[KrypteX]

I've been getting the following "ttraff: page allocation failure" errors while transferring data over FTP (on WiFi) from the USB-attached drive to my Netgear WNDR3800, build 26866 :

<4>[291236.810000] ttraff: page allocation failure: order:4, mode:0x200020
<4>[291236.810000] CPU: 0 PID: 1103 Comm: ttraff Not tainted 3.10.77 #796
<4>[291236.810000] Stack : 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
<4>[291236.810000] 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
<4>[291236.810000] 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
<4>[291236.810000] 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
<4>[291236.810000] 00000000 00000000 00000000 00000000 00000000 00000000 00000000 86dddc48
<4>[291236.810000] ...
<4>[291236.810000] Call Trace:[<8006cb04>] 0x8006cb04
<4>[291236.810000] [<8006cb04>] 0x8006cb04
<4>[291236.810000] [<800afe20>] 0x800afe20
<4>[291236.810000] [<800b1fdc>] 0x800b1fdc
<4>[291236.810000] [<800d1750>] 0x800d1750
<4>[291236.810000] [<801d3f20>] 0x801d3f20
<4>[291236.810000] [<800d1b68>] 0x800d1b68
<4>[291236.810000] [<801906d4>] 0x801906d4
<4>[291236.810000] [<800b7254>] 0x800b7254
<4>[291236.810000] [<80090470>] 0x80090470
<4>[291236.810000] [<800e93f0>] 0x800e93f0
<4>[291236.810000] [<800e975c>] 0x800e975c
<4>[291236.810000] [<800e9614>] 0x800e9614
<4>[291236.810000] [<801191fc>] 0x801191fc
<4>[291236.810000] [<800e68f4>] 0x800e68f4
<4>[291236.810000] [<8010f5b8>] 0x8010f5b8
<4>[291236.810000] [<800e7308>] 0x800e7308
<4>[291236.810000] [<800d535c>] 0x800d535c
<4>[291236.810000] [<800d6824>] 0x800d6824
<4>[291236.810000] [<800d60f8>] 0x800d60f8
<4>[291236.810000] [<800d4874>] 0x800d4874
<4>[291236.810000] [<800d5abc>] 0x800d5abc
<4>[291236.810000] [<8008d49c>] 0x8008d49c
<4>[291236.810000] [<800621b0>] 0x800621b0
<4>[291236.810000] [<800d4bf4>] 0x800d4bf4
<4>[291236.810000]
<4>[291236.810000] Mem-Info:
<4>[291236.810000] Normal per-cpu:
<4>[291236.810000] CPU 0: hi: 42, btch: 7 usd: 9
<4>[291236.810000] active_anon:717 inactive_anon:0 isolated_anon:0
<4>[291236.810000] active_file:3840 inactive_file:18908 isolated_file:0
<4>[291236.810000] unevictable:101 dirty:1760 writeback:1721 unstable:0
<4>[291236.810000] free:1502 slab_reclaimable:3231 slab_unreclaimable:1406
<4>[291236.810000] mapped:545 shmem:0 pagetables:66 bounce:0
<4>[291236.810000] free_cma:0
<4>[291236.810000] Normal free:6008kB min:4096kB low:5120kB high:6144kB active_anon:2868kB inactive_anon:0kB active_file:15360kB inactive_file:75632kB unevictable:404kB isolated(anon):0kB isolated(file):0kB present:131072kB managed:126504kB mlocked:0kB dirty:7040kB writeback:6884kB mapped:2180kB shmem:0kB slab_reclaimable:12924kB slab_unreclaimable:5624kB kernel_stack:320kB pagetables:264kB unstable:0kB bounce:0kB free_cma:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no
<4>[291236.810000] lowmem_reserve[]: 0 0
<4>[291236.810000] Normal: 174*4kB (UEM) 162*8kB (UE) 107*16kB (UEM) 72*32kB (M) 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 6008kB
<4>[291236.810000] 22863 total pagecache pages
<4>[291236.810000] 0 pages in swap cache
<4>[291236.810000] Swap cache stats: add 0, delete 0, find 0/0
<4>[291236.810000] Free swap = 0kB
<4>[291236.810000] Total swap = 0kB
<4>[291236.810000] SLAB: Unable to allocate memory on node 0 (gfp=0x20)
<4>[291236.810000] cache: kmalloc-65536, object size: 65536, order: 4
<4>[291236.810000] node 0: slabs: 6/6, objs: 6/6, free: 0

This gets repeated periodically, every 58-59 seconds during FTP transfer from USB (over WiFi).
AFAICT, these errors didn't happen on previous build 26653.
What could be the issue here ? I would appreciate any good explanation or related info.

Edit: I've opened a ticket http://svn.dd-wrt.com/ticket/4103

[mrjcd]

I don't know much about this but have seen such on other devices/other applications. Basically looks to be fragmentation. Don't know why 26866 is causing it but seems to be some blocks running over each other with others not knowing where the heck to go.
Have you got important stuff off?
I don't use USB attached to the router so I might not be very helpful but just wondering if you have another USB device handy and does it give same errors?

[KrypteX]

I formatted the USB drive and doesn't happen anymore. Very strange...

[mrjcd]

Probably be alright now. If router caused the memory frag it likely did it when you installed 26866.... when it first loaded. This is interesting. Let us know if it pops up again. I see some others have problems with USB on different routers with new build.....on broadcom too.

[KrypteX]

A couple of things to notice:

1) ttraff is a Traffic/Bandwidth Logger created by Eko 3 years ago, hasn't been updated since: https://github.com/iamacarpet/ttraff
I started getting these errors after checking out the Bandwidth page in DD-WRT *while* tranferring from USB by FTP. Of course the bandwidth was up @ 80-100 Mb/s for the WiFi interface and the CPU usage higher than 1. I guess the ttraff code got stuck or something due to the relatively heavy usage/traffic & monitoring at the same time.

2) Build 26866 is based on Kernel 3.10.77. I've just checked the changelog for the newly released 3.10.79 and I see this memory-related fix, which might have to do something with these errors I was getting:
https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/?id=0073e613da6fb1f39ab75f0373a63ee30d51d635

So I'm looking forward to next public build with 3.10.79 included, it might just be a memory bug/leak in K3.10.77.