[dpdk-stable] [PATCH] eal/mp: remove rte_panic and profanity

[dpdk-stable] [PATCH] eal/mp: remove rte_panic and profanity

[Anatoly Burakov]

EAL should not crash when setting alarm fails. Also, remove the
profanity in error message.

Fixes: daf9bfca717e ("ipc: remove thread for async requests")

Cc: stable@dpdk.org
Cc: stephen@networkplumber.org

Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
 lib/librte_eal/common/eal_common_proc.c | 28 ++++++++++++++++++-------
 1 file changed, 21 insertions(+), 7 deletions(-)

diff --git a/lib/librte_eal/common/eal_common_proc.c b/lib/librte_eal/common/eal_common_proc.c
index 97663d3ba..d17131296 100644
--- a/lib/librte_eal/common/eal_common_proc.c
+++ b/lib/librte_eal/common/eal_common_proc.c
@@ -827,6 +827,27 @@ mp_request_async(const char *dst, struct rte_mp_msg *req,
 		goto fail;
 	}
 
+	/*
+	 * set the alarm before sending message. there are two possible error
+	 * scenarios to consider here:
+	 *
+	 * - if the alarm set fails, we free the memory right there
+	 * - if the alarm set succeeds but sending message fails, then the alarm
+	 *   will trigger and clean up the memory
+	 *
+	 * Even if the alarm triggers too early (i.e. immediately), we're still
+	 * holding the lock to pending requests queue, so the interrupt thread
+	 * will just spin until we release the lock, and either release the
+	 * memory, or doesn't find any pending requests in the queue because we
+	 * never added any due to send message failure.
+	 */
+	if (rte_eal_alarm_set(ts->tv_sec * 1000000 + ts->tv_nsec / 1000,
+			      async_reply_handle, pending_req) < 0) {
+		RTE_LOG(ERR, EAL, "Fail to set alarm for request %s:%s\n",
+			dst, req->name);
+		goto fail;
+	}
+
 	ret = send_msg(dst, req, MP_REQ);
 	if (ret < 0) {
 		RTE_LOG(ERR, EAL, "Fail to send request %s:%s\n",
@@ -841,13 +862,6 @@ mp_request_async(const char *dst, struct rte_mp_msg *req,
 
 	param->user_reply.nb_sent++;
 
-	if (rte_eal_alarm_set(ts->tv_sec * 1000000 + ts->tv_nsec / 1000,
-			      async_reply_handle, pending_req) < 0) {
-		RTE_LOG(ERR, EAL, "Fail to set alarm for request %s:%s\n",
-			dst, req->name);
-		rte_panic("Fix the above shit to properly free all memory\n");
-	}
-
 	return 0;
 fail:
 	free(pending_req);
-- 
2.17.1

Re: [dpdk-stable] [PATCH] eal/mp: remove rte_panic and profanity

[Thomas Monjalon]

26/10/2018 16:55, Anatoly Burakov:
> --- a/lib/librte_eal/common/eal_common_proc.c
> +++ b/lib/librte_eal/common/eal_common_proc.c
> +	/*
> +	 * set the alarm before sending message. there are two possible error
> +	 * scenarios to consider here:
> +	 *
> +	 * - if the alarm set fails, we free the memory right there
> +	 * - if the alarm set succeeds but sending message fails, then the alarm
> +	 *   will trigger and clean up the memory
> +	 *
> +	 * Even if the alarm triggers too early (i.e. immediately), we're still
> +	 * holding the lock to pending requests queue, so the interrupt thread
> +	 * will just spin until we release the lock, and either release the
> +	 * memory, or doesn't find any pending requests in the queue because we
> +	 * never added any due to send message failure.
> +	 */
> +	if (rte_eal_alarm_set(ts->tv_sec * 1000000 + ts->tv_nsec / 1000,
> +			      async_reply_handle, pending_req) < 0) {
> +		RTE_LOG(ERR, EAL, "Fail to set alarm for request %s:%s\n",
> +			dst, req->name);
> +		goto fail;
> +	}

ret variable is not set and not initialized.

Re: [dpdk-stable] [PATCH] eal/mp: remove rte_panic and profanity

[Burakov, Anatoly]

On 26-Oct-18 9:41 PM, Thomas Monjalon wrote:
> 26/10/2018 16:55, Anatoly Burakov:
>> --- a/lib/librte_eal/common/eal_common_proc.c
>> +++ b/lib/librte_eal/common/eal_common_proc.c
>> +	/*
>> +	 * set the alarm before sending message. there are two possible error
>> +	 * scenarios to consider here:
>> +	 *
>> +	 * - if the alarm set fails, we free the memory right there
>> +	 * - if the alarm set succeeds but sending message fails, then the alarm
>> +	 *   will trigger and clean up the memory
>> +	 *
>> +	 * Even if the alarm triggers too early (i.e. immediately), we're still
>> +	 * holding the lock to pending requests queue, so the interrupt thread
>> +	 * will just spin until we release the lock, and either release the
>> +	 * memory, or doesn't find any pending requests in the queue because we
>> +	 * never added any due to send message failure.
>> +	 */
>> +	if (rte_eal_alarm_set(ts->tv_sec * 1000000 + ts->tv_nsec / 1000,
>> +			      async_reply_handle, pending_req) < 0) {
>> +		RTE_LOG(ERR, EAL, "Fail to set alarm for request %s:%s\n",
>> +			dst, req->name);
>> +		goto fail;
>> +	}
> 
> ret variable is not set and not initialized.
> 

Oh, right. Apologies. Will send a v2.

-- 
Thanks,
Anatoly

[dpdk-stable] [PATCH v2] eal/mp: remove rte_panic and profanity

[Anatoly Burakov]

EAL should not crash when setting alarm fails. Also, remove the
profanity in error message.

Fixes: daf9bfca717e ("ipc: remove thread for async requests")

Cc: stable@dpdk.org
Cc: stephen@networkplumber.org

Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
---

Notes:
    v2: fix uninitialized variable

 lib/librte_eal/common/eal_common_proc.c | 31 ++++++++++++++++++-------
 1 file changed, 23 insertions(+), 8 deletions(-)

diff --git a/lib/librte_eal/common/eal_common_proc.c b/lib/librte_eal/common/eal_common_proc.c
index 97663d3ba..f65ef56c0 100644
--- a/lib/librte_eal/common/eal_common_proc.c
+++ b/lib/librte_eal/common/eal_common_proc.c
@@ -800,7 +800,7 @@ mp_request_async(const char *dst, struct rte_mp_msg *req,
 {
 	struct rte_mp_msg *reply_msg;
 	struct pending_request *pending_req, *exist;
-	int ret;
+	int ret = -1;
 
 	pending_req = calloc(1, sizeof(*pending_req));
 	reply_msg = calloc(1, sizeof(*reply_msg));
@@ -827,6 +827,28 @@ mp_request_async(const char *dst, struct rte_mp_msg *req,
 		goto fail;
 	}
 
+	/*
+	 * set the alarm before sending message. there are two possible error
+	 * scenarios to consider here:
+	 *
+	 * - if the alarm set fails, we free the memory right there
+	 * - if the alarm set succeeds but sending message fails, then the alarm
+	 *   will trigger and clean up the memory
+	 *
+	 * Even if the alarm triggers too early (i.e. immediately), we're still
+	 * holding the lock to pending requests queue, so the interrupt thread
+	 * will just spin until we release the lock, and either release the
+	 * memory, or doesn't find any pending requests in the queue because we
+	 * never added any due to send message failure.
+	 */
+	if (rte_eal_alarm_set(ts->tv_sec * 1000000 + ts->tv_nsec / 1000,
+			      async_reply_handle, pending_req) < 0) {
+		RTE_LOG(ERR, EAL, "Fail to set alarm for request %s:%s\n",
+			dst, req->name);
+		ret = -1;
+		goto fail;
+	}
+
 	ret = send_msg(dst, req, MP_REQ);
 	if (ret < 0) {
 		RTE_LOG(ERR, EAL, "Fail to send request %s:%s\n",
@@ -841,13 +863,6 @@ mp_request_async(const char *dst, struct rte_mp_msg *req,
 
 	param->user_reply.nb_sent++;
 
-	if (rte_eal_alarm_set(ts->tv_sec * 1000000 + ts->tv_nsec / 1000,
-			      async_reply_handle, pending_req) < 0) {
-		RTE_LOG(ERR, EAL, "Fail to set alarm for request %s:%s\n",
-			dst, req->name);
-		rte_panic("Fix the above shit to properly free all memory\n");
-	}
-
 	return 0;
 fail:
 	free(pending_req);
-- 
2.17.1

Re: [dpdk-stable] [dpdk-dev] [PATCH v2] eal/mp: remove rte_panic and profanity

[Thomas Monjalon]

13/11/2018 19:03, Anatoly Burakov:
> EAL should not crash when setting alarm fails. Also, remove the
> profanity in error message.
> 
> Fixes: daf9bfca717e ("ipc: remove thread for async requests")
> 
> Cc: stable@dpdk.org
> Cc: stephen@networkplumber.org
> 
> Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
> Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>

Applied, thanks