Fix scheduling while atomic

In #33 (comment),
the following problem is reported:

[ 2351.274850] BUG: scheduling while atomic: swapper/1/0/0x10000500
[ 2351.274855] Modules linked in: 8723au(OF) cfg80211 dm_crypt vsock
ip6t_REJECT xt_hl ip6t_rt nf_conntrack_ipv6 nf_defrag_ipv6 ipt_REJECT
xt_LOG xt_limit xt_tcpudp xt_addrtype nf_conntrack_ipv4 nf_defrag_ipv4
xt_conntrack rts5139(C) ip6table_filter ip6_tables joydev
nf_conntrack_netbios_ns nf_conntrack_broadcast nf_nat_ftp nf_nat
nf_conntrack_ftp nf_conntrack hid_multitouch iptable_filter ip_tables
x_tables hid_sensor_hub parport_pc ppdev uvcvideo videobuf2_vmalloc
videobuf2_memops videobuf2_core videodev snd_hda_codec_hdmi
x86_pkg_temp_thermal snd_hda_codec_conexant rfcomm intel_powerclamp bnep
snd_hda_intel kvm_intel bluetooth kvm snd_hda_codec crct10dif_pclmul
crc32_pclmul snd_hwdep ghash_clmulni_intel snd_pcm aesni_intel aes_x86_64
lrw gf128mul glue_helper ablk_helper snd_page_alloc cryptd snd_seq_midi
snd_seq_midi_event snd_rawmidi snd_seq snd_seq_device snd_timer snd mei_me
psmouse soundcore lpc_ich mei serio_raw mac_hid nls_iso8859_1 coretemp lp
parport usbhid hid i915 microcode i2c_algo_bit drm_kms_helper ahci drm
libahci wmi video
[ 2351.274918] CPU: 1 PID: 0 Comm: swapper/1 Tainted: GF C O
3.11.0-19-generic #33-Ubuntu
[ 2351.274920] Hardware name: LENOVO 20175/INVALID, BIOS 66CN55WW 02/28/2013
[ 2351.274922] ffff88022f254580 ffff88022f243680 ffffffff816e87f1
ffff880223af2000
[ 2351.274926] ffff88022f243690 ffffffff816e2af1 ffff88022f2436f0
ffffffff816ed8d1
[ 2351.274929] ffff880223af3fd8 0000000000014580 ffff880223af3fd8
0000000000014580
[ 2351.274933] Call Trace:
[ 2351.274935] <IRQ> [<ffffffff816e87f1>] dump_stack+0x45/0x56
[ 2351.274944] [<ffffffff816e2af1>] __schedule_bug+0x4d/0x5b
[ 2351.274948] [<ffffffff816ed8d1>] __schedule+0x6d1/0x7e0
[ 2351.274953] [<ffffffff81092886>] __cond_resched+0x26/0x30
[ 2351.274956] [<ffffffff816eddda>] _cond_resched+0x3a/0x50
[ 2351.274961] [<ffffffff8118e6d8>] kmem_cache_alloc_trace+0x38/0x130
[ 2351.274973] [<ffffffffa061de9c>] ? rtw_addbareq_cmd+0x5e/0x122 [8723au]
[ 2351.274981] [<ffffffffa061de9c>] rtw_addbareq_cmd+0x5e/0x122 [8723au]
[ 2351.274992] [<ffffffffa062c790>] rtw_issue_addbareq_cmd+0x177/0x198
[8723au]
[ 2351.275006] [<ffffffffa0663b9d>] rtw_dump_xframe+0x5d/0x66a [8723au]
[ 2351.275019] [<ffffffffa06644ac>] rtl8723au_hal_xmit+0x143/0x1fb [8723au]
[ 2351.275033] [<ffffffffa0657ba1>] rtw_hal_xmit+0x1c/0x1f [8723au]
[ 2351.275046] [<ffffffffa064bf5c>] rtw_xmit+0x995/0x9fc [8723au]
[ 2351.275050] [<ffffffff815e1887>] ? kfree_skbmem+0x37/0x90
[ 2351.275061] [<ffffffffa0688eb6>] rtw_xmit_entry+0x1a1/0x2ab [8723au]
[ 2351.275065] [<ffffffff815f4368>] dev_hard_start_xmit+0x318/0x560
[ 2351.275068] [<ffffffff81611966>] sch_direct_xmit+0xe6/0x1c0
[ 2351.275071] [<ffffffff815f47b1>] dev_queue_xmit+0x201/0x4c0
[ 2351.275075] [<ffffffff815fbcfb>] neigh_resolve_output+0x11b/0x220
[ 2351.275079] [<ffffffff8162af51>] ip_finish_output+0x1b1/0x3b0
[ 2351.275083] [<ffffffff8162c448>] ip_output+0x58/0x90
[ 2351.275086] [<ffffffff8162bba5>] ip_local_out+0x25/0x30
[ 2351.275090] [<ffffffff8162befd>] ip_queue_xmit+0x13d/0x3e0
[ 2351.275093] [<ffffffff81642603>] tcp_transmit_skb+0x463/0x8c0
[ 2351.275097] [<ffffffff816453b4>] tcp_send_ack+0xa4/0xf0
[ 2351.275100] [<ffffffff81640781>] tcp_rcv_state_process+0xcf1/0xd00
[ 2351.275103] [<ffffffff81649378>] tcp_v4_do_rcv+0x268/0x470
[ 2351.275106] [<ffffffff8164b497>] tcp_v4_rcv+0x777/0x790
[ 2351.275110] [<ffffffff81626aa0>] ? ip_rcv_finish+0x350/0x350
[ 2351.275113] [<ffffffff816202f4>] ? nf_hook_slow+0x74/0x130
[ 2351.275116] [<ffffffff81626aa0>] ? ip_rcv_finish+0x350/0x350
[ 2351.275119] [<ffffffff81626b54>] ip_local_deliver_finish+0xb4/0x1f0
[ 2351.275122] [<ffffffff81626e28>] ip_local_deliver+0x48/0x80
[ 2351.275125] [<ffffffff816267cd>] ip_rcv_finish+0x7d/0x350
[ 2351.275128] [<ffffffff81627094>] ip_rcv+0x234/0x370
[ 2351.275131] [<ffffffff815f2656>] __netif_receive_skb_core+0x646/0x830
[ 2351.275134] [<ffffffff815f2858>] __netif_receive_skb+0x18/0x60
[ 2351.275137] [<ffffffff815f335d>] process_backlog+0xad/0x1a0
[ 2351.275139] [<ffffffff815f2bfc>] net_rx_action+0x11c/0x230
[ 2351.275143] [<ffffffff81067477>] __do_softirq+0xf7/0x240
[ 2351.275147] [<ffffffff816fa09c>] call_softirq+0x1c/0x30
[ 2351.275151] [<ffffffff81014bf5>] do_softirq+0x55/0x90
[ 2351.275153] [<ffffffff81067755>] irq_exit+0xb5/0xc0
[ 2351.275156] [<ffffffff816fa996>] do_IRQ+0x56/0xc0
[ 2351.275160] [<ffffffff816effed>] common_interrupt+0x6d/0x6d
[ 2351.275161] <EOI> [<ffffffff815a1812>] ? cpuidle_enter_state+0x52/0xc0
[ 2351.275170] [<ffffffff815a1808>] ? cpuidle_enter_state+0x48/0xc0
[ 2351.275173] [<ffffffff815a1949>] cpuidle_idle_call+0xc9/0x210
[ 2351.275177] [<ffffffff8101bafe>] arch_cpu_idle+0xe/0x30
[ 2351.275182] [<ffffffff810b5825>] cpu_startup_entry+0xe5/0x280
[ 2351.275185] [<ffffffff8103f187>] start_secondary+0x217/0x2c0

The problem was fixed by xhanging two kzalloc() calls in rtw_addbareq_cmd()
from GFP_KERNEL to GFP_ATOMIC().

Signed-off-by: Larry Finger <[email protected]>

core/rtw_cmd.c

@@ -1549,14 +1549,14 @@ u8 rtw_addbareq_cmd(struct rtw_adapter*padapter, u8 tid, u8 *addr)
 
 _func_enter_;
 
-	ph2c = (struct cmd_obj *)kzalloc(sizeof(struct cmd_obj), GFP_KERNEL);
+	ph2c = (struct cmd_obj *)kzalloc(sizeof(struct cmd_obj), GFP_ATOMIC);
 	if (!ph2c) {
 		res = _FAIL;
 		goto exit;
 	}
 
 	paddbareq_parm = (struct addBaReq_parm *)
-		kzalloc(sizeof(struct addBaReq_parm), GFP_KERNEL);
+		kzalloc(sizeof(struct addBaReq_parm), GFP_ATOMIC);
 	if (!paddbareq_parm) {
 		kfree(ph2c);
 		res= _FAIL;