Bugs item #2975903, was opened at 2010-03-24 14:47
Message generated for change (Comment added) made by stmane
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=2975903&group_id=56967
Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Core
Group: MonetDB5 CVS Head
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Stefan de Konink (skinkie)
Assigned to: Nobody/Anonymous (nobody)
Summary: sequencial scan, segv
Initial Comment:
I have created a table on a 32bits system having a boolean colum consisting of
268435456 true values and one false value.
create table test2 (mytest boolean);
Then I executed the following query:
select * from test2 where mytest = false;
This resulted in:
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb198eb70 (LWP 514)]
0xb7135e46 in seqscan_eq_chr_void_tloc_oid_vid (b=0x8066550, bn=0x8931e58,
tl=0x88f01d4, th=0x88f01d4, str_idx=0)
at ../../../src/gdk//gdk_scanselect_defs.mx:119
119 ptr v = b...@3(bi, p);
(gdb) bt
#0 0xb7135e46 in seqscan_eq_chr_void_tloc_oid_vid (b=0x8066550, bn=0x8931e58,
tl=0x88f01d4, th=0x88f01d4, str_idx=0)
at ../../../src/gdk//gdk_scanselect_defs.mx:119
#1 0xb76d4097 in BAT_scanselect (b=0x8066550, bn=0x8931e58, tl=0x88f01d4,
th=0x88f01d4, li=1 '\001', hi=1 '\001', equi=1, nequi=0, lval=1, hval=1,
nocheck=0) at ../../../src/gdk/gdk_scanselect.mx:347
#2 0xb773b057 in BAT_select_ (b=0x8066550, tl=0x88f01d4, th=0x88f01d4,
li=1 '\001', hi=1 '\001', tail=0 '\000', anti=0 '\000',
preserve_order=0 '\000') at ../../../src/gdk/gdk_batop.mx:1229
#3 0xb773b9d3 in BATuselect (b=0x8066550, h=0x88f01d4, t=0x0)
at ../../../src/gdk/gdk_batop.mx:1306
#4 0xb2ea5a7a in CMDuselect1 (result=0xb198bdc4, b=0x8066550, value=0x88f01d4)
at ../../../../src/modules/kernel/algebra.mx:1434
#5 0xb2eb1412 in ALGuselect1 (result=0x88f0254, bid=0x88f0204,
value=0x88f01d4) at ../../../../src/modules/kernel/algebra.mx:2544
#6 0xb7f34a58 in DFLOWstep (t=0x88d96e8, fs=0xb1a8c7e0)
at ../../../src/mal/mal_interpreter.mx:1862
#7 0xb7f3ab06 in runDFLOWworker (t=0x88d96e8)
at ../../../src/mal/mal_interpreter.mx:1094
#8 0xb7bc573c in wrapper_routine (data=0xb1a8c79c)
at ../../../src/gdk/gdk_posix.mx:1013
#9 0xb70728ff in start_thread () from /lib/libpthread.so.0
#10 0xb6df738e in clone () from /lib/libc.so.6
----------------------------------------------------------------------
>Comment By: Stefan Manegold (stmane)
Date: 2010-03-24 17:31
Message:
Yet another race condition: the BAT has suddenly disappeared after we have
scanned 94765057 of the 268435458 BUNs ...
We'd need to know what the other threads are during at time of (or
actually just before) the crash, in particular whether any is just busy
with BBPtrim(), GDKvmtrim(), or MT_mmap_trim() and/or running madvise() or
munmap() on (any heap of) that BAT ...
----------------------------------------------------------------------
Comment By: Stefan de Konink (skinkie)
Date: 2010-03-24 17:08
Message:
Missed q;
(gdb) print q
$6 = 268435458
----------------------------------------------------------------------
Comment By: Stefan de Konink (skinkie)
Date: 2010-03-24 17:07
Message:
Observation;
It seems to work on 'new' start sometimes. But when I insert an extra row
on the end it breaks again. Could it be an out of memory situation where a
malloc fails?
[Switching to Thread 0xb198eb70 (LWP 548)]
0xb7135e46 in seqscan_eq_chr_void_tloc_oid_vid (b=0x8066550,
bn=0x892a280,
tl=0x892a5bc, th=0x892a5bc, str_idx=0)
at ../../../src/gdk//gdk_scanselect_defs.mx:119
119 ptr v = b...@3(bi, p);
(gdb) print b
$1 = (BAT *) 0x8066550
(gdb) print p
$2 = 94765057
(gdb) print *b
$3 = {batCacheid = 0, H = 0x0, T = 0x0, P = 0x0, U = 0x0}
(gdb) print bi
$4 = {b = 0x8066550, hvid = 0, tvid = 0}
----------------------------------------------------------------------
Comment By: Stefan Manegold (stmane)
Date: 2010-03-24 15:17
Message:
likewise for b & q
----------------------------------------------------------------------
Comment By: Stefan Manegold (stmane)
Date: 2010-03-24 15:14
Message:
what are the values of bi & p at the time of the crash?
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=2975903&group_id=56967
------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Monetdb-bugs mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/monetdb-bugs
