Sendmail 8.8.5 and SCO 3.2v4.2

Sendmail 8.8.5 and SCO 3.2v4.2

Post by Jean-Pierre Radl » Fri, 21 Feb 1997 04:00:00



Elias Manesiotis's CPU sent forth this bytestream:
| Has anybody compiled and used succesfully sendmail 8.8.5 on SCO 3.2v4.2?
| Sendmail 8.8.4 and 8.8.5 dont close the connection leaving idle proccesses
| all over the place. Is there a patch for this?  

SCO just released oss443 on its web/ftp servers, and it's on
CompuServe too. It provides sendmail 8.8.5 (for OSR 5 systems).

--

 
 
 

Sendmail 8.8.5 and SCO 3.2v4.2

Post by Elias Manesiot » Fri, 21 Feb 1997 04:00:00


Has anybody compiled and used succesfully sendmail 8.8.5 on SCO 3.2v4.2?
Sendmail 8.8.4 and 8.8.5 dont close the connection leaving idle proccesses
all over the place. Is there a patch for this?  

Question No2: Has anybody compiled Berkeley DB on SCO 3.2v4.2? Any
particular spell? ;-)

Please answer via e-mail

Virtually


CompuLink Network S.A.  URL: http://www.compulink.gr

 
 
 

Sendmail 8.8.5 and SCO 3.2v4.2

Post by Thomas K. Wenri » Fri, 21 Feb 1997 04:00:00


    jpr> Elias Manesiotis's CPU sent forth this bytestream: | Has
    jpr> anybody compiled and used succesfully sendmail 8.8.5 on SCO
    jpr> 3.2v4.2?  | Sendmail 8.8.4 and 8.8.5 dont close the
    jpr> connection leaving idle proccesses | all over the place. Is
    jpr> there a patch for this?

    jpr> SCO just released oss443 on its web/ftp servers, and it's on
    jpr> CompuServe too. It provides sendmail 8.8.5 (for OSR 5
    jpr> systems).

You will note that the question was related to SCO 3.2v4.2 (!)

I have had the same troubles Elias mentions.

There seems to be a bug(?) within v4.2, which causes wait() to hang
while SIGCHLD is blocked. The child of the process which called both
SIG_BLOCK and wait() exits and becomes <defunct>, the father stands within
wait() and waits for the termination of the <defunct> process [?].

As I understand my POSIX book, there is no relation between SIGCHLD and
wait() [?].

Solution: When you comment the blocksignal(SIGCHLD) call within
          srvrsmtp.c/function runinchild() then no processes (neither
          waiting nor <defunct> are left.
          DISCLAIMER: I have no idea which other troubles you may run
                      into when doing so.

Thomas

 
 
 

Sendmail 8.8.5 and SCO 3.2v4.2

Post by Jean-Pierre Radl » Sat, 22 Feb 1997 04:00:00


Larry James's CPU sent forth this bytestream:

| : Elias Manesiotis's CPU sent forth this bytestream:
| : | Has anybody compiled and used succesfully sendmail 8.8.5 on SCO 3.2v4.2?
| : | Sendmail 8.8.4 and 8.8.5 dont close the connection leaving idle proccesses
| : | all over the place. Is there a patch for this?  
|
| : SCO just released oss443 on its web/ftp servers, and it's on
| : CompuServe too. It provides sendmail 8.8.5 (for OSR 5 systems).
|
|      Do you think this will work on SCO version 3.2v4.0
|

Assuredly not.

--

 
 
 

Sendmail 8.8.5 and SCO 3.2v4.2

Post by Larry Jam » Sat, 22 Feb 1997 04:00:00


: Elias Manesiotis's CPU sent forth this bytestream:
: | Has anybody compiled and used succesfully sendmail 8.8.5 on SCO 3.2v4.2?
: | Sendmail 8.8.4 and 8.8.5 dont close the connection leaving idle proccesses
: | all over the place. Is there a patch for this?  

: SCO just released oss443 on its web/ftp servers, and it's on
: CompuServe too. It provides sendmail 8.8.5 (for OSR 5 systems).

     Do you think this will work on SCO version 3.2v4.0

                                -- Larry