fs/fuse: fuse mount can cause panic with no memory numa node

fs/fuse: fuse mount can cause panic with no memory numa node

Orabug: 26591421

Host panics with the below stack trace:-

[ 2509.835362] BUG: unable to handle kernel paging request at
0000000000001e08
[ 2509.843155] IP: [<ffffffff8119e1a7>]
__alloc_pages_nodemask+0xb7/0xa10
[ 2509.850466] PGD 7f5bbf5067 PUD 7f5c492067 PMD 0
..
..
[ 2510.108923] Call Trace:
[ 2510.111658]  [<ffffffff81022075>] ? native_sched_clock+0x35/0xa0
[ 2510.118360]  [<ffffffff810220e9>] ? sched_clock+0x9/0x10
[ 2510.124291]  [<ffffffff810b9ad5>] ? sched_clock_cpu+0x85/0xc0
[ 2510.130692]  [<ffffffff810b6227>] ? try_to_wake_up+0x47/0x370
[ 2510.137108]  [<ffffffff811f0c43>] ? deactivate_slab+0x383/0x400
[ 2510.143712]  [<ffffffff811f1e37>] new_slab+0xa7/0x460
[ 2510.149350]  [<ffffffff81733cc0>] __slab_alloc+0x317/0x477
[ 2510.155473]  [<ffffffffa0380336>] ? fuse_conn_init+0x236/0x420 [fuse]
[ 2510.162664]  [<ffffffff8146184c>] ? extract_entropy+0x14c/0x330
[ 2510.169260]  [<ffffffff814622b0>] ? get_random_bytes+0x40/0xa0
[ 2510.175769]  [<ffffffff811f2c75>] kmem_cache_alloc_node_trace+0x95/0x2b0
[ 2510.183248]  [<ffffffffa0380336>] fuse_conn_init+0x236/0x420 [fuse]
[ 2510.190241]  [<ffffffffa03808c0>] fuse_fill_super+0x3a0/0x700 [fuse]
[ 2510.197330]  [<ffffffff811f30f6>] ? __kmalloc+0x266/0x2c0
[ 2510.203355]  [<ffffffff811a6dcc>] ? register_shrinker+0x3c/0xa0
[ 2510.209961]  [<ffffffff81216728>] ? sget+0x3b8/0x3f0
[ 2510.215500]  [<ffffffff81215880>] ? get_anon_bdev+0x120/0x120
[ 2510.221911]  [<ffffffffa0380520>] ? fuse_conn_init+0x420/0x420 [fuse]
[ 2510.229096]  [<ffffffff8121686d>] mount_nodev+0x4d/0xb0
[ 2510.234927]  [<ffffffffa037ec38>] fuse_mount+0x18/0x20 [fuse]
[ 2510.241337]  [<ffffffff812173e9>] mount_fs+0x39/0x180
[ 2510.246975]  [<ffffffff8123366b>] vfs_kern_mount+0x6b/0x110
[ 2510.253191]  [<ffffffff81236411>] do_mount+0x251/0xcf0
[ 2510.258922]  [<ffffffff812371f2>] SyS_mount+0xa2/0x110
[ 2510.264653]  [<ffffffff81028666>] ? syscall_trace_leave+0xc6/0x150
[ 2510.271551]  [<ffffffff8173d96e>] system_call_fastpath+0x12/0x71

This commit reverts the changes done in 937287c2e4ab ("fs/fuse: Fix for
correct number of numa nodes") and fixes the original problem by calling
kmalloc_node/kmem_cache_alloc_node, only on the online numa nodes. For the
offline nodes, kmalloc or kmem_cache_alloc will be used for allocation.
This problem happens only with slub allocator because it does not do
fallback allocation if node is offline, unlike slab. And for that reason,
the same behavior is emulated here.

Signed-off-by: Somasundaram Krishnasamy <somasundaram.krishnasamy@oracle.com>
Reviewed-by: Ashish Samant<ashish.samant@oracle.com>
Reviewed-by: Babu Moger <babu.moger@oracle.com>