ISE/EDK Kubuntu linux installation issues
MM
I was wondering if the ISE/EDK linux installation scripts are supposed to
create desktop icons and/or create appropriate entries in the program menu?
I have installed both tools on a Kubuntu system and none of these were
created...
Thanks,
/Mikhail
Ken Ryan
Nope, you have to do all that yourself.
Depending what version of Kubuntu you're running, you may have other
troubles. For example, if you get errors referencing GLIBC you need
to set the environment variable LD_ASSUME_KERNEL=2.4.7 (this is not a
problem in recent releases of ISE/EDK).
Getting a download cable to work is also a treat. I recommend ignoring
the Xilinx drivers and use the userspace driver from:
http://www.rmdir.de/~michael/xilinx/
I have it working for impact, chipscope analyzer, and XMD, at least
in all modes I actually use. I can elaborate further all the fun I
had getting it to work on a 64-bit system, if that's something you'll
need (post here but CC my email to be sure I see it; I don't check
every day).
In any case, it works great on my Linux boxen which is a relief since
EDK does not get along with my laptop at all (multiple applications
built on top of Cygwin are not playing nice at all).
Good luck, and drop me a note if you need more help.
ken
MM
> Depending what version of Kubuntu you're running, you may have other
> troubles. For example, if you get errors referencing GLIBC you need
> to set the environment variable LD_ASSUME_KERNEL=2.4.7 (this is not a
> problem in recent releases of ISE/EDK).
I am running Kubuntu 7.04, 32-bit version and the 9.1 version of the Xilinx
tools. I don't recall seeing this error...
> Getting a download cable to work is also a treat. I recommend ignoring
> the Xilinx drivers and use the userspace driver from:
>
> http://www.rmdir.de/~michael/xilinx/
That's really a great advice! I was just in the middle of the process of
trying to install Xilinx drivers :) Hopefully I haven't broken anything yet
by half-working scripts :)
> I have it working for impact, chipscope analyzer, and XMD, at least
> in all modes I actually use. I can elaborate further all the fun I
> had getting it to work on a 64-bit system, if that's something you'll
> need (post here but CC my email to be sure I see it; I don't check
> every day).
It's interesting. I was told in the past by Xilinx that EDK wasn't working
on 64-bit systems... Have they fixed it in 9.1 or have you found a
workaround?
> In any case, it works great on my Linux boxen which is a relief since
> EDK does not get along with my laptop at all (multiple applications
> built on top of Cygwin are not playing nice at all).
>
> Good luck, and drop me a note if you need more help.
Thanks again,
/Mikhail
Ken Ryan
> Thanks a lot Ken!
>
>> Depending what version of Kubuntu you're running, you may have other
>> troubles. For example, if you get errors referencing GLIBC you need
>> to set the environment variable LD_ASSUME_KERNEL=2.4.7 (this is not a
>> problem in recent releases of ISE/EDK).
>
> I am running Kubuntu 7.04, 32-bit version and the 9.1 version of the Xilinx
> tools. I don't recall seeing this error...
Good, that combination should be fine (one of my systems is Ubuntu 6.10,
I had no trouble).
>> Getting a download cable to work is also a treat. I recommend ignoring
>> the Xilinx drivers and use the userspace driver from:
>>
>> http://www.rmdir.de/~michael/xilinx/
>
> That's really a great advice! I was just in the middle of the process of
> trying to install Xilinx drivers :) Hopefully I haven't broken anything yet
> by half-working scripts :)
You'll still need the bit of the installation that handles usb firmware
loading (fxload) and permissions settings, but a variant of it.
At the moment, the only Xilinx-related stuff in my system directories
is a file /etc/udev/rules.d/xusbdfwu.rules containing the following:
----VVV
SYSFS{idVendor}=="03fd", SYSFS{idProduct}=="0008", NAME="windrvr6"
BUS=="usb", ACTION=="add", SYSFS{idVendor}=="03fd",
SYSFS{idProduct}=="0007", RUN+="/sbin/fxload -v -t fx2 -I
/usr/share/xusbdfwu.hex -D $TEMPNODE"
BUS=="usb", ACTION=="add", SYSFS{idVendor}=="03fd",
SYSFS{idProduct}=="0009", RUN+="/sbin/fxload -v -t fx2 -I
/usr/share/xusbdfwu.hex -D $TEMPNODE"
BUS=="usb", ACTION=="add", SYSFS{idVendor}=="03fd",
SYSFS{idProduct}=="000b", RUN+="/sbin/fxload -v -t fx2 -I
/usr/share/xusbdfwu.hex -D $TEMPNODE"
BUS=="usb", ACTION=="add", SYSFS{idVendor}=="03fd",
SYSFS{idProduct}=="000d", RUN+="/sbin/fxload -v -t fx2 -I
/usr/share/xusbdfwu.hex -D $TEMPNODE"
BUS=="usb", ACTION=="add", SYSFS{idVendor}=="03fd",
SYSFS{idProduct}=="000f", RUN+="/sbin/fxload -v -t fx2 -I
/usr/share/xusbdfwu.hex -D $TEMPNODE"
ACTION=="add", BUS=="usb", SYSFS{idVendor}=="03fd",
SYSFS{idProduct}=="0008", MODE="0666"
----^^^
Since the above is almost certainly mangled by my mailer, I separated
what should be single lines in the file with blank lines. In other
words, there is one line that starts with SYSFS{idVendor} and
ends with "windrvr6" (I think that line is actually not necessary,
but it's harmless), the next line starts with BUS=="usb" and ends
with $TEMPNODE, four more similar lines with different product IDs,
then a line that starts with ACTION and ends with "0666".
Note that the name of the file seems to be relavent - udev processes
rule files in order of filename, and getting the Xilinx stuff too early
seems to not work (i.e. I initially had this in a file "10-custom.rules"
but it didn't work there). I don't understand udev well enough to
know why that makes it break.
Anyway, since you're just starting to install the cable drivers I'll
follow up my own post with a description of how I got the userspace
shim to work transparently..
>> I have it working for impact, chipscope analyzer, and XMD, at least
>> in all modes I actually use. I can elaborate further all the fun I
>> had getting it to work on a 64-bit system, if that's something you'll
>> need (post here but CC my email to be sure I see it; I don't check
>> every day).
>
> It's interesting. I was told in the past by Xilinx that EDK wasn't working
> on 64-bit systems... Have they fixed it in 9.1 or have you found a
> workaround?
EDK does not work with a 64-bit ISE installation. I have both ISE and
EDK installed as 32-bit applications. The OS is the x86_64 version of
SuSE 9.3; the fact that all of the system binaries are 64 bit caused me
some pain getting it installed and getting the userspace cable driver to
work (to force the 32-bit ISE install, don't run setup at the top of the
installation directory; cd to bin/lin and run it from there. That is
also needed for all the service packs and IP updates). I have a 64-bit
ISE installation standing by should I need it, but I haven't yet had a
design big enough where it matters. [Note that running 32 bit will be a
bit faster, measurably so for long runs, due to the smaller pointer
sizes getting fetched more efficiently by the CPU].
Ken Ryan
> Anyway, since you're just starting to install the cable drivers I'll
> follow up my own post with a description of how I got the userspace
> shim to work transparently..
OK here goes...
Note this applies to the USB driver. Theoretically it also works with
the parallel cable, but I haven't tested it.
First, add the file /etc/udev/rules.d/xusbdfwu.rules as I described in
my previous post. That gets the firmware loaded and device permissions
set when you connect the usb cable. I *think* it's tailored narrowly
enough to not open unwise permissions, but if it matters to you have
someone with a clue about security look it over.
Next download the userspace shim sources from
http://www.rmdir.de/~michael/xilinx/ and compile them. For those on a
64-bit OS, edit the Makefile and add -m32 to the CFLAGS variable to
force a 32-bit compile.
Now, there seem to actually be only three binaries that try to
communicate with the cable hardware:
$XILINX/bin/lin/_impact iMPACT bitstream downloader (core)
$CHIPSCOPE/bin/lin/cse Chipscope cable server
$XILINX_EDK/bin/lin/xmd Software debugger server
To use the driver shim, they each need the libusb_driver.so prelinked
to intercept the windrvr calls. On a pure 32-bit platform you could
simply set the environment variable LD_PRELOAD to the full pathname of
the libusb_driver.so, but then the system attempts to prelink *all* of
your binaries to that library. On a 64-bit platform this fails miserably.
So I did the following:
1. Copy libusb_driver.so to $XILINX/bin/lin/ to ensure it's available
for all users of the system (in my case: me, myself and I but whatever).
2. Rename $XILINX/bin/lin/_impact to $XILINX/bin/lin/_impact.bin (note
'm renaming _impact not impact).
3. Create a shell script called $XILINX/bin/lin/_impact containing the
following:
---VVV
#!/bin/sh
#
# Preload driver binary
LD_PRELOAD=$XILINX/bin/lin/libusb-driver.so $0.bin $*
---^^^
4. Ensure the script permissions are set with chmod +x.
5. Repeat steps 2-4 for the Chipscope and XMD binaries, as applicable.
I use the identical script file in all three places; the $0.bin and
different name ensures the right thing gets run.
6. Enjoy!
Note that installing a service pack may overwrite this hack (the
EDK9.1.02 service pack did that to the xmd script). Just repeat steps
2-4 to fix it.
What's happening:
I initially just used a csh alias to include the LD_PRELOAD, but that
only worked for direct command-line invocations. Running any of these
from a GUI bypasses the shell. Similarly, putting a script in the
search path doesn't work, as the GUIs seem to call their helper programs
via a direct path. By putting an executable shell script directly in
the places where callers look for it, you can add the prelink and the
Linux system libaries will sort out whether it's executing a binary or a
shell script.
The impact program was a bit weird, and threw me for a little while on
my 64-bit box. The binary $XILINX/bin/lin/impact is what gets started,
but after doing some initial stuff it then calls _impact to do the heavy
lifting. I have no idea why that is so - perhaps _impact does the
downloading and something else does the prom file preparation, and
impact is just the front-end wrapper. It did have the bizarre behavior
of working properly if I just ran impact alone, but if I added any
command-line parameters it failed (the resulting message was highly
misleading as well - failed to find libusb-driver.so even though the
error message itself included the complete correct path...the clue was
the error message was from tcsh not impact, indicating impact actually
invoked my shell to execute _impact).
I have used impact, chipscope analyser.sh and xmd with this setup, both
direct from the command line and impact/xmd from within the xps GUI.
If I were Xilinx I would jump on this userspace driver and can the
kernel driver. Aside from debatable kernel module license issues, the
userspace driver means users are mucking with their system a whole lot
less, which *surely* has to reduce support calls! Three cheers to
Michael Gernoth for his outstanding work in solving this problem (not to
mention realizing it *could* be solved)!
ken
MM
I had actually got the cable working before you posted your detailed
instructions, but thanks a lot anyways, it's all very useful information!
I am now fighting the EDK, which doesn't want to start (the ISE seems to
work fine), and the issues of equalizing the environment for all the
possible ways of starting all of the Xilinx applications. I am just learning
about how the environment variables are handled in Ubuntu and it is
definitely not straightforward at all...At the moment my Impact works only
when I start it from the shell... But I believe I am on the right track,
just need a little bit of time to figure it all out...
> Three cheers to Michael Gernoth for his outstanding work in solving this
> problem (not to mention realizing it *could* be solved)!
That is indeed a great achievement!!!
Thanks,
/Mikhail
MM
says the following:
_xps: error while loading shared libraries: libPortability.so : cannot open
shared object file : no such file or directory.
If I set LD_ASSUME_KERNEL=2.4.7 (or 2.4.1) then it seems to break something
severely in bash so that I can't even do printenv and xps doesn't start
either... I suspect that the problem might have something to do with the
fact that I have both libstdc++5 and libstdc++6 installed. According to the
http://ubuntuforums.org:80/showthread.php?t=349336 only libstdc++5 is
required, however I am not sure now at which point libstdc++6 was installed
and what will I break if I try uninstalling it...
All other environment variables seem to be set correctly, at least in the
bash session I am using to launch the app...ISE launches fine as well as
Impact does...
Another thing I don't quite understand is why running
$XILINX_EDK/settings.sh script doesn's seem to have any effect on the
environment variables... So far I had to set some of them manually from bash
and some in /etc/environment...
Thanks,
/Mikhail
Andreas Ehliar
> Another thing I don't quite understand is why running
> $XILINX_EDK/settings.sh script doesn's seem to have any effect on the
> environment variables... So far I had to set some of them manually from bash
> and some in /etc/environment...
Just a quick comment regarding this..
Are you sourcing the settings.sh script as in the following?
> source $XILINX_EDK/settings.sh
Or are you running it as in
> sh $XILINX_EDK/settings.sh
If the later is true you will start a new subshell where the settings will
take effect and then happily disappear after the shell script has finished
running (which will occur in say a couple of microseconds after it has
started...)
If you are indeed sourcing the script I have no idea why it doesn't
work for you.
/Andreas
MM
news:5cb28eF...@mid.individual.net...
>
> Another thing I don't quite understand is why running
> $XILINX_EDK/settings.sh script doesn's seem to have any effect on the
> environment variables... So far I had to set some of them manually from
> bash and some in /etc/environment...
>
For other linux newbies out there I figured this one out... Bash needs to be
told that commands from the filename argument need to be executed in the
_current_ shell context. Otherwise it will create another context, which
will disappear as soon as the script ends... The way to tell this to bash is
to precede the filename with a dot. Thus, assuming the settings.sh is made
executable (and shebang added in the first line of the file) one has to do
the following:
$ . settings.sh
and not
$ settings.sh
/Mikhail
MM
Having another problem :( I've decided to try installing the latest EDK
service pack (SP2) and the installer refused to accept my Registration ID,
which it happily accepted during the original EDK installation...
/Mikhail
MM
> says the following:
>
> _xps: error while loading shared libraries: libPortability.so : cannot
> open shared object file : no such file or directory.
I found that this library is in fact in the ISE directory $XILINX/bin/lin. I
copied it into $XILINX_EDK/bin/lin (which is set to LD_LIBRARY_PATH) and it
complained about another library, which I also found in the ISE tree. After
I copied a few dozens of these libraries xps finally started, but with a
blank window :(
Arghh... What am I missing here?...
I am posting the same question on Ubuntu forum in the "Xilinx on Ubuntu"
thread:
http://ubuntuforums.org/showthread.php?p=2763922#post2763922
Thanks,
/Mikhail
MM
$ export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$XILINX/bin/lin
/Mikhail
svenand
I am running Ubuntu Linux 7.04 in VMware Fusion on my MacBook.
I have described the whole process of installing and using Xilinx
ISE/EDK 9.1 software on the ML403 evaluation board.
See my blog: www.fpgafromscratch.com
Sven