Linux FC3 - mprime v23.9 : Segmentation fault (core dumped) nohup ./mp -d

T.Rex · 2005-04-28, 16:15

Hi,

On Linux Fedora Core3, I have a Segmentation Fault with mprime v23.9

. It is easily reproduceable on my site.

Can someone provide an idea about the cause ?
Can someone try to reproduce the problem ?

Tony

# cat worktodo.ini
Test=34800697,68,0

Traces:
Starting factoring M34800697 to 2^68
[Apr 26 09:48] Factoring M34800697 to 2^63 is 7.41% complete. Time: 113.859 sec.
....
[Apr 27 18:22] Factoring M34800697 to 2^68 is 99.91% complete. Time: 139.419 sec.
M34800697 no factor to 2^68, WZ2: 4E02E79D
The program will now perform a self-test to make sure the
Lucas-Lehmer code is working properly on your computer.
This will take about an hour.
Test 1, 1900 Lucas-Lehmer iterations of M39151585 using 2048K FFT length.
.......
Test 21, 2500 Lucas-Lehmer iterations of M29311553 using 2048K FFT length.
Self-test 2048K passed!
Starting P-1 factoring on M34800697 with B1=385000, B2=6160000
Chance of finding a factor is an estimated 4.95%
P-1 on M34800697 with B1=385000, B2=6160000
[Apr 27 19:38] M34800697 stage 1 is 1.80% complete. Time: 861.220 sec.
.........
[Apr 28 08:33] M34800697 stage 1 is 99.03% complete. Time: 862.855 sec.
M34800697 stage 1 complete. 1110666 transforms. Time: 47821.138 sec.
Starting stage 1 GCD - please be patient.
Stage 1 GCD complete. Time: 189.480 sec.

SEGMENTATION FAULT HERE

If I relaunch mprime -d again the SegFault reappears :

Mersenne number primality test program version 23.9
Starting P-1 factoring on M34800697 with B1=385000, B2=6160000
Chance of finding a factor is an estimated 4.95%
P-1 on M34800697 with B1=385000, B2=6160000
M34800697 stage 1 complete. 19464 transforms. Time: 791.305 sec.
Starting stage 1 GCD - please be patient.
Stage 1 GCD complete. Time: 187.904 sec.
Mersenne number primality test program version 23.9
Starting P-1 factoring on M34800697 with B1=385000, B2=6160000
Chance of finding a factor is an estimated 4.95%
P-1 on M34800697 with B1=385000, B2=6160000
M34800697 stage 1 complete. 19464 transforms. Time: 810.450 sec.
Starting stage 1 GCD - please be patient.
Stage 1 GCD complete. Time: 188.289 sec.
Mersenne number primality test program version 23.9
Starting P-1 factoring on M34800697 with B1=385000, B2=6160000
Chance of finding a factor is an estimated 4.95%
P-1 on M34800697 with B1=385000, B2=6160000
M34800697 stage 1 complete. 19464 transforms. Time: 829.955 sec.
Starting stage 1 GCD - please be patient.
Stage 1 GCD complete. Time: 188.313 sec.

Here is what gdb says about the core.
It seems to be a problem with the allocation of memory.

Reading symbols from shared object read from target memory...(no debugging symbols found)...done.
Loaded system supplied DSO at 0xffffe000
Core was generated by `./mp -d'.
Program terminated with signal 11, Segmentation fault.

warning: svr4_current_sos: Can't read pathname for load map: Input/output error

Reading symbols from /lib/tls/libm.so.6...(no debugging symbols found)...done.
Loaded symbols for /lib/tls/libm.so.6
Reading symbols from /lib/tls/libc.so.6...(no debugging symbols found)...done.
Loaded symbols for /lib/tls/libc.so.6
Reading symbols from /lib/ld-linux.so.2...(no debugging symbols found)...done.
Loaded symbols for /lib/ld-linux.so.2
#0 0x4c861ab9 in free () from /lib/tls/libc.so.6
(gdb) where
#0 0x4c861ab9 in free () from /lib/tls/libc.so.6
#1 0x08055970 in ?? ()
#2 0xb7ce2b49 in ?? ()
#3 0x00000001 in ?? ()
#4 0xbfffdeb8 in ?? ()
#5 0x0806bf19 in ?? ()
#6 0x1af01573 in ?? ()
#7 0x00000000 in ?? ()

# cat local.ini
OldCpuType=12
OldCpuSpeed=2791
CPUHours=24
DayMemory=200
NightMemory=200
DayStartTime=840
DayEndTime=1840
Pid=3674
RollingStartTime=0
SelfTest1536Passed=1
RollingAverage=1153
SelfTest2048Passed=1

# cat prime.ini
UserPWD=
UserName=......
UserEmailAddr=........
Newsletters=0
UserID=
AskedAboutMemory=1
UsePrimenet=0
DialUp=0
DaysOfWork=7
WorkPreference=0
OutputIterations=10000
ResultsFileIterations=999999999
DiskWriteTime=30
TwoBackupFiles=1
SilentVictory=0

Prime95 · 2005-04-28, 21:35

Use 24.6. There is a not too uncommon memory corruption bug in the GCD code.

T.Rex · 2005-04-29, 07:22

OK, I've got it and will experiment with soon.

Huuum, there is no evident link to:
http://www.mersenne.org/gimps/
from:
http://www.mersenne.org/freesoft.htm

Maybe it could help to say that alpha versions can be found there.

Thanks,
Tony

blackguard · 2005-05-02, 12:08

yes, it would be nice if the download section of mersenne.org was kept a bit more up to date. v24.6 has been out for some time and seems to be the only solution to some of the problems that come up when running FC and mprime, but if people can't find it in the downloads list they may give up rather that dig through the forums for an answer.

T.Rex · 2005-05-02, 12:44

The SegFault with 23.9 appeared with another exponent.
And it disappeared when using 24.6 with both exponents !
Seems OK now.

devjonfos · 2005-06-22, 04:22

I had the same problem on Gentoo Linux 2.6.11-gentoo-r9. Upgrading mprime to 24.6 solved the problem. Thanks.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
mprime 28.9 - occassional segmentation fault during start	pessoft	Software	0	2016-06-13 20:58
odd segmentation fault	ChristianB	YAFU	4	2015-09-09 19:38
Segmentation fault	PhilF	Linux	5	2006-01-07 17:12
mprime segmentation fault on RHEL	bej	Software	28	2005-11-11 19:05
Segmentation Fault	sirius56	Software	2	2004-10-02 21:43

2005-04-28, 16:15	#1
T.Rex Feb 2004 France 3×311 Posts	Linux FC3 - mprime v23.9 : Segmentation fault (core dumped) nohup ./mp -d Hi, On Linux Fedora Core3, I have a Segmentation Fault with mprime v23.9 . It is easily reproduceable on my site. Can someone provide an idea about the cause ? Can someone try to reproduce the problem ? Tony # cat worktodo.ini Test=34800697,68,0 Traces: Starting factoring M34800697 to 2^68 [Apr 26 09:48] Factoring M34800697 to 2^63 is 7.41% complete. Time: 113.859 sec. .... [Apr 27 18:22] Factoring M34800697 to 2^68 is 99.91% complete. Time: 139.419 sec. M34800697 no factor to 2^68, WZ2: 4E02E79D The program will now perform a self-test to make sure the Lucas-Lehmer code is working properly on your computer. This will take about an hour. Test 1, 1900 Lucas-Lehmer iterations of M39151585 using 2048K FFT length. ....... Test 21, 2500 Lucas-Lehmer iterations of M29311553 using 2048K FFT length. Self-test 2048K passed! Starting P-1 factoring on M34800697 with B1=385000, B2=6160000 Chance of finding a factor is an estimated 4.95% P-1 on M34800697 with B1=385000, B2=6160000 [Apr 27 19:38] M34800697 stage 1 is 1.80% complete. Time: 861.220 sec. ......... [Apr 28 08:33] M34800697 stage 1 is 99.03% complete. Time: 862.855 sec. M34800697 stage 1 complete. 1110666 transforms. Time: 47821.138 sec. Starting stage 1 GCD - please be patient. Stage 1 GCD complete. Time: 189.480 sec. SEGMENTATION FAULT HERE If I relaunch mprime -d again the SegFault reappears : Mersenne number primality test program version 23.9 Starting P-1 factoring on M34800697 with B1=385000, B2=6160000 Chance of finding a factor is an estimated 4.95% P-1 on M34800697 with B1=385000, B2=6160000 M34800697 stage 1 complete. 19464 transforms. Time: 791.305 sec. Starting stage 1 GCD - please be patient. Stage 1 GCD complete. Time: 187.904 sec. Mersenne number primality test program version 23.9 Starting P-1 factoring on M34800697 with B1=385000, B2=6160000 Chance of finding a factor is an estimated 4.95% P-1 on M34800697 with B1=385000, B2=6160000 M34800697 stage 1 complete. 19464 transforms. Time: 810.450 sec. Starting stage 1 GCD - please be patient. Stage 1 GCD complete. Time: 188.289 sec. Mersenne number primality test program version 23.9 Starting P-1 factoring on M34800697 with B1=385000, B2=6160000 Chance of finding a factor is an estimated 4.95% P-1 on M34800697 with B1=385000, B2=6160000 M34800697 stage 1 complete. 19464 transforms. Time: 829.955 sec. Starting stage 1 GCD - please be patient. Stage 1 GCD complete. Time: 188.313 sec. Here is what gdb says about the core. It seems to be a problem with the allocation of memory. Reading symbols from shared object read from target memory...(no debugging symbols found)...done. Loaded system supplied DSO at 0xffffe000 Core was generated by `./mp -d'. Program terminated with signal 11, Segmentation fault. warning: svr4_current_sos: Can't read pathname for load map: Input/output error Reading symbols from /lib/tls/libm.so.6...(no debugging symbols found)...done. Loaded symbols for /lib/tls/libm.so.6 Reading symbols from /lib/tls/libc.so.6...(no debugging symbols found)...done. Loaded symbols for /lib/tls/libc.so.6 Reading symbols from /lib/ld-linux.so.2...(no debugging symbols found)...done. Loaded symbols for /lib/ld-linux.so.2 #0 0x4c861ab9 in free () from /lib/tls/libc.so.6 (gdb) where #0 0x4c861ab9 in free () from /lib/tls/libc.so.6 #1 0x08055970 in ?? () #2 0xb7ce2b49 in ?? () #3 0x00000001 in ?? () #4 0xbfffdeb8 in ?? () #5 0x0806bf19 in ?? () #6 0x1af01573 in ?? () #7 0x00000000 in ?? () # cat local.ini OldCpuType=12 OldCpuSpeed=2791 CPUHours=24 DayMemory=200 NightMemory=200 DayStartTime=840 DayEndTime=1840 Pid=3674 RollingStartTime=0 SelfTest1536Passed=1 RollingAverage=1153 SelfTest2048Passed=1 # cat prime.ini UserPWD= UserName=...... UserEmailAddr=........ Newsletters=0 UserID= AskedAboutMemory=1 UsePrimenet=0 DialUp=0 DaysOfWork=7 WorkPreference=0 OutputIterations=10000 ResultsFileIterations=999999999 DiskWriteTime=30 TwoBackupFiles=1 SilentVictory=0

2005-04-28, 21:35	#2
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 3²×907 Posts	Use 24.6. There is a not too uncommon memory corruption bug in the GCD code.

2005-04-29, 07:22	#3
T.Rex Feb 2004 France 3A5₁₆ Posts	OK, I've got it and will experiment with soon. Huuum, there is no evident link to: http://www.mersenne.org/gimps/ from: http://www.mersenne.org/freesoft.htm Maybe it could help to say that alpha versions can be found there. Thanks, Tony

2005-05-02, 12:08	#4
blackguard Jan 2005 Singapore 13₁₀ Posts	yes, it would be nice if the download section of mersenne.org was kept a bit more up to date. v24.6 has been out for some time and seems to be the only solution to some of the problems that come up when running FC and mprime, but if people can't find it in the downloads list they may give up rather that dig through the forums for an answer.

2005-05-02, 12:44	#5
T.Rex Feb 2004 France 3·311 Posts	Pb fixed with 24.6 The SegFault with 23.9 appeared with another exponent. And it disappeared when using 24.6 with both exponents ! Seems OK now.

2005-06-22, 04:22	#6
devjonfos Jun 2005 Portland Metro Area 3₈ Posts	I had the same problem on Gentoo Linux 2.6.11-gentoo-r9. Upgrading mprime to 24.6 solved the problem. Thanks.