Date: Mon, 8 May 2000 14:34:14 -0700 (PDT) From: Brian Buchanan <bbuchanan@desktop.com> To: stable@freebsd.org Subject: 1459 data byte TCP packets, pn driver, and 3.2-RELEASE Message-ID: <Pine.BSF.4.05.10005081347220.92815-100000@mail.desktop.com>
next in thread | raw e-mail | index | archive | help
Are there any known bugs in the pn driver in 3.2-STABLE that might cause a TCP packet with a 1459 byte payload to get dropped? I can reproduce this reliably on machines with the following configuration: FreeBSD b.proof.jumpdata.com 3.2-RELEASE FreeBSD 3.2-RELEASE #0: Mon Mar 27 17:53:57 PST 2000 root@z-proof.jumpdata.com:/usr/src/sys/compile/3.2R-20000327 i386 pn0: <82c169 PNIC 10/100BaseTX> rev 0x20 int a irq 11 on pci0.15.0 pn0: Ethernet address: 00:a0:cc:55:77:c6 pn0: autoneg complete, link status good (half-duplex, 100Mbps) <remote machine, any OS> dd if=/dev/random of=tmp.file bs=1 count=1459 netcat -p 9000 -f tmp.file <FreeBSD 3.2-RELEASE w/pn0> telnet <remote machine> 9000 ... Nothing <FreeBSD 3.2-RELEASE w/fxp0> telnet <remote machine> 9000 ... Contents of tmp.file This occurs both with version 1.6.2.10 and 1.6.2.13 of /sys/pci/if_pn.c Examination of tcpdump output revealed that packets with 1459 data bytes are sent by the sending end but never make it to tcpdump on the receiving end. Setting the MTU small enough that the packet had to be broken into 2 fragments caused the connection-hang symptom to go away, as expected. Brian Buchanan Desktop.com, Inc. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.05.10005081347220.92815-100000>