Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-05 Thread Johan Hovold
On Sat, May 04, 2013 at 07:39:57AM -0400, Peter Hurley wrote: > On 05/04/2013 07:15 AM, Johan Hovold wrote: > > On Sat, May 04, 2013 at 01:50:42AM +0400, Stas Sergeev wrote: > >> 04.05.2013 00:34, Greg KH пишет: > >>> On Fri, May 03, 2013 at 10:27:18PM +0400, Stas Sergeev wrote: > 03.05.2013 2

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-04 Thread Peter Hurley
On 05/04/2013 07:15 AM, Johan Hovold wrote: On Sat, May 04, 2013 at 01:50:42AM +0400, Stas Sergeev wrote: 04.05.2013 00:34, Greg KH пишет: On Fri, May 03, 2013 at 10:27:18PM +0400, Stas Sergeev wrote: 03.05.2013 21:16, Greg KH пишет: [...] There's no guarantee as to how long select or an i

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-04 Thread Stas Sergeev
04.05.2013 15:15, Johan Hovold пишет: The query takes longer than the transmit at decent baudrates (>=38k) and under the assumption that flow control isn't causing any delays. But you do have a point, and I have been meaning to look into whether the added overhead of checking the hardware buffer

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-04 Thread Johan Hovold
On Sat, May 04, 2013 at 01:50:42AM +0400, Stas Sergeev wrote: > 04.05.2013 00:34, Greg KH пишет: > > On Fri, May 03, 2013 at 10:27:18PM +0400, Stas Sergeev wrote: > >> 03.05.2013 21:16, Greg KH пишет: [...] > >>> There's no guarantee as to how long select or an ioctl will take, and > >>> now that

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-04 Thread Stas Sergeev
04.05.2013 00:34, Greg KH пишет: Don't call select for every character :) OK, let me draw an approximate workflow of the setup in question. Lets say we have 3 tty devices, A, B and C. A and B are blazingly fast, while C is a casual usb-serial chip. The app must establish a bidirectional relay be

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-03 Thread Stas Sergeev
04.05.2013 00:34, Greg KH пишет: On Fri, May 03, 2013 at 10:27:18PM +0400, Stas Sergeev wrote: 03.05.2013 21:16, Greg KH пишет: Sounds like an application is doing a foolish thing and should stop it. Its not. The app is quering only for _input_ (specifying only read fds to select). But the se

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-03 Thread Greg KH
On Fri, May 03, 2013 at 10:27:18PM +0400, Stas Sergeev wrote: > 03.05.2013 21:16, Greg KH пишет: > > >Sounds like an application is doing a foolish thing and should stop it. > Its not. > The app is quering only for _input_ (specifying only read fds > to select). But the select() in linux is implem

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-03 Thread Stas Sergeev
03.05.2013 21:16, Greg KH пишет: Sounds like an application is doing a foolish thing and should stop it. Its not. The app is quering only for _input_ (specifying only read fds to select). But the select() in linux is implemented the way that even when it polls for input, it will still call tty_

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-03 Thread Stas Sergeev
03.05.2013 20:52, Greg KH пишет: What exactly is the "problem" being seen? OK, the problem is that while select() "stalls", the entire output queue is transmitted, and when the app writes more, there is already a big pause. In fact, the app calls select() before writing every char, so then the d

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-03 Thread Greg KH
On Fri, May 03, 2013 at 10:05:55PM +0400, Stas Sergeev wrote: > 03.05.2013 20:52, Greg KH пишет: > >On Fri, May 03, 2013 at 09:38:50PM +0400, Stas Sergeev wrote: > >>03.05.2013 20:30, Greg KH пишет: > >>>We need some way to check the chars in the buffer, is the device you are > >>>using just very s

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-03 Thread Stas Sergeev
03.05.2013 20:52, Greg KH пишет: On Fri, May 03, 2013 at 09:38:50PM +0400, Stas Sergeev wrote: 03.05.2013 20:30, Greg KH пишет: We need some way to check the chars in the buffer, is the device you are using just very slow to respond to this request? How slow? Do you have a test case that we c

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-03 Thread Greg KH
On Fri, May 03, 2013 at 09:38:50PM +0400, Stas Sergeev wrote: > 03.05.2013 20:30, Greg KH пишет: > >We need some way to check the chars in the buffer, is the device you are > >using just very slow to respond to this request? How slow? Do you have > >a test case that we can see how it is affected?

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-03 Thread Stas Sergeev
03.05.2013 20:30, Greg KH пишет: We need some way to check the chars in the buffer, is the device you are using just very slow to respond to this request? How slow? Do you have a test case that we can see how it is affected? Greg, unfortunately, I do have nothing. The customer is in CC list, s

Re: Regression: ftdi_sio is slow (since Wed Oct 10 15:05:06 2012)

2013-05-03 Thread Greg KH
On Fri, May 03, 2013 at 07:02:46PM +0400, Stas Sergeev wrote: > Hi. > > We have a regression because of this patch: > http://lkml.indiana.edu/hypermail/linux/kernel/1210.1/01456.html > While it is arguably reasonable to have this for tcdrain or close, > it also slows down poll/select a lot because