esc rules.
90 prop fault.io.schizo@hostbridge (0)->
91 ereport.io.sch.saf.bca@hostbridge,
92 ereport.io.sch.saf.bcb@hostbridge,
93 ereport.io.sch.saf.ciq-to@hostbridge,
94 ereport.io.sch.saf.lpq-to@hostbridge,
95 ereport.io.sch.saf.sfpq-to@hostbridge,
96 ereport.io.sch.saf.ufpq-to@hostbridge,
97 ereport.io.sch.saf.ape@hostbridge,
98 ereport.io.sch.ecc.pue@hostbridge,
99 ereport.io.sch.ecc.s-pue@hostbridge,
100 error.io.sch.ecc.thresh@hostbridge,
101 error.io.sch.saf.to@hostbridge,
102 error.io.sch.saf.bus@hostbridge,
103 error.io.sch.saf.dstat@hostbridge;
105 prop error.io.sch.ecc.thresh@hostbridge (2)->
106 ereport.io.sch.ecc.thresh@hostbridge,
107 ereport.io.sch.ecc.pce@hostbridge;
109 prop error.io.sch.saf.to@hostbridge (2)->
110 ereport.io.sch.saf.to@hostbridge,
111 error.io.pci.device-ta@PCI_HB_DEV_PATH;
113 prop error.io.sch.saf.bus@hostbridge (2)->
114 ereport.io.sch.saf.bus@hostbridge,
115 error.io.pci.device-ta@PCI_HB_DEV_PATH;
117 prop error.io.sch.saf.dstat@hostbridge (1)->
118 ereport.io.sch.saf.dstat@hostbridge;
120 prop error.io.sch.saf.dstat@hostbridge (0)->
121 error.io.pci.device-ta@PCI_HB_DEV_PATH;
123 engine serd.io.schizo.ecc@hostbridge,
124 N=3, T=1day, method=persistent,
125 trip=ereport.io.sch.ecc.thresh@hostbridge;
127 event upset.io.schizo@hostbridge,
128 engine=serd.io.schizo.ecc@hostbridge;
131 * An upset schizo may cause:
133 * - pce: PIO correctable error.
135 prop upset.io.schizo@hostbridge (0)->
136 ereport.io.sch.ecc.pce@hostbridge;
140 event fault.io.datapath@cpu, FITrate=CPU_FIT, FRU=cpu, retire=0;
142 event error.io.cpu.ecc.thresh@cpu;
143 event ereport.io.sch.saf.ssm-dis@hostbridge{within(5s)};
144 event ereport.io.sch.saf.cpu0-par@hostbridge{within(5s)};
145 event ereport.io.sch.saf.cpu0-bidi@hostbridge{within(5s)};
146 event ereport.io.sch.saf.cpu1-par@hostbridge{within(5s)};
147 event ereport.io.sch.saf.cpu1-bidi@hostbridge{within(5s)};
150 * A faulty xcal CPU[0] may cause:
152 * - cpu0-par: parity error on the unidirectional signals.
153 * - cpu0-bidi: parity error on the bi-directional signals.
155 prop fault.io.datapath@cpu[0] (0)->
156 ereport.io.sch.saf.cpu0-par@hostbridge,
157 ereport.io.sch.saf.cpu0-bidi@hostbridge;
160 * A faulty xcal CPU[1] may cause:
162 * - cpu1-par: parity error on the unidirectional signals.
163 * - cpu1-bidi: parity error on the bidirectional signals.
165 prop fault.io.datapath@cpu[1] (0)->
166 ereport.io.sch.saf.cpu1-par@hostbridge,
167 ereport.io.sch.saf.cpu1-bidi@hostbridge;
170 * A faulty CPU may cause:
172 * - to: safari bus timeout.
173 * - bus: safari bus error.
174 * - dstat: incorrect dstat sent to hostbridge.
175 * - ssm-dis: ssm command sent to hostbridge when not enabled.
176 * - ape: safari address parity error.
177 * - pue: PIO uncorrectable error.
178 * - ecc: multiple PIO CEs.
180 prop fault.io.datapath@cpu (0)->
181 error.io.sch.saf.to@hostbridge,
182 error.io.sch.saf.bus@hostbridge,
183 error.io.sch.saf.dstat@hostbridge,
184 ereport.io.sch.saf.ssm-dis@hostbridge,
185 ereport.io.sch.saf.ape@hostbridge;
187 prop fault.io.datapath@cpu[cpuid] (0)->
188 ereport.io.sch.ecc.pue@hostbridge
189 {((payloadprop("ecc-afsr") >> AGENT_ID_SHIFT) & AGENT_ID_MASK) == cpuid};
191 prop fault.io.datapath@cpu (0)->
192 error.io.cpu.ecc.thresh@cpu;
194 prop error.io.cpu.ecc.thresh@cpu (1)->
195 ereport.io.sch.ecc.thresh@hostbridge<>;
197 prop error.io.cpu.ecc.thresh@cpu[cpuid] (1)->
198 ereport.io.sch.ecc.pce@hostbridge<>
199 {((payloadprop("ecc-afsr") >> AGENT_ID_SHIFT) & AGENT_ID_MASK) == cpuid};
201 event fault.io.hbus@hostbridge,
202 FITrate=HBUS_FIT, FRU=hostbridge, ASRU=hostbridge;
205 * A faulty host bus may cause:
207 * - ape: address parity error.
208 * - cpu0-par: parity error on the unidirectional signals.
209 * - cpu0-bidi: parity error on the bidirectional signals.
210 * - cpu1-par: parity error on the unidirectional signals.
211 * - cpu1-bidi: parity error on the bidirectional signals.
212 * - pue: PIO uncorrectable error.
213 * - s-pue: secondary PIO UE.
214 * - ecc: multiple PIO CEs.
216 prop fault.io.hbus@hostbridge (0)->
217 ereport.io.sch.saf.ape@hostbridge,
218 ereport.io.sch.saf.cpu0-par@hostbridge,
219 ereport.io.sch.saf.cpu0-bidi@hostbridge,
220 ereport.io.sch.saf.cpu1-par@hostbridge,
221 ereport.io.sch.saf.cpu1-bidi@hostbridge,
222 ereport.io.sch.ecc.pue@hostbridge,
223 ereport.io.sch.ecc.s-pue@hostbridge,
224 error.io.sch.ecc.thresh@hostbridge;
227 * A bad request from a downstream device/driver may cause
229 * - um: safari unmapped address error.
230 * - mmu: a iommu translation error.
232 event error.io.pci.badreq-pw-u@hostbridge/pcibus/pcidev/pcifn;
233 event error.io.pci.badreq-drw-u@hostbridge/pcibus/pcidev/pcifn;
235 event ereport.io.pci.rserr@hostbridge/pcibus/pcidev/pcifn{within(5s)};
236 event ereport.io.sch.mmu@hostbridge/pcibus/pcidev/pcifn{within(5s)};
237 event ereport.io.sch.saf.um@hostbridge{within(5s)};
239 prop error.io.pci.badreq-pw-u@hostbridge/pcibus/pcidev/pcifn (0)->
240 ereport.io.sch.saf.um@hostbridge;
242 prop error.io.pci.badreq-pw-u@hostbridge/pcibus/pcidev/pcifn (0)->
243 ereport.io.sch.mmu@PCI_HB_DEV_PATH;
245 prop error.io.pci.badreq-drw-u@hostbridge/pcibus/pcidev/pcifn (0)->
246 ereport.io.sch.saf.um@hostbridge;
248 prop error.io.pci.badreq-drw-u@hostbridge/pcibus/pcidev/pcifn (0)->
249 ereport.io.sch.mmu@PCI_HB_DEV_PATH;
254 event fault.io.pci.bus@hostbridge/pcibus,
255 FITrate=PCI_BUS_FIT, FRU=pcibus, ASRU=pcibus;
257 event ereport.io.sch.bu@hostbridge/pcibus/pcidev/pcifn{within(5s)};
258 event ereport.io.sch.s-bu@hostbridge/pcibus/pcidev/pcifn{within(5s)};
261 * A faulty PCI bus may cause:
263 * - bu: PCI bus unusable error.
264 * - s-bu: secondary PCI bus unusable error.
266 prop fault.io.pci.bus@hostbridge/pcibus (0)->
267 ereport.io.sch.bu@PCI_HB_DEV_PATH,
268 ereport.io.sch.s-bu@PCI_HB_DEV_PATH;
271 asru pcibus/pcidev/pcifn;
273 event fault.io.pci.device-interr@hostbridge/pcibus/pcidev/pcifn,
274 FITrate=PCI_DEV_FIT, FRU=pcibus/pcidev, ASRU=pcibus/pcidev/pcifn;
276 event fault.io.pci.device-interr@pcibus/pcidev/pcifn,
277 FITrate=PCI_DEV_FIT, FRU=pcibus/pcidev, ASRU=pcibus/pcidev/pcifn;
279 event error.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn;
280 event error.io.sch.pbm.rl@pcibus/pcidev/pcifn;
281 event error.io.sch.pbm.rl@pcibus/pcidev/pcifn/pcibus/pcidev/pcifn;
282 event error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn;
283 event error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn/pcibus/pcidev/pcifn;
284 event error.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn;
285 event error.io.sch.pbm.target-tto@hostbridge/pcibus/pcidev/pcifn;
286 event error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn;
287 event error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn/pcibus/pcidev/pcifn;
288 event error.sch.cpu.berr@cpu;
289 event error.io.pci.ma-u@hostbridge/pcibus/pcidev/pcifn;
290 event error.io.pci.perr-pw-u@hostbridge/pcibus/pcidev/pcifn;
291 event error.io.pci.perr-dw-u@hostbridge/pcibus/pcidev/pcifn;
292 event error.io.pci.dpdata-dr-u@hostbridge/pcibus/pcidev/pcifn;
293 event error.io.pci.ta-u@hostbridge/pcibus/pcidev/pcifn;
294 event error.io.pci.serr-u@hostbridge/pcibus/pcidev/pcifn;
295 event error.io.pci.retry-to-d@hostbridge/pcibus/pcidev/pcifn;
297 event ereport.io.sch.sbh@hostbridge/pcibus/pcidev/pcifn{within(5s)};
298 event ereport.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn{within(5s)};
299 event ereport.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn{within(5s)};
300 event ereport.io.sch.pbm.s-rl@hostbridge/pcibus/pcidev/pcifn{within(5s)};
301 event ereport.io.sch.pbm.s-tto@hostbridge/pcibus/pcidev/pcifn{within(5s)};
302 event ereport.io.sch.pbm.s-ma@hostbridge/pcibus/pcidev/pcifn{within(5s)};
303 event ereport.io.sch.pbm.s-rta@hostbridge/pcibus/pcidev/pcifn{within(5s)};
304 event ereport.io.sch.pbm.s-mdpe@hostbridge/pcibus/pcidev/pcifn{within(5s)};
305 event ereport.io.sch.pbm.target-rl@pcibus/pcidev/pcifn{within(5s)};
306 event ereport.io.sch.pbm.target-tto@pcibus/pcidev/pcifn{within(5s)};
307 event ereport.io.pci.sserr@hostbridge/pcibus/pcidev/pcifn{within(5s)};
308 event ereport.cpu.ultraSPARC-III.berr@cpu{within(5s)};
309 event ereport.cpu.ultraSPARC-IIIplus.berr@cpu{within(5s)};
310 event ereport.cpu.ultraSPARC-IV.berr@cpu{within(5s)};
311 event ereport.cpu.ultraSPARC-IVplus.berr@cpu{within(5s)};
314 * A faulty PCI device may cause:
316 * - sbh: a streaming byte hole error.
317 * - rl: it to exceed the number retriesfor a transaction.
318 * - tto: it to not assert trdy# within the alloted timeout.
320 * For rl and tto, there may be a target- ereport on a child device. For rl,
321 * there may also be an associated dto - the retry-to-d error propagates into
322 * the pci.esc rules to handle this.
324 prop fault.io.pci.device-interr@hostbridge/pcibus/pcidev/pcifn (0)->
325 ereport.io.sch.sbh@PCI_HB_DEV_PATH;
327 prop fault.io.pci.device-interr@pcibus/pcidev[fromdev]/pcifn (0)->
328 error.io.sch.pbm.rl@pcibus/pcidev<todev>/pcifn {
329 fromdev == todev && fromdev != 32 },
330 error.io.sch.pbm.target-rl@pcibus/pcidev<todev>/pcifn {
331 fromdev == todev && fromdev != 32 };
333 prop error.io.sch.pbm.rl@pcibus/pcidev/pcifn/pcibus/pcidev/pcifn (1)->
334 error.io.sch.pbm.rl@pcibus/pcidev/pcifn;
336 prop error.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn (1)->
337 ereport.io.sch.pbm.rl@PCI_HB_DEV_PATH,
338 ereport.io.sch.pbm.s-rl@PCI_HB_DEV_PATH;
340 prop error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn (1)->
341 error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn/pcibus<>/pcidev<>/pcifn<>;
343 prop error.io.sch.pbm.target-rl@pcibus/pcidev/pcifn (0)->
344 ereport.io.sch.pbm.target-rl@pcibus/pcidev/pcifn;
346 prop error.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn (0)->
347 error.io.pci.retry-to-d@hostbridge/pcibus/pcidev/pcifn;
349 prop error.io.sch.pbm.rl@hostbridge/pcibus/pcidev/pcifn (0)->
350 error.sch.cpu.berr@cpu;
352 prop fault.io.pci.device-interr@hostbridge/pcibus/pcidev[fromdev]/pcifn (0)->
353 error.io.sch.pbm.tto@hostbridge/pcibus/pcidev<todev>/pcifn {
354 fromdev == todev && fromdev != 32 };
356 prop error.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn (1)->
357 ereport.io.sch.pbm.tto@PCI_HB_DEV_PATH,
358 ereport.io.sch.pbm.s-tto@PCI_HB_DEV_PATH;
360 prop error.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn (1)->
361 error.io.sch.pbm.target-tto@hostbridge/pcibus/pcidev/pcifn;
363 prop error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn (0)->
364 ereport.io.sch.pbm.target-tto@pcibus/pcidev/pcifn;
366 prop error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn (1)->
367 error.io.sch.pbm.target-tto@pcibus/pcidev/pcifn/pcibus<>/pcidev<>/pcifn<>;
369 prop error.io.sch.pbm.tto@hostbridge/pcibus/pcidev/pcifn (1)->
370 ereport.io.sch.bu@PCI_HB_DEV_PATH;
373 * Need to add the following schizo specific propagations to complete the PCI
374 * fault tree. These are to allow propagations to secondary errors and cpu
375 * bus errors, and to represent the way the chip can raise both rserr and sserr
376 * on detection of SERR#
378 prop error.io.pci.ma-u@hostbridge/pcibus/pcidev/pcifn (0)->
379 ereport.io.sch.pbm.s-ma@PCI_HB_DEV_PATH;
381 prop error.io.pci.ta-u@hostbridge/pcibus/pcidev/pcifn (0)->
382 ereport.io.sch.pbm.s-rta@PCI_HB_DEV_PATH;
384 prop error.io.pci.perr-pw-u@hostbridge/pcibus/pcidev/pcifn (0)->
385 ereport.io.sch.pbm.s-mdpe@PCI_HB_DEV_PATH;
387 prop error.io.pci.perr-dw-u@hostbridge/pcibus/pcidev/pcifn (0)->
388 ereport.io.sch.pbm.s-mdpe@PCI_HB_DEV_PATH;
390 prop error.io.pci.dpdata-dr-u@hostbridge/pcibus/pcidev/pcifn (0)->
391 ereport.io.sch.pbm.s-mdpe@PCI_HB_DEV_PATH;
393 prop error.io.pci.ta-u@hostbridge/pcibus/pcidev/pcifn (0)->
394 error.sch.cpu.berr@cpu;
396 prop error.io.pci.dpdata-dr-u@hostbridge/pcibus/pcidev/pcifn (0)->
397 error.sch.cpu.berr@cpu;
399 prop error.io.pci.ma-u@hostbridge/pcibus/pcidev/pcifn (0)->
400 error.sch.cpu.berr@cpu;
402 prop error.io.pci.serr-u@hostbridge/pcibus/pcidev/pcifn (1)->
403 ereport.io.pci.rserr@PCI_HB_DEV_PATH;
405 prop error.io.pci.serr-u@hostbridge/pcibus/pcidev/pcifn (0)->
406 ereport.io.pci.sserr@PCI_HB_DEV_PATH;
408 prop error.sch.cpu.berr@cpu (1)->
409 ereport.cpu.ultraSPARC-III.berr@cpu,
410 ereport.cpu.ultraSPARC-IIIplus.berr@cpu,
411 ereport.cpu.ultraSPARC-IV.berr@cpu,
412 ereport.cpu.ultraSPARC-IVplus.berr@cpu;
414 event error.io.sch.ecc.drue@hostbridge;
415 event ereport.io.sch.ecc.drue@hostbridge{within(5s)};
416 event ereport.io.sch.nodiag@hostbridge;
419 * Upset used to hide ereports that can not be currently diagnosed.
421 * The drue error can cause a target abort to be sent onto the
422 * pci bus in response to a dma request. We represent this using a device-ta
423 * error to propagate into the generic pci.esc rules.
425 engine serd.io.sch.nodiag@hostbridge,
426 N=1000, T=1hour, method=persistent,
427 trip=ereport.io.sch.nodiag@hostbridge;
429 event upset.io.sch.nodiag@hostbridge,
430 engine=serd.io.sch.nodiag@hostbridge;
432 prop upset.io.sch.nodiag@hostbridge (0)->
433 ereport.io.sch.ecc.s-pce@hostbridge,
434 error.io.sch.ecc.drue@hostbridge,
435 ereport.io.sch.nodiag@hostbridge;
437 prop error.io.sch.ecc.drue@hostbridge (1)->
438 ereport.io.sch.ecc.drue@hostbridge;
440 prop error.io.sch.ecc.drue@hostbridge (0)->
441 error.io.pci.device-ta@PCI_HB_DEV_PATH;