gmane.comp.hardware.libftdi.general

Re: Reading/Writing FT4232 EEPROM

Anders Larsen — 2013-06-18T11:13:35

Hi,

On 2013-05-28 10:25, Thomas Jarosch wrote:

the HEAD version indeed works; FT4232 support was added last year.

You can use ftdi_eeprom from HEAD (or v1.0) to program the chip
and still use 0.19/0.20 from your application as long as the application
doesn't use any ftdi*eeprom*() function.

(sorry for joining the discussion so late - I was away on vacation)

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

Re: Reading/Writing FT4232 EEPROM

Andreas Helmcke — 2013-06-18T08:23:20

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Am 10.06.2013 13:33, schrieb Ori Idan:

Having had the same problem I have modified the eeprom example (see attached program). For my setting it is sufficient
to read and write binary data, since I only have to set a known configuration.
To get the binary data I used the windows configuration program from ftdichip
(http://www.ftdichip.com/Support/Utilities.htm#FT_Prog) to program the eeprom and read it using eeprom-set on linux.

Hope that helps

Greetings
  Andreas

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlHAGO8ACgkQDVaV4kk68F8s9gCgrOo4ze0R0EKWTSQELF3/CckV
TD0AoIAzbaRIJxDZHhZtOtfBDweDN5Sz
=GAB9
-----END PGP SIGNATURE-----


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

Re: Reading/Writing FT4232 EEPROM

Ori Idan — 2013-06-10T11:33:49

I still did not find a solution how to program EEPROM on FT4232
I am using Ubuntu 12.10 with libftdi1 0.20

Re: Re: data corruption at lower latency

Thomas Jarosch — 2013-06-05T09:30:26

I was thinking about stacking an ext4 filesystem on top of a loop device
which resides in a tmpfs. This is of course overkill and only needed
if you need to tune the filesystem f.e. to contain 30.000.000 files.

A simple "tmpfs" will do for your case. Or as Rogier pointed out,
just append to a log file might work, too. Still I would put it in a tmpfs.

Thomas


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

Re: data corruption at lower latency

mav — 2013-06-03T22:16:26

Thank you all for the suggestions!

Thomas,
Sorry, I could not understand the ext4 + tmpfs trick, can you please elaborate?

thanks,
mav

On Sun, Jun 2, 2013 at 4:28 AM, Thomas Jarosch
<thomas.jarosch< at >intra2net.com> wrote:

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

Re:

Thomas Jarosch — 2013-06-03T12:48:34

Hi,

On Monday, 3. June 2013 07:45:30 xantares 09 wrote:

applied, thanks.

Thomas


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

(unknown)

xantares 09 — 2013-06-03T07:45:30

Hi

Here is a cleaner way to build the python install path, accounting for dist-specific path:
https://gitorious.org/libftdi/libftdi/commit/0fba0d248235be294b2e3a547e6bc7894af01d21

M.
       

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

Re: [patch] fix compile with boost

Thomas Jarosch — 2013-06-03T07:22:31

Hi John,

On Tuesday, 28. May 2013 07:48:17 John Hein wrote:

ah sorry, scrap my idea, you are right about the _in-tree_ check.
Also your patch is for Makefile.am which is no longer part
of libftdi 1.x. We completely switched to cmake.

Since your patch does not hurt, I've applied it to libftdi 0.x.

Thanks,
Thomas


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

Re: data corruption at lower latency

Thomas Jarosch — 2013-06-02T08:28:24

ext4 + tmpfs might do the trick ;) Though you are right.


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

Re: data corruption at lower latency

Rogier Wolff — 2013-06-01T10:57:14


I would NOT do the new file trick if you're doing low-latency work. 
creating a new file is going to cost significant time after a few 
files. 

Roger.

Re: data corruption at lower latency

Thomas Jarosch — 2013-06-01T10:38:50

Hi mav,

On 05/31/2013 08:54 PM, mav wrote:

take a look at the ftdi_read_xxx() functions:

The status bytes are sent with every incoming usb packet
of the read request and then filtered by the code.

May be the filtering has a bug or the status bytes
appear at an unexpected position.

I would start by adding raw logging of the data
to disc and then examine it, may be create a new
file for every incoming packet so it's nicely partitioned.

Thomas


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

Re: data corruption at lower latency

mav — 2013-05-31T18:54:45

Not yet, testing on a PC is my next plan. The hardware may not be
providing required processing power, but so far I can't conclude so.
It may get clarified by executing the test on a PC setup.

The modem communicates well if FTDI chip and its corresponding USB
chain is bypassed. When FTDI and USB are brought in, the serial lines
sporadically gets 2 unwanted bytes which corrupts the data stream. The
two erroneous bytes are always printed as 2` which if mapped to the
FTDI  Status Bytes will translate to [0011 0010 0110 0000]; this
indicates THRE, BI and OE bits are set.

The same 2 bytes are also seen on the second FTDI chip which is left
to run at the chip's default latency (16ms) and is running slower
< at >19200 baud (the first line is < at >115200 baud). I suspect the erroneous
bytes are coming from the underlying USB libraries, and I am still
looking for ways to confirm this.

thanks,
mav

On Fri, May 31, 2013 at 11:26 AM, Jim Paris <jim< at >jtan.com> wrote:

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

Re: data corruption at lower latency

Jim Paris — 2013-05-31T15:26:05

Did you also try on a pc?  Your hardware might have problems.

-jim

mav wrote:

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

Re: data corruption at lower latency

mav — 2013-05-31T14:56:14

Here are the different combinations I tried. Data corruption occurs on all.

------
Set_A:
------
libftdi-0.20.tar.gz (Released 2012_03_19)
libusb-1.0.9.tar.bz2 (Released 2012_04_20)
libusb-compat-0.1.4.tar.bz2 (Released 2012_04_24)

Result: Data Corruption.

------
Set_B:
------
libftdi1-1.0.tar.bz2 (Released 2013_01_29)
libusb-1.0.16-rc10.tar.bz2 (Released 2013_05_06)
libconfuse-2.6.tar.gz (Released 2007_12_29)

Result: Underlying system does not support netlink sockets, which
breaks the compilation of libusb-1.0.16-rc10. The "--disable-udev" was
used with ./configure as udev is not supported.

------
Set_C:
------
libftdi1-1.0.tar.bz2 (Released 2013_01_29)
libusb-1.0.9.tar.bz2 (Released 2012_04_20)
libusb-compat-0.1.4.tar.bz2 (Released 2012_04_24)

Result: Data Corruption.

------
Set_D:
------
libftdi1-1.0.tar.bz2 (Released 2013_01_29)
libusb-1.0.9.tar.bz2 (Released 2012_04_20)
libusb-compat-0.1.5.tar.bz2 (Released 2013_05_20)

Result: Data Corruption.

thanks,
mav

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

Re: Reading/Writing FT4232 EEPROM

Ori Idan — 2013-05-28T17:12:46



I have to program the EEPROM and tried using ftdi_eeprom but it did not
work.
That's what started this whole conversation.

Re: [patch] fix compile with boost

John Hein — 2013-05-28T13:48:17

Thomas Jarosch wrote at 14:45 +0200 on May 28, 2013:
 > On Tuesday, 28. May 2013 05:43:33 John Hein wrote:
 > > When building examples/find_all_cpp.o, the attached patch will add
 > > -I flags for boost header files.
 > >
 > > Without it, you can get this when building examples/find_all_pp...
 > >
 > > In file included from find_all_pp.cpp:8:
 > > .../ftdipp/ftdi.hpp:34:32: error: boost/shared_ptr.hpp: No such file or
 > > directory In file included from find_all_pp.cpp:8:
 > > .../ftdipp/ftdi.hpp:144: error: 'boost' has not been declared
 > > .../ftdipp/ftdi.hpp:144: error: ISO C++ forbids declaration of
 > > 'shared_ptr' with no type
 >
 > thanks for you patch. I'm not sure this is the correct fix:
 > The pkgconfig stuff for libftdipp should provide the boost includes
 > automatically. Otherwise every program using it
 > must manually specify the BOOST_CFLAGS...
 >
 > What do you think?

Do you mean the build should use its own in-tree copy of libftdipp.pc
to set -I flags when building libftdi itself?

libftdipp.pc.in would have to be modified to include -I flags for
boost, but I can't think of a great way to have autoconf use
pkg-config to look for the in-tree .pc file at the moment.  Is there a
simple autoconf-ish way to do that? Seems much easier to just include
BOOST_CPPFLAGS in Makefile.am.

Modifying libftdipp.pc.in to include -I /path/to/boost/include
certainly seems reasonable for consumers of libftdi that might include
ftdi.hpp (which currently requires boost).

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

Re: [patch] fix compile with boost

Thomas Jarosch — 2013-05-28T12:45:06

Hi John,

On Tuesday, 28. May 2013 05:43:33 John Hein wrote:

thanks for you patch. I'm not sure this is the correct fix:
The pkgconfig stuff for libftdipp should provide the boost includes
automatically. Otherwise every program using it
must manually specify the BOOST_CFLAGS...

What do you think?

Thomas


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

Re: Reading/Writing FT4232 EEPROM

Uffe Jakobsen — 2013-05-28T12:41:15



On 2013-05-28 14:29, Ori Idan wrote:

The output shows that you are running libftdi version 0.20 on a 32bit 
platform (i386)

PS: I haven't followed your conversation deeply on the list but it 
should be fine - if you are aiming for the functionality in the libftdi 
version 0.x series releases.
If you need something newer eg libftdi 1.x functionality - you need to 
build the library yourself from the ftdi github trunk.

/Uffe





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

Re: Reading/Writing FT4232 EEPROM

Ori Idan — 2013-05-28T12:29:59



Here is what I got:
 dpkg -l | grep -i ftdi
ii  ftdi-eeprom                                   0.3-2
                        i386         Tool for reading/erasing/flashing FTDI
USB chip eeproms
ii  libftdi-dev                                   0.20-1
                         i386         Development files for libftdi
ii  libftdi1:i386                                 0.20-1
                         i386         Library to control and program the
FTDI USB controller

Re: Reading/Writing FT4232 EEPROM

Uffe Jakobsen — 2013-05-28T12:03:09



On 2013-05-28 13:56, Ori Idan wrote:

Look up "libftdi" in your package manager GUI...

OR

do something like this from a terminal:

# dpkg -l | grep -i ftdi

My Ubuntu 12.10 reports libftdi as version 0.20

/Uffe





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

Re: Reading/Writing FT4232 EEPROM

Ori Idan — 2013-05-28T11:56:33

How do I know the version shipped with Ubuntu?