You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@harmony.apache.org by Evgueni Brevnov <ev...@gmail.com> on 2007/08/24 13:30:33 UTC
[drlvm][gc_gen] Does anybody else observe failures in gc_gen?
HI all,
Today, I had several crashes in gc_gen module on my P4 HT Win2003. It
happened once while running kernel test. It almost always fails on
specjbb2005 at 8 warehouses. No special options were provided. Gc_gen
fails on assertion with the following stack trace. Could it be a
regression?
> msvcr71d.dll!_assert(const char * expr=0x0141a0f8, const char * filename=0x0141a110, unsigned int lineno=178) Line 295 C
> gc_gen.dll!blocked_space_extend(Blocked_Space * space=0x0155a728, unsigned int changed_size=327483392) Line 178 + 0x1c C++
> gc_gen.dll!gc_gen_adjust_heap_size(GC_Gen * gc=0x014348f8, __int64 pause_time=6594046) Line 428 + 0xd C++
> gc_gen.dll!gc_adjust_heap_size(GC * gc=0x014348f8, __int64 pause_time=6594046) Line 279 + 0x11 C++
> gc_gen.dll!gc_reclaim_heap(GC * gc=0x014348f8, unsigned int gc_cause=1) Line 344 + 0x11 C++
> gc_gen.dll!fspace_alloc(unsigned int size=2412, Allocator * allocator=0x01644970) Line 75 + 0xe C++
> gc_gen.dll!nos_alloc(unsigned int size=2412, Allocator * allocator=0x01644970) Line 281 + 0x10 C++
> gc_gen.dll!gc_alloc(unsigned int size=2412, unsigned int ah=101208304, void * unused_gc_tls=0x05bea3dc) Line 91 + 0xd C++
> harmonyvm.dll!vm_new_vector_using_vtable_and_thread_pointer(int length=600, unsigned int vector_handle=101208304, void * tp=0x05bea3dc) Line 219 + 0x14 C++
> harmonyvm.dll!vm_rt_new_vector_using_vtable_and_thread_pointer(int length=600, unsigned int vector_handle=101208304, void * tp=0x05bea3dc) Line 113 + 0x11 C++
Thanks
Evgueni
Re: [drlvm][gc_gen] Does anybody else observe failures in gc_gen?
Posted by Mikhail Fursov <mi...@gmail.com>.
My Win32 and Win64 copies run without crashes. Tested both debug and release
versions several times.
On 8/24/07, Alexey Varlamov <al...@gmail.com> wrote:
>
> During last 3 days I've seen 4-5 intermittent failures, all different,
> on different workloads and platforms, below some I could dig out
> quickly:
>
> Linux/IPF:
> SIGABRT in VM code.
> java:
> /export/users/cc/build/checkouts/drlvm/vm/gc_gen/src/common/gc_for_class.h:208:
> GC_VTable_Info* vtable_get_gcvt_raw(Partial_Reveal_VTable*): Assertion
> `vt && vt->gcvt' failed.
>
>
> Linux/x86:
> SIGSEGV in VM code.
> Stack trace:
> 0: oh_enumerate_handles(ObjectHandles*) (??:-1)
> 1: vm_enumerate_root_set_single_thread_on_stack(StackIterator*) (??:-1)
> 2: vm_enumerate_thread(VM_thread*) (??:-1)
> 3: vm_enumerate_root_set_all_threads (??:-1)
> 4: gc_reclaim_heap(GC*, unsigned int) (??:-1)
> 5: fspace_alloc(unsigned int, Allocator*) (??:-1)
> 6: nos_alloc(unsigned int, Allocator*) (??:-1)
> 7: gc_alloc (??:-1)
> 8: vm_new_vector_using_vtable_and_thread_pointer(int, unsigned int,
> void*) (??:-1)
> 9: vm_rt_new_vector_using_vtable_and_thread_pointer(int, unsigned int,
> void*) (??:-1)
> 10: 0x415B5A35 <Generated stub>
> 11: ?? (??:-1)
> 12: java/lang/String.<init>([CII)V (String.java:452)
> 13: java/lang/String.<init>([C)V (String.java:-1)
> 14: java/lang/String.<init>([C)V (String.java:426)
>
>
> 2007/8/24, Evgueni Brevnov <ev...@gmail.com>:
> > HI all,
> >
> > Today, I had several crashes in gc_gen module on my P4 HT Win2003. It
> > happened once while running kernel test. It almost always fails on
> > specjbb2005 at 8 warehouses. No special options were provided. Gc_gen
> > fails on assertion with the following stack trace. Could it be a
> > regression?
> >
> > > msvcr71d.dll!_assert(const char * expr=0x0141a0f8, const char *
> filename=0x0141a110, unsigned int lineno=178) Line 295 C
> > > gc_gen.dll!blocked_space_extend(Blocked_Space * space=0x0155a728,
> unsigned int changed_size=327483392) Line 178 + 0x1c C++
> > > gc_gen.dll!gc_gen_adjust_heap_size(GC_Gen * gc=0x014348f8, __int64
> pause_time=6594046) Line 428 + 0xd C++
> > > gc_gen.dll!gc_adjust_heap_size(GC * gc=0x014348f8, __int64
> pause_time=6594046) Line 279 + 0x11 C++
> > > gc_gen.dll!gc_reclaim_heap(GC * gc=0x014348f8, unsigned int
> gc_cause=1) Line 344 + 0x11 C++
> > > gc_gen.dll!fspace_alloc(unsigned int size=2412, Allocator *
> allocator=0x01644970) Line 75 + 0xe C++
> > > gc_gen.dll!nos_alloc(unsigned int size=2412, Allocator *
> allocator=0x01644970) Line 281 + 0x10 C++
> > > gc_gen.dll!gc_alloc(unsigned int size=2412, unsigned int ah=101208304,
> void * unused_gc_tls=0x05bea3dc) Line 91 + 0xd C++
> > > harmonyvm.dll!vm_new_vector_using_vtable_and_thread_pointer(int
> length=600, unsigned int vector_handle=101208304, void *
> tp=0x05bea3dc) Line 219 + 0x14 C++
> > > harmonyvm.dll!vm_rt_new_vector_using_vtable_and_thread_pointer(int
> length=600, unsigned int vector_handle=101208304, void *
> tp=0x05bea3dc) Line 113 + 0x11 C++
> >
> >
> > Thanks
> > Evgueni
> >
>
--
Mikhail Fursov
Re: [drlvm][gc_gen] Does anybody else observe failures in gc_gen?
Posted by Alexey Varlamov <al...@gmail.com>.
During last 3 days I've seen 4-5 intermittent failures, all different,
on different workloads and platforms, below some I could dig out
quickly:
Linux/IPF:
SIGABRT in VM code.
java: /export/users/cc/build/checkouts/drlvm/vm/gc_gen/src/common/gc_for_class.h:208:
GC_VTable_Info* vtable_get_gcvt_raw(Partial_Reveal_VTable*): Assertion
`vt && vt->gcvt' failed.
Linux/x86:
SIGSEGV in VM code.
Stack trace:
0: oh_enumerate_handles(ObjectHandles*) (??:-1)
1: vm_enumerate_root_set_single_thread_on_stack(StackIterator*) (??:-1)
2: vm_enumerate_thread(VM_thread*) (??:-1)
3: vm_enumerate_root_set_all_threads (??:-1)
4: gc_reclaim_heap(GC*, unsigned int) (??:-1)
5: fspace_alloc(unsigned int, Allocator*) (??:-1)
6: nos_alloc(unsigned int, Allocator*) (??:-1)
7: gc_alloc (??:-1)
8: vm_new_vector_using_vtable_and_thread_pointer(int, unsigned int,
void*) (??:-1)
9: vm_rt_new_vector_using_vtable_and_thread_pointer(int, unsigned int,
void*) (??:-1)
10: 0x415B5A35 <Generated stub>
11: ?? (??:-1)
12: java/lang/String.<init>([CII)V (String.java:452)
13: java/lang/String.<init>([C)V (String.java:-1)
14: java/lang/String.<init>([C)V (String.java:426)
2007/8/24, Evgueni Brevnov <ev...@gmail.com>:
> HI all,
>
> Today, I had several crashes in gc_gen module on my P4 HT Win2003. It
> happened once while running kernel test. It almost always fails on
> specjbb2005 at 8 warehouses. No special options were provided. Gc_gen
> fails on assertion with the following stack trace. Could it be a
> regression?
>
> > msvcr71d.dll!_assert(const char * expr=0x0141a0f8, const char * filename=0x0141a110, unsigned int lineno=178) Line 295 C
> > gc_gen.dll!blocked_space_extend(Blocked_Space * space=0x0155a728, unsigned int changed_size=327483392) Line 178 + 0x1c C++
> > gc_gen.dll!gc_gen_adjust_heap_size(GC_Gen * gc=0x014348f8, __int64 pause_time=6594046) Line 428 + 0xd C++
> > gc_gen.dll!gc_adjust_heap_size(GC * gc=0x014348f8, __int64 pause_time=6594046) Line 279 + 0x11 C++
> > gc_gen.dll!gc_reclaim_heap(GC * gc=0x014348f8, unsigned int gc_cause=1) Line 344 + 0x11 C++
> > gc_gen.dll!fspace_alloc(unsigned int size=2412, Allocator * allocator=0x01644970) Line 75 + 0xe C++
> > gc_gen.dll!nos_alloc(unsigned int size=2412, Allocator * allocator=0x01644970) Line 281 + 0x10 C++
> > gc_gen.dll!gc_alloc(unsigned int size=2412, unsigned int ah=101208304, void * unused_gc_tls=0x05bea3dc) Line 91 + 0xd C++
> > harmonyvm.dll!vm_new_vector_using_vtable_and_thread_pointer(int length=600, unsigned int vector_handle=101208304, void * tp=0x05bea3dc) Line 219 + 0x14 C++
> > harmonyvm.dll!vm_rt_new_vector_using_vtable_and_thread_pointer(int length=600, unsigned int vector_handle=101208304, void * tp=0x05bea3dc) Line 113 + 0x11 C++
>
>
> Thanks
> Evgueni
>
Re: [drlvm][gc_gen] Does anybody else observe failures in gc_gen?
Posted by Mikhail Fursov <mi...@gmail.com>.
If you can reproduce the same error in -Xem:jet mode the JIT team can relax
- 99% this is not a JIT issue.
On 8/25/07, Evgueni Brevnov <ev...@gmail.com> wrote:
>
> Oh, I fogot to mention that. I run with -Xem:opt.
>
> Evgueni
>
> On 8/25/07, Mikhail Fursov <mi...@gmail.com> wrote:
> > Evgueni,
> > Do you can reproduce the problem in both -Xem:opt and -Xem:jet modes?
> >
> >
> > On 8/24/07, Evgueni Brevnov <ev...@gmail.com> wrote:
> > >
> > > I can easily reproduce that on my local PC.
> > >
> > > Evgueni
> > >
> > > On 8/24/07, Xiao-Feng Li <xi...@gmail.com> wrote:
> > > > I didn't have the same failure locally. Probably it's another
> Win2003
> > > > related problem. Will check on Win2003.
> > > >
> > > > Thanks,
> > > > xiaofeng
> > > >
> > > > On 8/24/07, Evgueni Brevnov <ev...@gmail.com> wrote:
> > > > > HI all,
> > > > >
> > > > > Today, I had several crashes in gc_gen module on my P4 HT Win2003.
> It
> > > > > happened once while running kernel test. It almost always fails on
> > > > > specjbb2005 at 8 warehouses. No special options were provided.
> Gc_gen
> > > > > fails on assertion with the following stack trace. Could it be a
> > > > > regression?
> > > > >
> > > > > > msvcr71d.dll!_assert(const char * expr=0x0141a0f8, const char *
> > > filename=0x0141a110, unsigned int lineno=178) Line 295 C
> > > > > > gc_gen.dll!blocked_space_extend(Blocked_Space *
> space=0x0155a728,
> > > unsigned int changed_size=327483392) Line 178 + 0x1c C++
> > > > > > gc_gen.dll!gc_gen_adjust_heap_size(GC_Gen * gc=0x014348f8,
> __int64
> > > pause_time=6594046) Line 428 + 0xd C++
> > > > > > gc_gen.dll!gc_adjust_heap_size(GC * gc=0x014348f8, __int64
> > > pause_time=6594046) Line 279 + 0x11 C++
> > > > > > gc_gen.dll!gc_reclaim_heap(GC * gc=0x014348f8, unsigned int
> > > gc_cause=1) Line 344 + 0x11 C++
> > > > > > gc_gen.dll!fspace_alloc(unsigned int size=2412, Allocator *
> > > allocator=0x01644970) Line 75 + 0xe C++
> > > > > > gc_gen.dll!nos_alloc(unsigned int size=2412, Allocator *
> > > allocator=0x01644970) Line 281 + 0x10 C++
> > > > > > gc_gen.dll!gc_alloc(unsigned int size=2412, unsigned int
> > > ah=101208304, void * unused_gc_tls=0x05bea3dc) Line 91 + 0xd C++
> > > > > > harmonyvm.dll!vm_new_vector_using_vtable_and_thread_pointer(int
> > > length=600, unsigned int vector_handle=101208304, void *
> > > tp=0x05bea3dc) Line 219 + 0x14 C++
> > > > > > harmonyvm.dll!vm_rt_new_vector_using_vtable_and_thread_pointer
> (int
> > > length=600, unsigned int vector_handle=101208304, void *
> > > tp=0x05bea3dc) Line 113 + 0x11 C++
> > > > >
> > > > >
> > > > > Thanks
> > > > > Evgueni
> > > > >
> > > >
> > > >
> > > > --
> > > > http://xiao-feng.blogspot.com
> > > >
> > >
> >
> >
> >
> > --
> > Mikhail Fursov
> >
>
--
Mikhail Fursov
Re: [drlvm][gc_gen] Does anybody else observe failures in gc_gen?
Posted by Evgueni Brevnov <ev...@gmail.com>.
Oh, I fogot to mention that. I run with -Xem:opt.
Evgueni
On 8/25/07, Mikhail Fursov <mi...@gmail.com> wrote:
> Evgueni,
> Do you can reproduce the problem in both -Xem:opt and -Xem:jet modes?
>
>
> On 8/24/07, Evgueni Brevnov <ev...@gmail.com> wrote:
> >
> > I can easily reproduce that on my local PC.
> >
> > Evgueni
> >
> > On 8/24/07, Xiao-Feng Li <xi...@gmail.com> wrote:
> > > I didn't have the same failure locally. Probably it's another Win2003
> > > related problem. Will check on Win2003.
> > >
> > > Thanks,
> > > xiaofeng
> > >
> > > On 8/24/07, Evgueni Brevnov <ev...@gmail.com> wrote:
> > > > HI all,
> > > >
> > > > Today, I had several crashes in gc_gen module on my P4 HT Win2003. It
> > > > happened once while running kernel test. It almost always fails on
> > > > specjbb2005 at 8 warehouses. No special options were provided. Gc_gen
> > > > fails on assertion with the following stack trace. Could it be a
> > > > regression?
> > > >
> > > > > msvcr71d.dll!_assert(const char * expr=0x0141a0f8, const char *
> > filename=0x0141a110, unsigned int lineno=178) Line 295 C
> > > > > gc_gen.dll!blocked_space_extend(Blocked_Space * space=0x0155a728,
> > unsigned int changed_size=327483392) Line 178 + 0x1c C++
> > > > > gc_gen.dll!gc_gen_adjust_heap_size(GC_Gen * gc=0x014348f8, __int64
> > pause_time=6594046) Line 428 + 0xd C++
> > > > > gc_gen.dll!gc_adjust_heap_size(GC * gc=0x014348f8, __int64
> > pause_time=6594046) Line 279 + 0x11 C++
> > > > > gc_gen.dll!gc_reclaim_heap(GC * gc=0x014348f8, unsigned int
> > gc_cause=1) Line 344 + 0x11 C++
> > > > > gc_gen.dll!fspace_alloc(unsigned int size=2412, Allocator *
> > allocator=0x01644970) Line 75 + 0xe C++
> > > > > gc_gen.dll!nos_alloc(unsigned int size=2412, Allocator *
> > allocator=0x01644970) Line 281 + 0x10 C++
> > > > > gc_gen.dll!gc_alloc(unsigned int size=2412, unsigned int
> > ah=101208304, void * unused_gc_tls=0x05bea3dc) Line 91 + 0xd C++
> > > > > harmonyvm.dll!vm_new_vector_using_vtable_and_thread_pointer(int
> > length=600, unsigned int vector_handle=101208304, void *
> > tp=0x05bea3dc) Line 219 + 0x14 C++
> > > > > harmonyvm.dll!vm_rt_new_vector_using_vtable_and_thread_pointer(int
> > length=600, unsigned int vector_handle=101208304, void *
> > tp=0x05bea3dc) Line 113 + 0x11 C++
> > > >
> > > >
> > > > Thanks
> > > > Evgueni
> > > >
> > >
> > >
> > > --
> > > http://xiao-feng.blogspot.com
> > >
> >
>
>
>
> --
> Mikhail Fursov
>
Re: [drlvm][gc_gen] Does anybody else observe failures in gc_gen?
Posted by Mikhail Fursov <mi...@gmail.com>.
Evgueni,
Do you can reproduce the problem in both -Xem:opt and -Xem:jet modes?
On 8/24/07, Evgueni Brevnov <ev...@gmail.com> wrote:
>
> I can easily reproduce that on my local PC.
>
> Evgueni
>
> On 8/24/07, Xiao-Feng Li <xi...@gmail.com> wrote:
> > I didn't have the same failure locally. Probably it's another Win2003
> > related problem. Will check on Win2003.
> >
> > Thanks,
> > xiaofeng
> >
> > On 8/24/07, Evgueni Brevnov <ev...@gmail.com> wrote:
> > > HI all,
> > >
> > > Today, I had several crashes in gc_gen module on my P4 HT Win2003. It
> > > happened once while running kernel test. It almost always fails on
> > > specjbb2005 at 8 warehouses. No special options were provided. Gc_gen
> > > fails on assertion with the following stack trace. Could it be a
> > > regression?
> > >
> > > > msvcr71d.dll!_assert(const char * expr=0x0141a0f8, const char *
> filename=0x0141a110, unsigned int lineno=178) Line 295 C
> > > > gc_gen.dll!blocked_space_extend(Blocked_Space * space=0x0155a728,
> unsigned int changed_size=327483392) Line 178 + 0x1c C++
> > > > gc_gen.dll!gc_gen_adjust_heap_size(GC_Gen * gc=0x014348f8, __int64
> pause_time=6594046) Line 428 + 0xd C++
> > > > gc_gen.dll!gc_adjust_heap_size(GC * gc=0x014348f8, __int64
> pause_time=6594046) Line 279 + 0x11 C++
> > > > gc_gen.dll!gc_reclaim_heap(GC * gc=0x014348f8, unsigned int
> gc_cause=1) Line 344 + 0x11 C++
> > > > gc_gen.dll!fspace_alloc(unsigned int size=2412, Allocator *
> allocator=0x01644970) Line 75 + 0xe C++
> > > > gc_gen.dll!nos_alloc(unsigned int size=2412, Allocator *
> allocator=0x01644970) Line 281 + 0x10 C++
> > > > gc_gen.dll!gc_alloc(unsigned int size=2412, unsigned int
> ah=101208304, void * unused_gc_tls=0x05bea3dc) Line 91 + 0xd C++
> > > > harmonyvm.dll!vm_new_vector_using_vtable_and_thread_pointer(int
> length=600, unsigned int vector_handle=101208304, void *
> tp=0x05bea3dc) Line 219 + 0x14 C++
> > > > harmonyvm.dll!vm_rt_new_vector_using_vtable_and_thread_pointer(int
> length=600, unsigned int vector_handle=101208304, void *
> tp=0x05bea3dc) Line 113 + 0x11 C++
> > >
> > >
> > > Thanks
> > > Evgueni
> > >
> >
> >
> > --
> > http://xiao-feng.blogspot.com
> >
>
--
Mikhail Fursov
Re: [drlvm][gc_gen] Does anybody else observe failures in gc_gen?
Posted by Evgueni Brevnov <ev...@gmail.com>.
I can easily reproduce that on my local PC.
Evgueni
On 8/24/07, Xiao-Feng Li <xi...@gmail.com> wrote:
> I didn't have the same failure locally. Probably it's another Win2003
> related problem. Will check on Win2003.
>
> Thanks,
> xiaofeng
>
> On 8/24/07, Evgueni Brevnov <ev...@gmail.com> wrote:
> > HI all,
> >
> > Today, I had several crashes in gc_gen module on my P4 HT Win2003. It
> > happened once while running kernel test. It almost always fails on
> > specjbb2005 at 8 warehouses. No special options were provided. Gc_gen
> > fails on assertion with the following stack trace. Could it be a
> > regression?
> >
> > > msvcr71d.dll!_assert(const char * expr=0x0141a0f8, const char * filename=0x0141a110, unsigned int lineno=178) Line 295 C
> > > gc_gen.dll!blocked_space_extend(Blocked_Space * space=0x0155a728, unsigned int changed_size=327483392) Line 178 + 0x1c C++
> > > gc_gen.dll!gc_gen_adjust_heap_size(GC_Gen * gc=0x014348f8, __int64 pause_time=6594046) Line 428 + 0xd C++
> > > gc_gen.dll!gc_adjust_heap_size(GC * gc=0x014348f8, __int64 pause_time=6594046) Line 279 + 0x11 C++
> > > gc_gen.dll!gc_reclaim_heap(GC * gc=0x014348f8, unsigned int gc_cause=1) Line 344 + 0x11 C++
> > > gc_gen.dll!fspace_alloc(unsigned int size=2412, Allocator * allocator=0x01644970) Line 75 + 0xe C++
> > > gc_gen.dll!nos_alloc(unsigned int size=2412, Allocator * allocator=0x01644970) Line 281 + 0x10 C++
> > > gc_gen.dll!gc_alloc(unsigned int size=2412, unsigned int ah=101208304, void * unused_gc_tls=0x05bea3dc) Line 91 + 0xd C++
> > > harmonyvm.dll!vm_new_vector_using_vtable_and_thread_pointer(int length=600, unsigned int vector_handle=101208304, void * tp=0x05bea3dc) Line 219 + 0x14 C++
> > > harmonyvm.dll!vm_rt_new_vector_using_vtable_and_thread_pointer(int length=600, unsigned int vector_handle=101208304, void * tp=0x05bea3dc) Line 113 + 0x11 C++
> >
> >
> > Thanks
> > Evgueni
> >
>
>
> --
> http://xiao-feng.blogspot.com
>
Re: [drlvm][gc_gen] Does anybody else observe failures in gc_gen?
Posted by Xiao-Feng Li <xi...@gmail.com>.
I didn't have the same failure locally. Probably it's another Win2003
related problem. Will check on Win2003.
Thanks,
xiaofeng
On 8/24/07, Evgueni Brevnov <ev...@gmail.com> wrote:
> HI all,
>
> Today, I had several crashes in gc_gen module on my P4 HT Win2003. It
> happened once while running kernel test. It almost always fails on
> specjbb2005 at 8 warehouses. No special options were provided. Gc_gen
> fails on assertion with the following stack trace. Could it be a
> regression?
>
> > msvcr71d.dll!_assert(const char * expr=0x0141a0f8, const char * filename=0x0141a110, unsigned int lineno=178) Line 295 C
> > gc_gen.dll!blocked_space_extend(Blocked_Space * space=0x0155a728, unsigned int changed_size=327483392) Line 178 + 0x1c C++
> > gc_gen.dll!gc_gen_adjust_heap_size(GC_Gen * gc=0x014348f8, __int64 pause_time=6594046) Line 428 + 0xd C++
> > gc_gen.dll!gc_adjust_heap_size(GC * gc=0x014348f8, __int64 pause_time=6594046) Line 279 + 0x11 C++
> > gc_gen.dll!gc_reclaim_heap(GC * gc=0x014348f8, unsigned int gc_cause=1) Line 344 + 0x11 C++
> > gc_gen.dll!fspace_alloc(unsigned int size=2412, Allocator * allocator=0x01644970) Line 75 + 0xe C++
> > gc_gen.dll!nos_alloc(unsigned int size=2412, Allocator * allocator=0x01644970) Line 281 + 0x10 C++
> > gc_gen.dll!gc_alloc(unsigned int size=2412, unsigned int ah=101208304, void * unused_gc_tls=0x05bea3dc) Line 91 + 0xd C++
> > harmonyvm.dll!vm_new_vector_using_vtable_and_thread_pointer(int length=600, unsigned int vector_handle=101208304, void * tp=0x05bea3dc) Line 219 + 0x14 C++
> > harmonyvm.dll!vm_rt_new_vector_using_vtable_and_thread_pointer(int length=600, unsigned int vector_handle=101208304, void * tp=0x05bea3dc) Line 113 + 0x11 C++
>
>
> Thanks
> Evgueni
>
--
http://xiao-feng.blogspot.com