Hi Laura,
I think you can get this error if you're attempting to run a 32-bit
binary on a 64-bit machine without the compatibility libraries. Have you
installed these? You can do so by running:
$ apt-get install ia32-libs
Best,
Rurik
On 2/17/12 1:22 PM, Laura Vertatschitsch wrote:
Hey Griffin,
It is definitely there and is executable. We tried simply running in
the command line the same syntax described in gen_prog_files right from
the directory with mkbof - still same error. I mentioned that we did
need to use the gmake fix described, not sure if the executable that was
built is just not compatible with the Ubuntu specs. I am a bit out of
my expertise there.
We still have Ubuntu set up and wouldn't mind getting to the bottom of
it, but have started to push forward with a RHEL build to sync up with
others and minimize the occurrence of completely unique errors.
--Laura
On Fri, Feb 17, 2012 at 2:18 AM, Griffin Foster
<[email protected] <mailto:[email protected]>> wrote:
Hi Laura, the line './gen_prog_files: line 3: ./mkbof: No such file or
directory' in your error log suggests that mkbof is missing, which is
pretty odd. Is the mkbof executable in the <model
name>/XPS_ROACH_base/ directory after you run the compile? The
original is in mlib_devel/xps_lib/XPS_ROACH_base/ . What happens when
you run mkbof outside the toolflow?
i.e. in <model name>/XPS_ROACH_base/ run
./mkbof -o implementation/system.bof -s core_info.tab -t 3
implementation/system.bin
-Griff
On Wed, Feb 15, 2012 at 10:35 PM, Laura Vertatschitsch
<[email protected] <mailto:[email protected]>> wrote:
> Thank you so much for your kind replies. For what it's worth, I will
> include the error at the end of the message. It boils down to
mkbof. I
> searched the archives and found a similar error that Louis
commented on, but
> didn't see a resolution in that thread. Louis was kind enough to
suggest
> changing permissions (tried, same error), seems consistent with
this error
> that John was seeing. Starting to suspect operating system
issues - perhaps
> even libc in 64 bit ubuntu (the web seems to see issues from
11.04 on).
>
> A "solution": I copied the file over to my windows machine and
just ran the
> windows executable, and no problems. It's not an elegant flow
by any
> means, but hey, we now have our first compiled bof file and the
Roach board
> is blinking as expected - an exciting little victory.
>
> We will grab our sys admin who is a linux guru to see if he can see a
> solution that we could share with the community in case anyone is
interested
> in using Ubuntu.
>
> While its set up here we will update to 12.4 as Alec suggests to
see what
> changes. After that we will just take a few hours and set up our
virtual
> machine with Red Hat so we can be synced up with the rest of the
community.
>
> --Laura Vertatschitsch
>
> :::::::::::::::
>
> Specs: Ubuntu 11.10, Xilinx 11.5 (Including system generator
11.5), and
> Matlab 2009a. We are operating in a virtual machine, and we set
up 11.5
> using the Linux XPS webpage, including copying the 3 files
necessary, and we
> did need to apply the gmake fix as well as a symbolic link fix
that we found
> in the listserve archive (sh-->dash changed to sh-->bash). And
of course
> added 11.5 to the case statements in bee_xps.
>
> :::::::::::::::
>
> [edited to show the final few lines]
>
> Running DRC.
> WARNING:PhysDesignRules:367 - The signal
<infrastructure_inst/dly_clk> is
> incomplete. The signal does not drive any load pins in the design.
> WARNING:PhysDesignRules:367 - The signal <sys_reset> is
incomplete. The
> signal
> does not drive any load pins in the design.
> DRC detected 0 errors and 2 warnings. Please see the previously
displayed
> individual error or warning messages for more details.
> Creating bit map...
> Saving bit stream in "system.bit".
> Saving bit stream in "system.bin".
> Bitstream generation is complete.
> No changes to be saved in MSS file
> Saved project XMP file
> ./gen_prog_files: line 3: ./mkbof: No such file or directory
> chmod: cannot access `implementation/system.bof': No such file or
directory
> cp: cannot stat `implementation/system.bof': No such file or
directory
>
> Error using ==> gen_xps_files at 702
> Programation files generation failed, EDK compilation probably
also failed.
>
>
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
>
>
> On Wed, Feb 15, 2012 at 12:01 AM, Jason Manley
<[email protected] <mailto:[email protected]>>
> wrote:
>>
>> I would encourage you to just use CentOS/Redhat. You have to change
>> various libraries in Ubuntu (to very old ones!) to make the 11
'flow work
>> natively. I seem to recall perl and gcc both giving trouble, for
example. At
>> KAT, we do run Ubuntu, but the 'flow executes in a chroot RHEL
environment
>> to keep their old libraries and executables. I've been led to
believe that
>> this won't be necessary with the v13 'flow but I haven't
actually got it
>> running yet so can't vouch for that. And I don't recommend you
start with 13
>> because you have to manually copy deprecated cores from v11 to
make it work.
>>
>> Many find Redhat annoying but if you're getting started, I
advise you use
>> rather bite the bullet and use it. You can always install it as
a virtual
>> machine initially if you don't want to disturb your primary OS.
Also, don't
>> install all the latest RHEL updates.
>>
>> Jason
>>
>>
>>
>> On 15 Feb 2012, at 09:40, Alec Rust wrote:
>>
>> > Hi Laura, here are instructions for setting up 12.4 labtools,
the same
>> > procedure will work for setting up the full version.
>> >
>> > Install Xilinx in /opt/Xilinx
>> > • Download Xilinx 12.4 Labtools. This can be downloaded from
>> > http://www.xilinx.com
>> > • Untar (tar -xvf Xilinx_LabTools_12.4_M.81d.2.0.tar)
and install
>> > by running “sudo ./xsetup” in the Xilinx directory.
>> > • Generate a webpack licence on the website, a lic file
will be
>> > emailed to you. During installation select the webpack licence
and follow
>> > instructions.
>> > • To run Xilinx Impact type the following in a terminal
(source
>> > runs a pearl script that sets up the Xilinx paths):
>> > source /opt/Xilinx/12.4/LabTools/settings64.sh
>> > impact
>> >
>> >
>> > Regards
>> > Alec
>> >
>> >
>> > On Wed, Feb 15, 2012 at 12:36 AM, Laura Vertatschitsch
>> > <[email protected] <mailto:[email protected]>> wrote:
>> > Hi guys, first post.
>> >
>> > We are setting up a machine from scratch. The 2011 tutorials
section
>> > seems to indicate the all the machines were set up with Xilinx
11.5 on
>> > machines running 64 bit Ubuntu. Are there instructions on how
this set up
>> > was achieved? Everything I read on the wiki seems to say that
deviating
>> > from the xilinx-blessed operating systems will result in bad
news. We got
>> > everything set up using the latest Ubuntu 11.10, Matlab 2009a,
Xilinx 11.5,
>> > by following the LINUX XPS page and introducing the simple
case statement in
>> > bee_xps to allow 11.5. We are getting some errors at the end
of the compile
>> > of tutorial 1 - but wanted to check if there was further
documentation on
>> > the setup that we could read first.
>> >
>> > Laura Vertatschitsch
>> > Electrical Engineering
>> > University of Washington
>> >
>>
>
