On Tue, 2009-12-22 at 16:57 +0800, Zhu Yi wrote:
> On Tue, 2009-12-22 at 00:40 +0800, Maxim Levitsky wrote:
> > On Wed, 2009-12-02 at 19:17 +0200, Maxim Levitsky wrote:
> > > On Wed, 2009-12-02 at 13:42 +0800, Zhu Yi wrote:
> > > > On Tue, 2009-12-01 at 17:28 +0800, Zhu Yi wrote:
> > > > > On Tue, 2009-12-01 at 06:35 +0800, Maxim Levitsky wrote:
> > > > > > 0x000000000001668e <iwl3945_rx_reply_tx+302>: lea 0x38(%r8),%rdi
> > > > > > 0x0000000000016692 <iwl3945_rx_reply_tx+306>: lea 0x4f(%r8),%rax
> > > > >
> > > > > When this happened, from your previous post, r8 is 0x0 and rdi is 0x38.
> > > > > Since "info" is %rdi (see below), this means
> > > > > txq->txb[txq->q.read_ptr].skb[0], aka. r8 is 0.
> > > > >
> > > > > > rate_idx = iwl3945_hwrate_to_plcp_idx(tx_resp->rate);
> > > > > >
> > > > > > 0x0000000000016696 <iwl3945_rx_reply_tx+310>: movb $0x0,0x9(%rdi) <---------- RIP
> > > > > > 0x000000000001669a <iwl3945_rx_reply_tx+314>: movb $0x0,0xc(%rdi)
> > > > > > 0x000000000001669e <iwl3945_rx_reply_tx+318>: movb $0x0,0xf(%rdi)
> > > > > > 0x00000000000166a2 <iwl3945_rx_reply_tx+322>: movb $0x0,0x12(%rdi)
> > > > > > 0x00000000000166a6 <iwl3945_rx_reply_tx+326>: movb $0x0,0x15(%rdi)
> > > > >
> > > > > This equals to below code in ieee80211_tx_info_clear_status(). "info" is
> > > > > %rdi, which is 0x38. That matches NULL pointer dereference at 0x41 in
> > > > > your oops header.
> > > > >
> > > > > for (i = 0; i < IEEE80211_TX_MAX_RATES; i++)
> > > > > info->status.rates[i].count = 0;
> > > > >
> > > > > I guess there is a race for txq->q.read_ptr somewhere. Haven't checked
> > > > > though.
> > > >
> > > > OK. 3945 updated write_ptr without regard to read_ptr on the Tx path.
> > > > This messes up our TFD on high load. The patch should fix your problem.
> > > >
> > > > Signed-off-by: Zhu Yi <yi.zhu@xxxxxxxxx>
> > > >
> > > > diff --git a/drivers/net/wireless/iwlwifi/iwl3945-base.c b/drivers/net/wireless/iwlwifi/iwl3945-base.c
> > > > index 994db4a..b31b34c 100644
> > > > --- a/drivers/net/wireless/iwlwifi/iwl3945-base.c
> > > > +++ b/drivers/net/wireless/iwlwifi/iwl3945-base.c
> > > > @@ -548,6 +548,9 @@ static int iwl3945_tx_skb(struct iwl_priv *priv, struct sk_buff *skb)
> > > > txq = &priv->txq[txq_id];
> > > > q = &txq->q;
> > > >
> > > > + if ((iwl_queue_space(q) < q->high_mark))
> > > > + goto drop;
> > > > +
> > > > spin_lock_irqsave(&priv->lock, flags);
> > > >
> > > > idx = get_cmd_index(q, q->write_ptr, 0);
> > > >
> > > >
> > Few days ago, I had an idea to reply here that I am sure that this
> > problem disappeared with this patch.
> >
> >
> > Today I got same kernel panic _with_ the patch applied....
>
> Looks like (all of) the root causes are still not found yet. The symptom
> is exactly the same as the previous one.
>
> One thing I found today is when txq read_ptr catches up to write_ptr
> (read_ptr == write_ptr), iwl_queue_used() will _always_ return TRUE!
> This will be a problem if the firmware sends us a wrong index
> (sequence), then we will fail the check in this condition. I'm not sure
> if firmware can really send us a wrong sequence. Can you please try this
> patch? Apply it on top of the previous one. If you do see the "FIRMWARE
> BUG" in dmesg, then I think we find the root cause.
Finally exploded....
<3>[ 330.761347] iwl3945 0000:06:00.0: Microcode SW error detected. Restarting 0x82000008.
<3>[ 330.761403] iwl3945 0000:06:00.0: Start IWL Error Log Dump:
<3>[ 330.761421] iwl3945 0000:06:00.0: Status: 0x000302E6, count: 1
<3>[ 330.761438] iwl3945 0000:06:00.0: Desc Time asrtPC blink2 ilink1 nmiPC Line
<3>[ 330.761683] iwl3945 0000:06:00.0: SYSASSERT (#5) 1035963230 0x008B6 0x017B8 0x0031C 0x00000 65
<3>[ 330.761692]
<3>[ 330.761837] iwl3945 0000:06:00.0: Start IWL Event Log Dump: display last 20 count
<3>[ 330.761895] iwl3945 0000:06:00.0: 1035963151 0x00000000 1106
<3>[ 330.761931] iwl3945 0000:06:00.0: 1035963155 0x00000000 1106
<3>[ 330.761967] iwl3945 0000:06:00.0: 1035963160 0x00000000 1106
<3>[ 330.762003] iwl3945 0000:06:00.0: 1035963164 0x00000000 1106
<3>[ 330.762039] iwl3945 0000:06:00.0: 1035963168 0x00000000 1106
<3>[ 330.762075] iwl3945 0000:06:00.0: 1035963173 0x00000000 1106
<3>[ 330.762112] iwl3945 0000:06:00.0: 1035963177 0x00000000 1106
<3>[ 330.762148] iwl3945 0000:06:00.0: 1035963181 0x00000000 1106
<3>[ 330.762184] iwl3945 0000:06:00.0: 1035963185 0x00000000 1106
<3>[ 330.762220] iwl3945 0000:06:00.0: 1035963190 0x00000001 1106
<3>[ 330.762256] iwl3945 0000:06:00.0: 1035963190 0x0000000f 1106
<3>[ 330.762292] iwl3945 0000:06:00.0: 1035963192 0x00000417 1105
<3>[ 330.762329] iwl3945 0000:06:00.0: 1035963195 0x000005a5 1115
<3>[ 330.762365] iwl3945 0000:06:00.0: 1035963199 0x00000001 0463
<3>[ 330.762401] iwl3945 0000:06:00.0: 1035963200 0x00000001 0462
<3>[ 330.762437] iwl3945 0000:06:00.0: 1035963212 0x000002b1 0603
<3>[ 330.762474] iwl3945 0000:06:00.0: 1035963217 0x04fb0077 0401
<3>[ 330.762510] iwl3945 0000:06:00.0: 1035963222 0x04dc0018 0401
<3>[ 330.762547] iwl3945 0000:06:00.0: 1035963229 0x00000000 0401
<3>[ 330.762583] iwl3945 0000:06:00.0: 1035963232 0x00000100 0125
<4>[ 330.762623] FIRMWARE BUG: index 220 is given while read_ptr is 252
<0>[ 330.762674] ------------[ cut here ]------------
<2>[ 330.762689] kernel BUG at /home/maxim/software/kernel/linux-2.6/drivers/net/wireless/iwlwifi/iwl-dev.h:717!
<0>[ 330.762715] invalid opcode: 0000 [#1] PREEMPT SMP
<0>[ 330.762735] last sysfs file: /sys/devices/platform/coretemp.1/temp1_input
<4>[ 330.762755] CPU 0
<4>[ 330.762765] Modules linked in: nvidia(P) af_packet nfsd exportfs nfs lockd nfs_acl auth_rpcgss sunrpc usb_storage usb_libusual cpufreq_powersave cpufreq_conservative snd_hda_codec_realtek cpufreq_userspace snd_hda_intel acpi_cpufreq uvcvideo iwl3945 snd_hda_codec coretemp videodev iwlcore v4l1_compat snd_hwdep joydev ohci1394 v4l2_compat_ioctl32 tg3 snd_pcm sbp2 mac80211 uhci_hcd psmouse video snd_page_alloc ieee1394 libphy ehci_hcd output ac battery lirc_ene0100 serio_raw evdev cfg80211 rfkill usbcore fuse lzo lzo_decompress lzo_compress
<6>[ 330.762987] Pid: 0, comm: swapper Tainted: P 2.6.32-wl #243 Aspire 5720
<6>[ 330.763011] RIP: 0010:[<ffffffffa01c3fe7>] [<ffffffffa01c3fe7>] iwl_tx_cmd_complete+0x397/0x3a0 [iwlcore]
<6>[ 330.763057] RSP: 0018:ffff880002203d20 EFLAGS: 00010286
<6>[ 330.763075] RAX: 0000000000000049 RBX: ffff88006fc09b40 RCX: 0000000000000102
<6>[ 330.763096] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000001
<6>[ 330.763118] RBP: ffff880002203d90 R08: 0000000000000002 R09: 0000000000000000
<6>[ 330.763139] R10: 0000000000000000 R11: 0000000000000001 R12: 00000000000000fc
<6>[ 330.763160] R13: 00000000000000dc R14: ffff88006fd6d460 R15: ffff88006a2a1600
<6>[ 330.763182] FS: 0000000000000000(0000) GS:ffff880002200000(0000) knlGS:0000000000000000
<6>[ 330.763206] CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b
<6>[ 330.763224] CR2: 00007f7ffbf95000 CR3: 0000000001001000 CR4: 00000000000006f0
<6>[ 330.763246] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
<6>[ 330.763267] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
<4>[ 330.763289] Process swapper (pid: 0, threadinfo ffffffff815ca000, task ffffffff815ceb60)
<0>[ 330.763310] Stack:
<4>[ 330.763318] ffff88006a2a9690 0000000000000046 ffff880002203d40 0000000000000046
<4>[ 330.763346] <0> 0000000000000282 0000000000000282 0000000000000282 ffff8800699eba00
<4>[ 330.763375] <0> ffff880002203d70 ffff88006a2a1600 ffff88006a2a2f10 ffff88006a673000
<0>[ 330.763406] Call Trace:
<0>[ 330.763416] <IRQ>
<4>[ 330.763436] [<ffffffffa0230d13>] iwl3945_irq_tasklet+0x673/0x1740 [iwl3945]
<4>[ 330.763469] [<ffffffffa01b683e>] ? iwl_isr_legacy+0x3e/0x600 [iwlcore]
<4>[ 330.763496] [<ffffffff81053f56>] tasklet_action+0x106/0x110
<4>[ 330.763516] [<ffffffff81055674>] __do_softirq+0x114/0x2c0
<4>[ 330.763536] [<ffffffff8109eb38>] ? handle_edge_irq+0x78/0x160
<4>[ 330.763558] [<ffffffff8100d25c>] call_softirq+0x1c/0x30
<4>[ 330.763577] [<ffffffff8100f1fd>] do_softirq+0x7d/0xb0
<4>[ 330.763595] [<ffffffff81055125>] irq_exit+0x95/0xa0
<4>[ 330.763615] [<ffffffff8140bddc>] do_IRQ+0x7c/0xf0
<4>[ 330.763634] [<ffffffff8100ca13>] ret_from_intr+0x0/0xf
<0>[ 330.763650] <EOI>
<4>[ 330.763664] [<ffffffff812a1d5b>] ? acpi_idle_enter_bm+0x275/0x2aa
<4>[ 330.763684] [<ffffffff812a1d65>] ? acpi_idle_enter_bm+0x27f/0x2aa
<4>[ 330.763705] [<ffffffff812a1d5b>] ? acpi_idle_enter_bm+0x275/0x2aa
<4>[ 330.763729] [<ffffffff8133227f>] ? cpuidle_idle_call+0x9f/0x160
<4>[ 330.763749] [<ffffffff8100b1df>] ? cpu_idle+0xaf/0x110
<4>[ 330.763769] [<ffffffff813f534a>] ? rest_init+0x7a/0x80
<4>[ 330.763789] [<ffffffff81640d01>] ? start_kernel+0x3ac/0x3b8
<4>[ 330.763808] [<ffffffff81640315>] ? x86_64_start_reservations+0x125/0x129
<4>[ 330.763831] [<ffffffff816403fd>] ? x86_64_start_kernel+0xe4/0xeb
<0>[ 330.763848] Code: c7 c6 48 86 1d a0 48 c7 c7 53 86 1d a0 e8 a2 b3 07 e1 e9 68 fe ff ff 44 89 e2 44 89 ee 48 c7 c7 80 5a 1d a0 31 c0 e8 c7 31 24 e1 <0f> 0b eb fe 0f 1f 44 00 00 55 48 89 e5 41 57 41 56 41 55 41 54
<1>[ 330.764008] RIP [<ffffffffa01c3fe7>] iwl_tx_cmd_complete+0x397/0x3a0 [iwlcore]
<4>[ 330.764042] RSP <ffff880002203d20>
<4>[ 330.764059] ---[ end trace e386a588d450d76f ]---
<0>[ 330.764073] Kernel panic - not syncing: Fatal exception in interrupt
<4>[ 330.764093] Pid: 0, comm: swapper Tainted: P D 2.6.32-wl #243
<4>[ 330.764111] Call Trace:
<4>[ 330.764120] <IRQ> [<ffffffff814070f1>] panic+0x82/0x13f
<4>[ 330.764148] [<ffffffff81010a52>] oops_end+0xe2/0xf0
<4>[ 330.764166] [<ffffffff81010c7b>] die+0x5b/0x90
<4>[ 330.764183] [<ffffffff8100db24>] do_trap+0xc4/0x170
<4>[ 330.764201] [<ffffffff8100e025>] do_invalid_op+0x95/0xb0
<4>[ 330.764230] [<ffffffffa01c3fe7>] ? iwl_tx_cmd_complete+0x397/0x3a0 [iwlcore]
<4>[ 330.764254] [<ffffffff8140a94d>] ? trace_hardirqs_off_thunk+0x3a/0x3c
<4>[ 330.764276] [<ffffffff8100cef5>] invalid_op+0x15/0x20
<4>[ 330.764304] [<ffffffffa01c3fe7>] ? iwl_tx_cmd_complete+0x397/0x3a0 [iwlcore]
<4>[ 330.764333] [<ffffffffa0230d13>] iwl3945_irq_tasklet+0x673/0x1740 [iwl3945]
<4>[ 330.764364] [<ffffffffa01b683e>] ? iwl_isr_legacy+0x3e/0x600 [iwlcore]
<4>[ 330.764388] [<ffffffff81053f56>] tasklet_action+0x106/0x110
<4>[ 330.764408] [<ffffffff81055674>] __do_softirq+0x114/0x2c0
<4>[ 330.764427] [<ffffffff8109eb38>] ? handle_edge_irq+0x78/0x160
<4>[ 330.764447] [<ffffffff8100d25c>] call_softirq+0x1c/0x30
<4>[ 330.764465] [<ffffffff8100f1fd>] do_softirq+0x7d/0xb0
<4>[ 330.764483] [<ffffffff81055125>] irq_exit+0x95/0xa0
<4>[ 330.764500] [<ffffffff8140bddc>] do_IRQ+0x7c/0xf0
<4>[ 330.764519] [<ffffffff8100ca13>] ret_from_intr+0x0/0xf
<4>[ 330.764534] <EOI> [<ffffffff812a1d5b>] ? acpi_idle_enter_bm+0x275/0x2aa
<4>[ 330.764562] [<ffffffff812a1d65>] ? acpi_idle_enter_bm+0x27f/0x2aa
<4>[ 330.764583] [<ffffffff812a1d5b>] ? acpi_idle_enter_bm+0x275/0x2aa
<4>[ 330.764605] [<ffffffff8133227f>] ? cpuidle_idle_call+0x9f/0x160
<4>[ 330.764625] [<ffffffff8100b1df>] ? cpu_idle+0xaf/0x110
<4>[ 330.764643] [<ffffffff813f534a>] ? rest_init+0x7a/0x80
<4>[ 330.764662] [<ffffffff81640d01>] ? start_kernel+0x3ac/0x3b8
<4>[ 330.764682] [<ffffffff81640315>] ? x86_64_start_reservations+0x125/0x129
<4>[ 330.764703] [<ffffffff816403fd>] ? x86_64_start_kernel+0xe4/0xeb
Note that these days I see the 'Microcode SW error detected' somewhat more often that I used to.
This is the combined patch I used:
diff --git a/drivers/net/wireless/iwlwifi/iwl-dev.h b/drivers/net/wireless/iwlwifi/iwl-dev.h
index 2673e9a..8efc469 100644
--- a/drivers/net/wireless/iwlwifi/iwl-dev.h
+++ b/drivers/net/wireless/iwlwifi/iwl-dev.h
@@ -711,7 +711,13 @@ extern void iwl_txq_ctx_stop(struct iwl_priv *priv);
extern int iwl_queue_space(const struct iwl_queue *q);
static inline int iwl_queue_used(const struct iwl_queue *q, int i)
{
- return q->write_ptr > q->read_ptr ?
+ if (q->write_ptr == q->read_ptr) {
+ printk("FIRMWARE BUG: index %d is given while read_ptr is %d\n",
+ i, q->read_ptr);
+ BUG();
+ }
+
+ return q->write_ptr >= q->read_ptr ?
(i >= q->read_ptr && i < q->write_ptr) :
!(i < q->read_ptr && i >= q->write_ptr);
}
diff --git a/drivers/net/wireless/iwlwifi/iwl3945-base.c b/drivers/net/wireless/iwlwifi/iwl3945-base.c
index 2a28a1f..a36de73 100644
--- a/drivers/net/wireless/iwlwifi/iwl3945-base.c
+++ b/drivers/net/wireless/iwlwifi/iwl3945-base.c
@@ -548,6 +548,9 @@ static int iwl3945_tx_skb(struct iwl_priv *priv, struct sk_buff *skb)
txq = &priv->txq[txq_id];
q = &txq->q;
+ if ((iwl_queue_space(q) < q->high_mark))
+ goto drop;
+
spin_lock_irqsave(&priv->lock, flags);
idx = get_cmd_index(q, q->write_ptr, 0);
Best regards,
Maxim Levitsky
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
