netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Andrew Lunn <andrew@lunn.ch>
To: "Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>
Cc: netdev <netdev@vger.kernel.org>,
	Lucas Stach <l.stach@pengutronix.de>,
	tyler.baker@linaro.org, fabio.estevam@freescale.com,
	shawn.guo@linaro.org, kernel@pengutronix.de
Subject: Re: [PATCHv6 RFT] net: fec: Ensure clocks are enabled while using mdio bus
Date: Mon, 3 Aug 2015 15:50:23 +0200	[thread overview]
Message-ID: <20150803135023.GC3467@lunn.ch> (raw)
In-Reply-To: <20150803131340.GA9999@pengutronix.de>

On Mon, Aug 03, 2015 at 03:13:40PM +0200, Uwe Kleine-König wrote:
> Hello,
> 
> On Mon, Aug 03, 2015 at 02:55:40PM +0200, Andrew Lunn wrote:
> > v5 was accepted and then reverted, because it broke imx6.  Lucas Stach
> > debugged what was going wrong and provided a patch. Thanks Lucas. This
> > has been squashed into v6.
> This fixed-for-i.MX6 patch was included for 4.2-rc5 as commit
> 8fff755e9f8d0f70a595e79f248695ce6aef5cc3. Unfortunately it breaks
> i.MX27 like so:
> 
> 	Unhandled fault: external abort on non-linefetch (0x808) at 0xf442b014
> 	pgd = c0004000
> 	[f442b014] *pgd=10000452(bad)
> 	Internal error: : 808 [#1] PREEMPT ARM
> 	Modules linked in:
> 	CPU: 0 PID: 5 Comm: kworker/0:0H Not tainted 4.2.0-rc5 #19
> 	Hardware name: Freescale i.MX27 (Device Tree Support)
> 	Workqueue: rpciod xs_tcp_setup_socket
> 	task: c788eb80 ti: c7896000 task.ti: c7896000
> 	PC is at fec_enet_start_xmit+0x39c/0xe04
> 	LR is at fec_enet_start_xmit+0x2e8/0xe04
> 	pc : [<c038ea88>]    lr : [<c038e9d4>]    psr: 80000013
> 	sp : c7897ad8  ip : 00000018  fp : c71637a0
> 	r10: c7b4f800  r9 : c781f000  r8 : 00000000
> 	r7 : 00000001  r6 : c88e5018  r5 : c7b4f800  r4 : c88e5018
> 	r3 : f442b014  r2 : 00000000  r1 : 00000000  r0 : a7164800
> 	Flags: Nzcv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
> 	Control: 0005317f  Table: a0004000  DAC: 00000017
> 	Process kworker/0:0H (pid: 5, stack limit = 0xc7896190)
> 	Stack: (0xc7897ad8 to 0xc7898000)
> 	7ac0:                                                       00000001 00000000
> 	7ae0: c7b4fcc0 c7001fcf 0000cc00 c781e000 00000003 00001800 c7007b37 00000000
> 	7b00: c7a91880 008976fc 000001a0 c88e5010 00000000 0000002a a7164800 000072d8
> 	7b20: 00007210 00b43ef2 00000004 00007eac 00000002 00000200 00004000 c70eed00
> 	7b40: c71637a0 c71637a0 0000002a c0766a2c 00000000 c7b4f800 c0766a34 c04a59a0
> 	7b60: 00000000 00000001 00000000 00000001 00000000 00000000 c7a91880 c7897bbc
> 	7b80: c07c1374 00000000 c71637a0 00000000 00000000 c70eed00 c71637a0 c7a91880
> 	7ba0: c7b4f800 00000000 00000000 00000001 00000000 c04c0004 c71637a0 00000010
> 	7bc0: c71637a0 c70eed00 c7b4f800 00000000 c7a91880 c04a5f60 c70eed68 00000001
> 	7be0: c70ee322 00000000 00000004 fffffff4 c71637a0 c71637a0 c7b4f800 0417a8c0
> 	7c00: f218a8c0 00000003 c7994300 c7296600 c07b66e0 c04faee0 f218a8c0 00000003
> 	7c20: c7994300 c7296600 c07b66e0 c04fb534 f218a8c0 00000000 c7b4bc48 c70ee200
> 	7c40: c7b4f800 0417a8c0 f218a8c0 c04fbc90 f218a8c0 00000000 c7b4bc48 00000000
> 	7c60: c7b4f944 c7292e88 000000a4 c70ee218 c7b4f944 c71638c0 c7163860 c04931a0
> 	7c80: 00000000 c7163860 c70ee200 c7292e88 c70ee218 c7b4f944 00000010 c04ad914
> 	7ca0: 000002c0 c70ee200 00000001 c04b06ac c7292e88 c70ee200 00000000 c70ee200
> 	7cc0: c7b4f944 c04b082c 00000000 c7b4f800 c70fa900 c7b4f800 c70fa900 c7292e88
> 	7ce0: c70ee200 c7b4f944 00000010 c04cfa34 00000000 0417a8c0 c07954b4 c7292e88
> 	7d00: c729f9e0 00000001 00000000 c7b4f800 c70fa900 c7296600 c07b66e0 c04d1d5c
> 	7d20: 00000000 c7ee5bf0 c7891388 00000020 c7292e88 c729f9e0 00000000 c729fbbc
> 	7d40: c7292e88 c729f9e0 00000000 c729fbbc 00a40000 c04cfddc 8d5d04da 00000000
> 	7d60: 8d4fdf3b c7292de0 8d5d04da c729f9e0 c7292e88 00000000 00000000 000005b4
> 	7d80: c729fd60 00000020 c07b66e0 c04e7e64 05b4000b 00000004 00000000 ffff8c9e
> 	7da0: 00000000 00000000 00000000 c729f9e0 c7292de0 c729fdb0 c729fd60 c7897de4
> 	7dc0: 00000000 c07707e0 c07b66e0 c04e9b10 00000001 c7897de4 00000000 c049d4c0
> 	7de0: 9c33a999 00000904 fe103136 c729f9e0 0417a8c0 00000002 c70fa900 c729fbbc
> 	7e00: c729fb78 c70fa900 00006f00 c04ec638 c7801280 c729100c 00000000 00000000
> 	7e20: 000000d0 00020000 00000000 00000000 00000000 6f00e494 f218a8c0 0417a8c0
> 	7e40: 00000000 00000000 00000000 00000000 00000000 00000000 c04e9200 c729f9e0
> 	7e60: 00000800 00000010 c729100c c74052a0 c74052a0 c07c0b7c c076d664 c0501888
> 	7e80: cccccccd 00000000 c729f9e0 c04dabf0 00000000 c0491a78 c07b66e0 00000002
> 	7ea0: c059a4d8 c74052a0 00000800 00000010 c729100c c74052a0 00000000 c07c0b7c
> 	7ec0: c076d664 c0501948 00000800 c729f9e0 c7291304 c05246d0 c7291000 c0526734
> 	7ee0: 00000004 00000001 0000003c 00000003 00000001 0002bf20 c7291304 c7804ca0
> 	7f00: 00000000 c7ee7c00 c076d664 c002e578 c076d664 c005c310 c077025c c7804ca0
> 	7f20: c076d664 c7804cb8 c7896000 c076d674 00000008 c07c0895 c076d664 c002e8f8
> 	7f40: c7896000 c076d7c4 c7804ca0 c781ddc0 00000000 c7804ca0 c002e8bc 00000000
> 	7f60: 00000000 00000000 00000000 c0033af0 0000ffff 00000000 0000ffff c7804ca0
> 	7f80: 00000000 c7897f84 c7897f84 00000000 c7897f90 c7897f90 c7897fac c781ddc0
> 	7fa0: c0033a30 00000000 00000000 c000a340 00000000 00000000 00000000 00000000
> 	7fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
> 	7fe0: 00000000 00000000 00000000 00000000 00000013 00000000 0000ffff 00aeffff
> 	[<c038ea88>] (fec_enet_start_xmit) from [<c04a59a0>] (dev_hard_start_xmit+0x244/0x490)
> 	[<c04a59a0>] (dev_hard_start_xmit) from [<c04c0004>] (sch_direct_xmit+0xe0/0x24c)
> 	[<c04c0004>] (sch_direct_xmit) from [<c04a5f60>] (__dev_queue_xmit+0x258/0x70c)
> 	[<c04a5f60>] (__dev_queue_xmit) from [<c04faee0>] (arp_xmit+0x7c/0x8c)
> 	[<c04faee0>] (arp_xmit) from [<c04fbc90>] (arp_solicit+0xc8/0x1a4)
> 	[<c04fbc90>] (arp_solicit) from [<c04ad914>] (neigh_probe+0x64/0xa0)
> 	[<c04ad914>] (neigh_probe) from [<c04b06ac>] (__neigh_event_send+0x278/0x2e4)
> 	[<c04b06ac>] (__neigh_event_send) from [<c04b082c>] (neigh_resolve_output+0x114/0x19c)
> 	[<c04b082c>] (neigh_resolve_output) from [<c04cfa34>] (ip_finish_output2+0x128/0x3a4)
> 	[<c04cfa34>] (ip_finish_output2) from [<c04d1d5c>] (ip_output+0xf0/0x108)
> 	[<c04d1d5c>] (ip_output) from [<c04cfddc>] (ip_queue_xmit+0x12c/0x370)
> 	[<c04cfddc>] (ip_queue_xmit) from [<c04e7e64>] (tcp_transmit_skb+0x470/0x9c0)
> 	[<c04e7e64>] (tcp_transmit_skb) from [<c04e9b10>] (tcp_connect+0x5c4/0x6f8)
> 	[<c04e9b10>] (tcp_connect) from [<c04ec638>] (tcp_v4_connect+0x27c/0x408)
> 	[<c04ec638>] (tcp_v4_connect) from [<c0501888>] (__inet_stream_connect+0x264/0x2f0)
> 	[<c0501888>] (__inet_stream_connect) from [<c0501948>] (inet_stream_connect+0x34/0x48)
> 	[<c0501948>] (inet_stream_connect) from [<c0526734>] (xs_tcp_setup_socket+0x60/0x444)
> 	[<c0526734>] (xs_tcp_setup_socket) from [<c002e578>] (process_one_work+0x144/0x488)
> 	[<c002e578>] (process_one_work) from [<c002e8f8>] (worker_thread+0x3c/0x520)
> 	[<c002e8f8>] (worker_thread) from [<c0033af0>] (kthread+0xc0/0xdc)
> 	[<c0033af0>] (kthread) from [<c000a340>] (ret_from_fork+0x14/0x34)
> 	Code: 03a02f7b 13a02014 e0833002 e3a02000 (e5832000) 
> 	---[ end trace 9bc5f5f5fa9af0cb ]---
> 	Kernel panic - not syncing: Fatal exception in interrupt
> 
> The problem is that on i.MX27 there are two clocks involved that both
> must be on to send a packet, while on i.MX6 it's only a single one
> (abstracted by having ipg-clock = ahb-clock). With the suggested patch
> only a single one is asserted to be on. This is enough for i.MX6 but
> it's not for i.MX27 (and from looking at the device trees also i.MX25,
> i.MX28, and i.MX35 are affected).

Hi Uwe

I don't think it is as simple as this. If you are sending a packet,
fec_enet_open() must of been called. This does a pm_runtime_get_sync()
to ensure the ipg-clock is ticking, and fec_enet_clk_enable() to
enable all other clocks.

Can you debug this further to find out which clock is off, and where
it gets turned off?

> As I think it would be bold to fix this commit once more for 4.2, I
> suggest to revert 8fff755e9f8d0f70a595e79f248695ce6aef5cc3.
> (To be honest I wonder why v5 of this commit was even taken for -rc3.)
> 
> The straight forward fix for the MDIO issue that was intended to be
> addressed by 8fff755e9f8d is to add clk_prepare_enable and matching
> _unprepare_disable calls to the mdio callbacks, right?

You mean v1 of the patch?

https://patchwork.ozlabs.org/patch/483668/

Or at least something similar. That idea got NACK because it is
thought to consume too much power.

	Andrew

  reply	other threads:[~2015-08-03 13:57 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-07-25 20:38 [PATCHv6 RFT] net: fec: Ensure clocks are enabled while using mdio bus Andrew Lunn
2015-07-26  0:14 ` Fabio Estevam
2015-07-26  1:26 ` Tyler Baker
2015-07-26 16:36 ` Andrew Lunn
2015-07-27  8:22 ` David Miller
2015-08-03 13:13 ` Uwe Kleine-König
2015-08-03 13:50   ` Andrew Lunn [this message]
2015-08-03 15:02     ` Uwe Kleine-König

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20150803135023.GC3467@lunn.ch \
    --to=andrew@lunn.ch \
    --cc=fabio.estevam@freescale.com \
    --cc=kernel@pengutronix.de \
    --cc=l.stach@pengutronix.de \
    --cc=netdev@vger.kernel.org \
    --cc=shawn.guo@linaro.org \
    --cc=tyler.baker@linaro.org \
    --cc=u.kleine-koenig@pengutronix.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).