Hello
I am trying to write a trap handler to couth the SIGILL and SIGSEVG signals; so if the user program didn't "mapp" the adr on the vmebus the program will print a message and continue instead of being stop and exit with a core dump.
This is what my program does :
(a) the "mmap" is done and success , but the physical adr is wrong (e.g. no card on this adr) - error=SIGILL
(a.1) on "write" to be card I get the msg from my trap handler as many times as I try to write
(a.2) on "read" - I get my trap handler msg in an endless loop and the program MUST be kill. I know the trap does not return to the program normal flow (e.g. 1 command after the read) since I have there a printout that it is never reach.
(b) if the "mmap" is NOT done at all - error=SIGSEV then I get the endless loop on read or write
Looking on the vmebus with a data analyzer I can see the bus is being reach every fix period - look like same body is trying to access the memory.
Any idea/help will be great
Thanks in advance
Galia Gitliz
Weizmann Institute of Science
Physics Department, E&DAQ Group
Tel 972-8-9342562
Fax 972-8-9344166
Email : galia.gitliz@weizmann.ac.il
|
Hello
I am trying to write a trap handler to couth the SIGILL and
SIGSEVG signals; so if the user program didn't "mapp" the adr on the
vmebus the program will print a message and continue instead of being stop and
exit with a core dump.
This is what my program does :
(a) the "mmap" is done and success , but the physical
adr is wrong (e.g. no card on this adr) - error=SIGILL
(a.1) on "write" to be card I get the msg from my trap handler as many times as
I try to write
(a.2) on "read" - I get my trap handler msg in an endless loop and the program
MUST be kill. I know the trap does not return to the program normal
flow (e.g. 1 command after the read) since I have there a printout that it is
never reach.
(b) if the "mmap" is NOT done at all - error=SIGSEV then
I get the endless loop on read or write
Looking on the vmebus with a data analyzer I can see the bus is being
reach every fix period - look like same body is trying to access the
memory.
Any idea/help will be great
Thanks in advance
Galia Gitliz
Weizmann Institute of Science Physics Department, E&DAQ Group Tel 972-8-9342562 Fax 972-8-9344166 Email : galia.gitliz@weizmann.ac.il |