Yuri's Status Report - #13 of 15

[Yuri Gonzaga]

Hi,

I don't know what happened but I am still unable to connect to remote
machine to work with m501 board.
So, I couldn't try more to get JtR compiled and running with even one single
core of eksblowfish loop.
In the same way, I owe you the patch to help me to solve this problem.

I gathered 4 cores with manager to synthesize but I am getting the following
error (in Xilinx ISE).
Maybe David can say anything about that.

ERROR:PhysDesignRules:2399 - The GTXE1 comp

PicoFramework/core/pcie_2_0_i/pcie_gt_i/gtx_v6_i/GTXD[0].GTX has
POWER_SAVE[4] set to an unsupported value and must be set to 1. Please
see Answer Record 39430 for more information.

As soon as possible, I want to keep working on JtR integration.
Meanwhile, I can try to validate multicore on pico e101 running another
piece of software than JtR.

Thanks.

Yuri

[Solar Designer]

Hi Yuri,

Post by Yuri Gonzaga
I don't know what happened but I am still unable to connect to remote
machine to work with m501 board.

Sorry about that. It's DEFCON/BlackHat time for David. ;-)

Post by Yuri Gonzaga
So, I couldn't try more to get JtR compiled

This is not entirely true: you could/can try to setup a similar
development environment on your own Linux system and get past the
compilation errors. You won't be able to actually run and debug the
code, at least not fully, but you'll get it to compile. Then when you
get access to the remote machine again, you'll proceed with running and
debugging right away, without having to spend time to fight source code
compilation issues.

Post by Yuri Gonzaga
and running with even one single core of eksblowfish loop.
In the same way, I owe you the patch to help me to solve this problem.

Yes, and you should be able to produce that "patch" on your own computer.

Alternatively, if you like, I may setup a Linux system for you to access
remotely. You'll need to install Pico's development stuff on it (you
got a copy of it, right?) It won't have an FPGA board in it, but you'll
be able to get stuff to compile, or show to me how it fails to compile.
Please let me know if you'd like to proceed with this approach. Also,
please let me know what Linux distro and version was installed on the
machine David provided you with access to. Perhaps we'll want to use a
similar one.

Note: you won't need to install Xilinx ISE there, nor Pico's FPGA board
drivers. Just the development libraries (that talk to the drivers).

Post by Yuri Gonzaga
I gathered 4 cores with manager to synthesize but I am getting the following
error (in Xilinx ISE).
Maybe David can say anything about that.
ERROR:PhysDesignRules:2399 - The GTXE1 comp

PicoFramework/core/pcie_2_0_i/pcie_gt_i/gtx_v6_i/GTXD[0].GTX has
POWER_SAVE[4] set to an unsupported value and must be set to 1. Please
see Answer Record 39430 for more information.

Yes, this will need to wait for David's comment, unless you figure it
out on your own (please try while we await David's response).

Post by Yuri Gonzaga
As soon as possible, I want to keep working on JtR integration.
Meanwhile, I can try to validate multicore on pico e101 running another
piece of software than JtR.

Sounds reasonable.

Thanks,

Alexander

[Yuri Gonzaga]

Post by Solar Designer
This is not entirely true: you could/can try to setup a similar
development environment on your own Linux system and get past the
compilation errors. You won't be able to actually run and debug the
code, at least not fully, but you'll get it to compile. Then when you
get access to the remote machine again, you'll proceed with running and
debugging right away, without having to spend time to fight source code
compilation issues.

Yes, you right. But, there is a little problem.
I don't have a copy of Pico's includes and libraries.
I was just using the files already installed on the remote machine.
The files that David sent to me don't include those ones.

Yes, and you should be able to produce that "patch" on your own computer.


The code that I produced is on that machine.

Alternatively, if you like, I may setup a Linux system for you to access

Post by Solar Designer
remotely. You'll need to install Pico's development stuff on it (you
got a copy of it, right?) It won't have an FPGA board in it, but you'll
be able to get stuff to compile, or show to me how it fails to compile.
Please let me know if you'd like to proceed with this approach. Also,
please let me know what Linux distro and version was installed on the
machine David provided you with access to. Perhaps we'll want to use a
similar one.

Thanks. But it's not necessary.
If I had those files now, I think I prefer using linux on my own computer.

You give me an idea: next time I have access to the remote machine, I will
copy the libraries to my distro to work on my own. So, I will need the
remote machine only when I will run on the board.

Thank you!

---
Yuri

[Solar Designer]

Yuri -

Post by Yuri Gonzaga
Meanwhile, I can try to validate multicore on pico e101 running another
piece of software than JtR.

I'm not sure if we pointed you at this before, but here are some pieces
of code written by David that I think interface to multiple crypto cores
in an FPGA on a Pico board:

http://openciphers.sourceforge.net/oc/download.php

Alexander