- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
In trying to get the new distro to load and run, we see boot errors about "attempting to access beyond end of device." We have a 16MB device on the board (8Mx16) that is directly supported by NIOS2, and the kernel + ROMfs use less than half of it. So, there isn't any reason for the code to think it is going past the end of the device. This error occurs if we use a new ROMfs generated by the new distro, or if we use a ROMfs generated by the previous distro. So, it's probably not the ROMfs that's the problem.
Here are the bootscreen and .config file. Can anyone spot the feature we need to turn on or off to get past this?
Much appreciated.
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
The MTD map for Altera's kits is for 8 bit flash, while your flash chip is 16 bit wide. I think this is the problem. There is another map for Microtronix kits, which features 16 bit flash chips. In order to use that, you have to set the platform as microtronix cyclone kit. You also need to change the size of the map, which right now is only 8M.
The changes in the map definition brought some inconveniences, but also some benefits, such as jffs2 support, etc.- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Any chance that these "changes" are documented somewhere? I suspect that more of these undocumented "changes" are causing our boot problems.
Note: changing the target board to the Altera 1C12 Board Support option made it almost, sorta boot. Once. Choosing the Microtronix boards did nothing. The one time it did sorta boot I set up the eth0 interface and pinged my router, which responded in 0ms. I was hopeful, but the level-3 MTD diagnostic messages were annoying - back to the IDE. I changed the configuration to eliminate the MTD debug messages (note - that's the only change I made.) After recompiling and uploading, I was dead in the water. And when I say "dead" I mean "not a single message" dead. Very bad. So I put the debug messages back in at Level-1 and reloaded. Now it booted, but crashed with an "attempt to access past end" message just after announcing the preemptive scheduler. So I put the debug messages back in at Level-3 (just like when it booted that one, lone time) and reloaded. Now it booted, but crashed with an "attempt to access past end" message just after announcing the preemptive scheduler - same symptom as the Level-1 debug messages. So, if anyone can provide a list of configuration changes, I think we could sort this out. Short of that, I have to give this release a huge thumbs down. It just doesn't work. When are some patches due?- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi thosburke,
When you choose altera 1c12 board, the microtronix mtd map is selected. It is located in kernel source drivers/mtd/maps/microtronix.c, and right now the romfs is allocated 6M, as follows,static struct mtd_partition microtronix_partitions = {
{
.name = "romfs",
.size = 0x600000,
.offset = 0x200000,
},{
.name = "loader/kernel",
.size = 0x200000,
.offset = 0,
}
};
I don't know if you have changed the size of romfs partition. Since you have 16M flash, you can increase that size and have a try. At the same time, I am tring to get the old romfs map to work, and will post that shortly.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I'll try bumping the size up, but 6M should suffice for a romfs that is smaller than 6M. I think it is something like 5.5M in this latest distro (when you check all the romfs options).

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page