Bug 134304

Summary: [cannaLE] random gtk2 delays, auxmenu defunct
Product: [Fedora] Fedora Reporter: Warren Togami <wtogami>
Component: im-sdkAssignee: Akira TAGOH <tagoh>
Status: CLOSED RAWHIDE QA Contact:
Severity: medium Docs Contact:
Priority: medium    
Version: rawhideCC: eng-i18n-bugs, wtogami
Target Milestone: ---Keywords: i18n
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: im-sdk-12.0.1-17.svn1994 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2004-10-19 02:20:07 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 125997, 130887    

Description Warren Togami 2004-10-01 07:45:56 UTC
Description of problem:
It seems that all newer snapshots of svn1943 have severe trouble with
canna.  All kinds of difficult behavior like:

* Sometimes gtk2 windows or tabs take 2 seconds to draw, spinning with
100% CPU usage.
* Sometimes gtk2 windows or tabs take 2-5 seconds to close, spinning
with 100% CPU usage.
* Sometimes applications like xchat, firefox, or gaim crash completely.
* Sometimes canna fails to activate at all.
* Sometimes canna activates, but fails to convert any input text.

considering to use lang ja
searching module... CannaLE.so
During window open, htt_server -d stuck here, then completes drawing
and outputs:
Create session context(8ba12f8)

#0  0xf6fe9782 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
#1  0xf6dcc444 in poll () from /lib/tls/libc.so.6
#2  0x08091bbe in IMSocketListen::accept (this=0x8b58748) at
IMUtil.cpp:632
#3  0x08072b61 in IIIMProtocol::accept (this=0x8b615d0, flags=0) at
IIIMProtocol.cpp:59
#4  0x0806e904 in IMScheduler_MTPC::start (this=0x8b50800) at
IMScheduler_MTPC.cpp:53
#5  0x0804da8a in IMSvr::start (this=0xfef0aa00) at IMSvr.cpp:82
#6  0x0804d44a in main (argc=2, argv=0xfef0ab74) at main.cpp:44
#7  0xf6d26af3 in __libc_start_main () from /lib/tls/libc.so.6
#8  0x0804d1f9 in _start ()

Above is a backtrace during this window open spinning.

Version-Release number of selected component (if applicable):
im-sdk-12.0.1-10.svn1943

Comment 1 Warren Togami 2004-10-01 07:52:34 UTC
warren   11509  7177  0 21:29 ?        00:00:00 [auxmenu] <defunct>
I forgot to mention that I have dozens of these defunct processes that
are unkillable with even kill -9.  tagoh thinks this is somehow
related to the trouble I am seeing.


Comment 2 Warren Togami 2004-10-01 08:06:00 UTC
(gdb) thread apply all bt full

Thread 2 (Thread -154874960 (LWP 11737)):
#0  0xf6fe9782 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
No symbol table info available.
#1  0xf6e694cb in __read_nocancel () from /lib/tls/libpthread.so.0
No symbol table info available.
#2  0x00bac1a3 in RkcRecvWReply () from /usr/lib/libcanna.so.1
No symbol table info available.
#3  0x00bacc59 in RkcSendWRequest () from /usr/lib/libcanna.so.1
No symbol table info available.
#4  0x00000005 in ?? ()
No symbol table info available.
#5  0x00000000 in ?? ()
No symbol table info available.

Thread 1 (Thread -154156288 (LWP 11734)):
#0  0xf6fe9782 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
No symbol table info available.
#1  0xf6dcc444 in poll () from /lib/tls/libc.so.6
No symbol table info available.
#2  0x08091bbe in IMSocketListen::accept (this=0x9d6f748) at
IMUtil.cpp:632
        fd = 8
        i = 1
        fds = (pollfd *) 0x9d7dd10
        it = {_M_current = 0x9d6f764}
        pims = (IMSocket *) 0x9d78678
        n = 1
#3  0x08072b61 in IIIMProtocol::accept (this=0x9d785d0, flags=0) at
IIIMProtocol.cpp:59
        pimt = (IIIMPTrans *) 0xfeedb194
        pims = (class IIIMP_IMState *) 0xf6e32ff4
        pimst = (class IMSocketTrans *) 0x0
#4  0x0806e904 in IMScheduler_MTPC::start (this=0x9d67800) at
IMScheduler_MTPC.cpp:53
        pims = (class IMState *) 0x0
        th = 4140092336
        attr = {__detachstate = 0, __schedpolicy = 0, __schedparam =
{__sched_priority = 1}, __inheritsched = 4096,
  __scope = 0, __guardsize = 0, __stackaddr_set = 0, __stackaddr =
0x0, __stacksize = 0}
#5  0x0804da8a in IMSvr::start (this=0xfeedb020) at IMSvr.cpp:82
No locals.
---Type <return> to continue, or q <return> to quit---
#6  0x0804d44a in main (argc=2, argv=0xfeedb194) at main.cpp:44
        svr = {<IMAccept> = {_vptr.IMAccept = 0x8098610}, pcfg =
0xfeedb0b0, usermgr = {<IMAuth> = {
      _vptr.IMAuth = 0x8099250, ae_num = 3, alloced_ae_num = 10,
auth_type = 2, pae = 0x9d669e8, def_at = DENY, sm = {
        alloced_size = 10, size = 2, psym = 0x9d66990}, from_address =
0x0, from_address_bitlen = 0,
      from_hostname = 0x9d7a5e0 "localhost", command_name = 0x9d6fc30
"/usr/sbin/htt_server", domainname = 0x0},
    alloced_ue_num = 0, ue_num = 0, pue = 0x0, sysuser_at =
IMAuth::PASSWORD, usermap = {_M_t = {
        _M_impl = {<std::allocator<std::_Rb_tree_node<std::pair<const
u16string, IMUser*> > >> =
{<__gnu_cxx::new_allocator<std::_Rb_tree_node<std::pair<const
u16string, IMUser*> > >> = {<No data fields>}, <No data fields>},
          _M_key_compare =
{<std::binary_function<u16string,u16string,bool>> = {<No data
fields>}, <No data fields>},
          _M_header = {_M_color = std::_S_red, _M_parent = 0x9d7dd50,
_M_left = 0x9d7dd50, _M_right = 0x9d7dd50},
          _M_node_count = 1}}}}, plemgr = 0x9d7a340, pimprotocol =
0x9d785d0,
  addrvec =
{<std::_Vector_base<IMSocketAddress,std::allocator<IMSocketAddress> >> = {
      _M_impl = {<std::allocator<IMSocketAddress>> =
{<__gnu_cxx::new_allocator<IMSocketAddress>> = {<No data fields>}, <No
data fields>}, _M_start = 0x9d6f730, _M_finish = 0x9d6f744,
_M_end_of_storage = 0x9d6f744}}, <No data fields>},
  imclist =
{<std::_List_base<IMConnection*,std::allocator<IMConnection*> >> = {
      _M_impl = {<std::allocator<std::_List_node<IMConnection*> >> =
{<__gnu_cxx::new_allocator<std::_List_node<IMConnection*> >> = {<No
data fields>}, <No data fields>}, _M_node = {_M_next = 0x9d7de28,
_M_prev = 0x9d7de28}}}, <No data fields>},
  pkeyparser = 0x9d7a580}
        pimlog = (class IMLog *) 0x9d677e8
        pimsch = (class IMScheduler *) 0x9d67800
        pimobjmgr = (IMObjectMgr *) 0x9d67810
        exitcode = -152883212
        arg = {<IMSvrCfg> = {_vptr.IMSvrCfg = 0x8098968, pbase = 0x0,
popts = 0x9d66034, command_name = {
      static npos = 4294967295,
      _M_dataplus = {<std::allocator<char>> =
{<__gnu_cxx::new_allocator<char>> = {<No data fields>}, <No data fields>},
        _M_p = 0x9d66014 "/usr/sbin/htt_server"}}}, valid_flag = true,
pxmlsvrcfg = 0x9d6f8a0, pxmllecfg = 0x9d6f8f0,
  optmap = {_M_t = {
      _M_impl = {<std::allocator<std::_Rb_tree_node<std::pair<const
char* const, IMSvrArg::ArgVal> > >> =
{<__gnu_cxx::new_allocator<std::_Rb_tree_node<std::pair<const char*
const, IMSvrArg::ArgVal> > >> = {<No data fields>}, <No data fields>},
        _M_key_compare = {<No data fields>}, _M_header = {_M_color =
std::_S_red, _M_parent = 0x9d666c0,
          _M_left = 0x9d667e0, _M_right = 0x9d66960}, _M_node_count =
15}}}}
        pimsignal = (IMSignal *) 0x9d66fb8
        ptls = (class IMTLS *) 0x9d67810
#7  0xf6d26af3 in __libc_start_main () from /lib/tls/libc.so.6
No symbol table info available.
#8  0x0804d1f9 in _start ()
No symbol table info available.


Comment 3 Warren Togami 2004-10-06 08:35:26 UTC
12.0.1-8.svn1891 seems to have the same problem, although seemingly
less serious.  In any case, I consider this to be serious enough to be
a FC3 blocker.

Comment 4 Akira TAGOH 2004-10-06 13:49:01 UTC
12.0.1-13.svn1943 should be better than others. at least it's no
auxmenu defunct and no crashes for me. please try.

Comment 5 Warren Togami 2004-10-09 22:22:03 UTC
auxmenu still somehow becomes defunct, although not as easily. 
Crashes are indeed gone, but slowness is still apparent.

Comment 6 Akira TAGOH 2004-10-12 05:02:38 UTC
btw doesn't this problem appear on iiimf-le-chinput which also uses
the aux object?

Comment 7 Warren Togami 2004-10-13 05:08:38 UTC
The gtk2 delays seem gone with -14, and so far no auxmenu defunct
processes yet.  Please keep this open a day or two for verification. 
I will test with cannaLE, and llim should test with S Chinese.

Comment 8 Akira TAGOH 2004-10-18 09:54:34 UTC
any progress for testing?

Comment 9 Warren Togami 2004-10-18 09:57:20 UTC
Good news: No defunct processes.
Bad  news: Slow gtk2 window closing when cannaLE is the current LE.


Comment 10 Lawrence Lim 2004-10-18 10:07:17 UTC
Nothing significant for both cannaLE and chinput LE. auxmenu does not
appear anymore. I will leave it turn on tonight and see if anything
will happen. :)

However, I did notice the gtk2 window closing issue on my box as well.


Comment 11 Lawrence Lim 2004-10-19 02:20:07 UTC
Confirmed fixed in im-sdk-12.0.1-17.svn1994. Resolving the bug for now.