Most of the instructions can be found in the Intel Quark Board Support Package (BSP). The prerequisites, as stated in the BSP, are:
- A host PC running Linux. 64-bit Linux system is recommended.
- An internet connection to download some third party sources.
- At least 30GB of free disk space.
- A way to program the processor, in this case you’ll be using your Galileo board.
- The BSP sources.
The latest BSP is 1.2.0, but for Galileo it is recommended to use 1.0.1, given that this one is more suitable to the board. Here you can find the link to the BSP Build and Software Guide 1.0.1, http://download.intel.com/support/processors/quark/sb/quark_bsp_buildandswuserguide_329687_006.pdf. That document also provides the link to the BSP sources, but you’ll be redirected to the BSP 1.2.0 files, for older files and packages check this link https://downloadcenter.intel.com/download/23823.
Thanks Pablo for your reply, I have a few additional questions:
I would like to build Yocto image for Galileo Gen2 that include the SQLite3 DB, for a moment I thought it was already there on
Linux galileo 3.8.7-yocto-standard #1 Tue Sep 29 22:16:33 GMT 2015 i586 GNU/Linux, since I found sqlite3.h in the /usr/include directory..but I could not run sqlite3.
1.Can I use Intel Firmware Engine to create this custom yocto image with SQLite3(any)db for Galileo?
2.Can I use(fairly easily) a Windows PC with Linux running in VMware or Oracle VM VirtualBox to build the image?
3. What is the fastest/easiest way to do it?
OK so I'm getting quite far, but hitting a road block.
When I get to the stage "Creating a file system and building the kernel using Yocto" and run ./setup.sh I get the following fatal error:
Repository not found. This is when it tries to git clone git://github.com/openembedded/meta-oe meta-oe having a look on github it looks like that project has renamed a few things, or moved them or something. They do have a meta-oe project but it is at github.com/openembedded/meta-openembedded/meta-oe so I'm not sure if I can just update the path in setup/gitsetup.py or not.
JadeDragon – Both option are feasible, but I was thinking that you could also download SQLite from the AlexT repository, have you considered that option? I believe that would be the easiest option as you won’t need to build the image, just get the package. Take a look at this thread for some more information https://communities.intel.com/thread/58427.
SpiderKenny – Did you already try changing the path just like you said? Did the error disappear? If not, I would suggest you to try this and see what results you get.