我想通过＆＃34; dev_queue_xmit＆＃34;发送sk_buff，当我发送2个数据包时，网卡可能会挂起。 我想知道如何调试这个条件。
[root@10g-host2 test]# tail -f /var/log/messages
Sep 29 10:38:22 10g-host2 acpid: waiting for events: event logging is off
Sep 29 10:38:23 10g-host2 acpid: client connected from 2018[68:68]
Sep 29 10:38:23 10g-host2 acpid: 1 client rule loaded
Sep 29 10:38:24 10g-host2 automount[2210]: lookup_read_master: lookup(nisplus): couldn't locate nis+ table auto.master
Sep 29 10:38:24 10g-host2 mcelog: failed to prefill DIMM database from DMI data
Sep 29 10:38:24 10g-host2 xinetd[2246]: xinetd Version 2.3.14 started with libwrap loadavg labeled-networking options compiled in.
Sep 29 10:38:24 10g-host2 xinetd[2246]: Started working: 0 available services
Sep 29 10:38:25 10g-host2 abrtd: Init complete, entering main loop
Sep 29 10:39:41 10g-host2 kernel: vmalloc mmap_buf=ffffc90016e29000 mmap_size=4096
Sep 29 10:39:41 10g-host2 kernel: insmod module wsmmap successfully!
Sep 29 10:39:49 10g-host2 kernel: mmap_buf + 1024 is ffffc90016e29400
Sep 29 10:39:49 10g-host2 kernel: data ffffc90016e2942a, len is 42
Sep 29 10:39:49 10g-host2 kernel: udp data ffffc90016e29422
Sep 29 10:39:49 10g-host2 kernel: ip data ffffc90016e2940e
Sep 29 10:39:49 10g-host2 kernel: eth data ffffc90016e29400
Sep 29 10:39:49 10g-host2 kernel: h_source is ffffc90016e29406, dev_addr is ffff880c235c4750, len is 6result is 0
Sep 29 10:39:50 10g-host2 kernel: mmap_buf + 1024 is ffffc90016e29400
Sep 29 10:39:50 10g-host2 kernel: data ffffc90016e2942a, len is 42
Sep 29 10:39:50 10g-host2 kernel: udp data ffffc90016e29422
Sep 29 10:39:50 10g-host2 kernel: ip data ffffc90016e2940e
Sep 29 10:39:50 10g-host2 kernel: eth data ffffc90016e29400
Sep 29 10:39:50 10g-host2 kernel: h_source is ffffc90016e29406, dev_addr is ffff880c235c4750, len is 6result is 0
Sep 29 10:39:52 10g-host2 kernel: ixgbe 0000:03:00.0: eth2: Detected Tx Unit Hang
Sep 29 10:39:52 10g-host2 kernel: Tx Queue <11>
Sep 29 10:39:52 10g-host2 kernel: TDH, TDT <0>, <5>
Sep 29 10:39:52 10g-host2 kernel: next_to_use <5>
Sep 29 10:39:52 10g-host2 kernel: next_to_clean <0>
Sep 29 10:39:52 10g-host2 kernel: ixgbe 0000:03:00.0: eth2: tx_buffer_info[next_to_clean]
Sep 29 10:39:52 10g-host2 kernel: time_stamp <fffd3dd8>
Sep 29 10:39:52 10g-host2 kernel: jiffies <fffd497f>
Sep 29 10:39:52 10g-host2 kernel: ixgbe 0000:03:00.0: eth2: tx hang 1 detected on queue 11, resetting adapter
Sep 29 10:39:52 10g-host2 kernel: ixgbe 0000:03:00.0: eth2: Reset adapter
Sep 29 10:39:52 10g-host2 kernel: ixgbe 0000:03:00.0: master disable timed out
Sep 29 10:39:53 10g-host2 kernel: ixgbe 0000:03:00.0: eth2: detected SFP+: 5
Sep 29 10:39:54 10g-host2 kernel: ixgbe 0000:03:00.0: eth2: NIC Link is Up 10 Gbps, Flow Control: RX/TX
我的电脑的一些信息是：
ethtool -i eth2
driver: ixgbe
version: 3.21.2
firmware-version: 0x1bab0001
bus-info: 0000:03:00.0
supports-statistics: yes
supports-test: yes
supports-eeprom-access: yes
supports-register-dump: yes
supports-priv-flags: no
LSB Version: :base-4.0-amd64:base-4.0-noarch:core-4.0-amd64:core-4.0-noarch:graphics-4.0-amd64:graphics-4.0-noarch:printing-4.0-amd64:printing-4.0-noarch
Distributor ID: CentOS
Description: CentOS release 6.5 (Final)
Release: 6.5
Codename: Final
内核版本是：2.6.32-431.el6.x86_64
感谢您的帮助。
答案 0 :(得分：0)
我使用vmalloc（）为skb-＆gt;数据分配内存，因此这会让NIC失效。我通过使用kmalloc（）修复它。