DPDK patches and discussions
 help / color / mirror / Atom feed
From: Bruce Richardson <bruce.richardson@intel.com>
To: dev@dpdk.org
Cc: thomas@monjalon.net, david.marchand@redhat.com,
	Bruce Richardson <bruce.richardson@intel.com>
Subject: [dpdk-dev] [PATCH 3/4] eal: don't load drivers from insecure paths
Date: Thu, 18 Jun 2020 14:50:48 +0100	[thread overview]
Message-ID: <20200618135049.489773-4-bruce.richardson@intel.com> (raw)
In-Reply-To: <20200618135049.489773-1-bruce.richardson@intel.com>

Any paths on the system which are world-writable are insecure and should
not be used for loading drivers. Therefore check each driver path before
loading it and error out on insecure ones.

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
---
 lib/librte_eal/common/eal_common_options.c | 75 ++++++++++++++++++++--
 1 file changed, 69 insertions(+), 6 deletions(-)

diff --git a/lib/librte_eal/common/eal_common_options.c b/lib/librte_eal/common/eal_common_options.c
index 7aef6df4c..2a62a1342 100644
--- a/lib/librte_eal/common/eal_common_options.c
+++ b/lib/librte_eal/common/eal_common_options.c
@@ -15,6 +15,7 @@
 #include <getopt.h>
 #ifndef RTE_EXEC_ENV_WINDOWS
 #include <dlfcn.h>
+#include <libgen.h>
 #endif
 #include <sys/types.h>
 #include <sys/stat.h>
@@ -368,10 +369,74 @@ eal_plugindir_init(const char *path)
 	return (dent == NULL) ? 0 : -1;
 }
 
+#ifdef RTE_EXEC_ENV_WINDOWS
+int
+eal_plugins_init(void)
+{
+	return 0;
+}
+#else
+
+static int
+verify_perms(const char *dirpath)
+{
+	struct stat st;
+
+	/* if not root, check down one level first */
+	if (strcmp(dirpath, "/") != 0) {
+		char copy[PATH_MAX];
+
+		strlcpy(copy, dirpath, PATH_MAX);
+		if (verify_perms(dirname(copy)) != 0)
+			return -1;
+	}
+
+	/* call stat to check for permissions and ensure not world writable */
+	if (stat(dirpath, &st) != 0) {
+		RTE_LOG(ERR, EAL, "Error with stat on %s, %s\n",
+				dirpath, strerror(errno));
+		return -1;
+	}
+	if (st.st_mode & S_IWOTH) {
+		RTE_LOG(ERR, EAL,
+				"Error, directory path %s is world-writable and insecure\n",
+				dirpath);
+		return -1;
+	}
+
+	return 0;
+}
+
+static void *
+eal_dlopen(const char *pathname)
+{
+	void *retval = NULL;
+	char *realp = realpath(pathname, NULL);
+
+	if (realp == NULL) {
+		RTE_LOG(ERR, EAL, "Error with realpath, %s\n", strerror(errno));
+		goto out;
+	}
+	if (strnlen(realp, PATH_MAX) == PATH_MAX) {
+		RTE_LOG(ERR, EAL, "Error, driver path greater than PATH_MAX\n");
+		goto out;
+	}
+
+	/* do permissions checks */
+	if (verify_perms(realp) != 0)
+		goto out;
+
+	retval = dlopen(realp, RTLD_NOW);
+	if (retval == NULL)
+		RTE_LOG(ERR, EAL, "%s\n", dlerror());
+out:
+	free(realp);
+	return retval;
+}
+
 int
 eal_plugins_init(void)
 {
-#ifndef RTE_EXEC_ENV_WINDOWS
 	struct shared_driver *solib = NULL;
 	struct stat sb;
 
@@ -391,17 +456,15 @@ eal_plugins_init(void)
 		} else {
 			RTE_LOG(DEBUG, EAL, "open shared lib %s\n",
 				solib->name);
-			solib->lib_handle = dlopen(solib->name, RTLD_NOW);
-			if (solib->lib_handle == NULL) {
-				RTE_LOG(ERR, EAL, "%s\n", dlerror());
+			solib->lib_handle = eal_dlopen(solib->name);
+			if (solib->lib_handle == NULL)
 				return -1;
-			}
 		}
 
 	}
 	return 0;
-#endif
 }
+#endif
 
 /*
  * Parse the coremask given as argument (hexadecimal string) and fill
-- 
2.25.1


  parent reply	other threads:[~2020-06-18 13:51 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-18 13:50 [dpdk-dev] [PATCH 0/4] improve runtime loading of shared drivers Bruce Richardson
2020-06-18 13:50 ` [dpdk-dev] [PATCH 1/4] eal: remove unnecessary null-termination Bruce Richardson
2020-06-18 13:50 ` [dpdk-dev] [PATCH 2/4] eal: only load shared libs from driver plugin directory Bruce Richardson
2020-06-18 13:50 ` Bruce Richardson [this message]
2020-06-18 13:50 ` [dpdk-dev] [PATCH 4/4] eal: cache last directory permissions checked Bruce Richardson
2020-06-22 14:33 ` [dpdk-dev] [PATCH v2 0/4] improve runtime loading of shared drivers Bruce Richardson
2020-06-22 14:33   ` [dpdk-dev] [PATCH v2 1/4] eal: remove unnecessary null-termination Bruce Richardson
2020-06-22 14:33   ` [dpdk-dev] [PATCH v2 2/4] eal: only load shared libs from driver plugin directory Bruce Richardson
2020-06-22 14:33   ` [dpdk-dev] [PATCH v2 3/4] eal: don't load drivers from insecure paths Bruce Richardson
2020-06-22 14:33   ` [dpdk-dev] [PATCH v2 4/4] eal: cache last directory permissions checked Bruce Richardson
2020-07-02 21:13   ` [dpdk-dev] [PATCH v2 0/4] improve runtime loading of shared drivers Thomas Monjalon
2020-07-03 10:25     ` Bruce Richardson
2020-07-02 21:16   ` Thomas Monjalon
2020-07-03  8:33     ` Bruce Richardson
2020-07-03 10:23 ` [dpdk-dev] [PATCH v3 " Bruce Richardson
2020-07-03 10:23   ` [dpdk-dev] [PATCH v3 1/4] eal: remove unnecessary null-termination in plugin path Bruce Richardson
2020-07-03 10:23   ` [dpdk-dev] [PATCH v3 2/4] eal: load only shared libs from driver plugin directories Bruce Richardson
2020-07-03 10:23   ` [dpdk-dev] [PATCH v3 3/4] eal: forbid loading drivers from insecure paths Bruce Richardson
2020-07-03 10:23   ` [dpdk-dev] [PATCH v3 4/4] eal: cache last directory permissions checked Bruce Richardson
2020-07-05 17:50   ` [dpdk-dev] [PATCH v3 0/4] improve runtime loading of shared drivers Thomas Monjalon

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=20200618135049.489773-4-bruce.richardson@intel.com \
    --to=bruce.richardson@intel.com \
    --cc=david.marchand@redhat.com \
    --cc=dev@dpdk.org \
    --cc=thomas@monjalon.net \
    /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).