[Cs-dev] Serious bug (I think in CsoundQt) on WIndows
Date | 2013-08-14 12:06 |
From | Michael Gogins |
Subject | [Cs-dev] Serious bug (I think in CsoundQt) on WIndows |
Attachments | None None |
When running the latest build of CsoundQt-d on Windows with the Stockhausen-Studie example:
warning: HEAP[CsoundQt-d.exe]: warning: Heap block at 0453ED08 modified at 0453ED21 past requested size of 11
Program received signal SIGTRAP, Trace/breakpoint trap. [Switching to Thread 5848.0x454] 0x777bc156 in ?? () (gdb) bt #0 0x777bc156 in ?? ()
#1 0x777a095b in ?? () #2 0x777719ff in ?? () #3 0x777bd434 in ?? () #4 0x7776c67c in ?? () #5 0x7774980d in ?? () #6 0x765dcf54 in realloc () from C:\Windows\SysWOW64\msvcrt.dll
#7 0x01390000 in ?? () #8 0x6c5126b9 in mrealloc (csound=csound@entry=0x51236e0, oldp=0x453ed18, size=size@entry=85) at c:/Users/mkg.sorabji/csound-csound6-git/Engine/memalloc.c:186
#9 0x6c552e96 in strcat_opcode (csound=0x51236e0, p=0x4394a2c) at c:/Users/mkg.sorabji/csound-csound6-git/OOps/str_ops.c:254 #10 0x6c510d2c in useropcdset (csound=0x51236e0, p=0x547c1b8) at c:/Users/mkg.sorabji/csound-csound6-git/Engine/insert.c:1223
#11 0x6c50f938 in insert (csound=csound@entry=0x51236e0, insno=insno@entry=1, newevtp=newevtp@entry=0x5126898) at c:/Users/mkg.sorabji/csound-csound6-git/Engine/insert.c:269 #12 0x6c51466e in process_score_event (csound=csound@entry=0x51236e0, evt=evt@entry=0x5126898, rtEvt=rtEvt@entry=0)
at c:/Users/mkg.sorabji/csound-csound6-git/Engine/musmon.c:752 #13 0x6c5159fa in sensevents (csound=csound@entry=0x51236e0) at c:/Users/mkg.sorabji/csound-csound6-git/Engine/musmon.c:941 #14 0x6c5d3e59 in csoundPerformKsmps (csound=0x51236e0) at c:/Users/mkg.sorabji/csound-csound6-git/Top/csound.c:1606
#15 0x6b8895ee in CsoundPerformanceThread::Perform (this=0x44689c0) at c:/Users/mkg.sorabji/csound-csound6-git/interfaces/csPerfThread.cpp:300 #16 0x649458cb in pthread_create_wrapper () from c:\Users\mkg.sorabji\csound-csound6-git\libwinpthread-1.dll
#17 0x765e906a in ui64tow () from C:\Windows\SysWOW64\msvcrt.dll #18 0x765e9147 in msvcrt!iswalnum () from C:\Windows\SysWOW64\msvcrt.dll #19 0x771b850d in KERNEL32!BaseThreadInitThunk () from C:\Windows\SysWOW64\kernel32.dll
#20 0x7773bf39 in ?? () #21 0x7773bf0c in ?? () #22 0x00000000 in ?? () (gdb) And for Trapped in Convert: Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 4576.0x1c18] 0x6c51ff83 in spoutsf (csound=0x52d07c0) at c:/Users/mkg.sorabji/csound-csound6-git/InOut/libsnd.c:106 106 } (gdb) bt #0 0x6c51ff83 in spoutsf (csound=0x52d07c0) at c:/Users/mkg.sorabji/csound-csound6-git/InOut/libsnd.c:106
#1 0x6c5d3927 in kperf (csound=csound@entry=0x52d07c0) at c:/Users/mkg.sorabji/csound-csound6-git/Top/csound.c:1554 #2 0x6c5d3e68 in csoundPerformKsmps (csound=0x52d07c0) at c:/Users/mkg.sorabji/csound-csound6-git/Top/csound.c:1612
#3 0x6b8895ee in CsoundPerformanceThread::Perform (this=0x42d5198) at c:/Users/mkg.sorabji/csound-csound6-git/interfaces/csPerfThread.cpp:300 #4 0x649458cb in pthread_create_wrapper () from c:\Users\mkg.sorabji\csound-csound6-git\libwinpthread-1.dll
#5 0x765e906a in ui64tow () from C:\Windows\SysWOW64\msvcrt.dll #6 0x765e9147 in msvcrt!iswalnum () from C:\Windows\SysWOW64\msvcrt.dll #7 0x771b850d in KERNEL32!BaseThreadInitThunk () from C:\Windows\SysWOW64\kernel32.dll
#8 0x7773bf39 in ?? () #9 0x7773bf0c in ?? () #10 0x00000000 in ?? () (gdb) The Boulanger piece renders fine in external Csound, so I'm guessing this is a CsoundQt bug.
Regards, Mike =========================== Michael GoginsIrreducible Productions http://michaelgogins.tumblr.com Michael dot Gogins at gmail dot com |
Date | 2013-08-14 22:55 |
From | Andres Cabrera |
Subject | Re: [Cs-dev] Serious bug (I think in CsoundQt) on WIndows |
Attachments | None None |
Hi, I'm not sure what could be causing this. Do you have any information on other threads running at the time?Cheers, Andrés On Wed, Aug 14, 2013 at 4:06 AM, Michael Gogins <michael.gogins@gmail.com> wrote:
|
Date | 2013-08-15 02:47 |
From | jpff@cs.bath.ac.uk |
Subject | Re: [Cs-dev] Serious bug (I think in CsoundQt) on WIndows |
Sounds very like a generic bug we had in string assignment, but that is fixed > Hi, > > I'm not sure what could be causing this. Do you have any information on > other threads running at the time? > > Can you disable realtime recording, in and out channels and console > messages in the preferences? It's the only three things that could > interfere with Csound. > > Cheers, > Andrés > > > On Wed, Aug 14, 2013 at 4:06 AM, Michael Gogins > |
Date | 2013-08-15 03:00 |
From | Michael Gogins |
Subject | Re: [Cs-dev] Serious bug (I think in CsoundQt) on WIndows |
Attachments | None None |
I'm seeing write out of bounds errors in msvcrt realloc called in compiling the orc and I think specifically in str_cat. Only noticed in python and csound qt, not in the Csound command. On Aug 14, 2013 9:48 PM, <jpff@cs.bath.ac.uk> wrote:
Sounds very like a generic bug we had in string assignment, but that is fixed |
Date | 2013-08-15 08:47 |
From | Victor Lazzarini |
Subject | Re: [Cs-dev] Serious bug (I think in CsoundQt) on WIndows |
valgrind tells us there is still something not quite right there in strcpy and strcat. It does not explain the problems with trapped. I saw some errors in valgrind with ascii graphs, maybe you could add a -d and see if it makes any difference for trapped. On 15 Aug 2013, at 03:00, Michael Gogins wrote: > I'm seeing write out of bounds errors in msvcrt realloc called in compiling the orc and I think specifically in str_cat. Only noticed in python and csound qt, not in the Csound command. > > On Aug 14, 2013 9:48 PM, |
Date | 2013-08-15 09:27 |
From | Victor Lazzarini |
Subject | Re: [Cs-dev] Serious bug (I think in CsoundQt) on WIndows |
OK, strcat is fixed in GIT and valgrind reports all's OK with it. Victor On 15 Aug 2013, at 08:47, Victor Lazzarini wrote: > valgrind tells us there is still something not quite right there in strcpy and strcat. It does not explain the problems with trapped. I saw some errors in valgrind with > ascii graphs, maybe you could add a -d and see if it makes any difference for trapped. > On 15 Aug 2013, at 03:00, Michael Gogins wrote: > >> I'm seeing write out of bounds errors in msvcrt realloc called in compiling the orc and I think specifically in str_cat. Only noticed in python and csound qt, not in the Csound command. >> >> On Aug 14, 2013 9:48 PM, |