Hi Mahesh,
Thanks for your comments. I will send new version after Lennart comments.
I have checked your idea about multiple thread test case. The idea may be same
with patch that I sent it before.
Regards,
Canh.
-----Original Message-----
From: A V Mahesh [mailto:mahesh.va...@oracle.com]
Sent: Tuesday, June 21, 2016 11:47 AM
To: Canh Truong; vu.m.ngu...@dektech.com.au; lennart.l...@ericsson.com
Cc: opensaf-devel@lists.sourceforge.net
Subject: Re: [PATCH 1 of 1] log: fix issue with priority of messages adding to
mailbox in the server[1396]
Hi Canh,
It will be good to send V2 patch with Lennart comments and adding new test
cases.
-AVM
On 6/21/2016 9:46 AM, A V Mahesh wrote:
> Hi Canh,
>
> Some how we were not able to open your attachment , so my self
> written
>
> a multi threaded test case ( attached `thread_LogInitializeFinalize.c`
> ) and reviewed/tested it is working fine.
>
> ACK from me with following :
>
> - Please add such as attached multi threaded test case to logtest,
> before pushing
>
> - add # before ticket number in commit message #1396
>
> -AVM
>
>
> On 6/14/2016 2:47 PM, Canh Truong wrote:
>> Hi Mahesh,
>>
>> I have created new test case for multiple threads. Please help check
>> new patch in attachment.
>>
>> Thanks,
>> Canh.
>>
>> -----Original Message-----
>> From: A V Mahesh [mailto:mahesh.va...@oracle.com]
>> Sent: Tuesday, June 14, 2016 11:03 AM
>> To: Canh Van Truong; vu.m.ngu...@dektech.com.au;
>> lennart.l...@ericsson.com
>> Cc: opensaf-devel@lists.sourceforge.net
>> Subject: Re: [PATCH 1 of 1] log: fix issue with priority of messages
>> adding to mailbox in the server[1396]
>>
>> Hi Canh Van Truong,
>>
>> I hope this patch will work for multiple threads doing
>> saLogInitialize() , saLogStreamOpen_2() and saLogFinalize()
>> concurrently in a single process and reaches lga_use_count = 0 ,
>> againg restarting
>> saLogInitialize() , saLogStreamOpen_2() and
>> saLogFinalize()
>> without process exist.
>>
>> If so let us have a test case like for multiple threads doing
>> saLogInitialize() , saLogStreamOpen_2() and saLogFinalize()
>> concurrently
>>
>> -AVM
>>
>>
>> On 6/13/2016 12:37 PM, Canh Van Truong wrote:
>>> osaf/libs/agents/saf/lga/lga_api.c | 3 ++-
>>> osaf/libs/agents/saf/lga/lga_util.c | 21 ++++++++++++---------
>>> tests/logsv/logtest.h | 1 +
>>> tests/logsv/tet_saLogStreamOpen_2.c | 29
>>> +++++++++++++++++++++++++++++
>>> 4 files changed, 44 insertions(+), 10 deletions(-)
>>>
>>>
>>> Sometimes, Log service can handle LGSV_INITIALIZE_REQ before handle
>>> LGSV_LGS_EVT_LGA_DOWN, although LGSV_LGS_EVT_LGA_DOWN was sent first
>>> whenever
>>> finalize() is call for last client. Because lgs can not distinguish
>>> MDS address between new connection (initializing new client) and old
>>> connection (finalizing last client before). This patch fixes to keep
>>> MDS connection when last client finalized, and re-use it when next
>> initialize() called.
>>> diff --git a/osaf/libs/agents/saf/lga/lga_api.c
>>> b/osaf/libs/agents/saf/lga/lga_api.c
>>> --- a/osaf/libs/agents/saf/lga/lga_api.c
>>> +++ b/osaf/libs/agents/saf/lga/lga_api.c
>>> @@ -41,7 +41,8 @@
>>> /* The main controle block */
>>> lga_cb_t lga_cb = {
>>> .cb_lock = PTHREAD_MUTEX_INITIALIZER,
>>> - .lgs_state = LGS_START
>>> + .lgs_state = LGS_START,
>>> + .mds_hdl = 0
>>> };
>>> static bool is_well_know_stream(const char* dn) diff --git
>>> a/osaf/libs/agents/saf/lga/lga_util.c
>>> b/osaf/libs/agents/saf/lga/lga_util.c
>>> --- a/osaf/libs/agents/saf/lga/lga_util.c
>>> +++ b/osaf/libs/agents/saf/lga/lga_util.c
>>> @@ -308,21 +308,17 @@ unsigned int lga_startup(lga_cb_t *cb)
>>> osaf_mutex_lock_ordie(&lga_lock);
>>> TRACE_ENTER2("lga_use_count: %u", lga_use_count);
>>> - if (lga_use_count > 0) {
>>> - /* Already created, just increment the use_count */
>>> - lga_use_count++;
>>> - goto done;
>>> - } else {
>>> +
>>> + if (cb->mds_hdl == 0) {
>>> if ((rc = ncs_agents_startup()) != NCSCC_RC_SUCCESS) {
>>> TRACE("ncs_agents_startup FAILED");
>>> goto done;
>>> }
>>> if ((rc = lga_create()) != NCSCC_RC_SUCCESS) {
>>> + cb->mds_hdl = 0;
>>> ncs_agents_shutdown();
>>> goto done;
>>> - } else {
>>> - lga_use_count = 1;
>>> }
>>> /* Agent has successfully been started including
>> communication @@
>>> -331,6 +327,9 @@ unsigned int lga_startup(lga_cb_t *cb)
>>> set_lga_state(LGA_NORMAL);
>>> }
>>> + /* Increase the use_count */
>>> + lga_use_count++;
>>> +
>>> done:
>>> osaf_mutex_unlock_ordie(&lga_lock);
>>> @@ -361,8 +360,12 @@ unsigned int lga_shutdown_after_last_cli
>>> /* Users still exist, just decrement the use count */
>>> lga_use_count--;
>>> } else if (lga_use_count == 1) {
>>> - lga_destroy();
>>> - rc = ncs_agents_shutdown();
>>> + /*
>>> + * Disable MDS_UNINSTALL and ncs agent shutdown to keep MDS
>> connection.
>>> + * Fix ticket 1396. Msg NCSMDS_DOWN and Initialize are out
>> of order in Mbx.
>>> + */
>>> + /* lga_destroy(); */
>>> + /* rc = ncs_agents_shutdown(); */
>>> lga_use_count = 0;
>>> }
>>> diff --git a/tests/logsv/logtest.h b/tests/logsv/logtest.h
>>> --- a/tests/logsv/logtest.h
>>> +++ b/tests/logsv/logtest.h
>>> @@ -27,6 +27,7 @@
>>> #include <utest.h>
>>> #include <util.h>
>>> #include <sys/wait.h>
>>> +#include <unistd.h>
>>> #include <osaf_time.h>
>>> #include <logtrace.h>
>>> diff --git a/tests/logsv/tet_saLogStreamOpen_2.c
>>> b/tests/logsv/tet_saLogStreamOpen_2.c
>>> --- a/tests/logsv/tet_saLogStreamOpen_2.c
>>> +++ b/tests/logsv/tet_saLogStreamOpen_2.c
>>> @@ -704,6 +704,34 @@ done:
>>> }
>>> }
>>> +/*
>>> + * Ticket 1396
>>> + * Verify that saLogInitialize() then saLogFinalize() multiple time
>>> +OK */ void saLogMultipleInitialize(void) {
>>> + SaLogStreamHandleT logStreamHandle1;
>>> + int count = 100;
>>> + SaAisErrorT rc = SA_AIS_OK;
>>> +
>>> + while (count--) {
>>> + rc = saLogInitialize(&logHandle, &logCallbacks,
>> &logVersion);
>>> + if (rc != SA_AIS_OK)
>>> + break;
>>> +
>>> + rc = saLogStreamOpen_2(logHandle, &app1StreamName,
>> &appStream1LogFileCreateAttributes,
>>> + SA_LOG_STREAM_CREATE,
>> SA_TIME_ONE_SECOND, &logStreamHandle1);
>>> + safassert(rc, SA_AIS_OK);
>>> +
>>> + rc = saLogFinalize(logHandle);
>>> + if (rc != SA_AIS_OK)
>>> + break;
>>> +
>>> + }
>>> + test_validate(rc, SA_AIS_OK);
>>> +
>>> +}
>>> +
>>> extern void saLogStreamOpenAsync_2_01(void);
>>> extern void saLogStreamOpenCallbackT_01(void);
>>> extern void saLogWriteLog_01(void); @@ -790,5 +818,6 @@ extern
>>> void saLogStreamClose_01(void);
>>> test_case_add(2, verFixLogRec_Max_Err, "saLogStreamOpen_2
>>> with
>> maxLogRecordSize > MAX_RECSIZE, ERR");
>>> test_case_add(2, verFixLogRec_Min_Err, "saLogStreamOpen_2
>>> with
>> maxLogRecordSize < 150, ERR");
>>> test_case_add(2, saLogStreamOpen_2_50, "saLogStreamOpen_2
>>> with stream number out of the limitation, ERR");
>>> + test_case_add(2, saLogMultipleInitialize, "saLogInitialize()
>>> + then
>>> + saLogFinalize() multiple time. keep MDS connection, OK");
>>> }
>
# HG changeset patch
# User Canh Van Truong <canh.v.tru...@dektech.com.au>
# Date 1465545815 -25200
# Fri Jun 10 15:03:35 2016 +0700
# Node ID 4f31418251bf58bfa282abcd1ce1afabc32b6037
# Parent f30bc7992ce3343a427b6abf93e3f15ab1606329
log: fix issue with priority of messages adding to mailbox in the server[#1396]
Sometimes, Log service can handle LGSV_INITIALIZE_REQ before handle
LGSV_LGS_EVT_LGA_DOWN, although LGSV_LGS_EVT_LGA_DOWN was sent first whenever
finalize() is call for last client. Because lgs can not distinguish MDS
address between new connection (initializing new client) and old connection
(finalizing last client before). This patch fixes to keep MDS connection
when last client finalized, and re-use it when next initialize() called.
diff --git a/osaf/libs/agents/saf/lga/lga_api.c
b/osaf/libs/agents/saf/lga/lga_api.c
--- a/osaf/libs/agents/saf/lga/lga_api.c
+++ b/osaf/libs/agents/saf/lga/lga_api.c
@@ -41,7 +41,8 @@
/* The main controle block */
lga_cb_t lga_cb = {
.cb_lock = PTHREAD_MUTEX_INITIALIZER,
- .lgs_state = LGS_START
+ .lgs_state = LGS_START,
+ .mds_hdl = 0
};
static bool is_well_know_stream(const char* dn)
diff --git a/osaf/libs/agents/saf/lga/lga_util.c
b/osaf/libs/agents/saf/lga/lga_util.c
--- a/osaf/libs/agents/saf/lga/lga_util.c
+++ b/osaf/libs/agents/saf/lga/lga_util.c
@@ -308,21 +308,17 @@ unsigned int lga_startup(lga_cb_t *cb)
osaf_mutex_lock_ordie(&lga_lock);
TRACE_ENTER2("lga_use_count: %u", lga_use_count);
- if (lga_use_count > 0) {
- /* Already created, just increment the use_count */
- lga_use_count++;
- goto done;
- } else {
+
+ if (cb->mds_hdl == 0) {
if ((rc = ncs_agents_startup()) != NCSCC_RC_SUCCESS) {
TRACE("ncs_agents_startup FAILED");
goto done;
}
if ((rc = lga_create()) != NCSCC_RC_SUCCESS) {
+ cb->mds_hdl = 0;
ncs_agents_shutdown();
goto done;
- } else {
- lga_use_count = 1;
}
/* Agent has successfully been started including communication
@@ -331,6 +327,9 @@ unsigned int lga_startup(lga_cb_t *cb)
set_lga_state(LGA_NORMAL);
}
+ /* Increase the use_count */
+ lga_use_count++;
+
done:
osaf_mutex_unlock_ordie(&lga_lock);
@@ -361,8 +360,12 @@ unsigned int lga_shutdown_after_last_cli
/* Users still exist, just decrement the use count */
lga_use_count--;
} else if (lga_use_count == 1) {
- lga_destroy();
- rc = ncs_agents_shutdown();
+ /*
+ * Disable MDS_UNINSTALL and ncs agent shutdown to keep MDS
connection.
+ * Fix ticket 1396. Msg NCSMDS_DOWN and Initialize are out of
order in Mbx.
+ */
+ /* lga_destroy(); */
+ /* rc = ncs_agents_shutdown(); */
lga_use_count = 0;
}
diff --git a/tests/logsv/logtest.h b/tests/logsv/logtest.h
--- a/tests/logsv/logtest.h
+++ b/tests/logsv/logtest.h
@@ -27,6 +27,7 @@
#include <utest.h>
#include <util.h>
#include <sys/wait.h>
+#include <unistd.h>
#include <osaf_time.h>
#include <logtrace.h>
diff --git a/tests/logsv/tet_saLogStreamOpen_2.c
b/tests/logsv/tet_saLogStreamOpen_2.c
--- a/tests/logsv/tet_saLogStreamOpen_2.c
+++ b/tests/logsv/tet_saLogStreamOpen_2.c
@@ -704,6 +704,89 @@ done:
}
}
+void *saLogInitialize_1 (void *arg)
+{
+ SaAisErrorT *rt = (SaAisErrorT *)arg;
+ SaLogStreamHandleT logStreamHandle1;
+ SaLogHandleT logHandle1;
+
+ *rt = saLogInitialize(&logHandle1, &logCallbacks, &logVersion);
+ if (*rt != SA_AIS_OK){
+ goto done;
+ }
+
+ *rt = saLogStreamOpen_2(logHandle1, &app1StreamName,
&appStream1LogFileCreateAttributes,
+ SA_LOG_STREAM_CREATE, SA_TIME_ONE_SECOND,
&logStreamHandle1);
+ if (*rt != SA_AIS_OK){
+ goto done;
+ }
+
+ *rt = saLogFinalize(logHandle1);
+ if (*rt != SA_AIS_OK){
+ goto done;
+ }
+done:
+ pthread_exit(NULL);
+}
+
+/*
+ * Ticket 1396
+ * Verify that saLogInitialize() then saLogFinalize() multiple times OK
+ */
+void saLogMultipleInitialize(void)
+{
+ SaLogStreamHandleT logStreamHandle1;
+ int count = 100;
+ SaAisErrorT rc = SA_AIS_OK;
+
+ while (count--) {
+ rc = saLogInitialize(&logHandle, &logCallbacks, &logVersion);
+ if (rc != SA_AIS_OK)
+ break;
+
+ rc = saLogStreamOpen_2(logHandle, &app1StreamName,
&appStream1LogFileCreateAttributes,
+ SA_LOG_STREAM_CREATE,
SA_TIME_ONE_SECOND, &logStreamHandle1);
+ safassert(rc, SA_AIS_OK);
+
+ rc = saLogFinalize(logHandle);
+ if (rc != SA_AIS_OK)
+ break;
+
+ }
+ test_validate(rc, SA_AIS_OK);
+
+}
+
+/*
+ * Ticket 1396
+ * Verify that saLogInitialize() then saLogFinalize() multiple times in
multiple threads OK
+ */
+void saLogMultiThreadMultiInit(void)
+{
+ int count = 100, i;
+ SaAisErrorT rt[count], rc = SA_AIS_OK;
+ pthread_t threads[count];
+
+ for(i = 0; i < count; i++ ){
+ safassert(pthread_create(&threads[i], NULL,
+ saLogInitialize_1, (void *) &rt[i]),
0);
+ }
+
+ /* Wait for all threads terminate */
+ for(i = 0; i < count; i++ ){
+ pthread_join(threads[i], NULL);
+ }
+
+ for(i = 0; i < count; i++ ){
+ if (rt[i] != SA_AIS_OK) {
+ rc = rt[i];
+ break;
+ }
+ }
+
+ test_validate(rc, SA_AIS_OK);
+}
+
extern void saLogStreamOpenAsync_2_01(void);
extern void saLogStreamOpenCallbackT_01(void);
extern void saLogWriteLog_01(void);
@@ -790,5 +873,7 @@ extern void saLogStreamClose_01(void);
test_case_add(2, verFixLogRec_Max_Err, "saLogStreamOpen_2 with
maxLogRecordSize > MAX_RECSIZE, ERR");
test_case_add(2, verFixLogRec_Min_Err, "saLogStreamOpen_2 with
maxLogRecordSize < 150, ERR");
test_case_add(2, saLogStreamOpen_2_50, "saLogStreamOpen_2 with stream
number out of the limitation, ERR");
+ test_case_add(2, saLogMultipleInitialize, "saLogInitialize() then
saLogFinalize() multiple times. keep MDS connection, OK");
+ test_case_add(2, saLogMultiThreadMultiInit, "saLogInitialize() then
saLogFinalize() multiple times in multiple threads, OK");
}
------------------------------------------------------------------------------
Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape
_______________________________________________
Opensaf-devel mailing list
Opensaf-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/opensaf-devel
