Article delegate-en/195 of [1-5169] on the server localhost:119
  upper oldest olders older1 this newer1 newers latest
search
[Top/Up] [oldest] - [Older+chunk] - [Newer+chunk] - [newest + Check]
[Reference:<19990101205452.346.qmail@www0f.netaddress.usa.net>]
Newsgroups: mail-lists.delegate-en

[DeleGate-En] Re: Fatal Error in Win95 version of 5.8.5]]
02 Jan 1999 01:14:30 GMT ysato@etl.go.jp (Yutaka Sato 佐藤豊)


In message <19990101205452.346.qmail@www0f.netaddress.usa.net>
on 01/02/99(05:54:52)
you Mark Peacock <px4aabdyi-dyd2yvhynoxr.ml@ml.delegate.org> wrote:
 |The fix worked fine!  The result in my environment looks exactly
 |like yours.
 |Indeed, I am using it as an HTTP proxy as I write this.

Then I suppose now you can use the DeleGate/5.8.5 without the fix
without the problem too, isn't it?

If so...

I suppose I can describe the problem as this: a handle value for a
socket assigned in a child process, which is local to each process
in Win95, can be, by chance, the same with the one of sockets handles
which are asigned in the parent process, then it is closed by child
process which result in error with:
>     FATAL: inherited handle[2] 21 is not a socket
This will not be caused in a case I reported in my former message as

 |(WIN) 15:24 [-239559] PPID=-246407/23 CLSOCK=[29->11]
 |(WIN) 15:24 [-239559] PPID=-246407/23 CLSOCK=[14->14]
 |(WIN) 15:24 [-239559] PPID=-246407/23 CLSOCK=[22->22]

"29" will be closed but it may not be a active handle in the child.
"14" and "22" are not closed because they are obviously the handle value
in the child.  But in a case like below

 |(WIN) 15:24 [-239559] PPID=-246407/23 CLSOCK=[22->11]
 |(WIN) 15:24 [-239559] PPID=-246407/23 CLSOCK=[14->14]
 |(WIN) 15:24 [-239559] PPID=-246407/23 CLSOCK=[29->22]

"22" will be closed at the first line, then cause the error message
>     FATAL: inherited handle[2] 21 is not a socket
at processing of the third line.

I'm not sure if this hypothesis is to the point, but this may happen
by chance.  So I'll adopt the fix in the next release...

Cheers,
Yutaka
--
Yutaka Sato <ysato@etl.go.jp> http://www.etl.go.jp/~ysato/   @ @ 
Computer Science Division, Electrotechnical Laboratory      ( - )
1-1-4 Umezono, Tsukuba, Ibaraki, 305-8568 Japan            _<   >_

  admin search upper oldest olders older1 this newer1 newers latest
[Top/Up] [oldest] - [Older+chunk] - [Newer+chunk] - [newest + Check]
@_@V