LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Ian Kent <raven@themaw.net>
To: Kernel Mailing List <linux-kernel@vger.kernel.org>
Cc: David Howells <dhowells@redhat.com>,
	Oleg Nesterov <onestero@redhat.com>,
	Trond Myklebust <trond.myklebust@primarydata.com>,
	"J. Bruce Fields" <bfields@fieldses.org>,
	Benjamin Coddington <bcodding@redhat.com>,
	Al Viro <viro@ZenIV.linux.org.uk>,
	Jeff Layton <jeff.layton@primarydata.com>,
	"Eric W. Biederman" <ebiederm@xmission.com>
Subject: [RFC PATCH 5 2/7] kmod - teach usermodehelper to use service workqueues
Date: Tue, 31 Mar 2015 11:14:50 +0800	[thread overview]
Message-ID: <20150331031449.10464.73117.stgit@pluto.fritz.box> (raw)
In-Reply-To: <20150331030340.10464.30272.stgit@pluto.fritz.box>

The call_usermodehelper() function executes all binaries in the
global "init" root context. This doesn't allow a binary to be run
within a namespace environment such as a container. To do this
the namespace environment of the contaner must be available to
provide the required execution environment.

This can be done by creating a service thread, identified by
issuing a token identifier, used when executing the helper
with a function that takes the token as a parameter.

Signed-off-by: Ian Kent <ikent@redhat.com>
Cc: Benjamin Coddington <bcodding@redhat.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: J. Bruce Fields <bfields@fieldses.org>
Cc: David Howells <dhowells@redhat.com>
Cc: Trond Myklebust <trond.myklebust@primarydata.com>
Cc: Oleg Nesterov <onestero@redhat.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Jeff Layton <jeff.layton@primarydata.com>
---
 include/linux/kmod.h |    5 +++++
 kernel/kmod.c        |   38 +++++++++++++++++++++++++++++++-------
 2 files changed, 36 insertions(+), 7 deletions(-)

diff --git a/include/linux/kmod.h b/include/linux/kmod.h
index fa46722..9a9fcb3 100644
--- a/include/linux/kmod.h
+++ b/include/linux/kmod.h
@@ -56,6 +56,7 @@ struct file;
 struct subprocess_info {
 	struct work_struct work;
 	struct completion *complete;
+	int wq_token;
 	char *path;
 	char **argv;
 	char **envp;
@@ -72,6 +73,10 @@ extern void umh_wq_put_token(int token);
 extern int
 call_usermodehelper(char *path, char **argv, char **envp, int wait);
 
+extern int
+call_usermodehelper_service(char *path, char **argv,
+			    char **envp, int token, int wait);
+
 extern struct subprocess_info *
 call_usermodehelper_setup(char *path, char **argv, char **envp, gfp_t gfp_mask,
 			  int (*init)(struct subprocess_info *info, struct cred *new),
diff --git a/kernel/kmod.c b/kernel/kmod.c
index 55d20ce..47c5ff5 100644
--- a/kernel/kmod.c
+++ b/kernel/kmod.c
@@ -712,6 +712,7 @@ EXPORT_SYMBOL(call_usermodehelper_setup);
  */
 int call_usermodehelper_exec(struct subprocess_info *sub_info, int wait)
 {
+	static struct workqueue_struct *wq;
 	DECLARE_COMPLETION_ONSTACK(done);
 	int retval = 0;
 
@@ -720,7 +721,8 @@ int call_usermodehelper_exec(struct subprocess_info *sub_info, int wait)
 		return -EINVAL;
 	}
 	helper_lock();
-	if (!khelper_wq || usermodehelper_disabled) {
+	wq = umh_find_wq(sub_info->wq_token, wait);
+	if (!wq || usermodehelper_disabled) {
 		retval = -EBUSY;
 		goto out;
 	}
@@ -732,7 +734,7 @@ int call_usermodehelper_exec(struct subprocess_info *sub_info, int wait)
 	sub_info->complete = (wait == UMH_NO_WAIT) ? NULL : &done;
 	sub_info->wait = wait;
 
-	queue_work(khelper_wq, &sub_info->work);
+	queue_work(wq, &sub_info->work);
 	if (wait == UMH_NO_WAIT)	/* task has freed sub_info */
 		goto unlock;
 
@@ -759,19 +761,21 @@ unlock:
 EXPORT_SYMBOL(call_usermodehelper_exec);
 
 /**
- * call_usermodehelper() - prepare and start a usermode application
+ * call_usermodehelper_service() - start a usermode application within
+ * 				a service environment.
  * @path: path to usermode executable
  * @argv: arg vector for process
  * @envp: environment for process
+ * @token: token corresponding to a service environment obtained by a
+ * 	   call to umh_wq_get_token().
  * @wait: wait for the application to finish and return status.
  *        when UMH_NO_WAIT don't wait at all, but you get no useful error back
  *        when the program couldn't be exec'ed. This makes it safe to call
  *        from interrupt context.
- *
- * This function is the equivalent to use call_usermodehelper_setup() and
- * call_usermodehelper_exec().
  */
-int call_usermodehelper(char *path, char **argv, char **envp, int wait)
+int call_usermodehelper_service(char *path,
+				char **argv, char **envp,
+				int token, int wait)
 {
 	struct subprocess_info *info;
 	gfp_t gfp_mask = (wait == UMH_NO_WAIT) ? GFP_ATOMIC : GFP_KERNEL;
@@ -780,9 +784,29 @@ int call_usermodehelper(char *path, char **argv, char **envp, int wait)
 					 NULL, NULL, NULL);
 	if (info == NULL)
 		return -ENOMEM;
+	info->wq_token = token;
 
 	return call_usermodehelper_exec(info, wait);
 }
+EXPORT_SYMBOL(call_usermodehelper_service);
+
+/**
+ * call_usermodehelper() - prepare and start a usermode application
+ * @path: path to usermode executable
+ * @argv: arg vector for process
+ * @envp: environment for process
+ * @wait: wait for the application to finish and return status.
+ *        when UMH_NO_WAIT don't wait at all, but you get no useful error back
+ *        when the program couldn't be exec'ed. This makes it safe to call
+ *        from interrupt context.
+ *
+ * This function is the equivalent to using call_usermodehelper_setup() and
+ * call_usermodehelper_exec().
+ */
+int call_usermodehelper(char *path, char **argv, char **envp, int wait)
+{
+	return call_usermodehelper_service(path, argv, envp, 0, wait);
+}
 EXPORT_SYMBOL(call_usermodehelper);
 
 static int proc_cap_handler(struct ctl_table *table, int write,


  parent reply	other threads:[~2015-03-31  3:15 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-31  3:14 [RFC PATCH v5 0/7] Another attempt at contained helper execution Ian Kent
2015-03-31  3:14 ` [RFC PATCH 5 1/7] kmod - add workqueue service thread store Ian Kent
2015-03-31 11:21   ` Jeff Layton
2015-03-31 12:59     ` Ian Kent
2015-03-31  3:14 ` Ian Kent [this message]
2015-03-31  3:14 ` [RFC PATCH 5 3/7] nfsd - use service thread if not executing in init namespace Ian Kent
2015-03-31 13:14   ` J. Bruce Fields
2015-04-01  0:22     ` Ian Kent
2015-04-02 15:59       ` J. Bruce Fields
2015-03-31  3:15 ` [RFC PATCH 5 4/7] nfs - cache_lib " Ian Kent
2015-03-31  3:15 ` [RFC PATCH 5 5/7] nfs - objlayout " Ian Kent
2015-03-31  3:15 ` [RFC PATCH 5 6/7] KEYS - use correct memory allocation flag in call_usermodehelper_keys() Ian Kent
2015-03-31  3:15 ` [RFC PATCH 5 7/7] KEYS: exec request key within service thread of key creator Ian Kent
2015-04-02 12:43 ` [RFC PATCH 5 1/7] kmod - add workqueue service thread store David Howells
2015-04-07  0:42   ` Ian Kent
2015-04-02 12:58 ` [RFC PATCH 5 7/7] KEYS: exec request key within service thread of key creator David Howells
2015-04-07  0:54   ` Ian Kent
2015-04-02 13:00 ` [RFC PATCH 5 6/7] KEYS - use correct memory allocation flag in call_usermodehelper_keys() David Howells

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=20150331031449.10464.73117.stgit@pluto.fritz.box \
    --to=raven@themaw.net \
    --cc=bcodding@redhat.com \
    --cc=bfields@fieldses.org \
    --cc=dhowells@redhat.com \
    --cc=ebiederm@xmission.com \
    --cc=jeff.layton@primarydata.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=onestero@redhat.com \
    --cc=trond.myklebust@primarydata.com \
    --cc=viro@ZenIV.linux.org.uk \
    --subject='Re: [RFC PATCH 5 2/7] kmod - teach usermodehelper to use service workqueues' \
    /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

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).