Re: [lxc-devel] lxc_monitord - monitor exiting
On Mon, 6 May 2013 19:08:08 -0500 Serge Hallyn wrote: > Quoting Dwight Engen (dwight.en...@oracle.com): > > On Mon, 6 May 2013 15:31:14 -0500 > > Serge Hallyn wrote: > > > > > Quoting Dwight Engen (dwight.en...@oracle.com): > > > > On Mon, 6 May 2013 13:06:43 -0400 > > > > Dwight Engen wrote: > > > > > > > > > Hi Çağlar, > > > > > > > > > > Thanks for the test program, I can sort of recreate it here > > > > > with that, although once I lxc-stop them all, lxc-monitord > > > > > does go away. I put a debug in lxc_wait() to see that the > > > > > client always close the fd to the monitord and they all were > > > > > closed so I'm not sure why lxc-monitord isn't seeing the > > > > > accepted fd coming back from epoll to close. Still > > > > > investigating... > > > > > > > > Okay, so I debugged this and the problem is basically down to > > > > lxc not being thread aware. With your test program we get > > > > multiple threads in lxcapi_start() simultaneously in the > > > > daemonize case. One of them forks while another one has the > > > > client fd to the monitor open and thus the fd gets duped by the > > > > fork and that is the client fd that holds lxc-monitord open > > > > until the container shuts down. > > > > > > > > Çağlar you could try out the following patch, it essentially > > > > serializes container startup from a thread perspective. I > > > > haven't tested it thoroughly, but it did fix the problem here. > > > > Right now lxc doesn't support threaded use, so you may run into > > > > other things as well. Depending on our stance on thread support > > > > in lxc, you may need to do the serialization in the threaded > > > > app. I guess another alternative is that initially we could > > > > just thread serialize at the API (big lxc lock). > > > > > > It sounds like lxcapi_start should be locking c->slock? The > > > priv_lock is to protect the struct lxccontainer in memory (for > > > when you do c = lxc_container_new(); then clone a new thread), > > > while the slock is to protect the container data. It's being > > > taken at create, info, destroy, etc, but not at start. > > > > Hi Serge, thanks for pointing those out, lxc is more thread aware > > than I realized :) Unfortunately I don't think either lock will > > help here as both these locks are per-container and the data we > > need to synchronize is per-calling process (ie. which fd's are open > > at time of fork). To put it another way, this problem happens even > > when each c is different. > > Right you are. I was so busy worrying about how to protect the > on-disk container and in-memory container_struct I wasn't thinking > about the shared fdtable. > > Should we have a single mutex around all fd modifications? What else? I think that will certainly work here, I haven't considered other places. Do we need to think about sighand and umask? Would you prefer to use the semaphores lxc already has (ie lxc_newlock() with no name) over pthread_mutex? Since Çağlar verified the test fix worked for him also I'd like to make a real fix. -- Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. This 200-page book is written by three acclaimed leaders in the field. The early access version is available now. Download your free book today! http://p.sf.net/sfu/neotech_d2d_may ___ Lxc-devel mailing list Lxc-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lxc-devel
Re: [lxc-devel] lxc_monitord - monitor exiting
Quoting Dwight Engen (dwight.en...@oracle.com): > On Mon, 6 May 2013 19:08:08 -0500 > Serge Hallyn wrote: > > > Quoting Dwight Engen (dwight.en...@oracle.com): > > > On Mon, 6 May 2013 15:31:14 -0500 > > > Serge Hallyn wrote: > > > > > > > Quoting Dwight Engen (dwight.en...@oracle.com): > > > > > On Mon, 6 May 2013 13:06:43 -0400 > > > > > Dwight Engen wrote: > > > > > > > > > > > Hi Çağlar, > > > > > > > > > > > > Thanks for the test program, I can sort of recreate it here > > > > > > with that, although once I lxc-stop them all, lxc-monitord > > > > > > does go away. I put a debug in lxc_wait() to see that the > > > > > > client always close the fd to the monitord and they all were > > > > > > closed so I'm not sure why lxc-monitord isn't seeing the > > > > > > accepted fd coming back from epoll to close. Still > > > > > > investigating... > > > > > > > > > > Okay, so I debugged this and the problem is basically down to > > > > > lxc not being thread aware. With your test program we get > > > > > multiple threads in lxcapi_start() simultaneously in the > > > > > daemonize case. One of them forks while another one has the > > > > > client fd to the monitor open and thus the fd gets duped by the > > > > > fork and that is the client fd that holds lxc-monitord open > > > > > until the container shuts down. > > > > > > > > > > Çağlar you could try out the following patch, it essentially > > > > > serializes container startup from a thread perspective. I > > > > > haven't tested it thoroughly, but it did fix the problem here. > > > > > Right now lxc doesn't support threaded use, so you may run into > > > > > other things as well. Depending on our stance on thread support > > > > > in lxc, you may need to do the serialization in the threaded > > > > > app. I guess another alternative is that initially we could > > > > > just thread serialize at the API (big lxc lock). > > > > > > > > It sounds like lxcapi_start should be locking c->slock? The > > > > priv_lock is to protect the struct lxccontainer in memory (for > > > > when you do c = lxc_container_new(); then clone a new thread), > > > > while the slock is to protect the container data. It's being > > > > taken at create, info, destroy, etc, but not at start. > > > > > > Hi Serge, thanks for pointing those out, lxc is more thread aware > > > than I realized :) Unfortunately I don't think either lock will > > > help here as both these locks are per-container and the data we > > > need to synchronize is per-calling process (ie. which fd's are open > > > at time of fork). To put it another way, this problem happens even > > > when each c is different. > > > > Right you are. I was so busy worrying about how to protect the > > on-disk container and in-memory container_struct I wasn't thinking > > about the shared fdtable. > > > > Should we have a single mutex around all fd modifications? What else? > > I think that will certainly work here, I haven't considered other > places. Do we need to think about sighand and umask? Would you prefer Yeah we should. (And I need to decide how to fix the fact that c->slock is just called the container name, regardless the lxcpath) > to use the semaphores lxc already has (ie lxc_newlock() with no name) Actually we're doing different things there, so I'm happy to use pthread_mutex, unless it'll cause trouble for people using different toolchains (i.e. android/bionic). The semaphores we're using are to protect the container_struct itself, and to protect the on-disk container data. For protecting the task data itself, pthread_mutex seems perfect. (And no name is needed.) > over pthread_mutex? Since Çağlar verified the test fix worked for him > also I'd like to make a real fix. Should I push the patch from yesterday as a first step, or would you prefer to do all at once with one patch? I'm fine either way. thanks, -serge PS - I'm in meetings all week (a very rare thing for me), so apologies if I'm slow in responding. -- Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. This 200-page book is written by three acclaimed leaders in the field. The early access version is available now. Download your free book today! http://p.sf.net/sfu/neotech_d2d_may ___ Lxc-devel mailing list Lxc-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lxc-devel
[lxc-devel] [PATCH 1/6] coverity: free malloc'ed memory in error case
Signed-off-by: Dwight Engen
---
src/lxc/bdev.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/src/lxc/bdev.c b/src/lxc/bdev.c
index 35351a9..1a611f9 100644
--- a/src/lxc/bdev.c
+++ b/src/lxc/bdev.c
@@ -1203,6 +1203,8 @@ static int overlayfs_clonepaths(struct bdev *orig, struct
bdev *new, const char
return -ENOMEM;
}
if (do_rsync(odelta, ndelta) < 0) {
+ free(osrc);
+ free(ndelta);
ERROR("copying overlayfs delta");
return -1;
}
--
1.8.1.4
--
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and
their applications. This 200-page book is written by three acclaimed
leaders in the field. The early access version is available now.
Download your free book today! http://p.sf.net/sfu/neotech_d2d_may
___
Lxc-devel mailing list
Lxc-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-devel
[lxc-devel] [PATCH 2/6] coverity: open can return 0 as an fd, change error check to < 0
Signed-off-by: Dwight Engen --- src/lxc/bdev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/lxc/bdev.c b/src/lxc/bdev.c index 1a611f9..181e98e 100644 --- a/src/lxc/bdev.c +++ b/src/lxc/bdev.c @@ -90,7 +90,7 @@ static int blk_getsize(const char *path, unsigned long *size) int fd, ret; fd = open(path, O_RDONLY); - if (!fd) + if (fd < 0) return -1; ret = ioctl(fd, BLKGETSIZE64, size); close(fd); -- 1.8.1.4 -- Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. This 200-page book is written by three acclaimed leaders in the field. The early access version is available now. Download your free book today! http://p.sf.net/sfu/neotech_d2d_may ___ Lxc-devel mailing list Lxc-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lxc-devel
[lxc-devel] [PATCH 4/6] coverity: clonetest: check correct container is cloned
Signed-off-by: Dwight Engen
---
src/tests/clonetest.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/tests/clonetest.c b/src/tests/clonetest.c
index fcb5ea6..5644405 100644
--- a/src/tests/clonetest.c
+++ b/src/tests/clonetest.c
@@ -70,7 +70,7 @@ int main(int argc, char *argv[])
goto out;
}
- if (!c2->is_defined(c)) {
+ if (!c2->is_defined(c2)) {
fprintf(stderr, "%d: %s not defined after clone\n", __LINE__,
MYNAME2);
goto out;
}
--
1.8.1.4
--
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and
their applications. This 200-page book is written by three acclaimed
leaders in the field. The early access version is available now.
Download your free book today! http://p.sf.net/sfu/neotech_d2d_may
___
Lxc-devel mailing list
Lxc-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-devel
[lxc-devel] [PATCH 3/6] coverity: condition already checked for
Signed-off-by: Dwight Engen
---
src/lxc/bdev.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/src/lxc/bdev.c b/src/lxc/bdev.c
index 181e98e..fcde16b 100644
--- a/src/lxc/bdev.c
+++ b/src/lxc/bdev.c
@@ -1151,8 +1151,7 @@ static int overlayfs_clonepaths(struct bdev *orig, struct
bdev *new, const char
if (strcmp(orig->type, "dir") == 0) {
char *delta;
int ret, len;
- if (!snap)
- return -1;
+
// if we have /var/lib/lxc/c2/rootfs, then delta will be
///var/lib/lxc/c2/delta0
delta = strdup(new->dest);
--
1.8.1.4
--
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and
their applications. This 200-page book is written by three acclaimed
leaders in the field. The early access version is available now.
Download your free book today! http://p.sf.net/sfu/neotech_d2d_may
___
Lxc-devel mailing list
Lxc-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-devel
[lxc-devel] [PATCH 6/6] coverity: fix potential dereference NULL returned from malloc
Signed-off-by: Dwight Engen
---
src/lxc/network.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/src/lxc/network.c b/src/lxc/network.c
index 93fc169..d1ccc0d 100644
--- a/src/lxc/network.c
+++ b/src/lxc/network.c
@@ -783,8 +783,11 @@ static int ifa_get_local_ip(int family, struct ip_req
*ip_info, void** res) {
/* We might have found an IFA_ADDRESS before,
* which we now overwrite with an IFA_LOCAL. */
- if (!*res)
+ if (!*res) {
*res = malloc(addrlen);
+ if (!*res)
+ return -1;
+ }
memcpy(*res, RTA_DATA(rta), addrlen);
--
1.8.1.4
--
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and
their applications. This 200-page book is written by three acclaimed
leaders in the field. The early access version is available now.
Download your free book today! http://p.sf.net/sfu/neotech_d2d_may
___
Lxc-devel mailing list
Lxc-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-devel
[lxc-devel] [PATCH 5/6] coverity: check return from waitpid
Signed-off-by: Dwight Engen
---
src/lxc/monitor.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/src/lxc/monitor.c b/src/lxc/monitor.c
index 0521e9a..5648b86 100644
--- a/src/lxc/monitor.c
+++ b/src/lxc/monitor.c
@@ -207,7 +207,8 @@ int lxc_monitord_spawn(const char *lxcpath)
}
if (pid1) {
- waitpid(pid1, NULL, 0);
+ if (waitpid(pid1, NULL, 0) != pid1)
+ return -1;
return 0;
}
--
1.8.1.4
--
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and
their applications. This 200-page book is written by three acclaimed
leaders in the field. The early access version is available now.
Download your free book today! http://p.sf.net/sfu/neotech_d2d_may
___
Lxc-devel mailing list
Lxc-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-devel
Re: [lxc-devel] [PATCH 1/6] coverity: free malloc'ed memory in error case
Thanks, Dwight.
Acked-by: Serge E. Hallyn
to the whole set.
Quoting Dwight Engen (dwight.en...@oracle.com):
> Signed-off-by: Dwight Engen
> ---
> src/lxc/bdev.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/src/lxc/bdev.c b/src/lxc/bdev.c
> index 35351a9..1a611f9 100644
> --- a/src/lxc/bdev.c
> +++ b/src/lxc/bdev.c
> @@ -1203,6 +1203,8 @@ static int overlayfs_clonepaths(struct bdev *orig,
> struct bdev *new, const char
> return -ENOMEM;
> }
> if (do_rsync(odelta, ndelta) < 0) {
> + free(osrc);
> + free(ndelta);
> ERROR("copying overlayfs delta");
> return -1;
> }
> --
> 1.8.1.4
>
>
>
> --
> Learn Graph Databases - Download FREE O'Reilly Book
> "Graph Databases" is the definitive new guide to graph databases and
> their applications. This 200-page book is written by three acclaimed
> leaders in the field. The early access version is available now.
> Download your free book today! http://p.sf.net/sfu/neotech_d2d_may
> ___
> Lxc-devel mailing list
> Lxc-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/lxc-devel
--
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and
their applications. This 200-page book is written by three acclaimed
leaders in the field. The early access version is available now.
Download your free book today! http://p.sf.net/sfu/neotech_d2d_may
___
Lxc-devel mailing list
Lxc-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-devel
Re: [lxc-devel] [PATCH] lxc-monitor multiple paths
On Fri, 26 Apr 2013 09:31:08 -0500 Serge Hallyn wrote: > Quoting Dwight Engen (dwight.en...@oracle.com): > > >From c4af03c0e6b74e550161e4c8382fb9d0889e4fe3 Mon Sep 17 00:00:00 > > >2001 > > From: Dwight Engen > > Date: Thu, 25 Apr 2013 11:45:26 -0400 > > Subject: [PATCH] lxc-monitor multiple paths > > > > Signed-off-by: Dwight Engen > > Acked-by: Serge E. Hallyn > > I wonder if it's worth adding a 'numlxcpaths' to arguments, so that > lxc-attach and the others can set that to 1, and if a second lxcpath > argument is found, it immediately returns failure? Right now the > extra lxcpaths will simply be ignored. Hi Serge, this has not been applied yet, so I updated it against head and added a lxcpaths_additional to arguments like you suggested. I made it the count of additional -P arguments to accept so most programs don't have to initialize it. Patch to follow. -- Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. This 200-page book is written by three acclaimed leaders in the field. The early access version is available now. Download your free book today! http://p.sf.net/sfu/neotech_d2d_may ___ Lxc-devel mailing list Lxc-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lxc-devel
[lxc-devel] [PATCH v2] lxc-monitor multiple paths
Signed-off-by: Dwight Engen
---
doc/lxc-monitor.sgml.in | 7 ++
src/lxc/arguments.c | 34 +++--
src/lxc/arguments.h | 5 -
src/lxc/lxc.h| 31 +-
src/lxc/lxc_attach.c | 8 +++
src/lxc/lxc_cgroup.c | 6 ++---
src/lxc/lxc_checkpoint.c | 2 +-
src/lxc/lxc_console.c| 4 ++--
src/lxc/lxc_execute.c| 6 ++---
src/lxc/lxc_freeze.c | 4 ++--
src/lxc/lxc_info.c | 6 ++---
src/lxc/lxc_kill.c | 4 ++--
src/lxc/lxc_monitor.c| 35 +
src/lxc/lxc_restart.c| 6 ++---
src/lxc/lxc_start.c | 6 ++---
src/lxc/lxc_stop.c | 4 ++--
src/lxc/lxc_unfreeze.c | 4 ++--
src/lxc/lxc_wait.c | 5 +++--
src/lxc/monitor.c| 57 +++-
19 files changed, 165 insertions(+), 69 deletions(-)
diff --git a/doc/lxc-monitor.sgml.in b/doc/lxc-monitor.sgml.in
index 336061d..b460843 100644
--- a/doc/lxc-monitor.sgml.in
+++ b/doc/lxc-monitor.sgml.in
@@ -63,6 +63,13 @@ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
02111-1307 USA
to monitor all the containers, several of them or just one.
+
+ The -P, --lxcpath=PATH option may be specified multiple
+ times to monitor more than one container path. Note however that
+ containers with the same name in multiple paths will be
+ indistinguishable in the output.
+
+
&commonoptions;
diff --git a/src/lxc/arguments.c b/src/lxc/arguments.c
index f61c6eb..5f1c1af 100644
--- a/src/lxc/arguments.c
+++ b/src/lxc/arguments.c
@@ -150,13 +150,32 @@ See the %s man page for further information.\n\n",
exit(code);
}
+static int lxc_arguments_lxcpath_add(struct lxc_arguments *args,
+const char *lxcpath)
+{
+ if (args->lxcpath_additional != -1 &&
+ args->lxcpath_cnt > args->lxcpath_additional) {
+ fprintf(stderr, "This command only accepts %d -P,--lxcpath
arguments\n",
+ args->lxcpath_additional + 1);
+ exit(EXIT_FAILURE);
+ }
+
+ args->lxcpath = realloc(args->lxcpath, (args->lxcpath_cnt + 1) *
+sizeof(args->lxcpath[0]));
+ if (args->lxcpath == NULL) {
+ lxc_error(args, "no memory");
+ return ENOMEM;
+ }
+ args->lxcpath[args->lxcpath_cnt++] = lxcpath;
+ return 0;
+}
+
extern int lxc_arguments_parse(struct lxc_arguments *args,
int argc, char * const argv[])
{
char shortopts[256];
int ret = 0;
- args->lxcpath = default_lxc_path();
ret = build_shortopts(args->options, shortopts, sizeof(shortopts));
if (ret < 0) {
lxc_error(args, "build_shortopts() failed : %s",
@@ -176,7 +195,11 @@ extern int lxc_arguments_parse(struct lxc_arguments *args,
case 'l': args->log_priority = optarg; break;
case 'c': args->console = optarg; break;
case 'q': args->quiet = 1; break;
- case 'P': args->lxcpath = optarg; break;
+ case 'P':
+ ret = lxc_arguments_lxcpath_add(args, optarg);
+ if (ret < 0)
+ return ret;
+ break;
case OPT_USAGE: print_usage(args->options, args);
case '?': print_help(args, 1);
case 'h': print_help(args, 0);
@@ -195,6 +218,13 @@ extern int lxc_arguments_parse(struct lxc_arguments *args,
args->argv = &argv[optind];
args->argc = argc - optind;
+ /* If no lxcpaths were given, use default */
+ if (!args->lxcpath_cnt) {
+ ret = lxc_arguments_lxcpath_add(args, default_lxc_path());
+ if (ret < 0)
+ return ret;
+ }
+
/* Check the command options */
if (!args->name) {
diff --git a/src/lxc/arguments.h b/src/lxc/arguments.h
index 6f6826b..002a919 100644
--- a/src/lxc/arguments.h
+++ b/src/lxc/arguments.h
@@ -47,7 +47,10 @@ struct lxc_arguments {
const char *console;
const char *console_log;
const char *pidfile;
- const char *lxcpath;
+ const char **lxcpath;
+ int lxcpath_cnt;
+ /* set to 0 to accept only 1 lxcpath, -1 for unlimited */
+ int lxcpath_additional;
/* for lxc-checkpoint/restart */
const char *statefile;
diff --git a/src/lxc/lxc.h b/src/lxc/lxc.h
index db921f0..9057757 100644
--- a/src/lxc/lxc.h
+++ b/src/lxc/lxc.h
@@ -28,6 +28,7 @@ extern "C" {
#endif
#include
+#include
#include
struct lxc_msg;
@@ -77,17 +78,37 @@ extern int lxc_execute(const char *name, char *const
argv[], int quiet,
extern int lxc_monitor_open(const char *lxcpath);
/*
- * Read the state of the container if this one has changed
- * The
Re: [lxc-devel] [PATCH v2] lxc-monitor multiple paths
Quoting Dwight Engen (dwight.en...@oracle.com):
> Signed-off-by: Dwight Engen
Acked-by: Serge E. Hallyn
> ---
> doc/lxc-monitor.sgml.in | 7 ++
> src/lxc/arguments.c | 34 +++--
> src/lxc/arguments.h | 5 -
> src/lxc/lxc.h| 31 +-
> src/lxc/lxc_attach.c | 8 +++
> src/lxc/lxc_cgroup.c | 6 ++---
> src/lxc/lxc_checkpoint.c | 2 +-
> src/lxc/lxc_console.c| 4 ++--
> src/lxc/lxc_execute.c| 6 ++---
> src/lxc/lxc_freeze.c | 4 ++--
> src/lxc/lxc_info.c | 6 ++---
> src/lxc/lxc_kill.c | 4 ++--
> src/lxc/lxc_monitor.c| 35 +
> src/lxc/lxc_restart.c| 6 ++---
> src/lxc/lxc_start.c | 6 ++---
> src/lxc/lxc_stop.c | 4 ++--
> src/lxc/lxc_unfreeze.c | 4 ++--
> src/lxc/lxc_wait.c | 5 +++--
> src/lxc/monitor.c| 57
> +++-
> 19 files changed, 165 insertions(+), 69 deletions(-)
>
> diff --git a/doc/lxc-monitor.sgml.in b/doc/lxc-monitor.sgml.in
> index 336061d..b460843 100644
> --- a/doc/lxc-monitor.sgml.in
> +++ b/doc/lxc-monitor.sgml.in
> @@ -63,6 +63,13 @@ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
> 02111-1307 USA
>to monitor all the containers, several of them or just one.
>
>
> +
> + The -P, --lxcpath=PATH option may be specified
> multiple
> + times to monitor more than one container path. Note however that
> + containers with the same name in multiple paths will be
> + indistinguishable in the output.
> +
> +
>
>
>&commonoptions;
> diff --git a/src/lxc/arguments.c b/src/lxc/arguments.c
> index f61c6eb..5f1c1af 100644
> --- a/src/lxc/arguments.c
> +++ b/src/lxc/arguments.c
> @@ -150,13 +150,32 @@ See the %s man page for further information.\n\n",
> exit(code);
> }
>
> +static int lxc_arguments_lxcpath_add(struct lxc_arguments *args,
> + const char *lxcpath)
> +{
> + if (args->lxcpath_additional != -1 &&
> + args->lxcpath_cnt > args->lxcpath_additional) {
> + fprintf(stderr, "This command only accepts %d -P,--lxcpath
> arguments\n",
> + args->lxcpath_additional + 1);
> + exit(EXIT_FAILURE);
> + }
> +
> + args->lxcpath = realloc(args->lxcpath, (args->lxcpath_cnt + 1) *
> + sizeof(args->lxcpath[0]));
> + if (args->lxcpath == NULL) {
> + lxc_error(args, "no memory");
> + return ENOMEM;
> + }
> + args->lxcpath[args->lxcpath_cnt++] = lxcpath;
> + return 0;
> +}
> +
> extern int lxc_arguments_parse(struct lxc_arguments *args,
> int argc, char * const argv[])
> {
> char shortopts[256];
> int ret = 0;
>
> - args->lxcpath = default_lxc_path();
> ret = build_shortopts(args->options, shortopts, sizeof(shortopts));
> if (ret < 0) {
> lxc_error(args, "build_shortopts() failed : %s",
> @@ -176,7 +195,11 @@ extern int lxc_arguments_parse(struct lxc_arguments
> *args,
> case 'l': args->log_priority = optarg; break;
> case 'c': args->console = optarg; break;
> case 'q': args->quiet = 1; break;
> - case 'P': args->lxcpath = optarg; break;
> + case 'P':
> + ret = lxc_arguments_lxcpath_add(args, optarg);
> + if (ret < 0)
> + return ret;
> + break;
> case OPT_USAGE: print_usage(args->options, args);
> case '?': print_help(args, 1);
> case 'h': print_help(args, 0);
> @@ -195,6 +218,13 @@ extern int lxc_arguments_parse(struct lxc_arguments
> *args,
> args->argv = &argv[optind];
> args->argc = argc - optind;
>
> + /* If no lxcpaths were given, use default */
> + if (!args->lxcpath_cnt) {
> + ret = lxc_arguments_lxcpath_add(args, default_lxc_path());
> + if (ret < 0)
> + return ret;
> + }
> +
> /* Check the command options */
>
> if (!args->name) {
> diff --git a/src/lxc/arguments.h b/src/lxc/arguments.h
> index 6f6826b..002a919 100644
> --- a/src/lxc/arguments.h
> +++ b/src/lxc/arguments.h
> @@ -47,7 +47,10 @@ struct lxc_arguments {
> const char *console;
> const char *console_log;
> const char *pidfile;
> - const char *lxcpath;
> + const char **lxcpath;
> + int lxcpath_cnt;
> + /* set to 0 to accept only 1 lxcpath, -1 for unlimited */
> + int lxcpath_additional;
>
> /* for lxc-checkpoint/restart */
> const char *statefile;
> diff --git a/src/lxc/lxc.h b/src/lxc/lxc.h
> index db921f0..9057757 100644
> --- a/src/lxc/lxc.h
> +++ b/src/lxc/lxc.h
> @@ -28,6 +28,7 @@ extern "C" {
> #endif
>
> #include
> +#include
>
[lxc-devel] [PATCH] lxc-ps: handle cgroup collisions
A few months ago cgroup handling in lxc was updated so that if
/sys/fs/cgroup/$cgroup/lxc/$container already exists (most often
due to another container by the same name under a different lxcpath),
then /sys/fs/cgroup/$cgroup/lxc/${container}-N would be used.
lxc-ps was never updated to handle this. Fix that.
(Note, the ns cgroup is being special cased there, but I don't
really believe ns cgroup works any more.)
It would be preferable to rewrite lxc-ps in python or in C, but
this at least makes the basic lxc-ps work in the case of multiple
containers with the same name.
Signed-off-by: Serge Hallyn
---
src/lxc/lxc-ps.in | 18 +-
1 file changed, 13 insertions(+), 5 deletions(-)
diff --git a/src/lxc/lxc-ps.in b/src/lxc/lxc-ps.in
index 55a05ce..eaf5002 100644
--- a/src/lxc/lxc-ps.in
+++ b/src/lxc/lxc-ps.in
@@ -17,9 +17,11 @@
# License along with this library; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+. @DATADIR@/lxc/lxc.functions
+
usage()
{
-echo "usage: $(basename $0) [--lxc | --host | --name NAME] [--]
[PS_OPTIONS...]" >&2
+echo "usage: $(basename $0) [-P PATH] [--lxc | --host | --name NAME] [[--]
[PS_OPTIONS...]" >&2
}
help() {
@@ -31,6 +33,7 @@ help() {
echo " --hostshow processes not related to any container, i.e. to
the host" >&2
echo " --name NAME show processes in the specified container" >&2
echo " (multiple containers can be separated by commas)"
>&2
+echo " -P PATH show container in lxcpath PATH" >&2
echo " PS_OPTIONSps command options (see \`ps --help')" >&2
}
@@ -65,7 +68,7 @@ get_parent_cgroup()
# (do not append '/lxc' if the hierarchy contains the 'ns' subsystem)
case ",$subsystems," in
*,ns,*) parent_cgroup="${mountpoint}${init_cgroup%/}";;
-*) parent_cgroup="${mountpoint}${init_cgroup%/}/lxc";;
+*) parent_cgroup="${mountpoint}${init_cgroup%/}";;
esac
break
done
@@ -83,6 +86,8 @@ while true; do
list_container_processes=1; shift;;
--host)
list_container_processes=-1; shift;;
+-P|--lxcpath)
+lxc_path=$2; shift 2;;
--)
shift; break;;
*)
@@ -111,9 +116,12 @@ for container in ${containers}; do
container_field_width=${#container}
fi
-if [ -f "$parent_cgroup/$container/tasks" ]; then
-tasks_files="$tasks_files $parent_cgroup/$container/tasks"
-fi
+if lxc-info -P $lxc_path -t RUNNING -n z1; then
+initpid=`lxc-info -P $lxc_path -p -n $container | awk -F: '{ print $2
}' | awk '{ print $1 }'`
+cgroup=`head -n 1 /proc/$initpid/cgroup | awk -F: '{ print $3}'`
+if [ -f "$parent_cgroup/$cgroup/tasks" ]; then
+tasks_files="$tasks_files $parent_cgroup$cgroup/tasks"
+fi
done
# first file is stdin, the rest are the container tasks
--
1.8.1.2
--
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and
their applications. This 200-page book is written by three acclaimed
leaders in the field. The early access version is available now.
Download your free book today! http://p.sf.net/sfu/neotech_d2d_may
___
Lxc-devel mailing list
Lxc-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-devel
