From mboxrd@z Thu Jan 1 00:00:00 1970
Return-Path: <stable-bounces@dpdk.org>
Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124])
by inbox.dpdk.org (Postfix) with ESMTP id D7E0EA054F
for <public@inbox.dpdk.org>; Tue, 6 Sep 2022 04:48:41 +0200 (CEST)
Received: from [217.70.189.124] (localhost [127.0.0.1])
by mails.dpdk.org (Postfix) with ESMTP id CACBF40E25;
Tue, 6 Sep 2022 04:48:41 +0200 (CEST)
Received: from mga17.intel.com (mga17.intel.com [192.55.52.151])
by mails.dpdk.org (Postfix) with ESMTP id 280BB40041;
Tue, 6 Sep 2022 04:48:38 +0200 (CEST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple;
d=intel.com; i=@intel.com; q=dns/txt; s=Intel;
t=1662432519; x=1693968519;
h=from:to:cc:subject:date:message-id:mime-version:
content-transfer-encoding;
bh=uGJgAVHRst1D+4TzMtMOWicRNvMZL3OTIDZcr4E13PI=;
b=DxjdlAJojFF6k8RByCA+Op1wM0KNIMHfsQ56uUSH1Zf+VsnApp6tPOun
o/3EmhTy9vnACLgJwYxbbSlBlZ++s2e9baGplHM//NGyciSZ8ioACR7f8
dy01KXFVTAEY3peEYtoodFTE572inVCNW15gXTo4odh3yZIV4fJhBCWTb
pmC4GZ4gfVqQgZNVYLCVvX1v6cK4eTFXx0bn4WAXKXw8ef8U1YoTKYdD2
qqqQ600iO1qLL/X4h/8U+KJrljtKbo6slVR788qvDt0UsTyLquudnxlF+
p1e7A+/l6aavWW3BdFZ9Pg2zow5ZnT8YZqAti9on71rWF12DoGnW7vF2l w==;
X-IronPort-AV: E=McAfee;i="6500,9779,10461"; a="276881057"
X-IronPort-AV: E=Sophos;i="5.93,292,1654585200"; d="scan'208";a="276881057"
Received: from fmsmga005.fm.intel.com ([10.253.24.32])
by fmsmga107.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384;
05 Sep 2022 19:48:38 -0700
X-IronPort-AV: E=Sophos;i="5.93,292,1654585200"; d="scan'208";a="942285101"
Received: from unknown (HELO localhost.localdomain) ([10.239.252.103])
by fmsmga005-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384;
05 Sep 2022 19:48:35 -0700
From: Zhichao Zeng <zhichaox.zeng@intel.com>
To: dev@dpdk.org
Cc: david.marchand@redhat.com, qiming.yang@intel.com, yidingx.zhou@intel.com,
Zhichao Zeng <zhichaox.zeng@intel.com>, stable@dpdk.org,
Bruce Richardson <bruce.richardson@intel.com>,
Harman Kalra <hkalra@marvell.com>
Subject: [PATCH v5] lib/eal: fix segfaults in exiting
Date: Tue, 6 Sep 2022 10:49:50 +0800
Message-Id: <20220906024950.160740-1-zhichaox.zeng@intel.com>
X-Mailer: git-send-email 2.25.1
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
X-BeenThere: stable@dpdk.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: patches for DPDK stable branches <stable.dpdk.org>
List-Unsubscribe: <https://mails.dpdk.org/options/stable>,
<mailto:stable-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://mails.dpdk.org/archives/stable/>
List-Post: <mailto:stable@dpdk.org>
List-Help: <mailto:stable-request@dpdk.org?subject=help>
List-Subscribe: <https://mails.dpdk.org/listinfo/stable>,
<mailto:stable-request@dpdk.org?subject=subscribe>
Errors-To: stable-bounces@dpdk.org
The 'eal-intr-thread' is not closed before memory cleanup in the process of
exiting. There is a small chance that the 'eal-intr-thread' is about to use
some pointers, the memory was just cleaned, which causes segfaults
caught by ASan.
This patch closes the 'eal-intr-thread' before memory cleanup in
'rte_eal_cleanup' to avoid segfaults, and adds a flag to avoid executing
'rte_eal_cleanup' in the child process which is forked to execut some
test cases(e.g. debug_autotest of dpdk-test).
Bugzilla ID: 1006
Cc: stable@dpdk.org
Signed-off-by: Zhichao Zeng <zhichaox.zeng@intel.com>
---
v2: add same API for FreeBSD
---
v3: fix rte_eal_cleanup crash in debug_autotest
---
v4: shorten the prompt message and optimize the commit log
---
v5: simplify patch
---
lib/eal/common/eal_private.h | 7 +++++++
lib/eal/freebsd/eal.c | 13 +++++++++++++
lib/eal/freebsd/eal_interrupts.c | 12 ++++++++++++
lib/eal/linux/eal.c | 13 +++++++++++++
lib/eal/linux/eal_interrupts.c | 12 ++++++++++++
5 files changed, 57 insertions(+)
diff --git a/lib/eal/common/eal_private.h b/lib/eal/common/eal_private.h
index 44d14241f0..7adf41b7d7 100644
--- a/lib/eal/common/eal_private.h
+++ b/lib/eal/common/eal_private.h
@@ -152,6 +152,13 @@ int rte_eal_tailqs_init(void);
*/
int rte_eal_intr_init(void);
+/**
+ * Destroy interrupt handling thread.
+ *
+ * This function is private to EAL.
+ */
+void rte_eal_intr_destroy(void);
+
/**
* Close the default log stream
*
diff --git a/lib/eal/freebsd/eal.c b/lib/eal/freebsd/eal.c
index 26fbc91b26..f27c1d9f97 100644
--- a/lib/eal/freebsd/eal.c
+++ b/lib/eal/freebsd/eal.c
@@ -72,6 +72,8 @@ struct lcore_config lcore_config[RTE_MAX_LCORE];
/* used by rte_rdtsc() */
int rte_cycles_vmware_tsc_map;
+/* used to judge if is forked */
+static int is_forked;
int
eal_clean_runtime_dir(void)
@@ -574,6 +576,11 @@ static void rte_eal_init_alert(const char *msg)
RTE_LOG(ERR, EAL, "%s\n", msg);
}
+static void mark_forked(void)
+{
+ is_forked++;
+}
+
/* Launch threads, called at application init(). */
int
rte_eal_init(int argc, char **argv)
@@ -883,16 +890,22 @@ rte_eal_init(int argc, char **argv)
eal_mcfg_complete();
+ pthread_atfork(NULL, NULL, mark_forked);
+
return fctret;
}
int
rte_eal_cleanup(void)
{
+ if (is_forked)
+ return 0;
+
struct internal_config *internal_conf =
eal_get_internal_configuration();
rte_service_finalize();
rte_mp_channel_cleanup();
+ rte_eal_intr_destroy();
rte_trace_save();
eal_trace_fini();
/* after this point, any DPDK pointers will become dangling */
diff --git a/lib/eal/freebsd/eal_interrupts.c b/lib/eal/freebsd/eal_interrupts.c
index 9f720bdc8f..cac3859b06 100644
--- a/lib/eal/freebsd/eal_interrupts.c
+++ b/lib/eal/freebsd/eal_interrupts.c
@@ -648,6 +648,18 @@ rte_eal_intr_init(void)
return ret;
}
+void
+rte_eal_intr_destroy(void)
+{
+ /* cancel the host thread to wait/handle the interrupt */
+ pthread_cancel(intr_thread);
+ pthread_join(intr_thread, NULL);
+
+ /* close kqueue */
+ close(kq);
+ kq = -1;
+}
+
int
rte_intr_rx_ctl(struct rte_intr_handle *intr_handle,
int epfd, int op, unsigned int vec, void *data)
diff --git a/lib/eal/linux/eal.c b/lib/eal/linux/eal.c
index 37d29643a5..ea38ec183d 100644
--- a/lib/eal/linux/eal.c
+++ b/lib/eal/linux/eal.c
@@ -76,6 +76,8 @@ struct lcore_config lcore_config[RTE_MAX_LCORE];
/* used by rte_rdtsc() */
int rte_cycles_vmware_tsc_map;
+/* used to judge if is forked */
+static int is_forked;
int
eal_clean_runtime_dir(void)
@@ -954,6 +956,11 @@ eal_worker_thread_create(unsigned int lcore_id)
return ret;
}
+static void mark_forked(void)
+{
+ is_forked++;
+}
+
/* Launch threads, called at application init(). */
int
rte_eal_init(int argc, char **argv)
@@ -1324,6 +1331,8 @@ rte_eal_init(int argc, char **argv)
eal_mcfg_complete();
+ pthread_atfork(NULL, NULL, mark_forked);
+
return fctret;
}
@@ -1347,6 +1356,9 @@ mark_freeable(const struct rte_memseg_list *msl, const struct rte_memseg *ms,
int
rte_eal_cleanup(void)
{
+ if (is_forked)
+ return 0;
+
/* if we're in a primary process, we need to mark hugepages as freeable
* so that finalization can release them back to the system.
*/
@@ -1362,6 +1374,7 @@ rte_eal_cleanup(void)
vfio_mp_sync_cleanup();
#endif
rte_mp_channel_cleanup();
+ rte_eal_intr_destroy();
rte_trace_save();
eal_trace_fini();
/* after this point, any DPDK pointers will become dangling */
diff --git a/lib/eal/linux/eal_interrupts.c b/lib/eal/linux/eal_interrupts.c
index d52ec8eb4c..7e9853e8e7 100644
--- a/lib/eal/linux/eal_interrupts.c
+++ b/lib/eal/linux/eal_interrupts.c
@@ -1199,6 +1199,18 @@ rte_eal_intr_init(void)
return ret;
}
+void
+rte_eal_intr_destroy(void)
+{
+ /* cancel the host thread to wait/handle the interrupt */
+ pthread_cancel(intr_thread);
+ pthread_join(intr_thread, NULL);
+
+ /* close the pipe used by epoll */
+ close(intr_pipe.writefd);
+ close(intr_pipe.readfd);
+}
+
static void
eal_intr_proc_rxtx_intr(int fd, const struct rte_intr_handle *intr_handle)
{
--
2.25.1