home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.unix.ultrix:6246 comp.unix.wizards:3617
- Path: sparky!uunet!tymix!tardis!olivea!decwrl!pa.dec.com!decprl!decprl!boyd
- From: boyd@prl.dec.com (Boyd Roberts)
- Newsgroups: comp.unix.ultrix,comp.unix.wizards
- Subject: Re: A confused tar process...
- Message-ID: <1992Aug14.192122.6911@prl.dec.com>
- Date: 14 Aug 92 19:21:22 GMT
- References: <Bsxn95.HKJ@breeze.rsre.mod.uk> <1992Aug14.153439.12182@eagle.lerc.nasa.gov>
- Sender: news@prl.dec.com (USENET News System)
- Organization: Digital Equipment Corporation - Paris Research Laboratory
- Lines: 24
- Nntp-Posting-Host: prl313.prl.dec.com
-
- In article <1992Aug14.153439.12182@eagle.lerc.nasa.gov>, drich@sandman.lerc.nasa.gov (Daniel Rich) writes:
- >
- > The next step is the part that I am not sure how to do on Ultrix. You
- > need to enter a kernel debugger, and force a return from the routine
- > that is waiting. Since you have probably already sent numorous kills
- > to the process, this should force it to terminate.
- >
-
- I glad that you've given _precise_ instructions.
-
- Waking up processes with a kernel debugger (or adb) is just plain
- dangerous; it may not even be possible. I hope we've all read
- the comment above sleep() which states that the caller should be
- prepared for the sleep to return prematurely. For this reason,
- patching the wchan to `lbolt' (or something similar) may not work
- and the machine will panic if you forget to patch the wchan
- hash table/chain. Patching its p_pri may not work either.
-
- In short: Don't try this at home kids...
-
-
- Boyd Roberts boyd@prl.dec.com
-
- ``When the going gets wierd, the weird turn pro...''
-