rasdaemon: fix excessive block event messages
Only track block error events instead of logging sucessful ones. The upstream patch (9c86f6255f67a8) had to be modified since it relies on the kernel version to determine if the new trace point is available or not. Such tests are meaningless in RHEL as we backport a lot of patches while keeping the kernel version intact. Resolves: RHEL-8708 Signed-off-by: Aristeu Rozanski <arozansk@redhat.com>
This commit is contained in:
Aristeu Rozanski
parent
c355651dbd
commit
5cde7234ff
22
885e546add918457c453bd3f753ac7df90b39e36.patch
Normal file
22
885e546add918457c453bd3f753ac7df90b39e36.patch
Normal file
@ -0,0 +1,22 @@
|
||||
commit 885e546add918457c453bd3f753ac7df90b39e36
|
||||
Author: weidongkl <weidongkl@sina.com>
|
||||
Date: Tue Sep 19 16:29:21 2023 +0800
|
||||
|
||||
Add a space between "diskerror_event" and "store"
|
||||
|
||||
Signed-off-by: weidongkl <weidongkl@sina.com>
|
||||
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
|
||||
|
||||
diff --git a/ras-record.c b/ras-record.c
|
||||
index a5f99ae..6b050bb 100644
|
||||
--- a/ras-record.c
|
||||
+++ b/ras-record.c
|
||||
@@ -484,7 +484,7 @@ int ras_store_diskerror_event(struct ras_events *ras, struct diskerror_event *ev
|
||||
|
||||
if (!priv || !priv->stmt_diskerror_event)
|
||||
return 0;
|
||||
- log(TERM, LOG_INFO, "diskerror_eventstore: %p\n", priv->stmt_diskerror_event);
|
||||
+ log(TERM, LOG_INFO, "diskerror_event store: %p\n", priv->stmt_diskerror_event);
|
||||
|
||||
sqlite3_bind_text(priv->stmt_diskerror_event, 1, ev->timestamp, -1, NULL);
|
||||
sqlite3_bind_text(priv->stmt_diskerror_event, 2, ev->dev, -1, NULL);
|
||||
24
9bd84aef87978b806178a73ed33c39d6c442fc1f.patch
Normal file
24
9bd84aef87978b806178a73ed33c39d6c442fc1f.patch
Normal file
@ -0,0 +1,24 @@
|
||||
commit 9bd84aef87978b806178a73ed33c39d6c442fc1f
|
||||
Author: weidong <weidongkl@sina.com>
|
||||
Date: Tue Aug 8 08:59:12 2023 +0000
|
||||
|
||||
add ':' before error output
|
||||
|
||||
All prints except disk are preceded by a colon
|
||||
|
||||
Signed-off-by: weidong <weidongkl@sina.com>
|
||||
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
|
||||
|
||||
diff --git a/util/ras-mc-ctl.in b/util/ras-mc-ctl.in
|
||||
index dc326d3..13078c2 100755
|
||||
--- a/util/ras-mc-ctl.in
|
||||
+++ b/util/ras-mc-ctl.in
|
||||
@@ -1469,7 +1469,7 @@ sub errors
|
||||
$out .= "\n";
|
||||
}
|
||||
if ($out ne "") {
|
||||
- print "Disk errors\n$out\n";
|
||||
+ print "Disk errors:\n$out\n";
|
||||
} else {
|
||||
print "No disk errors.\n\n";
|
||||
}
|
||||
117
9c86f6255f67a8bae28cd46c54500fc16bfc7a30.patch
Normal file
117
9c86f6255f67a8bae28cd46c54500fc16bfc7a30.patch
Normal file
@ -0,0 +1,117 @@
|
||||
commit 9c86f6255f67a8bae28cd46c54500fc16bfc7a30
|
||||
Author: Yang Shi <shy828301@gmail.com>
|
||||
Date: Mon Apr 4 16:34:05 2022 -0700
|
||||
|
||||
rasdaemon: use the new block_rq_error tracepoint
|
||||
|
||||
Since Linux 5.18-rc1 a new block tracepoint called block_rq_error is
|
||||
available for tracing disk error events dedicatedly. Currently
|
||||
rasdaemon is using block_rq_complete which also traces successful cases.
|
||||
It incurs excessive tracing logs and somehow overhead since the event is
|
||||
triggered quite often.
|
||||
|
||||
Use the new tracepoint for disk error reporting, and the new trace point
|
||||
has the same format as block_rq_complete.
|
||||
|
||||
Signed-off-by: Yang Shi <shy828301@gmail.com>
|
||||
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
|
||||
|
||||
---
|
||||
ras-events.c | 53 ++++++++++-------------------------------------------
|
||||
ras-record.c | 2 +-
|
||||
2 files changed, 11 insertions(+), 44 deletions(-)
|
||||
|
||||
--- rasdaemon-0.6.7.orig/ras-events.c 2024-05-14 11:05:40.020599541 -0400
|
||||
+++ rasdaemon-0.6.7/ras-events.c 2024-05-14 11:06:38.831067957 -0400
|
||||
@@ -27,6 +27,7 @@ * Foundation, Inc., 51 Franklin Street,
|
||||
#include <sys/poll.h>
|
||||
#include <signal.h>
|
||||
#include <sys/signalfd.h>
|
||||
+#include <linux/version.h>
|
||||
#include "libtrace/kbuffer.h"
|
||||
#include "libtrace/event-parse.h"
|
||||
#include "ras-mc-handler.h"
|
||||
@@ -229,7 +230,7 @@ if (rc < 0) {
|
||||
#endif
|
||||
|
||||
#ifdef HAVE_DISKERROR
|
||||
- rc |= __toggle_ras_mc_event(ras, "block", "block_rq_complete", enable);
|
||||
+ rc |= __toggle_ras_mc_event(ras, "block", "block_rq_error", enable);
|
||||
#endif
|
||||
|
||||
#ifdef HAVE_MEMORY_FAILURE
|
||||
@@ -241,37 +242,6 @@ free_ras:
|
||||
return rc;
|
||||
}
|
||||
|
||||
-/*
|
||||
- * Set kernel filter. libtrace doesn't provide an API for setting filters
|
||||
- * in kernel, we have to implement it here.
|
||||
- */
|
||||
-static int filter_ras_mc_event(struct ras_events *ras, char *group, char *event,
|
||||
- const char *filter_str)
|
||||
-{
|
||||
- int fd, rc;
|
||||
- char fname[MAX_PATH + 1];
|
||||
-
|
||||
- snprintf(fname, sizeof(fname), "events/%s/%s/filter", group, event);
|
||||
- fd = open_trace(ras, fname, O_RDWR | O_APPEND);
|
||||
- if (fd < 0) {
|
||||
- log(ALL, LOG_WARNING, "Can't open filter file\n");
|
||||
- return errno;
|
||||
- }
|
||||
-
|
||||
- rc = write(fd, filter_str ,strlen(filter_str));
|
||||
- if (rc < 0) {
|
||||
- log(ALL, LOG_WARNING, "Can't write to filter file\n");
|
||||
- close(fd);
|
||||
- return rc;
|
||||
- }
|
||||
- close(fd);
|
||||
- if (!rc) {
|
||||
- log(ALL, LOG_WARNING, "Nothing was written on filter file\n");
|
||||
- return EIO;
|
||||
- }
|
||||
-
|
||||
- return 0;
|
||||
-}
|
||||
|
||||
/*
|
||||
* Tracing read code
|
||||
@@ -901,17 +871,14 @@ (void)open("/sys/kernel/debug/ras/daemon
|
||||
#endif
|
||||
|
||||
#ifdef HAVE_DISKERROR
|
||||
- rc = filter_ras_mc_event(ras, "block", "block_rq_complete", "error != 0");
|
||||
- if (!rc) {
|
||||
- rc = add_event_handler(ras, pevent, page_size, "block",
|
||||
- "block_rq_complete", ras_diskerror_event_handler,
|
||||
- NULL, DISKERROR_EVENT);
|
||||
- if (!rc)
|
||||
- num_events++;
|
||||
- else
|
||||
- log(ALL, LOG_ERR, "Can't get traces from %s:%s\n",
|
||||
- "block", "block_rq_complete");
|
||||
- }
|
||||
+ rc = add_event_handler(ras, pevent, page_size, "block",
|
||||
+ "block_rq_error", ras_diskerror_event_handler,
|
||||
+ NULL, DISKERROR_EVENT);
|
||||
+ if (!rc)
|
||||
+ num_events++;
|
||||
+ else
|
||||
+ log(ALL, LOG_ERR, "Can't get traces from %s:%s\n",
|
||||
+ "block", "block_rq_error");
|
||||
#endif
|
||||
|
||||
#ifdef HAVE_MEMORY_FAILURE
|
||||
--- rasdaemon-0.6.7.orig/ras-record.c 2024-05-14 11:07:24.573654494 -0400
|
||||
+++ rasdaemon-0.6.7/ras-record.c 2024-05-14 11:07:07.626807674 -0400
|
||||
@@ -456,7 +456,7 @@ return 0;
|
||||
#endif
|
||||
|
||||
/*
|
||||
- * Table and functions to handle block:block_rq_complete
|
||||
+ * Table and functions to handle block:block_rq_error
|
||||
*/
|
||||
|
||||
#ifdef HAVE_DISKERROR
|
||||
@ -1,6 +1,6 @@
|
||||
Name: rasdaemon
|
||||
Version: 0.6.7
|
||||
Release: 10%{?dist}
|
||||
Release: 11%{?dist}
|
||||
Summary: Utility to receive RAS error tracings
|
||||
License: GPL-2.0-only
|
||||
URL: http://git.infradead.org/users/mchehab/rasdaemon.git
|
||||
@ -33,6 +33,9 @@ Patch24: 1f74a59ee33b7448b00d7ba13d5ecd4918b9853c.patch
|
||||
Patch25: 2d15882a0cbfce0b905039bebc811ac8311cd739.patch
|
||||
Patch26: c785d309dcbdeb7ecd219975244f3944a8d047e9.patch
|
||||
Patch27: b6a64416ab31b66ce92cabcc7fa1f3c5e9db2e87.patch
|
||||
Patch28: 9c86f6255f67a8bae28cd46c54500fc16bfc7a30.patch
|
||||
Patch29: 9bd84aef87978b806178a73ed33c39d6c442fc1f.patch
|
||||
Patch30: 885e546add918457c453bd3f753ac7df90b39e36.patch
|
||||
|
||||
ExcludeArch: s390 s390x
|
||||
BuildRequires: make
|
||||
@ -95,6 +98,9 @@ an utility for reporting current error counts from the EDAC sysfs files.
|
||||
%patch25 -p1
|
||||
%patch26 -p1
|
||||
%patch27 -p1
|
||||
%patch28 -p1
|
||||
%patch29 -p1
|
||||
%patch30 -p1
|
||||
|
||||
# The tarball is locked in time the first time aclocal was ran and will keep
|
||||
# requiring an older version of automake
|
||||
@ -130,6 +136,9 @@ sed -i "s/^PAGE_CE_ACTION=.*/PAGE_CE_ACTION=account/" %{buildroot}/%{_sysconfdir
|
||||
%{_sysconfdir}/sysconfig/rasdaemon
|
||||
|
||||
%changelog
|
||||
* Wed May 08 2024 Aristeu Rozanski <aris@redhat.com> 0.6.7-11
|
||||
- Fix excessive block messages [RHEL-8708]
|
||||
|
||||
* Wed Jan 10 2024 Aristeu Rozanski <aris@redhat.com> 0.6.7-10
|
||||
- Update License string to use SPDX [RHELMISC-1262]
|
||||
|
||||
|
||||
Loading…
Reference in New Issue
Block a user