Re: [PATCH v5 00/11] serial: 8250: split LPSS to 8250_lpss, enable DMA on Quark

[Bryan O'Donoghue <pure.logic@nexus-software.ie>]

On Tue, 2016-06-07 at 13:04 +0300, Andy Shevchenko wrote:
> > I'll find some time to look at both a short stream and a long
> stream,
> > based on the branch you've indicated and maybe against the latest
> > stable kernel.
> 
> Please, do. I'm planning to resend new version soon (this week or
> lately beginning of the next one).

So.

Generally DMA appears to make little-to-no difference here. With small
data sets, 128 bytes in the test case - there appears to be a slight
performance gain at lower BAUD rates, however for small data-sets and
large data sets 32k for the large test case, there appears to be a
performance penalty at higher BAUD rates.

One thing I have noticed is that the following never completes and
can't be reset back to mode=0 or mode=3 to restore functionality -
definitely worth looking at if you are spinning another patch.

### never completes ####
root@galileo:~# echo 1 > /sys/class/tty/ttyS0/dma_mode
root@galileo:~# stty -F /dev/ttyS0 921600
root@galileo:~# time cat tester.txt > /dev/ttyS0
### never completes ####

Performance data set

root@galileo:~# dmesg > tester.txt
root@galileo:~# ls -las tester.txt 
32 -rw-r--r-- 1 root root 32300
Jan  1  2001 tester.txt


#### With DMA on @ 926100 32k file ####

root@galileo:~# echo 0 > /sys/class/tty/ttyS0/dma_mode
root@galileo:~# stty -F /dev/ttyS0 921600
root@galileo:~# time cat tester.txt > /dev/ttyS0

real    0m0.421s
user    0m0.010s
sys     0m0.000s


### With NoDMA on @ 926100 32k file ####

root@galileo:~# echo 3 > /sys/class/tty/ttyS0/dma_mode
root@galileo:~# stty -F /dev/ttyS0 921600
root@galileo:~# time cat tester.txt > /dev/ttyS0

real    0m0.421s
user    0m0.020s
sys     0m0.000s

#### With DMA on @ 9600 32k file ####
root@galileo:~# echo 0 > /sys/class/tty/ttyS0/dma_mode
root@galileo:~# stty -F /dev/ttyS0 9600  
root@galileo:~# time cat tester.txt > /dev/ttyS0

real    0m34.381s
user    0m0.000s
sys     0m0.010s

#### With NoDMA on @ 9600 32k file #####
root@galileo:~# echo 3 > /sys/class/tty/ttyS0/dma_mode
root@galileo:~# stty -F /dev/ttyS0 9600
root@galileo:~# time cat tester.txt > /dev/ttyS0

real    0m34.391s
user    0m0.010s
sys     0m0.000s


root@galileo:~# ls -las tester_small.txt 
4 -rw-r--r-- 1 root root 128 Jan  1 00:06 tester_small.txt

### With DMA on @ 921600 128 byte test file ###
root@galileo:~# stty -F /dev/ttyS0 921600
root@galileo:~# echo 0 > /sys/class/tty/ttyS0/dma_mode
root@galileo:~# time cat tester_small.txt > /dev/ttyS0

real    0m0.031s
user    0m0.010s
sys     0m0.000s

### With NoDMA on @ 921600 128 byte test file ###
root@galileo:~# stty -F /dev/ttyS0 921600
root@galileo:~# echo 3 > /sys/class/tty/ttyS0/dma_mode
root@galileo:~# time cat tester_small.txt > /dev/ttyS0

real    0m0.013s
user    0m0.010s
sys     0m0.000s

### With DMA on @ 9600 128 byte test file ###
root@galileo:~# stty -F /dev/ttyS0 9600
root@galileo:~# echo 0 > /sys/class/tty/ttyS0/dma_mode
root@galileo:~# time cat tester_small.txt > /dev/ttyS0

real    0m0.161s
user    0m0.000s
sys     0m0.010s

### With NoDMA on @ 9600 128 byte test file ###
root@galileo:~# stty -F /dev/ttyS0 9600
root@galileo:~# echo 3 > /sys/class/tty/ttyS0/dma_mode
root@galileo:~# time cat tester_small.txt > /dev/ttyS0

real    0m0.161s
user    0m0.000s
sys     0m0.000s

---
bod