I've seen this before and wasn't sure what to do or think about it. I'm busy compiling the PTX dist from http://www.oselas.org/oselas/bsp/pengutronix/mini2440_bsp_en.html When I ran the ./configure script I got the following check: "checking whether we are cross compiling... no" I had a similar response when compiling QT 4.6.2 for the mini2440. Shouldn't the answer to this check = yes ??? Or what should answer be for the following scenarios: 1.) Compiling a bootloader? 2.) Compiling a kernel? 3.) Compiling a FS? 4.) Compiling a cross compiler? 3.) Compiling QT embedded? Thank you
Running ./configure script.... are we cross compiling?
I don't think you just run ./configure. Start on page 8, extracting sources <ptxdist setup> page 10 2.3.3 2.3.4 Protecting the toolchain (haven't done this myself ... yet) 3.1 3.2.2.1 3.3 just a check 3.4 just a check 3.6 <ptxdist go> 3.7 <ptxdist images> 3.7.1 Deploying the Mini2440
1.) Compiling a bootloader? 2.) Compiling a kernel? 3.) Compiling a FS? 4.) Compiling a cross compiler? 3.) Compiling QT embedded? 1 and 2 run on the target so they need to be cross-compiled. For 3 the term usually seems to be "build a rootfs", it runs on the target so I guess you can say it is cross-compiled. 4 the cross-compiler is complied to run on the host, so i386 or i686 (not cross-compiled) 5 QT embedded has to be cross-compiled.
Hi dave, I didn't just run ./configure. I followed the procedure in the Mini2440 Quickstart PDF (extract, make, make install, setup project directories) I then extracted the toolchain and selected th ptx config ptxdist select ptxconfigs/arm-v4t-linux-gnueabi_gcc-4.5.2_glibc-2.13_binutils-2.21_kernel-2.6.3 6-sanitized.ptxconfig but that is where it stopped when I ran in to the following problem: ptxdist: error: md5sum for 'glibc-ports' (/home/jdoe/local/OSELAS.Toolchain-2011.03.0/src/glibc-ports-2.13.tar.bz2) missing ......sigh, the problems, the problems. I'm stuck again because I do not know how to resolve this problem. I tried to manually download the archive and place it in the src folder but that didn't help. I do not know whye its getting a MD5 checksum error or how does it know there is an error if it does not have MD5 checksum to verify against. I've noticed that in most cases the repositories has a MD5 checksum next to the respective archive, but in this case there is no MD5 checksum for the archive. Do you have any ideas?
Ah, read the issues at the bottom of the page: http://www.oselas.org/oselas/bsp/pengutronix/mini2440_bsp_en.html I built my toolchain with 2011.05.0 then up-graded to 2011.06.0 Just start with 2011.05.0 and upgrade after you have had a play with that.
10.04LTS all current updates. I will be very interested if you are able to tick the Qt and Arora boxes :)
May I ask why you are so interested in Arora? I'm about to compile the rootfs. On p.15 , 3.2.2.1 Bootloader barebox, I've edited the barebox-NAND-128.config to my requirements, but they state: "Don’t forget the setting of an unique MAC address. At least the entry eth0.ethaddr must be set." I cant find this variable in the default config file. Should I add it manually? How did you do it?
Because Arora has a full screen mode and the POST command works. For my home automation system. I run a webserver on an ATmega328 and I want to use Arora for local control. I set up the host up with a statically assigned IP then filled in all the relevant details in /env/config In fact, if you look in barebox-64M and the config file you will see my set-up. I am going from memory as my Linux setup is at home and I am at work. Which country are you in? NZ here.
Ahh, I see (env/config) is the place. I'm from ZA (South Africa). You said your host distro has all the latest updates. Everytime the update manager window pops up I avoid it because I fear the updates might affect my current setup and toolchains. Have you experienced any problems when doing updates?
You must be burning the midnight oil! Problems doing updates. Who knows, I just trust that they know what they are doing! Get the basic rootfs working and then if you have time before morning comes try Qt :)
Your project sound interesting. The Atmega328 is a small device. I would assume you use it at the various ethernet control points? Fell asleep before I got to start compiling the rootfs. Will try and compile the rootfs today, and then try Qt :)
Do you mean the "configure" from the PTXdist package? Or form any other package PTXdist tries to configure? If you mean the PTXdist package itself, then the answer must be 'no', as PTXdist is a host tool - it runs on you host. In PTXdist we have three types of tools: - host: Thy run on your host and mostly create data for the host - cross: They run on your host, but create data for the target - target: They run on your target and create data for the target So, if you take a look into the logfile, you can see this answer changes on a per package base.
> ptxdist: error: md5sum for 'glibc-ports' >(/home/jdoe/local/OSELAS.Toolchain-2011.03.0/src/glibc-ports-2.13.tar.bz2) > missing Marc will provide us with a toolchain update, then also the toolchains can be built again with a more recent PTXdist. Until this happens, we must continue to use an older PTXdist version for the toolchain.
bluescape: You can switch off the MD5 checks Run "ptxdist setup" Source Download ---> archive check (always) ---> (X) for existing checksum or "(X) never". Its up to you.
bluscape, RS485 to interface to 5 or 7 outstations. Off-grid, so one for picohydro, one PV panels and one spare for wind. How's the roofs going?
bluscape, > "Don’t forget the setting of an unique MAC address. At least the entry > eth0.ethaddr must be set." > I cant find this variable in the default config file. Hmm, when I take a look into the "configs/platform-friendlyarm-mini2440/barebox-128m-env/config" I find this text block: [...] # or set your networking parameters here #eth0.ipaddr=a.b.c.d #eth0.netmask=a.b.c.d #eth0.gateway=a.b.c.d #eth0.serverip=a.b.c.d #eth0.ethaddr=a:b:b:d:e:f <---- Here! [...] If you need faster response you could also try our maillist.