I finished gpjtag on Aug 2016, so it is time to opensource.
All posts by yuliang
Enable Let’s Encrypt
I am using StartSSL free certificate before, but Chrome and Firefox begin to drop it.
So I change to Let’s Encrypt. It really work!!! Great!!!
And force all request to SSL.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 |
#http://gpjtag.com server { listen 80; server_name gpjtag.com; rewrite ^(.*) https://www.gpjtag.com$1 permanent; } #https://gpjtag.com server { listen 443; server_name gpjtag.com; ssl on; ssl_certificate ......; ssl_certificate_key ......; rewrite ^(.*) https://www.gpjtag.com$1 permanent; } #http://www.gpjtag.com server { listen 80; server_name www.gpjtag.com; rewrite ^(.*) https://www.gpjtag.com$1 permanent; } #https://blog.gpjtag.com server { listen 80; server_name blog.gpjtag.com; rewrite ^(.*) https://blog.gpjtag.com$1 permanent; } |
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License
Found final audit of Truecrypt report
I found the final report on opencryptoaudit.org, and the answer is Truecrypt is safe, great!
Or I think is safer than most open source crypto on the fly disk implement without UEFI.
Continue reading Found final audit of Truecrypt report
Create an open source vmnetcfg
VMware VMNetcfg is a useful tool, you can change IP, DNS, NAT quickly. But begin from VMware Player 3.x, it have been removed from then standard installer. You must extract from Workstation version, painful…
In the older days, you can extract from VMware Player installer, and it removed from the installer from VMware Player 5.x… But we still have vmnetui.dll,VMNetUI_ShowStandalone. But now is removed too… We have to extract from Workstation, but it really huge(500MB+).
So I decided to create an open source UI, like vmnetcfg.
Source push to Github: https://github.com/buaabyl/openvmnetcfg
Bug with GCC and Qt 5.7.0
When I build a Qt library from source, everything seems ok, but when the Qt example run, it show segmentation fault… Why do I compile from source, because my CPU on server not support SSE2 >_<
So I do some research, build it with difference version of GCC and CFLAGS.
This is the summary:
Building microblaze toolchians
Xilinx have a pre-compiled toolchain ship with ISE or Vivado, but I want to build it myself (just for fun).
My system with mingw32-w64 and build-essential
lsb_release -a
# Distributor ID: Ubuntu
# Description: Ubuntu 16.04 LTS
# Release: 16.04
# Codename: xenial
# Linux fish-vm 4.4.0-24-generic #43-Ubuntu SMP Wed Jun 8 19:27:37 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
# gcc version 5.3.1 20160413 (Ubuntu 5.3.1-14ubuntu2.1)
# Target: x86_64-linux-gnu
Build openssl for windows
First, download source code: https://www.openssl.org/source/openssl-1.0.2h.tar.gz
Second, download ActivePerl https://www.activestate.com/activeperl
Third, download mingw64 https://sourceforge.net/projects/mingw-w64/files/Toolchains%20targetting%20Win32/
|
1 2 3 4 5 6 7 |
set PATH=C:\Perl\bin;%PATH% perl Configure VC-WIN32 no-asm enable-static-engine --prefix=c:\openssl call ms\do_ms nmake -f ms\ntdll.mak nmake -f ms\nt.mak nmake -f ms\ntdll.mak install nmake -f ms\nt.mak install |
And I compile mingw version under xubuntu 16.04 with mingw32 installed.
|
1 2 3 4 5 6 7 |
tar xvf openssl-1.0.2h.tar.gz cd openssl-1.0.2h/ ./Configure mingw enable-static-engine --cross-compile-prefix=i686-w64-mingw32- \ --prefix=/home/fish/work/openssl-w32 make depend make make install |
Have fun:P
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License
Custom Ubuntu live CD
Ubuntu live cd can startup on u-disk, so just test with this media.
Read Spartan6LX DNA Serial Number
Spartan6 Contain DNA logic — DNA_PORT, We can let this as random seed:)
This is 57bits numbers.
DNA_PORT DNA_PORT_inst(
.DOUT(dna_dout),
.CLK(dna_clk),
.DIN(1'b0),
.READ(dna_read),
.SHIFT(dna_shift)
);
DNA_PORT is act as SPI Slave, capture at raise edge, and first MSB is drive out when trigger READ signal.
Then shift out one by one when SHIFT signal valid. It is better to save bits on negative edge.
Bug in data2mem…
I using data2mem to replace data in BlockRAM,
It confused me when I try to replace BlockRAM data in Spartan6 Devices.
There are correct in BRAM16 with 8bits and 16bits, only error BRAM18 with 18bits.
Update: bug exist in ISE 12.4, 13.4, 14.7.