X-Git-Url: http://ftp.safe.ca/?a=blobdiff_plain;f=drivers%2Fedac%2Fedac_mc.c;h=3630308e7b811a66f398193eae71cc12f5d7383f;hb=67166af4abc11d9c0deb497ebe0b562f69c71942;hp=7c5fdd1ee972df4669c0674e651c62580c22e178;hpb=20bcb7a81dee21bfa3408f03f46b2891c9b5c84b;p=safe%2Fjmp%2Flinux-2.6 diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c index 7c5fdd1..3630308 100644 --- a/drivers/edac/edac_mc.c +++ b/drivers/edac/edac_mc.c @@ -34,10 +34,9 @@ #include "edac_core.h" #include "edac_module.h" - /* lock to memory controller's control array */ static DEFINE_MUTEX(mem_ctls_mutex); -static struct list_head mc_devices = LIST_HEAD_INIT(mc_devices); +static LIST_HEAD(mc_devices); #ifdef CONFIG_EDAC_DEBUG @@ -54,13 +53,11 @@ static void edac_mc_dump_csrow(struct csrow_info *csrow) { debugf4("\tcsrow = %p\n", csrow); debugf4("\tcsrow->csrow_idx = %d\n", csrow->csrow_idx); - debugf4("\tcsrow->first_page = 0x%lx\n", - csrow->first_page); + debugf4("\tcsrow->first_page = 0x%lx\n", csrow->first_page); debugf4("\tcsrow->last_page = 0x%lx\n", csrow->last_page); debugf4("\tcsrow->page_mask = 0x%lx\n", csrow->page_mask); debugf4("\tcsrow->nr_pages = 0x%x\n", csrow->nr_pages); - debugf4("\tcsrow->nr_channels = %d\n", - csrow->nr_channels); + debugf4("\tcsrow->nr_channels = %d\n", csrow->nr_channels); debugf4("\tcsrow->channels = %p\n", csrow->channels); debugf4("\tcsrow->mci = %p\n\n", csrow->mci); } @@ -75,12 +72,35 @@ static void edac_mc_dump_mci(struct mem_ctl_info *mci) debugf3("\tmci->nr_csrows = %d, csrows = %p\n", mci->nr_csrows, mci->csrows); debugf3("\tdev = %p\n", mci->dev); - debugf3("\tmod_name:ctl_name = %s:%s\n", - mci->mod_name, mci->ctl_name); + debugf3("\tmod_name:ctl_name = %s:%s\n", mci->mod_name, mci->ctl_name); debugf3("\tpvt_info = %p\n\n", mci->pvt_info); } -#endif /* CONFIG_EDAC_DEBUG */ +/* + * keep those in sync with the enum mem_type + */ +const char *edac_mem_types[] = { + "Empty csrow", + "Reserved csrow type", + "Unknown csrow type", + "Fast page mode RAM", + "Extended data out RAM", + "Burst Extended data out RAM", + "Single data rate SDRAM", + "Registered single data rate SDRAM", + "Double data rate SDRAM", + "Registered Double data rate SDRAM", + "Rambus DRAM", + "Unbuffered DDR2 RAM", + "Fully buffered DDR2", + "Registered DDR2 RAM", + "Rambus XDR", + "Unbuffered DDR3 RAM", + "Registered DDR3 RAM", +}; +EXPORT_SYMBOL_GPL(edac_mem_types); + +#endif /* CONFIG_EDAC_DEBUG */ /* 'ptr' points to a possibly unaligned item X such that sizeof(X) is 'size'. * Adjust 'ptr' so that its alignment is at least as stringent as what the @@ -89,7 +109,7 @@ static void edac_mc_dump_mci(struct mem_ctl_info *mci) * If 'size' is a constant, the compiler will optimize this whole function * down to either a no-op or the addition of a constant to the value of 'ptr'. */ -char * edac_align_ptr(void *ptr, unsigned size) +void *edac_align_ptr(void *ptr, unsigned size) { unsigned align, r; @@ -106,14 +126,14 @@ char * edac_align_ptr(void *ptr, unsigned size) else if (size > sizeof(char)) align = sizeof(short); else - return (char *) ptr; + return (char *)ptr; r = size % align; if (r == 0) - return (char *) ptr; + return (char *)ptr; - return (char *) (((unsigned long) ptr) + align - r); + return (void *)(((unsigned long)ptr) + align - r); } /** @@ -133,7 +153,7 @@ char * edac_align_ptr(void *ptr, unsigned size) * struct mem_ctl_info pointer */ struct mem_ctl_info *edac_mc_alloc(unsigned sz_pvt, unsigned nr_csrows, - unsigned nr_chans) + unsigned nr_chans, int edac_index) { struct mem_ctl_info *mci; struct csrow_info *csi, *csrow; @@ -141,30 +161,32 @@ struct mem_ctl_info *edac_mc_alloc(unsigned sz_pvt, unsigned nr_csrows, void *pvt; unsigned size; int row, chn; + int err; /* Figure out the offsets of the various items from the start of an mc * structure. We want the alignment of each item to be at least as * stringent as what the compiler would provide if we could simply * hardcode everything into a single struct. */ - mci = (struct mem_ctl_info *) 0; - csi = (struct csrow_info *)edac_align_ptr(&mci[1], sizeof(*csi)); - chi = (struct channel_info *) - edac_align_ptr(&csi[nr_csrows], sizeof(*chi)); + mci = (struct mem_ctl_info *)0; + csi = edac_align_ptr(&mci[1], sizeof(*csi)); + chi = edac_align_ptr(&csi[nr_csrows], sizeof(*chi)); pvt = edac_align_ptr(&chi[nr_chans * nr_csrows], sz_pvt); - size = ((unsigned long) pvt) + sz_pvt; + size = ((unsigned long)pvt) + sz_pvt; - if ((mci = kmalloc(size, GFP_KERNEL)) == NULL) + mci = kzalloc(size, GFP_KERNEL); + if (mci == NULL) return NULL; /* Adjust pointers so they point within the memory we just allocated * rather than an imaginary chunk of memory located at address 0. */ - csi = (struct csrow_info *) (((char *) mci) + ((unsigned long) csi)); - chi = (struct channel_info *) (((char *) mci) + ((unsigned long) chi)); - pvt = sz_pvt ? (((char *) mci) + ((unsigned long) pvt)) : NULL; + csi = (struct csrow_info *)(((char *)mci) + ((unsigned long)csi)); + chi = (struct channel_info *)(((char *)mci) + ((unsigned long)chi)); + pvt = sz_pvt ? (((char *)mci) + ((unsigned long)pvt)) : NULL; - memset(mci, 0, size); /* clear all fields */ + /* setup index and various internal pointers */ + mci->mc_idx = edac_index; mci->csrows = csi; mci->pvt_info = pvt; mci->nr_csrows = nr_csrows; @@ -184,20 +206,45 @@ struct mem_ctl_info *edac_mc_alloc(unsigned sz_pvt, unsigned nr_csrows, } } + mci->op_state = OP_ALLOC; + + /* + * Initialize the 'root' kobj for the edac_mc controller + */ + err = edac_mc_register_sysfs_main_kobj(mci); + if (err) { + kfree(mci); + return NULL; + } + + /* at this point, the root kobj is valid, and in order to + * 'free' the object, then the function: + * edac_mc_unregister_sysfs_main_kobj() must be called + * which will perform kobj unregistration and the actual free + * will occur during the kobject callback operation + */ return mci; } EXPORT_SYMBOL_GPL(edac_mc_alloc); /** - * edac_mc_free: Free a previously allocated 'mci' structure + * edac_mc_free + * 'Free' a previously allocated 'mci' structure * @mci: pointer to a struct mem_ctl_info structure */ void edac_mc_free(struct mem_ctl_info *mci) { - kfree(mci); + edac_mc_unregister_sysfs_main_kobj(mci); } EXPORT_SYMBOL_GPL(edac_mc_free); + +/* + * find_mci_by_dev + * + * scan list of controllers looking for the one that manages + * the 'dev' device + */ static struct mem_ctl_info *find_mci_by_dev(struct device *dev) { struct mem_ctl_info *mci; @@ -215,18 +262,149 @@ static struct mem_ctl_info *find_mci_by_dev(struct device *dev) return NULL; } +/* + * handler for EDAC to check if NMI type handler has asserted interrupt + */ +static int edac_mc_assert_error_check_and_clear(void) +{ + int old_state; + + if (edac_op_state == EDAC_OPSTATE_POLL) + return 1; + + old_state = edac_err_assert; + edac_err_assert = 0; + + return old_state; +} + +/* + * edac_mc_workq_function + * performs the operation scheduled by a workq request + */ +static void edac_mc_workq_function(struct work_struct *work_req) +{ + struct delayed_work *d_work = to_delayed_work(work_req); + struct mem_ctl_info *mci = to_edac_mem_ctl_work(d_work); + + mutex_lock(&mem_ctls_mutex); + + /* if this control struct has movd to offline state, we are done */ + if (mci->op_state == OP_OFFLINE) { + mutex_unlock(&mem_ctls_mutex); + return; + } + + /* Only poll controllers that are running polled and have a check */ + if (edac_mc_assert_error_check_and_clear() && (mci->edac_check != NULL)) + mci->edac_check(mci); + + mutex_unlock(&mem_ctls_mutex); + + /* Reschedule */ + queue_delayed_work(edac_workqueue, &mci->work, + msecs_to_jiffies(edac_mc_get_poll_msec())); +} + +/* + * edac_mc_workq_setup + * initialize a workq item for this mci + * passing in the new delay period in msec + * + * locking model: + * + * called with the mem_ctls_mutex held + */ +static void edac_mc_workq_setup(struct mem_ctl_info *mci, unsigned msec) +{ + debugf0("%s()\n", __func__); + + /* if this instance is not in the POLL state, then simply return */ + if (mci->op_state != OP_RUNNING_POLL) + return; + + INIT_DELAYED_WORK(&mci->work, edac_mc_workq_function); + queue_delayed_work(edac_workqueue, &mci->work, msecs_to_jiffies(msec)); +} + +/* + * edac_mc_workq_teardown + * stop the workq processing on this mci + * + * locking model: + * + * called WITHOUT lock held + */ +static void edac_mc_workq_teardown(struct mem_ctl_info *mci) +{ + int status; + + status = cancel_delayed_work(&mci->work); + if (status == 0) { + debugf0("%s() not canceled, flush the queue\n", + __func__); + + /* workq instance might be running, wait for it */ + flush_workqueue(edac_workqueue); + } +} + +/* + * edac_mc_reset_delay_period(unsigned long value) + * + * user space has updated our poll period value, need to + * reset our workq delays + */ +void edac_mc_reset_delay_period(int value) +{ + struct mem_ctl_info *mci; + struct list_head *item; + + mutex_lock(&mem_ctls_mutex); + + /* scan the list and turn off all workq timers, doing so under lock + */ + list_for_each(item, &mc_devices) { + mci = list_entry(item, struct mem_ctl_info, link); + + if (mci->op_state == OP_RUNNING_POLL) + cancel_delayed_work(&mci->work); + } + + mutex_unlock(&mem_ctls_mutex); + + + /* re-walk the list, and reset the poll delay */ + mutex_lock(&mem_ctls_mutex); + + list_for_each(item, &mc_devices) { + mci = list_entry(item, struct mem_ctl_info, link); + + edac_mc_workq_setup(mci, (unsigned long) value); + } + + mutex_unlock(&mem_ctls_mutex); +} + + + /* Return 0 on success, 1 on failure. * Before calling this function, caller must * assign a unique value to mci->mc_idx. + * + * locking model: + * + * called with the mem_ctls_mutex lock held */ -static int add_mc_to_global_list (struct mem_ctl_info *mci) +static int add_mc_to_global_list(struct mem_ctl_info *mci) { struct list_head *item, *insert_before; struct mem_ctl_info *p; insert_before = &mc_devices; - if (unlikely((p = find_mci_by_dev(mci->dev)) != NULL)) + p = find_mci_by_dev(mci->dev); + if (unlikely(p != NULL)) goto fail0; list_for_each(item, &mc_devices) { @@ -247,14 +425,14 @@ static int add_mc_to_global_list (struct mem_ctl_info *mci) fail0: edac_printk(KERN_WARNING, EDAC_MC, - "%s (%s) %s %s already assigned %d\n", p->dev->bus_id, - dev_name(p->dev), p->mod_name, p->ctl_name, p->mc_idx); + "%s (%s) %s %s already assigned %d\n", dev_name(p->dev), + edac_dev_name(mci), p->mod_name, p->ctl_name, p->mc_idx); return 1; fail1: edac_printk(KERN_WARNING, EDAC_MC, - "bug in low-level driver: attempt to assign\n" - " duplicate mc_idx %d in %s()\n", p->mc_idx, __func__); + "bug in low-level driver: attempt to assign\n" + " duplicate mc_idx %d in %s()\n", p->mc_idx, __func__); return 1; } @@ -264,16 +442,14 @@ static void complete_mc_list_del(struct rcu_head *head) mci = container_of(head, struct mem_ctl_info, rcu); INIT_LIST_HEAD(&mci->link); - complete(&mci->complete); } static void del_mc_from_global_list(struct mem_ctl_info *mci) { atomic_dec(&edac_handlers); list_del_rcu(&mci->link); - init_completion(&mci->complete); call_rcu(&mci->rcu, complete_mc_list_del); - wait_for_completion(&mci->complete); + rcu_barrier(); } /** @@ -284,7 +460,7 @@ static void del_mc_from_global_list(struct mem_ctl_info *mci) * * Caller must hold mem_ctls_mutex. */ -struct mem_ctl_info * edac_mc_find(int idx) +struct mem_ctl_info *edac_mc_find(int idx) { struct list_head *item; struct mem_ctl_info *mci; @@ -316,10 +492,10 @@ EXPORT_SYMBOL(edac_mc_find); */ /* FIXME - should a warning be printed if no error detection? correction? */ -int edac_mc_add_mc(struct mem_ctl_info *mci, int mc_idx) +int edac_mc_add_mc(struct mem_ctl_info *mci) { debugf0("%s()\n", __func__); - mci->mc_idx = mc_idx; + #ifdef CONFIG_EDAC_DEBUG if (edac_debug_level >= 3) edac_mc_dump_mci(mci); @@ -332,8 +508,8 @@ int edac_mc_add_mc(struct mem_ctl_info *mci, int mc_idx) edac_mc_dump_csrow(&mci->csrows[i]); for (j = 0; j < mci->csrows[i].nr_channels; j++) - edac_mc_dump_channel( - &mci->csrows[i].channels[j]); + edac_mc_dump_channel(&mci->csrows[i]. + channels[j]); } } #endif @@ -351,9 +527,19 @@ int edac_mc_add_mc(struct mem_ctl_info *mci, int mc_idx) goto fail1; } + /* If there IS a check routine, then we are running POLLED */ + if (mci->edac_check != NULL) { + /* This instance is NOW RUNNING */ + mci->op_state = OP_RUNNING_POLL; + + edac_mc_workq_setup(mci, edac_mc_get_poll_msec()); + } else { + mci->op_state = OP_RUNNING_INTERRUPT; + } + /* Report action taken */ - edac_mc_printk(mci, KERN_INFO, "Giving out device to %s %s: DEV %s\n", - mci->mod_name, mci->ctl_name, dev_name(mci->dev)); + edac_mc_printk(mci, KERN_INFO, "Giving out device to '%s' '%s':" + " DEV %s\n", mci->mod_name, mci->ctl_name, edac_dev_name(mci)); mutex_unlock(&mem_ctls_mutex); return 0; @@ -374,24 +560,35 @@ EXPORT_SYMBOL_GPL(edac_mc_add_mc); * * Return pointer to removed mci structure, or NULL if device not found. */ -struct mem_ctl_info * edac_mc_del_mc(struct device *dev) +struct mem_ctl_info *edac_mc_del_mc(struct device *dev) { struct mem_ctl_info *mci; - debugf0("MC: %s()\n", __func__); + debugf0("%s()\n", __func__); + mutex_lock(&mem_ctls_mutex); - if ((mci = find_mci_by_dev(dev)) == NULL) { + /* find the requested mci struct in the global list */ + mci = find_mci_by_dev(dev); + if (mci == NULL) { mutex_unlock(&mem_ctls_mutex); return NULL; } - edac_remove_sysfs_mci_device(mci); + /* marking MCI offline */ + mci->op_state = OP_OFFLINE; + del_mc_from_global_list(mci); mutex_unlock(&mem_ctls_mutex); + + /* flush workq processes and remove sysfs */ + edac_mc_workq_teardown(mci); + edac_remove_sysfs_mci_device(mci); + edac_printk(KERN_INFO, EDAC_MC, "Removed device %d for %s %s: DEV %s\n", mci->mc_idx, - mci->mod_name, mci->ctl_name, dev_name(mci->dev)); + mci->mod_name, mci->ctl_name, edac_dev_name(mci)); + return mci; } EXPORT_SYMBOL_GPL(edac_mc_del_mc); @@ -406,7 +603,7 @@ static void edac_mc_scrub_block(unsigned long page, unsigned long offset, debugf3("%s()\n", __func__); /* ECC error page was not in our memory. Ignore it. */ - if(!pfn_valid(page)) + if (!pfn_valid(page)) return; /* Find the actual page structure then map it and fix */ @@ -459,7 +656,7 @@ int edac_mc_find_csrow_by_page(struct mem_ctl_info *mci, unsigned long page) if (row == -1) edac_mc_printk(mci, KERN_ERR, "could not look up page error address %lx\n", - (unsigned long) page); + (unsigned long)page); return row; } @@ -468,8 +665,9 @@ EXPORT_SYMBOL_GPL(edac_mc_find_csrow_by_page); /* FIXME - setable log (warning/emerg) levels */ /* FIXME - integrate with evlog: http://evlog.sourceforge.net/ */ void edac_mc_handle_ce(struct mem_ctl_info *mci, - unsigned long page_frame_number, unsigned long offset_in_page, - unsigned long syndrome, int row, int channel, const char *msg) + unsigned long page_frame_number, + unsigned long offset_in_page, unsigned long syndrome, + int row, int channel, const char *msg) { unsigned long remapped_page; @@ -495,7 +693,7 @@ void edac_mc_handle_ce(struct mem_ctl_info *mci, return; } - if (edac_get_log_ce()) + if (edac_mc_get_log_ce()) /* FIXME - put in DIMM location */ edac_mc_printk(mci, KERN_WARNING, "CE page 0x%lx, offset 0x%lx, grain %d, syndrome " @@ -519,18 +717,18 @@ void edac_mc_handle_ce(struct mem_ctl_info *mci, * page - which can then be scrubbed. */ remapped_page = mci->ctl_page_to_phys ? - mci->ctl_page_to_phys(mci, page_frame_number) : - page_frame_number; + mci->ctl_page_to_phys(mci, page_frame_number) : + page_frame_number; edac_mc_scrub_block(remapped_page, offset_in_page, - mci->csrows[row].grain); + mci->csrows[row].grain); } } EXPORT_SYMBOL_GPL(edac_mc_handle_ce); void edac_mc_handle_ce_no_info(struct mem_ctl_info *mci, const char *msg) { - if (edac_get_log_ce()) + if (edac_mc_get_log_ce()) edac_mc_printk(mci, KERN_WARNING, "CE - no information available: %s\n", msg); @@ -540,8 +738,8 @@ void edac_mc_handle_ce_no_info(struct mem_ctl_info *mci, const char *msg) EXPORT_SYMBOL_GPL(edac_mc_handle_ce_no_info); void edac_mc_handle_ue(struct mem_ctl_info *mci, - unsigned long page_frame_number, unsigned long offset_in_page, - int row, const char *msg) + unsigned long page_frame_number, + unsigned long offset_in_page, int row, const char *msg) { int len = EDAC_MC_LABEL_LEN * 4; char labels[len + 1]; @@ -562,26 +760,26 @@ void edac_mc_handle_ue(struct mem_ctl_info *mci, } chars = snprintf(pos, len + 1, "%s", - mci->csrows[row].channels[0].label); + mci->csrows[row].channels[0].label); len -= chars; pos += chars; for (chan = 1; (chan < mci->csrows[row].nr_channels) && (len > 0); - chan++) { + chan++) { chars = snprintf(pos, len + 1, ":%s", - mci->csrows[row].channels[chan].label); + mci->csrows[row].channels[chan].label); len -= chars; pos += chars; } - if (edac_get_log_ue()) + if (edac_mc_get_log_ue()) edac_mc_printk(mci, KERN_EMERG, "UE page 0x%lx, offset 0x%lx, grain %d, row %d, " "labels \"%s\": %s\n", page_frame_number, - offset_in_page, mci->csrows[row].grain, row, labels, - msg); + offset_in_page, mci->csrows[row].grain, row, + labels, msg); - if (edac_get_panic_on_ue()) + if (edac_mc_get_panic_on_ue()) panic("EDAC MC%d: UE page 0x%lx, offset 0x%lx, grain %d, " "row %d, labels \"%s\": %s\n", mci->mc_idx, page_frame_number, offset_in_page, @@ -594,10 +792,10 @@ EXPORT_SYMBOL_GPL(edac_mc_handle_ue); void edac_mc_handle_ue_no_info(struct mem_ctl_info *mci, const char *msg) { - if (edac_get_panic_on_ue()) + if (edac_mc_get_panic_on_ue()) panic("EDAC MC%d: Uncorrected Error", mci->mc_idx); - if (edac_get_log_ue()) + if (edac_mc_get_log_ue()) edac_mc_printk(mci, KERN_WARNING, "UE - no information available: %s\n", msg); mci->ue_noinfo_count++; @@ -605,16 +803,14 @@ void edac_mc_handle_ue_no_info(struct mem_ctl_info *mci, const char *msg) } EXPORT_SYMBOL_GPL(edac_mc_handle_ue_no_info); - /************************************************************* * On Fully Buffered DIMM modules, this help function is * called to process UE events */ void edac_mc_handle_fbd_ue(struct mem_ctl_info *mci, - unsigned int csrow, - unsigned int channela, - unsigned int channelb, - char *msg) + unsigned int csrow, + unsigned int channela, + unsigned int channelb, char *msg) { int len = EDAC_MC_LABEL_LEN * 4; char labels[len + 1]; @@ -656,20 +852,21 @@ void edac_mc_handle_fbd_ue(struct mem_ctl_info *mci, /* Generate the DIMM labels from the specified channels */ chars = snprintf(pos, len + 1, "%s", mci->csrows[csrow].channels[channela].label); - len -= chars; pos += chars; + len -= chars; + pos += chars; chars = snprintf(pos, len + 1, "-%s", mci->csrows[csrow].channels[channelb].label); - if (edac_get_log_ue()) + if (edac_mc_get_log_ue()) edac_mc_printk(mci, KERN_EMERG, "UE row %d, channel-a= %d channel-b= %d " "labels \"%s\": %s\n", csrow, channela, channelb, labels, msg); - if (edac_get_panic_on_ue()) + if (edac_mc_get_panic_on_ue()) panic("UE row %d, channel-a= %d channel-b= %d " - "labels \"%s\": %s\n", csrow, channela, - channelb, labels, msg); + "labels \"%s\": %s\n", csrow, channela, + channelb, labels, msg); } EXPORT_SYMBOL(edac_mc_handle_fbd_ue); @@ -678,9 +875,7 @@ EXPORT_SYMBOL(edac_mc_handle_fbd_ue); * called to process CE events */ void edac_mc_handle_fbd_ce(struct mem_ctl_info *mci, - unsigned int csrow, - unsigned int channel, - char *msg) + unsigned int csrow, unsigned int channel, char *msg) { /* Ensure boundary values */ @@ -701,38 +896,15 @@ void edac_mc_handle_fbd_ce(struct mem_ctl_info *mci, return; } - if (edac_get_log_ce()) + if (edac_mc_get_log_ce()) /* FIXME - put in DIMM location */ edac_mc_printk(mci, KERN_WARNING, "CE row %d, channel %d, label \"%s\": %s\n", csrow, channel, - mci->csrows[csrow].channels[channel].label, - msg); + mci->csrows[csrow].channels[channel].label, msg); mci->ce_count++; mci->csrows[csrow].ce_count++; mci->csrows[csrow].channels[channel].ce_count++; } EXPORT_SYMBOL(edac_mc_handle_fbd_ce); - - -/* - * Iterate over all MC instances and check for ECC, et al, errors - */ -void edac_check_mc_devices(void) -{ - struct list_head *item; - struct mem_ctl_info *mci; - - debugf3("%s()\n", __func__); - mutex_lock(&mem_ctls_mutex); - - list_for_each(item, &mc_devices) { - mci = list_entry(item, struct mem_ctl_info, link); - - if (mci->edac_check != NULL) - mci->edac_check(mci); - } - - mutex_unlock(&mem_ctls_mutex); -}