From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.5) Gecko/20031208 Description of problem: Network just hangs after a while after suspend/resume (for eg: when downloading stuff with lftp) Version-Release number of selected component (if applicable): kernel-2.6.5-1.327 How reproducible: Always Steps to Reproduce: 1. Install FC2T3 on an IBM ThinkPad T40 2. Start up networking (eth0 / e1000) 3. suspend using: echo 3 > /proc/acpi/sleep 4. recover from suspend by pressing 'power' button 3. lftp mirror.... 4. mget *.rpm Actual Results: < netowk hangs after a while> Expected Results: no such hang Additional info: I've posted about this in fedora-test-list a while back with rawhide http://www.redhat.com/archives/fedora-test-list/2004-April/msg01857.html 02:01.0 Ethernet controller: Intel Corp. 82540EP Gigabit Ethernet Controller (Mobile) (rev 03) <When the hang happens - from /var/log/messages> Apr 28 19:57:48 asterix kernel: Restarting tasks... done Apr 28 19:57:48 asterix kernel: agpgart: Found an AGP 2.0 compliant device at 0000:00:00.0. Apr 28 18:46:14 asterix input.agent[2255]: ... no modules for INPUT product Apr 28 18:46:14 asterix input.agent[2264]: ... no modules for INPUT product Apr 28 19:57:48 asterix kernel: agpgart: Putting AGP V2 device at 0000:00:00.0 into 1x mode Apr 28 18:46:14 asterix input.agent[2273]: ... no modules for INPUT product 11/2/1/0 Apr 28 18:46:14 asterix kernel: agpgart: Putting AGP V2 device at 0000:01:00.0 into 1x mode Apr 28 18:46:14 asterix kernel: [drm] Loading R200 Microcode Apr 28 18:46:14 asterix kernel: input: PS/2 Generic Mouse on isa0060/serio1 Apr 28 18:46:15 asterix kernel: e1000: eth0 NIC Link is Up 100 Mbps Full Duplex Apr 28 18:47:00 asterix kernel: NETDEV WATCHDOG: eth0: transmit timed out Apr 28 18:47:02 asterix kernel: e1000: eth0 NIC Link is Up 100 Mbps Full Duplex Apr 28 18:47:37 asterix kernel: NETDEV WATCHDOG: eth0: transmit timed out ----------------- BTW: When recovering from suspend - I see some 'wierd chars' printed on the <suspend-console-screen>. Don't know how to get to this screen - as it just pops-up briefly during susend - and resume
The driver appears to work fine with APM (acpi=off). The same behavior happens with kernel-2.6.5-1.344
Identical behaviour with kernel 2.6.5-1.349, IBM ThinkPad A30p, e100 ; rmmod'ing and modprobing e100 and mii does not resolve the issue : eth0 NIC link detects 100 Mbps Full Duplex, but subsequent dhclient eth0 fails to communicate with DHCP server. Does this qualify for a bug #121020 blocker ?
Still present with kernel 2.6.6-1.427, on a Thinkpad T41 (2373 4HU). I had to rmmod ehci_hcd to get ACPI suspend to work on this kernel, but except for that and reloading it the sequence is the same as above. eth0 restarts OK: Jun 12 02:09:19 localhost kernel: e1000: eth0: e1000_watchdog: NIC Link is Up 100 Mbps Half Duplex but transfers hang after a few MB. Problem not seen until an ACPI suspend/resume has been performed. FWIW, I see earlier in the log that there are some issues with the parms for this driver: Jun 12 01:57:48 localhost kernel: e1000: Ignoring new-style parameters in presence of obsolete ones Jun 12 01:57:48 localhost kernel: Intel(R) PRO/1000 Network Driver - version 5.2.52-k4 Jun 12 01:57:48 localhost kernel: Copyright (c) 1999-2004 Intel Corporation. Jun 12 01:57:48 localhost kernel: e1000: eth0: e1000_probe: Intel(R) PRO/1000 Network Connection Jun 12 01:57:48 localhost kernel: e1000: eth0: e1000_watchdog: NIC Link is Up 100 Mbps Half Duplex
I tried with fc3t3/rawhide 2.6.8-1.624 kernel today - and I can't reporduce this anymore. (downloaded 700mb of stuff after ACPI suspend/resume) Perhpas this issue got fixed upstream - since the last time I tried it out.