From: "Rich" 

This is a multi-part message in MIME format.

------=_NextPart_000_034E_01C5C014.C8AF0780
Content-Type: text/plain;
        charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

   Almost.  You hit the nail on the head about there sometimes being no =
visible effect if blocked in a system call.  There is an effect though.  =
The process is killed and will terminate when it is unblocked.  Just =
because an API can be blocking doesn't mean it has this behavior.  Most =
blocking calls are interruptable.

Rich

  "Paul Ranson"  wrote in message =
news:4333e479$1{at}w3.nls.net...
  Things may have changed but I'm pretty sure that 'kill -9' can have no =

  effect in some conditions, if the process is blocked in some system =
calls=20
  for example. IOW the signal will be handled when the call returns, but =
the=20
  call never returns and isn't interruptible.

  If you get an NT process you cannot terminate try attaching to it with =
a=20
  debugger, that usually makes it see sense.

  Paul

  "Mike '/m'"  wrote in message=20
  news:l1b6j19bo6ummoit70h9snroig3mmjq5ui{at}4ax.com...
  > Should be:
  >
  > In the *nix's I have used, there is the regular kill (kill -15) that
  > nicely  asks the app to shutdown.  Then for those obstinate apps, =
there
  > is the kill option (kill -9) that pretty much just terminates the
  > process, releases memory, and brushes the crumbs off the table.
  > Obviously, you do not use kill -9 on a whim, but it is nice to know =
it
  > is there when needed.
  >
  > /m=20


------=_NextPart_000_034E_01C5C014.C8AF0780
Content-Type: text/html;
        charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable








  
Almost.  You hit the =
nail on the=20
head about there sometimes being no visible effect if blocked in a = system=20
call.  There is an effect though.  The process is killed
and = will=20
terminate when it is unblocked.  Just because an API can be = blocking=20
doesn't mean it has this behavior.  Most blocking calls are=20
interruptable.
 
Rich
 

  "Paul Ranson" <paul{at}barkto.com>">mailto:paul{at}barkto.com">paul{at}barkto.com>
wrote in =
message news:4333e479$1{at}w3.nls.net...Things=20
  may have changed but I'm pretty sure that 'kill -9' can have no =
effect in=20
  some conditions, if the process is blocked in some system calls =
for=20
  example. IOW the signal will be handled when the call returns, but the =

  call never returns and isn't
interruptible.If you get an =
NT=20
  process you cannot terminate try attaching to it with a debugger, =
that=20
  usually makes it see
sense.Paul"Mike '/m'"
<mike{at}barkto.com>">mailto:mike{at}barkto.com">mike{at}barkto.com>
wrote in =
message news:l1b6j19bo6u=
mmoit70h9snroig3mmjq5ui{at}4ax.com...>=20
  Should be:>> In the *nix's I have used,
there is the =
regular=20
  kill (kill -15) that> nicely  asks the app to =
shutdown.  Then=20
  for those obstinate apps, there> is the kill option (kill -9) =
that=20
  pretty much just terminates the> process, releases memory, and =
brushes=20
  the crumbs off the table.> Obviously, you do not use kill -9 on =
a whim,=20
  but it is nice to know it> is there when =
needed.>> /m=20
  

------=_NextPart_000_034E_01C5C014.C8AF0780--

--- BBBS/NT v4.01 Flag-5