QNX 4.25 dump file help

QNX Newsgroups (Archives) comp.os.qnx
1

We had a process crash and inside the dump file is assembly code that I
can’t really figure out. Assembly isn’t an area I have a lot of expertise in
so I thought I’d see if anyone else could give me some hints as to what went
wrong. The version of QNX is 4.25 and it has the latest patch installed (G I
believe). Below is a transcription of the Watcom debugger information I’m
given from the dump file… Watcom shows the line that caused the SIGSEGV as
“E4E9 ret”. Any ideas would be appreciated.

Assembly: kernel
_receive
E4B6 push ecx
E4B7 sub esp, 0000000C
E4BA mov ecx,eax
E4BC mov eax,edx
E4BE mov [esp], edx SS:00009D8C=0400C1C2
E4C1 xor edx,edx
E4C3 mov +6[esp],dx SS:00009D92=0000
E4C8 mov dx,ds
E4CA mov ah,01
E4CC mov +4[esp],dx SS:00009D90=0000
E4D1 mov +8[esp],ebx SS:00009D94=0000A573
E4D5 mov ebx,esp
E4D7 mov edx,ecx
E4D9 mov al,01
E4DB clc
E4DC int 0F2
E4DE jae 0000E4E5
E4ED call _kererr
E4E5 add esp,0000000C
E4E8 pop ecx
E4E9 ret <---------- hilighted line in watcom debugger


Registers
EAX: 000002A2 EBX: 00009D7C
ECX: 000002A2 EDX: 00000000
ESI: 00EFEA18 EDI: 00014C20
EBP: 00000000 ESP: 00009D8C
EIP: 0000E4E9 EFL: 00012216
C: 0 P:1
A:1 Z:0
S:0 I:1
D:0 O:0
DS:000F ES:000F
FS:0000 GS:0000
SS:000F CS:0007
IOP:1 NT:0
RF:1 VM:0


\

C. Scott

2

“C. Scott” <cscott@nospam.com> wrote in message
news:bu918q$952$1@nntp.qnx.com…

We had a process crash and inside the dump file is assembly code that I
can’t really figure out. Assembly isn’t an area I have a lot of expertise
in
so I thought I’d see if anyone else could give me some hints as to what
went
wrong. The version of QNX is 4.25 and it has the latest patch installed (G
I
believe). Below is a transcription of the Watcom debugger information I’m
given from the dump file… Watcom shows the line that caused the SIGSEGV
as
“E4E9 ret”. Any ideas would be appreciated.

al set to 1 indicate the kernel call by int f2 is a receivemx (Receive ends

up being a receivemx).
Since the ret seems to cause grief I think the problem is a corrupted
stack. My guess is the receive
area passed to receivemx is on the stack and is somehow corrupting it’s
content. The ret is probably
trying to return to a invalide address

Assembly: kernel
_receive
E4B6 push ecx
E4B7 sub esp, 0000000C
E4BA mov ecx,eax
E4BC mov eax,edx
E4BE mov [esp], edx SS:00009D8C=0400C1C2
E4C1 xor edx,edx
E4C3 mov +6[esp],dx SS:00009D92=0000
E4C8 mov dx,ds
E4CA mov ah,01
E4CC mov +4[esp],dx SS:00009D90=0000
E4D1 mov +8[esp],ebx SS:00009D94=0000A573
E4D5 mov ebx,esp
E4D7 mov edx,ecx
E4D9 mov al,01
E4DB clc
E4DC int 0F2
E4DE jae 0000E4E5
E4ED call _kererr
E4E5 add esp,0000000C
E4E8 pop ecx
E4E9 ret <---------- hilighted line in watcom debugger


Registers
EAX: 000002A2 EBX: 00009D7C
ECX: 000002A2 EDX: 00000000
ESI: 00EFEA18 EDI: 00014C20
EBP: 00000000 ESP: 00009D8C
EIP: 0000E4E9 EFL: 00012216
C: 0 P:1
A:1 Z:0
S:0 I:1
D:0 O:0
DS:000F ES:000F
FS:0000 GS:0000
SS:000F CS:0007
IOP:1 NT:0
RF:1 VM:0


\

C. Scott

3

Thanks for the info! I’m not sure how to chase down the root of the problem
but at least I know which module to look in.

C. Scott




“Mario Charest” postmaster@127.0.0.1 wrote in message
news:bu94cg$5oh$1@inn.qnx.com…

“C. Scott” <> cscott@nospam.com> > wrote in message
news:bu918q$952$> 1@nntp.qnx.com> …
We had a process crash and inside the dump file is assembly code that I
can’t really figure out. Assembly isn’t an area I have a lot of
expertise
in
so I thought I’d see if anyone else could give me some hints as to what
went
wrong. The version of QNX is 4.25 and it has the latest patch installed
(G
I
believe). Below is a transcription of the Watcom debugger information
I’m
given from the dump file… Watcom shows the line that caused the
SIGSEGV
as
“E4E9 ret”. Any ideas would be appreciated.

al set to 1 indicate the kernel call by int f2 is a receivemx (Receive
ends
up being a receivemx).
Since the ret seems to cause grief I think the problem is a corrupted
stack. My guess is the receive
area passed to receivemx is on the stack and is somehow corrupting it’s
content. The ret is probably
trying to return to a invalide address

Assembly: kernel
_receive
E4B6 push ecx
E4B7 sub esp, 0000000C
E4BA mov ecx,eax
E4BC mov eax,edx
E4BE mov [esp], edx SS:00009D8C=0400C1C2
E4C1 xor edx,edx
E4C3 mov +6[esp],dx SS:00009D92=0000
E4C8 mov dx,ds
E4CA mov ah,01
E4CC mov +4[esp],dx SS:00009D90=0000
E4D1 mov +8[esp],ebx SS:00009D94=0000A573
E4D5 mov ebx,esp
E4D7 mov edx,ecx
E4D9 mov al,01
E4DB clc
E4DC int 0F2
E4DE jae 0000E4E5
E4ED call _kererr
E4E5 add esp,0000000C
E4E8 pop ecx
E4E9 ret <---------- hilighted line in watcom debugger


Registers
EAX: 000002A2 EBX: 00009D7C
ECX: 000002A2 EDX: 00000000
ESI: 00EFEA18 EDI: 00014C20
EBP: 00000000 ESP: 00009D8C
EIP: 0000E4E9 EFL: 00012216
C: 0 P:1
A:1 Z:0
S:0 I:1
D:0 O:0
DS:000F ES:000F
FS:0000 GS:0000
SS:000F CS:0007
IOP:1 NT:0
RF:1 VM:0


\

C. Scott

\