|
Thank you for your information, I am busy with getting XMK working and
building an application with his own makefile (which specifies another start
address as the bootloader). At this moment one thing is not clear to me: how are
interrupts handled? The microblaze jumps to address 0x18, but in my case that's
the bootloader in block ram. And I want to handle the interrupts in my
application of coarse. Is that possible and if so,
how?
TIA,
Frank
Frank wrote:
Hello,
I have build a bootloader which is located in block ram. Now I want to
download my final application to sdram and execute it. If I'm correct,
I've
to make a linker script in order to make this
possible. Not necessarily. You can simply specify a different
start address for your boot loader and your application on the mb-gcc command
line. See the Makefile_mb.sh and other Makefile*.* files related to MicroBlaze
in the Xilkernel install area.
Besides this, I want
to use the xilkernel in my
application, but not in the bootloader, is this
possible? Yes.
I guess I have to convert the application.elf file
to a binary
file in order to be placed into sdram by the
bootloader?! Depends on your bootloader. If your bootloader
expects binary then you have to. I have seen bootloaders that use other
formats as well, such as SREC or some application specific encoding.
How are the
interrupts handled? Is the interrupt
handler from the bootloader used
(because it default jumps to address
0x18)? Can I install a new interrupt
handler which is located in my
application?!
a lot of questions, I searched at the forums, but there are not much
examples available. I'm sure there a people who already did this before.
There are some examples of Xilkernel usage in the
install area itself (search for print_thread.c).
Please help,
thanks
|
