Skip to contentSkip to conversation

EDAC mc numbers refactor 2-of-2

PATCH In review Doug Thompson 2 participants 2 messages Jun 26, 2006 – Jun 27, 2006 EDAC-core

lore
Share & tools
lore
#0 lore raw

PATCH 3/6 EDAC mc numbers refactor 2-of-2

From: Doug Thompson <[email protected]>

This is part 2 of a 2-part patch set.

1 Reimplement add_mc_to_global_list() with semantics that allow the caller to
determine the ID number for a mem_ctl_info structure. Then modify edac_mc_add_mc() so that the caller specifies the ID number for the new mem_ctl_info structure. Platform-specific code should be able to assign the ID numbers in a platform-specific manner. For instance, on Opteron it makes sense to have the ID of the mem_ctl_info structure match the ID of the node that the memory controller belongs to.

2 Modify callers of edac_mc_add_mc() so they use the new semantics.

---
 amd76x_edac.c  |    5 ++++-
 e752x_edac.c   |    5 ++++-
 e7xxx_edac.c   |    5 ++++-
 edac_mc.c      |   46 +++++++++++++++++++++++++++++++++++++++++++++-
 edac_mc.h      |    2 +-
 i82860_edac.c  |    5 ++++-
 i82875p_edac.c |    5 ++++-
 r82600_edac.c  |    5 ++++-
 8 files changed, 70 insertions(+), 8 deletions(-)

Index: linux-2.6.17-rc5/drivers/edac/amd76x_edac.c
===================================================================
--- linux-2.6.17-rc5.orig/drivers/edac/amd76x_edac.c 2006-05-25 14:01:24.000000000
-0600
+++ linux-2.6.17-rc5/drivers/edac/amd76x_edac.c 2006-05-25 14:06:48.000000000 -0600
@@ -257,7 +257,10 @@

amd76x_get_error_info(mci, &discard); /* clear counters */

- if (edac_mc_add_mc(mci)) {
+ /* Here we assume that we will never see multiple instances of this
+ * type of memory controller. The ID is therefore hardcoded to 0.
+ */
+ if (edac_mc_add_mc(mci,0)) {
debugf3("%s(): failed edac_mc_add_mc()\n", __func__); goto fail;
}
Index: linux-2.6.17-rc5/drivers/edac/e752x_edac.c
===================================================================
--- linux-2.6.17-rc5.orig/drivers/edac/e752x_edac.c 2006-05-25 14:01:24.000000000
-0600
+++ linux-2.6.17-rc5/drivers/edac/e752x_edac.c 2006-05-25 14:06:48.000000000 -0600
@@ -953,7 +953,10 @@
"tolm = %x, remapbase = %x, remaplimit = %x\n", pvt->tolm, pvt->remapbase, pvt->remaplimit);

- if (edac_mc_add_mc(mci)) {
+ /* Here we assume that we will never see multiple instances of this
+ * type of memory controller. The ID is therefore hardcoded to 0.
+ */
+ if (edac_mc_add_mc(mci,0)) {
debugf3("%s(): failed edac_mc_add_mc()\n", __func__); goto fail;
}
Index: linux-2.6.17-rc5/drivers/edac/e7xxx_edac.c
===================================================================
--- linux-2.6.17-rc5.orig/drivers/edac/e7xxx_edac.c 2006-05-25 14:01:24.000000000
-0600
+++ linux-2.6.17-rc5/drivers/edac/e7xxx_edac.c 2006-05-25 14:06:48.000000000 -0600
@@ -463,7 +463,10 @@
/* clear any pending errors, or initial state bits */ e7xxx_get_error_info(mci, &discard);

- if (edac_mc_add_mc(mci) != 0) {
+ /* Here we assume that we will never see multiple instances of this
+ * type of memory controller. The ID is therefore hardcoded to 0.
+ */
+ if (edac_mc_add_mc(mci,0)) {
debugf3("%s(): failed edac_mc_add_mc()\n", __func__); goto fail;
}

drivers/edac/edac_mc.c +45 -1
Index: linux-2.6.17-rc5/drivers/edac/edac_mc.c
===================================================================
@@ -1632,6 +1632,48 @@
 	return NULL;
 }
 
+/* Return 0 on success, 1 on failure.
+ * Before calling this function, caller must
+ * assign a unique value to mci->mc_idx.
+ */
+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))
+		goto fail0;
+
+	list_for_each(item, &mc_devices) {
+		p = list_entry(item, struct mem_ctl_info, link);
+
+		if (p->mc_idx >= mci->mc_idx) {
+			if (unlikely(p->mc_idx == mci->mc_idx))
+				goto fail1;
+
+			insert_before = item;
+			break;
+		}
+	}
+
+	list_add_tail_rcu(&mci->link, insert_before);
+	return 0;
+
+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);
+	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__);
+	return 1;
+}
+
 static void complete_mc_list_del(struct rcu_head *head)
 {
 	struct mem_ctl_info *mci;
@@ -1653,6 +1695,7 @@
  * edac_mc_add_mc: Insert the 'mci' structure into the mci global list and
  *                 create sysfs entries associated with mci structure
  * @mci: pointer to the mci structure to be added to the list
+ * @mc_idx: A unique numeric identifier to be assigned to the 'mci' structure.
  *
  * Return:
  *	0	Success
@@ -1660,9 +1703,10 @@
  */
 
 /* FIXME - should a warning be printed if no error detection? correction? */
-int edac_mc_add_mc(struct mem_ctl_info *mci)
+int edac_mc_add_mc(struct mem_ctl_info *mci, int mc_idx)
 {
 	debugf0("%s()\n", __func__);
+	mci->mc_idx = mc_idx;
 #ifdef CONFIG_EDAC_DEBUG
 	if (edac_debug_level >= 3)
 		edac_mc_dump_mci(mci);
drivers/edac/edac_mc.h +1 -1
Index: linux-2.6.17-rc5/drivers/edac/edac_mc.h
===================================================================
@@ -417,7 +417,7 @@
 void edac_mc_dump_csrow(struct csrow_info *csrow);
 #endif  /* CONFIG_EDAC_DEBUG */
 
-extern int edac_mc_add_mc(struct mem_ctl_info *mci);
+extern int edac_mc_add_mc(struct mem_ctl_info *mci,int mc_idx);
 extern struct mem_ctl_info * edac_mc_del_mc(struct device *dev);
 extern int edac_mc_find_csrow_by_page(struct mem_ctl_info *mci,
 					unsigned long page);

Index: linux-2.6.17-rc5/drivers/edac/i82860_edac.c
===================================================================
--- linux-2.6.17-rc5.orig/drivers/edac/i82860_edac.c 2006-05-25 14:01:24.000000000
-0600
+++ linux-2.6.17-rc5/drivers/edac/i82860_edac.c 2006-05-25 14:06:48.000000000 -0600
@@ -208,7 +208,10 @@

i82860_get_error_info(mci, &discard); /* clear counters */

- if (edac_mc_add_mc(mci)) {
+ /* Here we assume that we will never see multiple instances of this
+ * type of memory controller. The ID is therefore hardcoded to 0.
+ */
+ if (edac_mc_add_mc(mci,0)) {
debugf3("%s(): failed edac_mc_add_mc()\n", __func__); edac_mc_free(mci);
} else {
Index: linux-2.6.17-rc5/drivers/edac/i82875p_edac.c
===================================================================
--- linux-2.6.17-rc5.orig/drivers/edac/i82875p_edac.c 2006-05-25 14:01:24.000000000
-0600
+++ linux-2.6.17-rc5/drivers/edac/i82875p_edac.c 2006-05-25 14:06:48.000000000 -0600
@@ -390,7 +390,10 @@

i82875p_get_error_info(mci, &discard); /* clear counters */

- if (edac_mc_add_mc(mci)) {
+ /* Here we assume that we will never see multiple instances of this
+ * type of memory controller. The ID is therefore hardcoded to 0.
+ */
+ if (edac_mc_add_mc(mci,0)) {
debugf3("%s(): failed edac_mc_add_mc()\n", __func__); goto fail3;
}
Index: linux-2.6.17-rc5/drivers/edac/r82600_edac.c
===================================================================
--- linux-2.6.17-rc5.orig/drivers/edac/r82600_edac.c 2006-05-25 14:01:24.000000000
-0600
+++ linux-2.6.17-rc5/drivers/edac/r82600_edac.c 2006-05-25 14:06:48.000000000 -0600
@@ -304,7 +304,10 @@

r82600_get_error_info(mci, &discard); /* clear counters */

- if (edac_mc_add_mc(mci)) {
+ /* Here we assume that we will never see multiple instances of this
+ * type of memory controller. The ID is therefore hardcoded to 0.
+ */
+ if (edac_mc_add_mc(mci,0)) {
debugf3("%s(): failed edac_mc_add_mc()\n", __func__); goto fail;
}

#1 lore raw
Quoted 48 lines · Doug Thompson

From: Doug Thompson <[email protected]>

This is part 2 of a 2-part patch set.

1 Reimplement add_mc_to_global_list() with semantics that allow the caller to
determine the ID number for a mem_ctl_info structure. Then modify edac_mc_add_mc() so that the caller specifies the ID number for the new mem_ctl_info structure. Platform-specific code should be able to assign the ID numbers in a platform-specific manner. For instance, on Opteron it makes sense to have the ID of the mem_ctl_info structure match the ID of the node that the memory controller belongs to.

2 Modify callers of edac_mc_add_mc() so they use the new semantics.

...

drivers/edac/edac_mc.c +31 -0
+/* Return 0 on success, 1 on failure.
+ * Before calling this function, caller must
+ * assign a unique value to mci->mc_idx.
+ */
+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))
+		goto fail0;
+
+	list_for_each(item, &mc_devices) {
+		p = list_entry(item, struct mem_ctl_info, link);
+
+		if (p->mc_idx >= mci->mc_idx) {
+			if (unlikely(p->mc_idx == mci->mc_idx))
+				goto fail1;
+
+			insert_before = item;
+			break;
+		}
+	}
+
+	list_add_tail_rcu(&mci->link, insert_before);
+	return 0;
+
+fail0:
+	edac_printk(KERN_WARNING, EDAC_MC,

drivers/edac/edac_mc.c line 1675

Quoted 10 lines
drivers/edac/edac_mc.c +10 -0
+		    "%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);
+	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__);
+	return 1;
+}

I assume the caller must hold some lock to prevent the list from getting trashed, but I didn't locate it in (literally) a five-second search. I'd suggest that a comment describing the locking requirements be added to this function.