[PATCH v4 0/6] Touchscreen performance related fixes

[Vignesh R]

On Monday 24 November 2014 05:21 PM, Sebastian Andrzej Siewior wrote:
> * Vignesh R | 2014-11-14 10:37:25 [+0530]:
> 
>> This series of patches fix TSC defects related to lag in touchscreen
>> performance and cursor jump at touch release. The lag was result of
>> udelay in TSC interrupt handler. Cursor jump due to false pen-up event.
>> The patches implement Advisory 1.0.31 in silicon errata of am335x-evm
> am335x not -evm. The am335x-evm is a board (with its own advisory
> document) built around the SoC.
> 
> Just testing the v4. I can use now IIO and touchscren at the same time.
> back at v1 I reported that it does not work, this has been fixed now.
> I had it running for a few minutes, now I see one of WARN_ON() beeing
> triggered (I've cut a few numbers so don't wonder about PID 2 and so on):
> 
> |dmesg |grep WARNING | wc -l
> |10
> | dmesg |grep WARNING
> |[306.257995] WARNING: CPU: 0 PID: 97 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> |[365.469591] WARNING: CPU: 0 PID: 58 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> |[379.255904] WARNING: CPU: 0 PID: 24 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> |[426.230505] WARNING: CPU: 0 PID: 35 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> |[435.654091] WARNING: CPU: 0 PID: 28 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> |[438.897519] WARNING: CPU: 0 PID: 91 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> |[525.720193] WARNING: CPU: 0 PID: 88 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> |[527.644770] WARNING: CPU: 0 PID: 38 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> |[557.218349] WARNING: CPU: 0 PID: 56 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> |[610.077274] WARNING: CPU: 0 PID: 2 at mfd/ti_am335x_tscadc.c:94 am335x_tsc_se_set_once+0xf8/0x104
> 
> The complete trace:
> 
> |[610.110692] CPU: 0 PID: 4422 Comm: cat Tainted: G        W      3.18.0-rc6+ #1745
> |[610.118577] [<c00138ec>] (unwind_backtrace) from [<c0011544>] (show_stack+0x10/0x14)
> |[610.126772] [<c0011544>] (show_stack) from [<c003c9b0>] (warn_slowpath_common+0x68/0x88)
> |[610.135313] [<c003c9b0>] (warn_slowpath_common) from [<c003c9ec>] (warn_slowpath_null+0x1c/0x24)
> |[610.144586] [<c003c9ec>] (warn_slowpath_null) from [<bf00569c>] (am335x_tsc_se_set_once+0xf8/0x104 [ti_am335x_tscadc])
> |[610.155886] [<bf00569c>] (am335x_tsc_se_set_once [ti_am335x_tscadc]) from [<bf067494>] (tiadc_read_raw+0xbc/0x190 [ti_am335x_adc])
> |[610.168326] [<bf067494>] (tiadc_read_raw [ti_am335x_adc]) from [<bf02dccc>] (iio_read_channel_info+0x9c/0xa4 [industrialio])
> |[610.180191] [<bf02dccc>] (iio_read_channel_info [industrialio]) from [<c02a42d4>] (dev_attr_show+0x1c/0x48)
> |[610.190477] [<c02a42d4>] (dev_attr_show) from [<c013d544>] (sysfs_kf_seq_show+0x8c/0x110)
> |[610.199108] [<c013d544>] (sysfs_kf_seq_show) from [<c013c1c8>] (kernfs_seq_show+0x24/0x28)
> |[610.207833] [<c013c1c8>] (kernfs_seq_show) from [<c0102658>] (seq_read+0x1b4/0x47c)
> |[610.215922] [<c0102658>] (seq_read) from [<c00e6700>] (vfs_read+0x8c/0x148)
> |[610.223269] [<c00e6700>] (vfs_read) from [<c00e67fc>] (SyS_read+0x40/0x8c)
> |[610.230525] [<c00e67fc>] (SyS_read) from [<c000e640>] (ret_fast_syscall+0x0/0x30)
> 
> Could you please look at that one? (I tested it on am335x-evm btw).

I have tried running both IIO and TSC at the same time. But I have never
seen WARN_ON() even after running for close to 30 min. Can you send me
the exact script, so that it will be easy to reproduce?

Regards
Vignesh


> 
> Sebastian
>