On Wed, Nov 16, 2005 at 11:03:00AM +, Pawel Jakub Dawidek wrote:
[snip]
> - Exit with 2 on usage errors as documented.
Just curious as to why we are not using EX_USAGE (from ) here. Or
doesn't pkill qualify as a `system program'?
--
Jos Backus
jos at catnook.com
___
On Wed, Nov 16, 2005 at 03:06:35PM +0100, [EMAIL PROTECTED] wrote:
> The process didn't receive the signal, it died for itself. The new
> behaviour is identical to whether it died before the list was build.
> Warning about it or stopping with error just makes the two cases
> asymmetrical.
Ok, I ag
On Wed, Nov 16, 2005 at 03:41:43PM +0300, Andrey Chernov wrote:
+> On Wed, Nov 16, 2005 at 12:59:00PM +0100, Pawel Jakub Dawidek wrote:
+> > Let's say you call 'pkill sleep'. First it creates list of the matching
+> > processes and when the list is done, it starts killing processes from
+> > the li
On Wed, Nov 16, 2005 at 03:41:43PM +0300, Andrey Chernov wrote:
>
> 1) For SIGKILL, pkill should ignore disappeared precesses, as NetBSD does.
> 2) For any other signal, it should issue _warning_ and not error & stop.
The process didn't receive the signal, it died for itself. The new
behaviour is
On Wed, Nov 16, 2005 at 03:41:43PM +0300, Andrey Chernov wrote:
> 1) For SIGKILL, pkill should ignore disappeared precesses, as NetBSD does.
> 2) For any other signal, it should issue _warning_ and not error & stop.
Because SIGKILL can't be trapped and even not do any special action like
SIGILL (
On Wed, Nov 16, 2005 at 12:59:00PM +0100, Pawel Jakub Dawidek wrote:
> Let's say you call 'pkill sleep'. First it creates list of the matching
> processes and when the list is done, it starts killing processes from
> the list. When there is a race, so that process was inserted into the
> list, but
On Wed, Nov 16, 2005 at 02:24:19PM +0300, Andrey Chernov wrote:
+> On Wed, Nov 16, 2005 at 11:03:00AM +, Pawel Jakub Dawidek wrote:
+> > - If a process dissappears while we are signalling it, don't count it as
a
+> > match/error.
+>
+> I don't understand "match" part here, but can say s
On Wed, Nov 16, 2005 at 11:03:00AM +, Pawel Jakub Dawidek wrote:
> - If a process dissappears while we are signalling it, don't count it as a
> match/error.
I don't understand "match" part here, but can say something about "error"
part: while it looks reasonable, it is wrong direction m
pjd 2005-11-16 11:03:00 UTC
FreeBSD src repository
Modified files:
usr.bin/pkillpkill.c
Log:
Sync as close as possible with NetBSD.
This includes fixes and cleanups listed below:
- If a process dissappears while we are signalling it, don't count it as a
mat
