From: Dan Williams <dan.j.williams@intel.com>
To: Neil Brown <neilb@suse.de>
Cc: "Neubauer, Wojciech" <Wojciech.Neubauer@intel.com>,
Doug Ledford <dledford@redhat.com>,
"Ciechanowski, Ed" <ed.ciechanowski@intel.com>,
"Hawrylewicz Czarnowski,
Przemyslaw" <przemyslaw.hawrylewicz.czarnowski@intel.com>,
"Labun, Marcin" <Marcin.Labun@intel.com>,
linux-raid <linux-raid@vger.kernel.org>,
"Jiang, Dave" <dave.jiang@intel.com>
Subject: fixes for 3.1.3 (was: Re: [mdadm GIT PULL] rebuild checkpoints...)
Date: Tue, 06 Jul 2010 12:51:49 -0700 [thread overview]
Message-ID: <1278445909.3949.11.camel@dwillia2-linux> (raw)
In-Reply-To: <20100706145043.107f7f76@notabene.brown>
On Mon, 2010-07-05 at 21:50 -0700, Neil Brown wrote:
> I'm fairly seriously considering cutting a 3.1.3 shortly (Thursday?) so that
> I have a clean slate to build the policy frame work and aim it for 3.2.0.
>
> If you have anything that you would like to see included in (or addressed
> for) 3.1.3, please let me know.
>
The pthread_create() vs clone(2) issue is still being discussed, but in
the meantime using pthreads seems the safe thing to do. So, here is
that implementation and another trivial build-warning fixup.
The following changes since commit 1538aca5cbbd99be47657e0ca0b7e2186426a1b1:
NeilBrown (1):
Merge branch 'master' of git://github.com/djbw/mdadm
are available in the git repository at:
git://github.com/djbw/mdadm.git master
Dan Williams (2):
mdmon: satisfy glibc tls abi requirements with pthreads
imsm: fix a -O2 build warning
Makefile | 14 +++++++++++---
mdmon.c | 42 +++++++++++++++++++++++++++++++++++++++---
super-intel.c | 4 ++--
3 files changed, 52 insertions(+), 8 deletions(-)
commit f4190c2f12527e37304f7c185afa0449fa9dee1c
Author: Dan Williams <dan.j.williams@intel.com>
Date: Tue Jul 6 12:48:56 2010 -0700
mdmon: satisfy glibc tls abi requirements with pthreads
Setting up a proper tls descriptor is required to conform to the abi
[1]. Until it can be implemented in mdmon use pthreads instead of
clone(2) to let glibc handle the details. The old behaviour can be had
by un-defining USE_PTHREADS.
Note, the "O2" builds need LDFLAGS now to pick up the '-pthread' option.
[1]: http://people.redhat.com/drepper/tls.pdf
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
diff --git a/Makefile b/Makefile
index 237f4fc..0f42f88 100644
--- a/Makefile
+++ b/Makefile
@@ -75,6 +75,14 @@ ALTFLAGS = -DALT_RUN=\"$(ALT_RUN)\" -DALT_MAPFILE=\"$(ALT_MAPFILE)\"
VARFLAGS = -DVAR_RUN=\"$(VAR_RUN)\"
CFLAGS = $(CWFLAGS) $(CXFLAGS) -DSendmail=\""$(MAILCMD)"\" $(CONFFILEFLAGS) $(ALTFLAGS) $(VARFLAGS)
+# The glibc TLS ABI requires applications that call clone(2) to set up
+# TLS data structures, use pthreads until mdmon implements this support
+USE_PTHREADS = 1
+ifdef USE_PTHREADS
+CFLAGS += -DUSE_PTHREADS
+LDFLAGS += -pthread
+endif
+
# If you want a static binary, you might uncomment these
# LDFLAGS = -static
# STRIP = -s
@@ -149,13 +157,13 @@ mdadm.klibc : $(SRCS) mdadm.h
$(CC) -nostdinc -iwithprefix include -I$(KLIBC)/klibc/include -I$(KLIBC)/linux/include -I$(KLIBC)/klibc/arch/i386/include -I$(KLIBC)/klibc/include/bits32 $(CFLAGS) $(SRCS)
mdadm.Os : $(SRCS) mdadm.h
- $(CC) -o mdadm.Os $(CFLAGS) -DHAVE_STDINT_H -Os $(SRCS)
+ $(CC) -o mdadm.Os $(CFLAGS) $(LDFLAGS) -DHAVE_STDINT_H -Os $(SRCS)
mdadm.O2 : $(SRCS) mdadm.h mdmon.O2
- $(CC) -o mdadm.O2 $(CFLAGS) -DHAVE_STDINT_H -O2 -D_FORTIFY_SOURCE=2 $(SRCS)
+ $(CC) -o mdadm.O2 $(CFLAGS) $(LDFLAGS) -DHAVE_STDINT_H -O2 -D_FORTIFY_SOURCE=2 $(SRCS)
mdmon.O2 : $(MON_SRCS) mdadm.h mdmon.h
- $(CC) -o mdmon.O2 $(CFLAGS) -DHAVE_STDINT_H -O2 -D_FORTIFY_SOURCE=2 $(MON_SRCS)
+ $(CC) -o mdmon.O2 $(CFLAGS) $(LDFLAGS) -DHAVE_STDINT_H -O2 -D_FORTIFY_SOURCE=2 $(MON_SRCS)
# use '-z now' to guarantee no dynamic linker interactions with the monitor thread
mdmon : $(MON_OBJS)
diff --git a/mdmon.c b/mdmon.c
index 0c37426..c4c0181 100644
--- a/mdmon.c
+++ b/mdmon.c
@@ -58,8 +58,11 @@
#include <fcntl.h>
#include <signal.h>
#include <dirent.h>
-
+#ifdef USE_PTHREADS
+#include <pthread.h>
+#else
#include <sched.h>
+#endif
#include "mdadm.h"
#include "mdmon.h"
@@ -71,7 +74,39 @@ int mon_tid, mgr_tid;
int sigterm;
-int run_child(void *v)
+#ifdef USE_PTHREADS
+static void *run_child(void *v)
+{
+ struct supertype *c = v;
+
+ mon_tid = syscall(SYS_gettid);
+ do_monitor(c);
+ return 0;
+}
+
+static int clone_monitor(struct supertype *container)
+{
+ pthread_attr_t attr;
+ pthread_t thread;
+ int rc;
+
+ mon_tid = -1;
+ pthread_attr_init(&attr);
+ pthread_attr_setstacksize(&attr, 4096);
+ pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_DETACHED);
+ rc = pthread_create(&thread, &attr, run_child, container);
+ if (rc)
+ return rc;
+ while (mon_tid == -1)
+ usleep(10);
+ pthread_attr_destroy(&attr);
+
+ mgr_tid = syscall(SYS_gettid);
+
+ return mon_tid;
+}
+#else /* USE_PTHREADS */
+static int run_child(void *v)
{
struct supertype *c = v;
@@ -85,7 +120,7 @@ int __clone2(int (*fn)(void *),
int flags, void *arg, ...
/* pid_t *pid, struct user_desc *tls, pid_t *ctid */ );
#endif
- int clone_monitor(struct supertype *container)
+static int clone_monitor(struct supertype *container)
{
static char stack[4096];
@@ -103,6 +138,7 @@ int __clone2(int (*fn)(void *),
return mon_tid;
}
+#endif /* USE_PTHREADS */
static int make_pidfile(char *devname)
{
commit 569cc43ffb0634510defee91407d261555c7a991
Author: Dan Williams <dan.j.williams@intel.com>
Date: Tue Jul 6 12:48:59 2010 -0700
imsm: fix a -O2 build warning
super-intel.c: In function ‘imsm_add_spare’:
super-intel.c:4833: error: ‘array_start’ may be used uninitialized in this function
super-intel.c:4834: error: ‘array_end’ may be used uninitialized in this function
This is valid, if we don't find a spare candidate then array_{start,end}
will be uninitialized.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
diff --git a/super-intel.c b/super-intel.c
index daf811f..6826d9b 100644
--- a/super-intel.c
+++ b/super-intel.c
@@ -4830,8 +4830,8 @@ static struct dl *imsm_add_spare(struct intel_super *super, int slot,
struct extent *ex;
int i, j;
int found;
- __u32 array_start;
- __u32 array_end;
+ __u32 array_start = 0;
+ __u32 array_end = 0;
struct dl *dl;
for (dl = super->disks; dl; dl = dl->next) {
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2010-07-06 19:51 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-05-27 0:50 [mdadm GIT PULL] rebuild checkpoints, incremental assembly, volume delete/rename, and fixes Dan Williams
2010-05-31 1:37 ` Neil Brown
2010-06-11 6:42 ` Dan Williams
2010-06-16 6:33 ` Neil Brown
2010-07-02 0:56 ` Dan Williams
2010-07-06 4:50 ` Neil Brown
2010-07-06 19:51 ` Dan Williams [this message]
2010-07-21 18:04 ` fixes for 3.1.3 (was: Re: [mdadm GIT PULL] rebuild checkpoints...) Dan Williams
2010-07-22 7:47 ` Neil Brown
2010-07-06 21:43 ` [mdadm GIT PULL] rebuild checkpoints, incremental assembly, volume delete/rename, and fixes Doug Ledford
2010-07-06 22:17 ` Neil Brown
2010-07-07 14:03 ` Doug Ledford
2010-07-08 7:50 ` Neil Brown
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=1278445909.3949.11.camel@dwillia2-linux \
--to=dan.j.williams@intel.com \
--cc=Marcin.Labun@intel.com \
--cc=Wojciech.Neubauer@intel.com \
--cc=dave.jiang@intel.com \
--cc=dledford@redhat.com \
--cc=ed.ciechanowski@intel.com \
--cc=linux-raid@vger.kernel.org \
--cc=neilb@suse.de \
--cc=przemyslaw.hawrylewicz.czarnowski@intel.com \
/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).