MemeCode - Blog

Page: 0 1 2 3 4 5 6 7 8 ... 10 ... 15 ... 20 ... 25 ... 30 ... 35 ... 40 ... 45 ... 50 ... 55 ... 60 ... 65 ... 70 ... 75 ... 80 ... 85 ... 90 ... 95 ... 100

Portable OpenSSL for Linux
Date: 21/12/2015 Tags: shared-object linux openssl scribe	A year ago I worked out how to make a portable build of OpenSSL for Mac. And now with the imminent release of the Linux build of Scribe I need to do the same with the Linux build of OpenSSL. After downloading and unpacking OpenSSL I configured it with: ./config shared -DPURIFY The reason for adding the -DPURIFY #define is to suppress valgrind errors related to uninitialized memory usage. Once that is built I had a look at the shared object paths with ldd: matthew@ubuntu:~/Downloads/openssl$ ldd ./libssl.so.1.0.0 linux-vdso.so.1 => (0x00007ffc0e197000) libcrypto.so.1.0.0 => /lib/x86_64-linux-gnu/libcrypto.so.1.0.0 (0x00007f5c5ab6c000) libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007f5c5a7a7000) libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2 (0x00007f5c5a5a3000) /lib64/ld-linux-x86-64.so.2 (0x00007f5c5b1c1000) So it's not portable yet. I initially tried using the chrpath command but it can only modify existing RPATH records. My binary doesn't have any RPATH yet. So that didn't work. The next thing I tried was the patchelf command from here: http://nixos.org/releases/patchelf/patchelf-0.8/ After building and installing I issued this command in the OpenSSL folder: patchelf --set-rpath '$ORIGIN' ./libssl.so.1.0.0 Now to check if libssl pulls in the local libcrypto: matthew@ubuntu:~/Downloads/openssl$ ldd ./libssl.so.1.0.0 linux-vdso.so.1 => (0x00007ffcdce86000) libcrypto.so.1.0.0 => /home/matthew/Code/Scribe/trunk/./libcrypto.so.1.0.0 (0x00007f39b99d8000) libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007f39b9613000) libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2 (0x00007f39b940f000) /lib64/ld-linux-x86-64.so.2 (0x00007f39ba0a5000) Yes! Now this build can be installed in the same folder as the Scribe binary and Scribe will use it without interfering with the system OpenSSL which is often out of date, or not built with -DPURIFY or missing.
(0) Comments \| Add Comment

Portable OpenSSL for Linux

Date: 21/12/2015
Tags: shared-object linux openssl scribe

A year ago I worked out how to make a portable build of OpenSSL for Mac. And now with the imminent release of the Linux build of Scribe I need to do the same with the Linux build of OpenSSL.

After downloading and unpacking OpenSSL I configured it with:

./config shared -DPURIFY

The reason for adding the -DPURIFY #define is to suppress valgrind errors related to uninitialized memory usage.

Once that is built I had a look at the shared object paths with ldd:

matthew@ubuntu:~/Downloads/openssl$ ldd ./libssl.so.1.0.0 
	linux-vdso.so.1 =>  (0x00007ffc0e197000)
	libcrypto.so.1.0.0 => /lib/x86_64-linux-gnu/libcrypto.so.1.0.0 (0x00007f5c5ab6c000)
	libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007f5c5a7a7000)
	libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2 (0x00007f5c5a5a3000)
	/lib64/ld-linux-x86-64.so.2 (0x00007f5c5b1c1000)

So it's not portable yet. I initially tried using the chrpath command but it can only modify existing RPATH records. My binary doesn't have any RPATH yet. So that didn't work. The next thing I tried was the patchelf command from here:

http://nixos.org/releases/patchelf/patchelf-0.8/

After building and installing I issued this command in the OpenSSL folder:

patchelf --set-rpath '$ORIGIN' ./libssl.so.1.0.0

Now to check if libssl pulls in the local libcrypto:

matthew@ubuntu:~/Downloads/openssl$ ldd ./libssl.so.1.0.0 
	linux-vdso.so.1 =>  (0x00007ffcdce86000)
	libcrypto.so.1.0.0 => /home/matthew/Code/Scribe/trunk/./libcrypto.so.1.0.0 (0x00007f39b99d8000)
	libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007f39b9613000)
	libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2 (0x00007f39b940f000)
	/lib64/ld-linux-x86-64.so.2 (0x00007f39ba0a5000)

Yes! Now this build can be installed in the same folder as the Scribe binary and Scribe will use it without interfering with the system OpenSSL which is often out of date, or not built with -DPURIFY or missing.

(0) Comments | Add Comment

Scribe 2.1
Date: 27/10/2015 Tags: scribe	Well it's been most of a year since I said anything about Scribe v2.1, and the list of changes hasn't moved forward much in that time. There are a few reasons for the slow progress, but it mostly boils down to: Working on various music projects, including a audio/video mix down of a band that shot a live set earlier in the year and a cover of New Day. Working on the MC2 foot controller. And the old RSI is back to haunt me. Anyway there is some good news. The integration of Aspell is complete. Scribe v2.1 will come pre-loaded with a spell check and it will download and install dictionaries on the fly with no user intervention. Well you'll need an internet connection for that part, but once installed it'll work offline. Secondly I'm in the middle of doing a fairly thorough rewrite of the Help files. Which includes fixing some things in the HTML rendering. Reviewing all the parts of the application that have "Help" buttons (or should have "Help" buttons) and making sure they work correctly. Updating all the pages with current information. It's actually a big slab of work. Lots of cross referencing things and checking the UI and source code so I accurately describe the functionality. I'm not much of a tech writer so it's tedious for me. Anyway there is now a beta release online for v2.1. I look forward to getting feedback on what works and what doesn't.
(9) Comments \| Add Comment

MC1 flash procedure update
Date: 19/10/2015 Tags: mc1	The original MC1 flash procedure involved having to use an actual parallel port. For this reason I have kept an old PC running XP around. Just to flash the firmware from time to time. As it didn't work so good on Windows 7 or pretty much any Mac. Then a few months ago someone was asking about the possibility of using a USB in circuit programmer like this. So seeing as they are dirt cheap on ebay, I bought one to try out. Compiling the code into AVR instructions is accomplished with the WinAVR tool chain. I'm using the latest release. However there is one gotcha when installing. Do NOT install it into a folder with brackets in the name, like say: C:\Program Files (x86)\WinAVR This causes make.exe to crash. No I'm not kidding. The next issue is the Avrdude binary that comes with the WinAVR compile is just a little too old to know about the usbasp programmer. So that needs to be updated to something recent. Just unpack that file into the bin folder of the WinAVR install, overwriting the existing avrdude files. Check that the version available on the command line is correct by running it without any arguments. Next the usbasp needs a driver to work in Windows (7 in my case). You can get that here. I didn't have much luck with the installer, but if you go into the Device Manager, right click on the usbasp device and install drivers from that menu. Then select the folder where you unpacked the drivers. On my machine the device driver takes 10 seconds to connect so give it time before trying to flash the firmware. Finally the actual command you use to flash the firmware is: avrdude -c usbasp -p m128 -U flash:w:main.hex Where 'main.hex' is the compiled binary. It's faster and easier than the old PonyProg2000 method, and scriptable. Just put it in the makefile. Less mucking around. Now I'll have to find another excuse to have an old PC lying around. Maybe LAN games with the lad? :D
(0) Comments \| Add Comment

MC1 flash procedure update

Date: 19/10/2015
Tags: mc1

The original MC1 flash procedure involved having to use an actual parallel port. For this reason I have kept an old PC running XP around. Just to flash the firmware from time to time. As it didn't work so good on Windows 7 or pretty much any Mac. Then a few months ago someone was asking about the possibility of using a USB in circuit programmer like this. So seeing as they are dirt cheap on ebay, I bought one to try out.

Compiling the code into AVR instructions is accomplished with the WinAVR tool chain. I'm using the latest release. However there is one gotcha when installing. Do NOT install it into a folder with brackets in the name, like say:

C:\Program Files (x86)\WinAVR

This causes make.exe to crash. No I'm not kidding.

The next issue is the Avrdude binary that comes with the WinAVR compile is just a little too old to know about the usbasp programmer. So that needs to be updated to something recent. Just unpack that file into the bin folder of the WinAVR install, overwriting the existing avrdude files. Check that the version available on the command line is correct by running it without any arguments.

Next the usbasp needs a driver to work in Windows (7 in my case). You can get that here. I didn't have much luck with the installer, but if you go into the Device Manager, right click on the usbasp device and install drivers from that menu. Then select the folder where you unpacked the drivers. On my machine the device driver takes 10 seconds to connect so give it time before trying to flash the firmware.

Finally the actual command you use to flash the firmware is:

avrdude -c usbasp -p m128 -U flash:w:main.hex

Where 'main.hex' is the compiled binary. It's faster and easier than the old PonyProg2000 method, and scriptable. Just put it in the makefile. Less mucking around.

Now I'll have to find another excuse to have an old PC lying around. Maybe LAN games with the lad? :D

(0) Comments | Add Comment

Mc2 Progress #4
Date: 12/10/2015 Tags: mc2	The time comes to stop mucking around with software and start building the hardware. I'm no where near as good with the hardware things. Software is easy and fine... so how much did I break it? Turns out a fair bit. Lets take a look. First I started with an Inkscape file that described all the positions of the components and specifically the places where I have to drill or cut. I printed it out in A3 and glued it down to the top surface of the chassis. I started drilling out the holes and all the switch / neopixel holes went fine. Then came the LED rings and the paper started tearing everywhere. So I left the LED ring holes for a while and worked on the screen rectangle. I bought this tool called a "nibbler" to cut the hole. You drill a hole in the area to be removed, and insert the nibbler into the hole and start cutting. It can cut in a curve so I used that to align to the border. Then I came back and finished all the LED ring holes. However my accuracy was terrible and I don't feel all those holes are good enough. I've decided that this chassis is a bit of a write off. I have since gone and bought a new chassis (they are about $60 AUD inc shipping so not too bad). And now I'm looking for someone that can do CNC drilling. Moving on I tried test mounting some of the components, starting with the screen and Pi. Seems to fit quite nicely. So I tried to bend the holes in to accommodate the countersunk machine screws that I had bought. That bent the metal around the hole too much and deformed the chassis around the screen area. In hind sight I should've used a large drill bit instead of mechanical force. So because of the deformations in the metal when I tightened up the screws around the screen I heard a faint cracking sound. Yup I broke the screen just a little. That's just fantastic! :( Another thing I have to re-order. Similar price too... bah. There is some good news thought... the screen still works: Although I haven't had a chance to test the touch screen part. The PCBs and Neopixels need to be mounted behind the front of the chassis, so I went and bought some rectangles of 3mm perspex: The switches go through the perspex and keep it in place. I've started soldering some 3 wire ribbon cable between the Neopixels, not shown in this photo. It's slow and tedious work :( And the various glue electronics to interface the Pi with the buttons, encoders, expression pedal ports and neopixels is all mounted to that as well.
(0) Comments \| Add Comment

Why are my videos encoding at a snails pace?
Date: 8/10/2015 Tags: hardware	I'm currently rendering out composites for a music video and it's crawling along. Vegas is taking 40+ minutes for a 8 minute 1080p video? That doesn't sound right. My CPU is a i5-3570K and the machine has 16GiB of RAM (5 in use right now). So I started looking at the CPU to see if it's running OK. First stop CPU-Z: 2ghz? What? Where is my 3.8ghz the CPU is capable of? Hmpf. Well better check the temps then. Oh... oh dear. Well... I er... guess I need to replace that stock cooler then huh. Off the shops, will update with results. Edit: So I cleaned out the fan, it was clogged with dust. Reapplied thermal grease and reinstalled the heatsink/fan. I think it might've come loose and not have been making proper contact with the CPU? In any case: At least now I'm getting decent speed. The render time is down to 21 minutes or so. I wonder if I can overclock with a decent cooler?
(0) Comments \| Add Comment

MC2 Progress #3
Date: 23/9/2015 Tags: mc2	The last week or so has been spent a twisty maze of fonts, colour spaces, CPU endianess and cross platform C++. Anyway I won't regail you with all the fun I've had but I will show you some screenshots. Firstly because I have RGB LEDs for each Instant Access switch, I need a way for the user to select a colour: It remains to be seen how easy it is to use on the real hardware. But colour selection works either by hitting one of the palette entries or pushing the Red, Green and Blue sliders around. Also for entering details like the CC number and other textual information I probably need a touch screen keyboard: This is roughly copied off my Android phone's touch keyboard. Without Swype unfortunately ;) The screen to select a scene change for an IA button: And the page where the user picks an Axefx block:
(0) Comments \| Add Comment

Sitemap