Malloc() receives signals and the process hangs

Question

Hi, I am having a batch process which receives a signal (sometimes SIGSEGV and someother times SIGUSR) during malloc and hangs forever. Please refer the following pstack and let me know, if it is related to some patch issue or heap corruption issue.
Os Details
SunOS dxxxxx09 5.8 Generic_117350-46 sun4u sparc SUNW,Ultra-Enterprise-10000
Pstack
----------------- lwp# 1 / thread# 1 --------------------
ffffffff748a79bc lwp_sema_wait (1002666d0)
ffffffff74e0c3d0 _park (1002666d0, ffffffff74f20000, 0, 1002665c0, 2, ffffffff74f21c74) + 114
ffffffff74e0c09c _swtch (1002665c0, ffffffff749c0308, ffffffff74f20000, 0, 5, 1000) + 424
ffffffff74e0db80 mutexadaptive_lock (ffffffff749c0308, ffffffff74f2f5d8, 4c00, 1000, fffeffff, 1) + 164
ffffffff74e0d888 cmutexlock (ffffffff749c0308, ffffffff74f20000, ffffffff7f72e930, ffffffff7484ab5 4, ffffffff7a701d28, ffffffff7fff982c) + 84
ffffffff7484ab54 free (100272900, ffffffff76810b98, ffffffff768384d0, ffffffff7a701d28, ffffffff7f72 e930, ffffffff73d00838) + 18
ffffffff763179bc meFree (1, 100272900, ffffffff7fff9980, ffffffff768384d0, ffffffff72800000, a) + 1c
ffffffff7683348c shm_deallocate (100378b80, ffffffff7fffa0a8, ffffffff769525e0, ffffffff769525d8, ff ffffff7694a5f0, 1) + 24
ffffffff768384d0 discshm (0, c350, ffffffff7fffa0a8, ffffffff7682d6b8, ffffffff7694a5f0, ffffffff73d 00838) + 250
ffffffff76830ab0 tlDiscon (10030f980, a, ffffffff7681ccfc, ffffffff76826cb8, ffffffff7f72e930, fffff fff73d00838) + b4
ffffffff7682d6b8 slSQIdiscon (0, a, a, ffffffff7fffa0a8, ffffffffffffffff, 10037eea0) + 2b8
ffffffff76826cb8 pfDiscon (ffffffff7fffa0a8, a, ffffffff7fffa008, ffffffff7fffa0a8, 10030f980, 10037 ee00) + b8
ffffffff7681ccfc cmDiscon (a, 10030fa50, ffffffff7fffa008, ffffffff76814dc4, 0, 0) + ac
ffffffff7681198c ascAbort (ffffffff7fffa0a8, 10030fa50, ffffffff7fffa000, ffffffff7fffa0a8, ffffffff 7694a5f0, ffffffffffffff84) + 24
ffffffff76814dc4 asfExit (10030ff30, ffffffff7fffa0a8, a, 0, 0, 10030fa50) + 224
ffffffff76814b34 ASF_Call (a, ffffffff76950f18, f0, ffffffff7fffa000, ffffffff7fffa0a8, 10030fa50) + 3e4
ffffffff76a1388c asf_abort (10030f980, 2, 1002f6870, 2, ffffffff7a701680, 10037c816) + 84
ffffffff76a1546c disconnect (1002f6870, 2, 0, 1002b4ae0, 0, 1002f7140) + 184
ffffffff76a1579c disconnect_all (1002f6870, 0, ffffffff76b63a00, 2, 2, 1002f6870) + 13c
ffffffff76a15a74 sqli_connect_close_tcb (0, 100268610, 2, ffffffff76b5df48, 0, 2) + a4
ffffffff76a2b804 sqlexit (1002f6870, 0, 0, 0, 0, 0) + 34
0000000100077c40 __1cOsignal_handler6Fi_v_ (a, ffffffff7fffaf90, ffffffff7fffacb0, 0, 0, 0) + 10
ffffffff74e1edd8 __sighndlr (a, ffffffff7fffaf90, ffffffff7fffacb0, 100077c30, 8000, 0) + c
ffffffff74e1b668 sigacthandler (1002665c0, 0, 0, 0, ffffffff7fffacb0, a) + 71c
--- called from signal handler with signal 2516416 (SIG Unknown) ---
ffffffff7484aa60 t_splay (0, 100470ff0, ffffffff749b6f60, 140, 0, 2d60) + 110
ffffffff7484a3a4 realfree (100470fe0, ffffffff749c3780, 100470eb0, ffffffff749b6f60, 100470ea0, 141) + a0
ffffffff7484acb8 cleanfree (0, ffffffff749c3670, ffffffff749b6f60, 0, ffffffff749c3668, ffffffff749c 3768) + 60
ffffffff74849da0 mallocunlocked (a, 0, ffffffff749b6f60, 10, 5, 0) + f8
ffffffff74849c8c malloc (a, 100268610, 0, 0, 1002f6c18, 2000c) + 20
ffffffff74c08820 __1c2n6FL_pv_ (a, ffffffff74c07830, 0, 104274, ffffffff74d0ca70, 1002f6c18) + 28
00000001000ee2a8 __1cIpre_sfno6F_v_ (0, 0, ffffffff7fffba57, ffffffff7fffbce4, ffffffff7fffba6b, 2) + 40d0
00000001000679dc __1cTprocessStagingTable6F_i_ (0, ffffffff7fffc720, ffffffff7fffcd78, 20, 132, ffff ffff7fffc843) + 32c
0000000100076948 main (9, ffffffff7fffcf08, ffffffff7fffcf58, 0, 0, 100000000) + 21e8
000000010003cd1c _start (0, 0, 0, 0, 0, 0) + 17c
----------------- lwp# 2 / thread# 2 --------------------
ffffffff748a74c0 signotifywait ()
ffffffff74e11794 _dynamiclwps (ffffffff74f20000, ffffffff746044b6, ffffffff74b01698, ffffffff73d0083                                          8, 6d8b5, 0) + 24
ffffffff74e14d38 thr_yield (0, 0, 0, 0, 0, 0) + 8c
----------------- lwp# 4 / thread# 4 --------------------
ffffffff74e1c77c lwp_cond_wait (ffffffff74f21c20, ffffffff74f21c08, 0)
ffffffff74e0b684 cotimerset (ffffffff74f21c20, ffffffff74f20000, ffffffff74f21c08, 3, 0, 0) + a0
ffffffff74e1ecd8 threadstart (0, 0, 0, 0, 0, 0) + 40
----------------- lwp# 5 --------------------------------
ffffffff748a5534 door (0, 0, 0, 0, ffffffff73209ba0, 4)
ffffffff74e19c94 scdoor_func (ffffffff73209c50, ffffffff74f20000, 4, 3, ffffffff74f20000, 1) + 6c
ffffffff74e0d0c8 lwpstart (ffffffff73209c50, 0, 0, 0, 0, 0) + 18
ffffffff74e14d38 thr_yield (0, 0, 0, 0, 0, 0) + 8c
----------------- lwp# 6 / thread# 3 --------------------
ffffffff748a79bc lwp_sema_wait (ffffffff7360dd60)
ffffffff74e0c3d0 _park (ffffffff7360dd60, ffffffff74f20000, 0, ffffffff7360dc50, 1, ffffffff74f21c74                                          ) + 114
ffffffff74e0c09c _swtch (ffffffff7360dc50, ffffffff74f260c8, ffffffff74f20000, 0, 5, 1000) + 424
ffffffff74e10800 reapwait (ffffffff74f260c8, ffffffff74f2b618, 0, 0, 0, 0) + 38
ffffffff74e10550 _reaper (ffffffff74f21d18, ffffffff74f260c8, 0, ffffffff74f21cf0, 0, 1000) + 38
ffffffff74e1ecd8 threadstart (0, 0, 0, 0, 0, 0) + 40

Answer 1

Thanks a lot for your response.
Yes, I executed the program in dbx using RTC. There are couple of Read from uninitialized (rui) errors and Read from unallocated (rua). Surprisingly, the (rua) error comes from the location where I am using EXEC SQL FREE.
The following is the code inside DoPrepare():
db_count++;
EXEC SQL PREPARE curs_stm_1 FROM ' select * from eq_to_pr
where location_id = ? and e_trak = ?';
db_count++;
EXEC SQL DECLARE cc_bay CURSOR WITH HOLD FOR curs_stm_1;
EXEC SQL FREE curs_stm_1;
db_count++;
return true; /* SUCCESS*/
In the above code, dbx points Read from unallocated (rua) error from EXEC SQL FREE statement. When I commented out, the error is gone.
The samething with the following code too:
EXEC SQL CLOSE cc_bay;
EXEC SQL FREE cc_bay;
Note: I am using IBM Informix CSDK Version 2.80, IBM Informix-ESQL Version 9.52.FC2
Otherthan the above errors, RTC output was not having anything else.
I dont know why EXEC SQL FREE is being flaged.
The following is the dbx output:
<rtc> Read from unallocated (rua) on thread 1:
Attempting to read 4 bytes at address 0x1006a1dcc
which is 75 bytes past end of heap block of size 121 bytes at 0x1006a1d08
This block was allocated from:
[1] sqgdesc() at 0xffffffff5312af94
[2] process_msgs() at 0xffffffff5312f0d4
[3] sqrmessages() at 0xffffffff5312bccc
[4] _iqprepare() at 0xffffffff5311ae14
[5] sqli_prep_tcb() at 0xffffffff5311ab74
[6] sqli_prep() at 0xffffffff5311aa48
[7] DoPrepare() at line 275 in "oreo_eq_add.ec"
[8] oreo_eq_add() at line 390 in "oreo_eq_add.ec"
Location of error:
current thread: t@1
=>[1] sqli_curs_free_tcb(0x1005cbdf8, 0x1006a1db0, 0x2010c, 0x0, 0x1005a67d8, 0x0), at 0xffffffff531197dc
[2] DoPrepare(), line 284 in "oreo_eq_add.ec"
[3] oreo_eq_add(e_trak = 29365102, e_error1 = 0, process = 9, equip_id = 0), line 390 in "oreo_eq_add.ec"
[4] pre_bay_sc(), line 452 in "pre_bay_sc.ec"
[5] processStagingTable(), line 576 in "OreoDatabaseFunctions.ec"
[6] main(argc = 9, argv = 0xffffffff7fffccb8), line 327 in "OreoMain.ec"