Latest working kernel version: unknown Earliest failing kernel version: observed from 2.6.26.xx Distribution: Archlinux Hardware Environment: ASUS F3JC (1Gb of RAM) HP Compaq 6710b (2Gb of RAM) On both: Network controller: Intel Corporation PRO/Wireless 3945ABG Network Connection (rev 02) Both have i686 systems. Software Environment: modinfo iwl3945 | grep version version: 1.2.26ks iwlwifi-3945-ucode 15.28.1.6 on HP laptop iwlwifi-3945-ucode 15.28.2.8 on ASUS laptop Problem Description: Being online and using WiFi in ad-hoc mode for two of three days (with nightly suspends) system starts acting one by of two scenarios: 1. It begins swapping with increasing intensity. After several hours system stops responding insanely using hdd. (occurs more likely on ASUS laptop, since it has less RAM) 2. Oom killer wakes up and continuously kills applications (when I restart xserver for example, it is killed immediately). But according to /usr/bin/free there is a lot of free RAM (nearly 1GB on HP laptop, and 400Mb on ASUS laptop). If I use Wifi rarely, system on both laptops works longer. When I try to reload module (rmmod iwl3945; sleep 5; modprobe iwl3945) it either just doesn't help or leads to system freeze with blinking caps led. Some parts of /var/log/messages.log: Pid: 13218, comm: ifconfig Not tainted 2.6.27-ARCH #1 [<c016622c>] oom_kill_process+0x8c/0x210 [<c01e3f89>] security_capable+0x9/0x10 [<c016677c>] badness+0x14c/0x1d0 [<c016699c>] out_of_memory+0x19c/0x1e0 [<c01697a2>] __alloc_pages_internal+0x442/0x480 [<c0169841>] __get_free_pages+0x21/0x30 [<f91d4244>] iwl3945_tx_queue_init+0x84/0x1a0 [iwl3945] [<f91d6c3f>] iwl3945_hw_nic_init+0x89f/0x900 [iwl3945] [<f91cbb1c>] __iwl3945_up+0x8c/0x640 [iwl3945] [<c0211afd>] msi_set_enable+0x7d/0x90 [<c014d58c>] debug_mutex_add_waiter+0x1c/0x60 [<c0317873>] __mutex_lock_slowpath+0x1d3/0x250 [<f91d2f3f>] iwl3945_mac_start+0x41f/0x610 [iwl3945] [<c0173cb4>] handle_mm_fault+0x134/0x7f0 [<c0318f05>] _spin_unlock+0x5/0x20 [<c019d4c0>] dentry_iput+0xa0/0xb0 [<f91a3a77>] ieee80211_open+0x167/0x600 [mac80211] [<c0194e8b>] __link_path_walk+0x25b/0xe10 [<c029759e>] dev_open+0x7e/0xe0 [<c0318cca>] _spin_lock_bh+0x1a/0x20 [<c02947b0>] dev_set_rx_mode+0x20/0x40 [<c0296e95>] dev_change_flags+0x85/0x1c0 [<c02dffd5>] devinet_ioctl+0x6e5/0x700 [<c0294a95>] __dev_get_by_name+0x75/0x90 [<c028836b>] sock_ioctl+0x6b/0x250 [<c0288300>] sock_ioctl+0x0/0x250 [<c01980fb>] vfs_ioctl+0x2b/0x90 [<c01981c7>] do_vfs_ioctl+0x67/0x2f0 [<c01984b6>] sys_ioctl+0x66/0x70 [<c0103ef3>] sysenter_do_call+0x12/0x33 [<c0310000>] cache_add_dev+0x19a/0x44a ======================= Mem-Info: DMA per-cpu: CPU 0: hi: 0, btch: 1 usd: 0 CPU 1: hi: 0, btch: 1 usd: 0 Normal per-cpu: CPU 0: hi: 186, btch: 31 usd: 0 CPU 1: hi: 186, btch: 31 usd: 4 HighMem per-cpu: CPU 0: hi: 186, btch: 31 usd: 0 CPU 1: hi: 186, btch: 31 usd: 20 Active:89627 inactive:13724 dirty:1 writeback:0 unstable:0 free:220308 slab:164563 mapped:14376 pagetables:961 bounce:0 DMA free:3584kB min:64kB low:80kB high:96kB active:0kB inactive:0kB present:15868kB pages_scanned:0 all_unreclaimable? no lowmem_reserve[]: 0 873 2007 2007 Normal free:124652kB min:3744kB low:4680kB high:5616kB active:736kB inactive:528kB present:894080kB pages_scanned:0 all_unreclaimable? no lowmem_reserve[]: 0 0 9078 9078 HighMem free:752996kB min:512kB low:1728kB high:2944kB active:357772kB inactive:54368kB present:1161984kB pages_scanned:0 all_unreclaimable? no lowmem_reserve[]: 0 0 0 0 DMA: 8*4kB 6*8kB 73*16kB 5*32kB 4*64kB 1*128kB 1*256kB 1*512kB 1*1024kB 0*2048kB 0*4096kB = 3584kB Normal: 8724*4kB 8200*8kB 1480*16kB 3*32kB 0*64kB 1*128kB 1*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 124656kB HighMem: 5686*4kB 391*8kB 189*16kB 2150*32kB 1239*64kB 516*128kB 270*256kB 137*512kB 74*1024kB 36*2048kB 54*4096kB = 752992kB 47796 total pagecache pages 17237 pages in swap cache Swap cache stats: add 36646, delete 19409, find 14593/17223 Free swap = 2366216kB Total swap = 2441840kB 522160 pages RAM 292784 pages HighMem 5382 pages reserved 103384 pages shared 236842 pages non-shared I think similar issue was discussed here: http://lkml.org/lkml/2008/6/12/121 Steps to reproduce: 1. Start wifi in Ad-Hoc modes. 2. Proceed every-day work with low network activity suspending (to ram) systems. 3. Observer issue described above.
Could you please retest with 2.6.28?
No changes.
Created attachment 20614 [details] Use GFP_KERNEL Check the patch and apply it over latest kernel or compact wireless package.
Ramil, More than a month ago we asked you to try a patch. Are you able to do more testing? Are you still seeing the problem? If this is no longer an issue for you, please close this bug. If this is an issue, please help us to debug it. Thanks!
Sorry, I haven't been able to check the patch, although it was in my TODO. I'll check it this week probably.
Kernel: 2.6.29.1 (I've decided to test unpatched latest kernel first) Since 25.04 I'm using ad-hoc and the issue seems to be gone. System on HP laptop hang up yesterday, but I think its unrelated.