Bug#678461: linux-image-3.0.0-1-amd64: e1000e driver for onboard Intel Corporation

2012-06-21 Thread Attila Mate
Package: linux-2.6
Version: 3.0.0-3
Severity: normal

..
Dear Maintainer,   Thu Jun 21 07:22:32 PM EDT 2012
   Using the driver (# is the superuser prompt):

  # ethtool -i eth0
  driver: e1000e
  version: 1.3.10-k2
  firmware-version: 0.13-4
  bus-info: :00:19.0
  supports-statistics: yes
  supports-test: yes
  supports-eeprom-access: yes
  supports-register-dump: yes
  supports-priv-flags: no

Without tuning, it drops about 30-50% of packets when pinging
yahoo.com. The following tuning eliminates all packet loss:

  # ethtool -G eth0 rx 4096 tx 4096
  # ethtool -s eth0 speed 100 duplex full autoneg off

The problem is that this slows down my Internet connection more
than it should. I have nominal dowload speed with RCN cable
of 25 Megabit/sec. Behind my D-Link VWR-VD router my download
speed is at most 7 or 8 Megabit/sec, whereas another computer,
behind the same router, running Fedora 12 actually gets 22
Megabits/sec or even 25 Megabits/sec. In principle, the above
command slows down the Gigabit network adapter in my computer
to 100 Megabits/sec, so I should be able to get around 22
Megabits/sec download speed, which is what RCN guarantees.
(There is no problem with the upload speed; this is a little
over 2 Megabits/sec on both of my computers behind the D-Link
router.) It would be best if the network adapter worked without
any tuning or slowing down, but the packetloss is too high, and
it makes the network virtually unusable.
   The rx and tx setting do not seem to make much difference.
Packet loss is avoided even with the settings

  # ethtool -G eth0 rx 64 tx 64
  # ethtool -s eth0 speed 100 duplex full autoneg off

and different settings of rx and tx do not seem to make
much difference in download speeds.
..


-- Package-specific info:
** Version:
Linux version 3.0.0-1-amd64 (Debian 3.0.0-3) (b...@decadent.org.uk) (gcc 
version 4.5.3 (Debian 4.5.3-8) ) #1 SMP Sat Aug 27 16:21:11 UTC 2011

** Command line:
BOOT_IMAGE=/boot/vmlinuz-3.0.0-1-amd64 
root=UUID=9a46da86-1e6f-4a0f-af4e-76bad515efb3 ro quiet

** Tainted: W (512)
 * Taint on warning.

** Kernel log:
[   13.326131] RPC: Registered tcp NFSv4.1 backchannel transport module.
[   13.499231] FS-Cache: Loaded
[   13.519322] FS-Cache: Netfs 'nfs' registered for caching
[   13.554739] Installing knfsd (copyright (C) 1996 o...@monad.swb.de).
[   14.148303] fuse init (API version 7.16)
[   14.353971] e1000e: eth0 NIC Link is Up 100 Mbps Full Duplex, Flow Control: 
None
[   14.353975] e1000e :00:19.0: eth0: 10/100 speed: disabling TSO
[   14.354396] ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[   17.060281] Bluetooth: Core ver 2.16
[   17.060292] NET: Registered protocol family 31
[   17.060293] Bluetooth: HCI device and connection manager initialized
[   17.060295] Bluetooth: HCI socket layer initialized
[   17.060296] Bluetooth: L2CAP socket layer initialized
[   17.060324] Bluetooth: SCO socket layer initialized
[   17.075163] Bluetooth: RFCOMM TTY layer initialized
[   17.075167] Bluetooth: RFCOMM socket layer initialized
[   17.075168] Bluetooth: RFCOMM ver 1.11
[   17.120083] Bluetooth: BNEP (Ethernet Emulation) ver 1.3
[   17.120085] Bluetooth: BNEP filters: protocol multicast
[   17.405060] lp: driver loaded but no devices found
[   17.474111] ppdev: user-space parallel port driver
[   25.086162] eth0: no IPv6 routers present
[  100.713780] postgres (2365): /proc/2365/oom_adj is deprecated, please use 
/proc/2365/oom_score_adj instead.
[  102.620229] NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery 
directory
[  102.626680] NFSD: starting 90-second grace period
[  363.036698] [ cut here ]
[  363.036709] WARNING: at 
/build/buildd-linux-2.6_3.0.0-3-amd64-9ClimQ/linux-2.6-3.0.0/debian/build/source_amd64_none/net/sched/sch_generic.c:255
 dev_watchdog+0xe9/0x148()
[  363.036715] Hardware name: 
[  363.036719] NETDEV WATCHDOG: eth0 (e1000e): transmit queue 0 timed out
[  363.036721] Modules linked in: acpi_cpufreq mperf cpufreq_powersave 
cpufreq_userspace cpufreq_conservative cpufreq_stats parport_pc ppdev lp 
parport bnep rfcomm bluetooth rfkill binfmt_misc fuse nfsd nfs lockd fscache 
auth_rpcgss nfs_acl sunrpc loop snd_hda_codec_hdmi snd_hda_codec_realtek 
snd_hda_intel i915 snd_hda_codec snd_hwdep drm_kms_helper snd_pcm snd_seq 
snd_timer snd_seq_device drm snd i2c_algo_bit iTCO_wdt iTCO_vendor_support 
i2c_i801 i2c_core video processor soundcore snd_page_alloc shpchp pci_hotplug 
evdev pcspkr thermal_sys button psmouse serio_raw ext4 mbcache jbd2 crc16 
usbhid hid sr_mod cdrom sd_mod crc_t10dif ahci libahci libata xhci_hcd scsi_mod 
ehci_hcd usbcore e1000e [last unloaded: scsi_wait_scan]
[  363.036781] Pid: 0, comm: swapper Not tainted 3.0.0-1-amd64 #1
[  363.036784] Call Trace:
[  363.036785]  IRQ  

Bug#678461: linux-image-3.0.0-1-amd64: e1000e driver for onboard Intel Corporation

2012-06-21 Thread Ben Hutchings
On Thu, Jun 21, 2012 at 07:26:07PM -0400, Attila Mate wrote:
 Package: linux-2.6
 Version: 3.0.0-3

This is a year out of date; please upgrade to the current version
in testing or unstable.  You should install the metapackage
linux-image-amd64 to ensure that the kernel is kept up to date.

 Severity: normal
 
 ..
 Dear Maintainer,   Thu Jun 21 07:22:32 PM EDT 2012
Using the driver (# is the superuser prompt):
 
   # ethtool -i eth0
   driver: e1000e
   version: 1.3.10-k2
   firmware-version: 0.13-4
   bus-info: :00:19.0
   supports-statistics: yes
   supports-test: yes
   supports-eeprom-access: yes
   supports-register-dump: yes
   supports-priv-flags: no
 
 Without tuning, it drops about 30-50% of packets when pinging
 yahoo.com.
[...]
 [  363.036698] [ cut here ]
 [  363.036709] WARNING: at 
 /build/buildd-linux-2.6_3.0.0-3-amd64-9ClimQ/linux-2.6-3.0.0/debian/build/source_amd64_none/net/sched/sch_generic.c:255
  dev_watchdog+0xe9/0x148()
 [  363.036715] Hardware name: 
 [  363.036719] NETDEV WATCHDOG: eth0 (e1000e): transmit queue 0 timed out
[...]

I see a bigger problem...

Ben.

-- 
Ben Hutchings
We get into the habit of living before acquiring the habit of thinking.
  - Albert Camus



-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20120621235058.gs2...@decadent.org.uk



Bug#678461: linux-image-3.0.0-1-amd64: e1000e driver for onboard Intel Corporation

2012-06-21 Thread Stephen Hemminger
On Thu, 21 Jun 2012 19:26:07 -0400
Attila Mate m...@sci.brooklyn.cuny.edu wrote:

 Package: linux-2.6
 Version: 3.0.0-3
 Severity: normal
 
 ..

A lot of the E1000 issues are related to ASPM (power management)
kicking in and stopping the device. There are some fixes in 3.2 kernel to
deal with this. You can follow the details over on the e1000-devel mailing
list. The problem seems inter-related with BIOS and chip versions.
  


https://lwn.net/Articles/449448/



-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/20120621172851.342cf...@nehalam.linuxnetplumber.net