home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!gatech!darwin.sura.net!spool.mu.edu!nigel.msen.com!math.fu-berlin.de!news.netmbx.de!Germany.EU.net!Informatik.Uni-Dortmund.DE!trillian!rg
- From: rg@trillian.informatik.uni-dortmund.de (Rudolf Gross)
- Newsgroups: comp.os.linux
- Subject: Memory leaks in the TCP/IP-Code ?
- Date: 15 Nov 1992 17:07:43 GMT
- Organization: University of Dortmund, Germany
- Lines: 22
- Message-ID: <1e604vINNlqr@fbi-news.Informatik.Uni-Dortmund.DE>
- References: <1992Nov15.142934.29409@monu6.cc.monash.edu.au>
- NNTP-Posting-Host: trillian
-
- I've been compiling an LPmud driver using linux.
- For those who aren't familiar with it: The driver
- creates a socket, and accept()/select()s on it.
- The driver is monolithic and has NO childs or so.
-
- However, even if nothing happens, "free" reports
- that the memory is running out. In 4 blocks
- portions every 3-6 seconds or so. The driver itself
- keeps memory statistics and reports NO additional
- memory being allocated.
-
- Could it be that there is some obscure bug in the
- kernel ? The same code works 100% under SunOS and
- a VERY capable LPmud-driver-specialist assured me
- that the driver itself does NOT allocate memory
- like that.
-
- Does any kernel wizard have an idea where the memory
- leak could be ?
-
- Thanks in advance,
- Rudi Gross
-