On Thu, Nov 5, 2015 at 3:47 PM, Robert Nelson <[email protected]> wrote:
> The journey into spi & dma..
>
> So for awhile now, we've had a bug for spi users, where the system
> would hardlock at the "160" byte threashold:
>
> debian@beaglebone:~$ sudo dd if=/dev/zero of=/dev/spidev1.0 bs=159 count=1
> 1+0 records in
> 1+0 records out
> 159 bytes (159 B) copied, 0.000508833 s, 312 kB/s
> debian@beaglebone:/sys$ dd if=/dev/zero of=/dev/spidev1.0 bs=160 count=1
> ^C
>
> (hangs..)
>
> "160" happens to be a the magic number where DMA takes over from PIO mode...
>
> Well as of yesterday:
>
> debian@beaglebone:~$ sudo dd if=/dev/zero of=/dev/spidev1.0 bs=320 count=1
> 1+0 records in
> 1+0 records out
> 320 bytes (320 B) copied, 0.00266366 s, 120 kB/s
>
> So here's the deal, it looks to be a spi/overlay bug, as we don't seem
> to get the correct dma tx/rx channels..
>
> What interesting, when the "spi" node is enabled in the main dtb:
>
> https://github.com/RobertCNelson/dtb-rebuilder/blob/4.1-ti/src/arm/am335x-boneblack-spi0.dts#L120-L137
>
> it works fine in dma mode...
>
> So with "4.1.12-ti-r27" spi-dma is now disabled..
>
> *btw:
>
> Late last night i wrote a better fix (will be r28)
>
> where you can disable "spi-dma" via the dt option:
>
> ti,pio-mode;
>
> https://github.com/beagleboard/bb.org-overlays/commit/48d33d4d22f284103db83626b343724cf18c578d
>
> Regards,
>
It is very long time ago, but now I am running into performance issues
with SPI. I see the DMA is still disabled in current bb.org-overlays.
So is the bug for blocks > 160 still there (what about kernel 4.4.x?)

Thanks in advance

Andreas

-- 
For more options, visit http://beagleboard.org/discuss
--- 
You received this message because you are subscribed to the Google Groups 
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/beagleboard/CALbNGRSqXpYEd1p2bMKngZKZz-mGZOshv%3D3Os1WcdscvAh5dtg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to