Re: ftdi_write_data unstable on 2.6 Works perfect on 2.4

Sorry,

looks like I spoke too soon. The first write issue appears to have
nothing to do with the ftdi timeout values. If you want to experiment
anyway, just set the timeouts directly on your app's ftdi handle:

ftdic.usb_write_timeout = usecs;
ftdic.usb_read_timeout = usecs;

where usecs is some positive integer number of microseconds. The
default value is 5000. Not sure what a reasonable value is. In
libusb0.1 transfers are polled every 1ms.

Nath.

2009/7/31 J.C. Wren <jcwren <at> jcwren.com>:
> What's the correct way to set the timeout, what are the units, and what
> would be a reasonable value?

--
libftdi - see http://www.intra2net.com/en/developer/libftdi for details.
To unsubscribe send a mail to libftdi+unsubscribe <at> developer.intra2net.com   

About FT2232

Re: About FT2232

Hi wangqiang,

> I am using the ft2232 library on windows which is  provided by FTDI.
> My question is that is libftdi support the windows.

FTDI provides the D2XX drivers for windows and they support the FT2232H. 
libftdi is developed to work under linux. We have no experience using it under 
windows.

Another user, Uwe Bonnes, tried to use libftdi under windows some time ago 
using mingw. But I don't know if he is still using it and if there were any 
problems.

> I will use the FT2232H chip, so have the libftdi support this new chip?

The FT2232H is supported in current libftdi releases.

Kind regards,

Gerd

--
libftdi - see http://www.intra2net.com/en/developer/libftdi for details.
To unsubscribe send a mail to libftdi+unsubscribe <at> developer.intra2net.com   

RE: About FT2232

>> We have no experience using it under windows.

I can't speak for the "H" part, but the "FT2232D" part works just fine under
Win32 with MSVC.  I also built the same libftdi source for Linux with gcc.
I use direct/UART mode ("direct" meaning *not* VCP), as well as MPSSE/SPI
mode.  My team will likely move to the "H" part in a few months, so I'll
know more then.  One caveat is that Device Manager likes to give the device
to d2xx, and it takes some coaxing to make it take libusb-win32 (I sometimes
have to "update driver" a second time after the first unplug/replug...but
not always).

Michael

-----Original Message-----
From: Gerd v. Egidy [mailto:gerd.von.egidy <at> intra2net.com]
Sent: Wednesday, August 05, 2009 8:31 AM
To: libftdi <at> developer.intra2net.com
Cc: Qiang Wang
Subject: Re: About FT2232

Hi wangqiang,

> I am using the ft2232 library on windows which is  provided by FTDI.
> My question is that is libftdi support the windows.

FTDI provides the D2XX drivers for windows and they support the FT2232H.
libftdi is developed to work under linux. We have no experience using it
under
windows.

Another user, Uwe Bonnes, tried to use libftdi under windows some time ago
using mingw. But I don't know if he is still using it and if there were any
problems.

> I will use the FT2232H chip, so have the libftdi support this new chip?

The FT2232H is supported in current libftdi releases.

Kind regards,

Gerd

--
libftdi - see http://www.intra2net.com/en/developer/libftdi for details.
To unsubscribe send a mail to libftdi+unsubscribe <at> developer.intra2net.com

--
libftdi - see http://www.intra2net.com/en/developer/libftdi for details.
To unsubscribe send a mail to libftdi+unsubscribe <at> developer.intra2net.com   

Re: About FT2232

On Wed, Aug 5, 2009 at 9:45 PM, Michael Plante<michael.plante <at> gmail.com> wrote:
>>> We have no experience using it under windows.
>
> I can't speak for the "H" part, but the "FT2232D" part works just fine under
> Win32 with MSVC.  I also built the same libftdi source for Linux with gcc.
> I use direct/UART mode ("direct" meaning *not* VCP), as well as MPSSE/SPI
> mode.  My team will likely move to the "H" part in a few months, so I'll
> know more then.  One caveat is that Device Manager likes to give the device
> to d2xx, and it takes some coaxing to make it take libusb-win32 (I sometimes
> have to "update driver" a second time after the first unplug/replug...but
> not always).

According to the reports from OpenOCD 0.2.0 mailing list, libftdi
works fine under Windows. In fact the GPL license prevented new
binaries based on FTD2xx from being publicly distributed, so one can
only distribute libftdi based OpenOCD binary. Not so sure about
the "H" part, but it might work as well.

The performance of libftdi under Winodws still lags behind ftd2xx
even though the performance of libftdi under Linux (with async
mode) matches the performance of ftd2xx.
https://lists.berlios.de/pipermail/openocd-development/2009-June/008193.html
https://lists.berlios.de/pipermail/openocd-development/2009-June/008846.html

libusb-win32 actually supports async mode, so does libusb-1.0.
Right now the async mode of libftdi seems to directly manipulate
sysfs and it is only for Linux.

--

-- 
Xiaofan http://mcuee.blogspot.com

--
libftdi - see http://www.intra2net.com/en/developer/libftdi for details.
To unsubscribe send a mail to libftdi+unsubscribe <at> developer.intra2net.com   

Re: About FT2232

> > I can't speak for the "H" part, but the "FT2232D" part works just fine
> > under Win32 with MSVC.  I also built the same libftdi source for Linux
> > with gcc. I use direct/UART mode ("direct" meaning *not* VCP), as well as
> > MPSSE/SPI mode.  My team will likely move to the "H" part in a few
> > months, so I'll know more then.

Ok, nice to know that it works.

> > One caveat is that Device Manager likes
> > to give the device to d2xx, and it takes some coaxing to make it take
> > libusb-win32 (I sometimes have to "update driver" a second time after the
> > first unplug/replug...but not always).

Do you use the default vendor-/product-id? Maybe it would help if you use a 
different product id that the d2xx-driver doesn't know about.

You can ask FTDIchip, they will usually provide you with a unique product id 
out of their range for free.

> libusb-win32 actually supports async mode, so does libusb-1.0.
> Right now the async mode of libftdi seems to directly manipulate
> sysfs and it is only for Linux.

The async-mode of libftdi is an evil hack which I wrote when libusb-1.0 wasn't 
available yet. It would be nice to replace it with something based upon 
libusb-1.0. I currently don't have the time to do that, but if someone wants 
to work on it we could certainly provide some help and feedback. 

Actually Jie Zhang wrote that he is working on it some days ago, you can 
search for his post the mailinglist-archive.

Kind regards,

Gerd

--
libftdi - see http://www.intra2net.com/en/developer/libftdi for details.
To unsubscribe send a mail to libftdi+unsubscribe <at> developer.intra2net.com   

RE: About FT2232

>> Do you use the default vendor-/product-id? Maybe it would help if you use
a
>> different product id that the d2xx-driver doesn't know about.

I do.  I'll look into that if my group starts complaining, but, in the end,
the win32 platform is primarily for initial testing/development, and the
fact that the finished product seems to work just as well there is merely a
(big) convenience.  Just trying to point out something Qiang might run into.

>> [...] replace it with something based upon libusb-1.0 [...]
>>
>> Actually Jie Zhang wrote that he is working on it some days ago, you can
>> search for his post the mailinglist-archive.

Okay, now I'm just hoping the new version will degrade nicely (as opposed to
just failing) with only 0.1.12-win32 available. ;)

Thanks,
Michael

-----Original Message-----
From: Gerd v. Egidy [mailto:gerd.von.egidy <at> intra2net.com]
Sent: Wednesday, August 05, 2009 9:42 AM
To: libftdi <at> developer.intra2net.com
Cc: Xiaofan Chen
Subject: Re: About FT2232

> > I can't speak for the "H" part, but the "FT2232D" part works just fine
> > under Win32 with MSVC.  I also built the same libftdi source for Linux
> > with gcc. I use direct/UART mode ("direct" meaning *not* VCP), as well
as
> > MPSSE/SPI mode.  My team will likely move to the "H" part in a few
> > months, so I'll know more then.

Ok, nice to know that it works.

> > One caveat is that Device Manager likes
> > to give the device to d2xx, and it takes some coaxing to make it take
> > libusb-win32 (I sometimes have to "update driver" a second time after
the
> > first unplug/replug...but not always).

Do you use the default vendor-/product-id? Maybe it would help if you use a
different product id that the d2xx-driver doesn't know about.

You can ask FTDIchip, they will usually provide you with a unique product id
out of their range for free.

> libusb-win32 actually supports async mode, so does libusb-1.0.
> Right now the async mode of libftdi seems to directly manipulate
> sysfs and it is only for Linux.

The async-mode of libftdi is an evil hack which I wrote when libusb-1.0
wasn't
available yet. It would be nice to replace it with something based upon
libusb-1.0. I currently don't have the time to do that, but if someone wants
to work on it we could certainly provide some help and feedback.

Actually Jie Zhang wrote that he is working on it some days ago, you can
search for his post the mailinglist-archive.

Kind regards,

Gerd

--
libftdi - see http://www.intra2net.com/en/developer/libftdi for details.
To unsubscribe send a mail to libftdi+unsubscribe <at> developer.intra2net.com

--
libftdi - see http://www.intra2net.com/en/developer/libftdi for details.
To unsubscribe send a mail to libftdi+unsubscribe <at> developer.intra2net.com   

Permissions issues?

RE: Permissions issues?

RE: Permissions issues?