On 1/30/17 7:28 PM, Joe Perches wrote:
On Mon, 2017-01-30 at 18:25 -0800, Alexei Starovoitov wrote:
under memory pressure 'ethtool -S' command may warn:
[ 2374.385195] ethtool: page allocation failure: order:4, mode:0x242c0c0
[ 2374.405573] CPU: 12 PID: 40211 Comm: ethtool Not tainted
[ 2374.423071] Call Trace:
[ 2374.423076] [<ffffffff8148cb29>] dump_stack+0x4d/0x64
[ 2374.423080] [<ffffffff811667cb>] warn_alloc_failed+0xeb/0x150
[ 2374.423082] [<ffffffff81169cd3>] ? __alloc_pages_direct_compact+0x43/0xf0
[ 2374.423084] [<ffffffff8116a25c>] __alloc_pages_nodemask+0x4dc/0xbf0
[ 2374.423091] [<ffffffffa0023dc2>] ? cmd_exec+0x722/0xcd0 [mlx5_core]
[ 2374.423095] [<ffffffff811b3dcc>] alloc_pages_current+0x8c/0x110
[ 2374.423097] [<ffffffff81168859>] alloc_kmem_pages+0x19/0x90
[ 2374.423099] [<ffffffff81186e5e>] kmalloc_order_trace+0x2e/0xe0
[ 2374.423101] [<ffffffff811c0084>] __kmalloc+0x204/0x220
[ 2374.423105] [<ffffffff816c269e>] dev_ethtool+0xe4e/0x1f80
[ 2374.423106] [<ffffffff816b967e>] ? dev_get_by_name_rcu+0x5e/0x80
[ 2374.423108] [<ffffffff816d6926>] dev_ioctl+0x156/0x560
[ 2374.423111] [<ffffffff811d4c68>] ? mem_cgroup_commit_charge+0x78/0x3c0
[ 2374.423117] [<ffffffff8169d542>] sock_do_ioctl+0x42/0x50
[ 2374.423119] [<ffffffff8169d9c3>] sock_ioctl+0x1b3/0x250
[ 2374.423121] [<ffffffff811f0f42>] do_vfs_ioctl+0x92/0x580
[ 2374.423123] [<ffffffff8100222b>] ? do_audit_syscall_entry+0x4b/0x70
[ 2374.423124] [<ffffffff8100287c>] ? syscall_trace_enter_phase1+0xfc/0x120
[ 2374.423126] [<ffffffff811f14a9>] SyS_ioctl+0x79/0x90
[ 2374.423127] [<ffffffff81002bb0>] do_syscall_64+0x50/0xa0
[ 2374.423129] [<ffffffff817e19bc>] entry_SYSCALL64_slow_path+0x25/0x25
~1160 mlx5 counters ~= order 4 allocation which is unlikely to succeed
under memory pressure. Convert them to vzalloc() as ethtool_get_regs() does.
Also take care of drivers without counters similar to
commit 67ae7cf1eeda ("ethtool: Allow zero-length register dumps again")
and reduce warn_on to warn_on_once.
I think this is generally not a good idea as
most uses already fit fine in a kcalloc.
most nics have large numbers of counters that don't fit into one page.
that's already pushing mm.
especially in this case control plane apps call 'ethtool -S'
periodically.
Maybe use Michal Hocko's kvmalloc changes.
https://lkml.org/lkml/2017/1/30/120
v1 discussion here
http://patchwork.ozlabs.org/patch/721122/
as I mentioned there long term I don't mind using kvmalloc,
but the issue has to be fixed now. Either via vzalloc or nowarn+noretry.
My stress testing with memory hog shows that kmalloc fails
quite often, thankfully user space daemon is ready for failures,
whereas vzalloc approach works all the time and no extra headaches
for user space.
