Re: [PATCH] spi: expose spi_master and spi_device statistics via sysfs

[Martin Sperl <kernel@xxxxxxxxxxxxxxxx>]

> On 19.06.2015, at 11:53, Mark Brown <broonie@xxxxxxxxxx> wrote:
> 
> On Fri, Jun 19, 2015 at 11:35:39AM +0200, Jakub Kiciński wrote:
> 
>> How about putting this file in debugfs after all?  Mark suggested it
>> initially but the conversation went in other direction.  Martin said
>> that his main purpose is to help *debugging* performance of RPi dma,
>> so it's a perfect candidate for debugfs.  I can't in my mind see any
>> use for this information on a production system to be honest...
> 
> That's the solution I'd been expecting.

I got a patch in the meantime exposing the following files in
/sys/devices/platform/soc/20204000.spi/spi_master/spi32766/statistics/:
bytes
bytes_rx
bytes_tx
bytes_transfer_histo_0-1
bytes_transfer_histo_1024-2047
bytes_transfer_histo_128-255
bytes_transfer_histo_16-31
bytes_transfer_histo_16384-32767
bytes_transfer_histo_2048-4095
bytes_transfer_histo_2-3
bytes_transfer_histo_256-511
bytes_transfer_histo_32-63
bytes_transfer_histo_32768-65535
bytes_transfer_histo_4096-8191
bytes_transfer_histo_4-7
bytes_transfer_histo_512-1023
bytes_transfer_histo_64-127
bytes_transfer_histo_65536+
bytes_transfer_histo_8-15
bytes_transfer_histo_8192-16383
errors
messages
spi_async
spi_sync
spi_sync_immediate
timedout
transfers

You want me to move the bytes_transfer_histo_* now into debugfs instead?

If so then I would post a “basic” patch for the other statistics
and have debugfs for the histogram as a separate patch.

Martin--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html