From: Vaibhav Jain <vaibhav@linux.vnet.ibm.com>
To: linuxppc-dev@lists.ozlabs.org, Russell Currey <ruscur@russell.cc>,
frederic.barrat@fr.ibm.com
Cc: Vaibhav Jain <vaibhav@linux.vnet.ibm.com>,
Andrew Donnellan <andonnel@au1.ibm.com>,
Ian Munsie <ianmunsi@au1.ibm.com>,
Christophe Lombard <christophe_lombard@fr.ibm.com>,
Philippe Bergheaud <philippe.bergheaud@fr.ibm.com>,
Gregory Kurz <KURZGREG@fr.ibm.com>,
Gavin Shan <gwshan@linux.vnet.ibm.com>
Subject: [RFC 1/3] powerpc/eeh: Refactor eeh_pe_update_time_stamp to update freeze_count
Date: Tue, 28 Feb 2017 12:32:20 +0530 [thread overview]
Message-ID: <20170228070222.21126-2-vaibhav@linux.vnet.ibm.com> (raw)
In-Reply-To: <20170228070222.21126-1-vaibhav@linux.vnet.ibm.com>
This patch introduces a new function named
eeh_pe_update_freeze_counter replacing existing function
eeh_pe_update_time_stamp. The new function also manages the value of
freeze_count along with tstamp to track the number of times the PE
froze in last one hour and if the freeze_count > eeh_max_freezes then
reports an error(-ENOTRECOVERABLE) to indicate that the PE should be
permanently disabled.
This patch should not introduce any behavioral change.
Signed-off-by: Vaibhav Jain <vaibhav@linux.vnet.ibm.com>
---
arch/powerpc/include/asm/eeh.h | 2 +-
arch/powerpc/kernel/eeh_driver.c | 20 +++-------------
arch/powerpc/kernel/eeh_pe.c | 50 ++++++++++++++++++++++++++--------------
3 files changed, 37 insertions(+), 35 deletions(-)
diff --git a/arch/powerpc/include/asm/eeh.h b/arch/powerpc/include/asm/eeh.h
index 8e37b71..68806be 100644
--- a/arch/powerpc/include/asm/eeh.h
+++ b/arch/powerpc/include/asm/eeh.h
@@ -265,7 +265,7 @@ struct eeh_pe *eeh_phb_pe_get(struct pci_controller *phb);
struct eeh_pe *eeh_pe_get(struct eeh_dev *edev);
int eeh_add_to_parent_pe(struct eeh_dev *edev);
int eeh_rmv_from_parent_pe(struct eeh_dev *edev);
-void eeh_pe_update_time_stamp(struct eeh_pe *pe);
+int eeh_pe_update_freeze_counter(struct eeh_pe *pe);
void *eeh_pe_traverse(struct eeh_pe *root,
eeh_traverse_func fn, void *flag);
void *eeh_pe_dev_traverse(struct eeh_pe *root,
diff --git a/arch/powerpc/kernel/eeh_driver.c b/arch/powerpc/kernel/eeh_driver.c
index b948871..326b4e4 100644
--- a/arch/powerpc/kernel/eeh_driver.c
+++ b/arch/powerpc/kernel/eeh_driver.c
@@ -739,10 +739,9 @@ static void eeh_handle_normal_event(struct eeh_pe *pe)
return;
}
- eeh_pe_update_time_stamp(pe);
- pe->freeze_count++;
- if (pe->freeze_count > eeh_max_freezes)
- goto excess_failures;
+ /* Update freeze counters and see if we have tripped max-freeze limit */
+ if (eeh_pe_update_freeze_counter(pe) < 0)
+ goto perm_error;
pr_warn("EEH: This PCI device has failed %d times in the last hour\n",
pe->freeze_count);
@@ -872,19 +871,6 @@ static void eeh_handle_normal_event(struct eeh_pe *pe)
return;
-excess_failures:
- /*
- * About 90% of all real-life EEH failures in the field
- * are due to poorly seated PCI cards. Only 10% or so are
- * due to actual, failed cards.
- */
- pr_err("EEH: PHB#%x-PE#%x has failed %d times in the\n"
- "last hour and has been permanently disabled.\n"
- "Please try reseating or replacing it.\n",
- pe->phb->global_number, pe->addr,
- pe->freeze_count);
- goto perm_error;
-
hard_fail:
pr_err("EEH: Unable to recover from failure from PHB#%x-PE#%x.\n"
"Please try reseating or replacing it\n",
diff --git a/arch/powerpc/kernel/eeh_pe.c b/arch/powerpc/kernel/eeh_pe.c
index cc4b206..cf70a8b 100644
--- a/arch/powerpc/kernel/eeh_pe.c
+++ b/arch/powerpc/kernel/eeh_pe.c
@@ -504,30 +504,46 @@ int eeh_rmv_from_parent_pe(struct eeh_dev *edev)
}
/**
- * eeh_pe_update_time_stamp - Update PE's frozen time stamp
+ * eeh_pe_update_freeze_counter - Update PE's frozen time stamp
+ * and freeze counter
* @pe: EEH PE
*
- * We have time stamp for each PE to trace its time of getting
- * frozen in last hour. The function should be called to update
- * the time stamp on first error of the specific PE. On the other
- * handle, we needn't account for errors happened in last hour.
+ * We have a freeze-counter and time stamp for each PE to trace
+ * number of times the PE was frozen in last one hour. This function
+ * updates the PE's freeze counter and if its > eeh_max_freezes then
+ * returns an error. The function should be called to once every-time
+ * a specific PE freezes.
*/
-void eeh_pe_update_time_stamp(struct eeh_pe *pe)
+int eeh_pe_update_freeze_counter(struct eeh_pe *pe)
{
struct timeval tstamp;
- if (!pe) return;
+ if (!pe)
+ return -EINVAL;
- if (pe->freeze_count <= 0) {
- pe->freeze_count = 0;
- do_gettimeofday(&pe->tstamp);
- } else {
- do_gettimeofday(&tstamp);
- if (tstamp.tv_sec - pe->tstamp.tv_sec > 3600) {
- pe->tstamp = tstamp;
- pe->freeze_count = 0;
- }
- }
+ do_gettimeofday(&tstamp);
+ if (pe->freeze_count <= 0 || tstamp.tv_sec - pe->tstamp.tv_sec > 3600) {
+ pe->tstamp = tstamp;
+ pe->freeze_count = 1;
+
+ } else if (pe->freeze_count >= eeh_max_freezes) {
+ pe->freeze_count++;
+ /*
+ * About 90% of all real-life EEH failures in the field
+ * are due to poorly seated PCI cards. Only 10% or so are
+ * due to actual, failed cards.
+ */
+ pr_err("EEH: PHB#%x-PE#%x has failed %d times in the\n"
+ "last hour and has been permanently disabled.\n"
+ "Please try reseating or replacing it.\n",
+ pe->phb->global_number, pe->addr,
+ pe->freeze_count);
+ return -ENOTRECOVERABLE;
+
+ } else
+ pe->freeze_count++;
+
+ return 0;
}
/**
--
2.9.3
next prev parent reply other threads:[~2017-02-28 7:03 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-02-28 7:02 [RFC 0/3] cxl: Reset freeze counter for the adapter before PERST Vaibhav Jain
2017-02-28 7:02 ` Vaibhav Jain [this message]
2017-02-28 23:44 ` [RFC 1/3] powerpc/eeh: Refactor eeh_pe_update_time_stamp to update freeze_count Russell Currey
2017-02-28 7:02 ` [RFC 2/3] powerpc/eeh: Introduce function eeh_pe_reset_freeze_counter Vaibhav Jain
2017-02-28 23:54 ` Russell Currey
2017-02-28 7:22 ` [RFC 3/3] cxl: Reset freeze counters before adapter PERST for flashing new image Vaibhav Jain
2017-03-01 0:06 ` Russell Currey
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170228070222.21126-2-vaibhav@linux.vnet.ibm.com \
--to=vaibhav@linux.vnet.ibm.com \
--cc=KURZGREG@fr.ibm.com \
--cc=andonnel@au1.ibm.com \
--cc=christophe_lombard@fr.ibm.com \
--cc=frederic.barrat@fr.ibm.com \
--cc=gwshan@linux.vnet.ibm.com \
--cc=ianmunsi@au1.ibm.com \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=philippe.bergheaud@fr.ibm.com \
--cc=ruscur@russell.cc \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).