[PATCH v2] spi: atmel: Refactor spi-atmel to use SPI framework queue

[Måns Rullgård]

"Yang, Wenyou" <Wenyou.Yang at atmel.com> writes:

> Hello Måns,
>
> Sorry for late answer.
>
>> -----Original Message-----
>> From: Måns Rullgård [mailto:mans at mansr.com]
>> Sent: 2016年1月6日 3:28
>> To: Yang, Wenyou <Wenyou.Yang at atmel.com>
>> Cc: broonie at kernel.org; richard.genoud at gmail.com; Ferre, Nicolas
>> <Nicolas.FERRE at atmel.com>; linux-kernel at vger.kernel.org; linux-
>> spi at vger.kernel.org; linux-arm-kernel at lists.infradead.org
>> Subject: Re: [PATCH v2] spi: atmel: Refactor spi-atmel to use SPI framework
>> queue
>> 
>> Wenyou Yang <wenyou.yang at atmel.com> writes:
>> 
>> > Replace the deprecated master->transfer with transfer_one_message()
>> > and allow the SPI subsystem handle all the queuing of messages.
>> >
>> > Signed-off-by: Wenyou Yang <wenyou.yang at atmel.com>
>> > Tested-by: Richard Genoud <richard.genoud at gmail.com>
>> > ---
>> > Hi Mark,
>> >
>> > Thanks for your feedback.
>> >
>> > According to your advice, I prepared this version 2.
>> >
>> > The patch is based on the for-next branch of
>> > 	git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi.git
>> >
>> > I tested it on sama5d3xek and at91sam9m10g45ek board.
>> > Richard helped tested it on at91sam9g35-ek board.
>> >
>> > Best Regards,
>> > Wenyou Yang
>> >
>> > v2 changelog:
>> >  1./ Rebase the latest for-next branch included the patch from Richard.
>> >  2./ remove the lock on complete(&as->xfer_completion).
>> >  3./ rework the xfer->cs_change is true.
>> >  4./ remove the member stopping of struct atmel_spi.
>> >
>> >  drivers/spi/spi-atmel.c |  678
>> > +++++++++++++++--------------------------------
>> >  1 file changed, 220 insertions(+), 458 deletions(-)
>> 
>> This patch broke the display control on the AT32STK1000 board.  I don't know
>> which part of the patch is causing the problem.  This shows up in the kernel log
>> (with debug messages enabled):
>> 
>> atmel_spi atmel_spi.0: version: 0x171
>> atmel_spi atmel_spi.0: Atmel SPI Controller at 0xffe00000 (irq 3) spi spi0.0: setup:
>> bpw 8 mode 0x1 -> csr0 00000008 spi spi0.1: setup: bpw 8 mode 0x3 -> csr1
>> 00000009 ltv350qv spi0.1: new message 93c29cb4 submitted for spi0.1 ltv350qv
>> spi0.1: activate 1, mr 000e0011
>> ltv350qv spi0.1:   start xfer 93c29c78: len 3 tx 93c69ccc/13c69ccc rx   (null)/ffffffff
>> ------------[ cut here ]------------
>> WARNING: at drivers/spi/spi-atmel.c:1351
>> CPU: 0 PID: 1 Comm: swapper Not tainted 4.4.0-rc8+ #14
>> task: 93c1e000 ti: 93c28000 task.ti: 93c28000 PC is at
>> atmel_spi_transfer_one_message+0x33a/0x72c
>> LR is at schedule_timeout+0x72/0xc0
>> pc : [<90139a26>]    lr : [<901d44b6>]    Not tainted
>> sp : 93c29b2c  r12: 00000000  r11: 93d1adb8
>> r10: 93c1e364  r9 : 00000200  r8 : 00400022
>> r7 : 93d12dfc  r6 : 93d12dfc  r5 : 93c29c78  r4 : 93c29bb8
>> r3 : 93c29cb4  r2 : 93d12c00  r1 : 00000000  r0 : 00000000
>> Flags: qvnZc
>> Mode bits: hjmde....G
>> CPU Mode: Supervisor
>> Call trace:
>>  [<900ebfea>] ida_get_new_above+0x5a/0x148  [<900ec07e>]
>> ida_get_new_above+0xee/0x148  [<900efa96>] put_dec+0xa2/0xbc  [<900efdea>]
>> number+0x2e6/0x2f4  [<90137e6a>] __spi_pump_messages+0x21e/0x304
>> [<90137910>] spi_queued_transfer+0x0/0x8  [<90138046>] __spi_sync+0xf6/0xf8
>> [<90137910>] spi_queued_transfer+0x0/0x8  [<90138056>] spi_sync+0x6/0x8
>> [<9010a39e>] ltv350qv_write_reg+0x100/0x106  [<90080000>]
>> __generic_block_fiemap+0x78/0x250  [<90138150>] spi_complete+0x0/0x8
>> [<9010a3cc>] ltv350qv_power+0x28/0x1e4  [<9010a5f6>]
>> ltv350qv_probe+0x4e/0x68  [<90136b68>] spi_drv_probe+0x20/0x22
>> [<901336e6>] driver_probe_device+0xe6/0x188  [<90133826>]
>> __device_attach_driver+0x4a/0x54  [<901327ae>] bus_for_each_drv+0x32/0x54
>> [<901337dc>] __device_attach_driver+0x0/0x54  [<901338a2>]
>> __device_attach+0x5a/0x7c  [<901338ca>] device_initial_probe+0x6/0x8
>> [<90132e2c>] bus_probe_device+0x58/0x5c  [<90131d94>]
>> device_add+0x1fc/0x344  [<9013823c>] spi_setup+0x88/0xf8  [<90139e18>]
>> atmel_spi_setup+0x0/0x160  [<901386e8>] spi_add_device+0x70/0xf0
>> [<9013881e>] spi_new_device+0x56/0x6c  [<9013884e>]
>> spi_match_master_to_boardinfo+0x1a/0x30
>>  [<90138938>] spi_register_master+0xd4/0x22c  [<90138aac>]
>> devm_spi_register_master+0x1c/0x44
>>  [<90138faa>] atmel_spi_probe+0x29a/0x38c  [<900b06c0>]
>> sysfs_do_create_link_sd+0x34/0x60  [<90134430>]
>> platform_drv_probe+0x14/0x40  [<901336e6>] driver_probe_device+0xe6/0x188
>> [<900ecbca>] kobject_add_internal+0x56/0x1c8  [<901337d8>]
>> __driver_attach+0x50/0x54  [<90000418>] repair_env_string+0x0/0x58
>> [<9013280e>] bus_for_each_dev+0x3e/0x58  [<90133788>]
>> __driver_attach+0x0/0x54  [<90133508>] driver_attach+0x10/0x14  [<90132cb2>]
>> bus_add_driver+0xf6/0x15c  [<90133aae>] driver_register+0x3e/0x98
>> [<901343b6>] __platform_driver_register+0x22/0x28
>>  [<9000aba8>] atmel_spi_driver_init+0x0/0x10  [<9000abb2>]
>> atmel_spi_driver_init+0xa/0x10  [<9000023e>] do_one_initcall+0xaa/0x130
>> [<90027ae2>] parse_args+0x19e/0x27c  [<90000418>]
>> repair_env_string+0x0/0x58  [<90000374>] kernel_init_freeable+0x88/0x12c
>> [<90000382>] kernel_init_freeable+0x96/0x12c  [<90000418>]
>> repair_env_string+0x0/0x58  [<901d2588>] kernel_init+0x8/0xa4  [<90013130>]
>> syscall_return+0x0/0x12  [<901d2580>] kernel_init+0x0/0xa4  [<90013130>]
>> syscall_return+0x0/0x12
>> 
>> ---[ end trace 6b6391f8ee24f957 ]---
>> ltv350qv spi0.1: spi transfer timeout
>> atmel_spi atmel_spi.0: overrun (2/3 remaining) atmel_spi atmel_spi.0: timeout
>> waiting for TXEMPTY
>
> I tried to reproduce this issue, but I failed.
>
> Could you send me your .config file? Thank you.

I can do better yet.  I just sent a patch for the issue.

-- 
Måns Rullgård