We've been tracking a memory leak on some of our machines, and have
begun using kmemleak. kmemleak is flagging things related to the ixgbe
driver:
unreferenced object 0xffff8810298dc4c0 (size 40):
comm "softirq", pid 0, jiffies 4296008800 (age 437.153s)
hex dump (first 32 bytes):
81 c6 8d 29 10 88 ff ff 00 00 00 00 00 00 00 00 ...)............
00 00 00 00 00 00 00 00 2a b6 0f 00 00 00 00 00 ........*.......
backtrace:
[<ffffffff816f5f78>] kmemleak_alloc+0x28/0x50
[<ffffffff811c36af>] kmem_cache_alloc+0xdf/0x260
[<ffffffff81420f3a>] alloc_iova_mem+0x1a/0x20
[<ffffffff81421272>] alloc_iova+0x32/0x80
[<ffffffff8142c5bb>] intel_alloc_iova+0xcb/0x100
[<ffffffff8142f963>] __intel_map_single+0xb3/0x1e0
[<ffffffff8142fac9>] intel_map_page+0x39/0x40
[<ffffffffa03b2568>] ixgbe_alloc_mapped_page+0xc8/0x170 [ixgbe]
[<ffffffffa03b2668>] ixgbe_alloc_rx_buffers+0x58/0xd0 [ixgbe]
[<ffffffffa03b4421>] ixgbe_clean_rx_irq+0x141/0x2a0 [ixgbe]
[<ffffffffa03b4712>] ixgbe_poll+0xc2/0x270 [ixgbe]
[<ffffffff815de6bc>] napi_poll+0xcc/0x1c0
[<ffffffff815de83d>] net_rx_action+0x8d/0x1b0
[<ffffffff8107ade1>] __do_softirq+0xe1/0x290
[<ffffffff8107b086>] irq_exit+0xa6/0xb0
[<ffffffff817027f5>] do_IRQ+0x65/0xf0
Is there an issue here? I don't really have any sort of reproduction
steps right now.
We're using the 4.3.0 kernel, and the in-kernel driver:
driver: ixgbe
version: 4.0.1-k
firmware-version: 0x800006da
bus-info: 0000:81:00.1
supports-statistics: yes
supports-test: yes
supports-eeprom-access: yes
supports-register-dump: yes
supports-priv-flags: no
------------------------------------------------------------------------------
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel® Ethernet, visit
http://communities.intel.com/community/wired
