linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Zach Levis <zml@linux.vnet.ibm.com>
To: linux-fsdevel@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, viro@zeniv.linux.org.uk,
	akpm@linux-foundation.org, Zach Levis <zml@linux.vnet.ibm.com>,
	Zach Levis <zach@zachsthings.com>
Subject: [PATCH 2/3] fs/binfmts: Better handling of binfmt loops
Date: Thu, 25 Jul 2013 08:40:44 -0700	[thread overview]
Message-ID: <1374766845-13565-2-git-send-email-zml@linux.vnet.ibm.com> (raw)
In-Reply-To: <1374766845-13565-1-git-send-email-zml@linux.vnet.ibm.com>

With these changes, when a binfmt loop is encountered,
the ELOOP will propogate back to the 0 depth. At this point the
argv and argc values will be reset to what they were originally and an
attempt is made to continue with the following binfmt handlers.

Caveats:
- binfmt_misc is skipped as a whole. To allow for a more generic
  solution that works for any binfmt without a lot of duplicated code
  and added complexity, the error detection code is applied on the
  binfmt level.
- This is a fallback solution. It attempts to restore the state to allow
  executing most binaries without side effects, but it may not work for
  everything and should not be depended on for regular usage.

My (rough, but functional) test scripts for this issue are available at:
    https://gist.github.com/zml2008/6075418

Signed-off-by: Zach Levis <zach@zachsthings.com>
Signed-off-by: Zach Levis <zml@linux.vnet.ibm.com>
---
 fs/exec.c               |   30 +++++++++++++++++++++++++++++-
 include/linux/binfmts.h |    5 ++++-
 2 files changed, 33 insertions(+), 2 deletions(-)

diff --git a/fs/exec.c b/fs/exec.c
index ffd7a81..fa59f05 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -1401,13 +1401,40 @@ int search_binary_handler(struct linux_binprm *bprm)
 			if (!try_module_get(fmt->module))
 				continue;
 			read_unlock(&binfmt_lock);
+			bprm->previous_binfmts[1] = bprm->previous_binfmts[0];
+			bprm->previous_binfmts[0] = fmt;
+
 			bprm->recursion_depth = depth + 1;
 			retval = fn(bprm);
 			bprm->recursion_depth = depth;
+			if (retval == -ELOOP && depth == 0) { /* cur, previous */
+				pr_err("Too much recursion with binfmts (0:%s, -1:%s) in file %s, skipping (base %s).\n",
+						bprm->previous_binfmts[0]->name,
+						bprm->previous_binfmts[1]->name,
+						bprm->filename,
+						fmt->name);
+
+				/* Put argv back in its place */
+				while (bprm->argc > 0) {
+					retval = remove_arg_zero(bprm);
+					if (retval)
+						return retval;
+				}
+
+				copy_strings(bprm->argc_orig, *((struct user_arg_ptr *) bprm->argv_orig), bprm);
+				bprm->argc = bprm->argc_orig;
+				retval = -ENOEXEC;
+				continue;
+			}
+
 			if (retval >= 0) {
 				if (depth == 0) {
 					trace_sched_process_exec(current, old_pid, bprm);
 					ptrace_event(PTRACE_EVENT_EXEC, old_vpid);
+					/* Successful execution, now null out the cached argv
+					 * (we don't want to access it later)
+					 * */
+					bprm->argv_orig = NULL;
 				}
 				put_binfmt(fmt);
 				allow_write_access(bprm->file);
@@ -1515,7 +1542,7 @@ static int do_execve_common(const char *filename,
 	if (retval)
 		goto out_file;
 
-	bprm->argc = count(argv, MAX_ARG_STRINGS);
+	bprm->argc_orig = bprm->argc = count(argv, MAX_ARG_STRINGS);
 	if ((retval = bprm->argc) < 0)
 		goto out;
 
@@ -1539,6 +1566,7 @@ static int do_execve_common(const char *filename,
 	retval = copy_strings(bprm->argc, argv, bprm);
 	if (retval < 0)
 		goto out;
+	bprm->argv_orig = &argv;
 
 	retval = search_binary_handler(bprm);
 	if (retval < 0)
diff --git a/include/linux/binfmts.h b/include/linux/binfmts.h
index 402a74a..2e3dc66 100644
--- a/include/linux/binfmts.h
+++ b/include/linux/binfmts.h
@@ -32,11 +32,14 @@ struct linux_binprm {
 	unsigned int taso:1;
 #endif
 	unsigned int recursion_depth;
+	/* current binfmt at 0 and previous binfmt at 1 */
+	struct linux_binfmt *previous_binfmts[2];
 	struct file * file;
 	struct cred *cred;	/* new credentials */
 	int unsafe;		/* how unsafe this exec is (mask of LSM_UNSAFE_*) */
 	unsigned int per_clear;	/* bits to clear in current->personality */
-	int argc, envc;
+	int argc, argc_orig, envc;
+	void *argv_orig;
 	const char * filename;	/* Name of binary as seen by procps */
 	const char * interp;	/* Name of the binary really executed. Most
 				   of the time same as filename, but could be
-- 
1.7.1

  reply	other threads:[~2013-07-25 15:40 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-07-25 15:40 [PATCH 1/3] fs/binfmts: Add a name field to the binfmt struct Zach Levis
2013-07-25 15:40 ` Zach Levis [this message]
2013-07-30 21:04   ` [PATCH 2/3] fs/binfmts: Better handling of binfmt loops Andrew Morton
2013-07-30 23:16     ` Zach Levis
2013-07-30 23:26       ` Andrew Morton
2013-07-31 16:17         ` Zach Levis
2013-07-25 15:40 ` [PATCH 3/3] fs/binfmts: Whitespace fixes with scripts/cleanfile Zach Levis
2013-08-02 22:12 ` [PATCH v2 0/3] fs/binfmts: Improve handling of loops Zach Levis
2013-08-02 22:12 ` [PATCH v2 1/3] fs/binfmts: Add a name field to the binfmt struct Zach Levis
2013-08-02 22:12 ` [PATCH v2 2/3] fs/binfmts: Better handling of binfmt loops Zach Levis
2013-08-02 22:49   ` Zach Levis
2013-08-02 22:12 ` [PATCH v2 3/3] fs/binfmts: Whitespace fixes with scripts/cleanfile Zach Levis
2013-08-02 23:21 ` [PATCH v3 0/3] fs/binfmts: Improve handling of loops Zach Levis
2013-08-06 21:11   ` Kees Cook
2013-08-07 23:30     ` Zach Levis
2013-08-02 23:21 ` [PATCH v3 1/3] fs/binfmts: Add a name field to the binfmt struct Zach Levis
2013-08-03 16:41   ` Oleg Nesterov
2013-08-02 23:21 ` [PATCH v3 2/3] fs/binfmts: Better handling of binfmt loops Zach Levis
2013-08-03 16:42   ` Oleg Nesterov
2013-08-02 23:21 ` [PATCH v3 3/3] fs/binfmts: Whitespace fixes with scripts/cleanfile Zach Levis
2013-08-03 16:51   ` Oleg Nesterov
2013-08-14 16:31 ` [PATCH v4 0/3] fs/binfmts: Improve handling of loops Zach Levis
2013-08-14 16:31 ` [PATCH v4 1/3] fs/binfmts: Add a name field to the binfmt struct Zach Levis
2013-08-14 16:31 ` [PATCH v4 2/3] fs/binfmts: Better handling of binfmt loops Zach Levis
2013-08-14 17:50   ` Oleg Nesterov
2013-08-15 16:26     ` Zach L
2013-08-16 12:23       ` Oleg Nesterov
2013-08-14 18:16   ` Oleg Nesterov
2013-08-14 16:31 ` [PATCH v4 3/3] fs/binfmts: Whitespace fixes with scripts/cleanfile Zach Levis
2013-08-15 16:20 ` [PATCH v5 0/3] fs/binfmts: Improve handling of loops Zach Levis
2013-08-15 16:20 ` [PATCH v5 1/3] fs/binfmts: Add a name field to the binfmt struct Zach Levis
2013-08-15 17:06   ` Kees Cook
2013-08-15 16:20 ` [PATCH v5 2/3] fs/binfmts: Better handling of binfmt loops Zach Levis
2013-08-16 13:15   ` Oleg Nesterov
2013-08-15 16:20 ` [PATCH v5 3/3] fs/binfmts: Whitespace fixes with scripts/cleanfile Zach Levis

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=1374766845-13565-2-git-send-email-zml@linux.vnet.ibm.com \
    --to=zml@linux.vnet.ibm.com \
    --cc=akpm@linux-foundation.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=viro@zeniv.linux.org.uk \
    --cc=zach@zachsthings.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).