/* * Copyright (c) 2006 IronPort Systems Inc. * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. 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. * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. */ /* $FreeBSD: src/sys/dev/ipmi/ipmi.c,v 1.16 2011/11/07 15:43:11 ed Exp $ */ /* * Copyright 2012, Joyent, Inc. All rights reserved. * Copyright 2013 Nexenta Systems, Inc. All rights reserved. */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "ipmivars.h" /* * Request management. */ /* Allocate a new request with request and reply buffers. */ struct ipmi_request * ipmi_alloc_request(struct ipmi_device *dev, long msgid, uint8_t addr, uint8_t command, size_t requestlen, size_t replylen) { struct ipmi_request *req; req = kmem_zalloc(sizeof (struct ipmi_request) + requestlen + replylen, KM_SLEEP); req->ir_sz = sizeof (struct ipmi_request) + requestlen + replylen; req->ir_owner = dev; req->ir_msgid = msgid; req->ir_addr = addr; req->ir_command = command; if (requestlen) { req->ir_request = (uchar_t *)&req[1]; req->ir_requestlen = requestlen; } if (replylen) { req->ir_reply = (uchar_t *)&req[1] + requestlen; req->ir_replybuflen = replylen; } cv_init(&req->ir_cv, NULL, CV_DEFAULT, NULL); req->ir_status = IRS_ALLOCATED; return (req); } /* Free a request no longer in use. */ void ipmi_free_request(struct ipmi_request *req) { if (req == NULL) return; cv_destroy(&req->ir_cv); kmem_free(req, req->ir_sz); } /* Store a processed request on the appropriate completion queue. */ /*ARGSUSED*/ void ipmi_complete_request(struct ipmi_softc *sc, struct ipmi_request *req) { struct ipmi_device *dev; IPMI_LOCK_ASSERT(sc); if (req->ir_status == IRS_CANCELED) { ASSERT(req->ir_owner == NULL); ipmi_free_request(req); return; } req->ir_status = IRS_COMPLETED; /* * Anonymous requests (from inside the driver) always have a * waiter that we awaken. */ if (req->ir_owner == NULL) { cv_signal(&req->ir_cv); } else { dev = req->ir_owner; TAILQ_INSERT_TAIL(&dev->ipmi_completed_requests, req, ir_link); pollwakeup(dev->ipmi_pollhead, POLLIN | POLLRDNORM); dev->ipmi_status &= ~IPMI_BUSY; if (dev->ipmi_status & IPMI_CLOSING) cv_signal(&dev->ipmi_cv); } } /* * Enqueue an internal driver request and wait until it is completed. */ static int ipmi_submit_driver_request(struct ipmi_softc *sc, struct ipmi_request **preq, int timo) { int error; struct ipmi_request *req = *preq; ASSERT(req->ir_owner == NULL); IPMI_LOCK(sc); error = sc->ipmi_enqueue_request(sc, req); if (error != 0) { IPMI_UNLOCK(sc); return (error); } while (req->ir_status != IRS_COMPLETED && error >= 0) if (timo == 0) cv_wait(&req->ir_cv, &sc->ipmi_lock); else error = cv_timedwait(&req->ir_cv, &sc->ipmi_lock, ddi_get_lbolt() + timo); switch (req->ir_status) { case IRS_QUEUED: TAILQ_REMOVE(&sc->ipmi_pending_requests, req, ir_link); req->ir_status = IRS_CANCELED; error = EWOULDBLOCK; break; case IRS_PROCESSED: req->ir_status = IRS_CANCELED; error = EWOULDBLOCK; *preq = NULL; break; case IRS_COMPLETED: error = req->ir_error; break; default: panic("IPMI: Invalid request status"); break; } IPMI_UNLOCK(sc); return (error); } /* * Helper routine for polled system interfaces that use * ipmi_polled_enqueue_request() to queue requests. This request * waits until there is a pending request and then returns the first * request. If the driver is shutting down, it returns NULL. */ struct ipmi_request * ipmi_dequeue_request(struct ipmi_softc *sc) { struct ipmi_request *req; IPMI_LOCK_ASSERT(sc); while (!sc->ipmi_detaching && TAILQ_EMPTY(&sc->ipmi_pending_requests)) cv_wait(&sc->ipmi_request_added, &sc->ipmi_lock); if (sc->ipmi_detaching) return (NULL); req = TAILQ_FIRST(&sc->ipmi_pending_requests); TAILQ_REMOVE(&sc->ipmi_pending_requests, req, ir_link); req->ir_status = IRS_PROCESSED; if (req->ir_owner != NULL) req->ir_owner->ipmi_status |= IPMI_BUSY; return (req); } int ipmi_polled_enqueue_request(struct ipmi_softc *sc, struct ipmi_request *req) { IPMI_LOCK_ASSERT(sc); TAILQ_INSERT_TAIL(&sc->ipmi_pending_requests, req, ir_link); req->ir_status = IRS_QUEUED; cv_signal(&sc->ipmi_request_added); return (0); } void ipmi_shutdown(struct ipmi_softc *sc) { taskq_destroy(sc->ipmi_kthread); cv_destroy(&sc->ipmi_request_added); mutex_destroy(&sc->ipmi_lock); } boolean_t ipmi_startup(struct ipmi_softc *sc) { struct ipmi_request *req; int error, i; /* Initialize interface-independent state. */ mutex_init(&sc->ipmi_lock, NULL, MUTEX_DEFAULT, NULL); cv_init(&sc->ipmi_request_added, NULL, CV_DEFAULT, NULL); TAILQ_INIT(&sc->ipmi_pending_requests); /* Initialize interface-dependent state. */ error = sc->ipmi_startup(sc); if (error) { cmn_err(CE_WARN, "Failed to initialize interface: %d", error); return (B_FALSE); } /* Send a GET_DEVICE_ID request. */ req = ipmi_alloc_driver_request(IPMI_ADDR(IPMI_APP_REQUEST, 0), IPMI_GET_DEVICE_ID, 0, 15); error = ipmi_submit_driver_request(sc, &req, MAX_TIMEOUT); if (error == EWOULDBLOCK) { cmn_err(CE_WARN, "Timed out waiting for GET_DEVICE_ID"); ipmi_free_request(req); return (B_FALSE); } else if (error) { cmn_err(CE_WARN, "Failed GET_DEVICE_ID: %d", error); ipmi_free_request(req); return (B_FALSE); } else if (req->ir_compcode != 0) { cmn_err(CE_WARN, "Bad completion code for GET_DEVICE_ID: %d", req->ir_compcode); ipmi_free_request(req); return (B_FALSE); } else if (req->ir_replylen < 5) { cmn_err(CE_WARN, "Short reply for GET_DEVICE_ID: %d", req->ir_replylen); ipmi_free_request(req); return (B_FALSE); } cmn_err(CE_CONT, "!device rev. %d, firmware rev. %d.%d%d, " "version %d.%d", req->ir_reply[1] & 0x0f, req->ir_reply[2] & 0x7f, req->ir_reply[3] >> 4, req->ir_reply[3] & 0x0f, req->ir_reply[4] & 0x0f, req->ir_reply[4] >> 4); ipmi_free_request(req); req = ipmi_alloc_driver_request(IPMI_ADDR(IPMI_APP_REQUEST, 0), IPMI_CLEAR_FLAGS, 1, 0); if ((error = ipmi_submit_driver_request(sc, &req, 0)) != 0) { cmn_err(CE_WARN, "Failed to clear IPMI flags: %d\n", error); ipmi_free_request(req); return (B_FALSE); } /* Magic numbers */ if (req->ir_compcode == 0xc0) { cmn_err(CE_NOTE, "!Clear flags is busy"); } if (req->ir_compcode == 0xc1) { cmn_err(CE_NOTE, "!Clear flags illegal"); } ipmi_free_request(req); for (i = 0; i < 8; i++) { req = ipmi_alloc_driver_request(IPMI_ADDR(IPMI_APP_REQUEST, 0), IPMI_GET_CHANNEL_INFO, 1, 0); req->ir_request[0] = (uchar_t)i; if (ipmi_submit_driver_request(sc, &req, 0) != 0) { ipmi_free_request(req); break; } if (req->ir_compcode != 0) { ipmi_free_request(req); break; } ipmi_free_request(req); } cmn_err(CE_CONT, "!number of channels %d", i); /* probe for watchdog */ req = ipmi_alloc_driver_request(IPMI_ADDR(IPMI_APP_REQUEST, 0), IPMI_GET_WDOG, 0, 0); if ((error = ipmi_submit_driver_request(sc, &req, 0)) != 0) { cmn_err(CE_WARN, "Failed to check IPMI watchdog: %d\n", error); ipmi_free_request(req); return (B_FALSE); } if (req->ir_compcode == 0x00) { cmn_err(CE_CONT, "!watchdog supported"); /* * Here is where we could register a watchdog event handler. * See ipmi_wd_event() in the FreeBSD code. */ } ipmi_free_request(req); return (B_TRUE); }