[acc25ee] | 1 | The code in this directory has been taken WITH PERMISSION from |
---|
| 2 | Gabriel Paubert, paubert@iram.es. The main reason for having |
---|
| 3 | a separate bootloader for PreP compliant firmware is that the |
---|
| 4 | initial code is relocated by firmware at an unknow address |
---|
| 5 | (actually 0x5000 on motorola MCP750) and that as Gabriel I |
---|
| 6 | think having a relocatable bootloder code is a must. |
---|
| 7 | |
---|
| 8 | So the way of building a binary executable that can be booted via |
---|
| 9 | hard disk or network boot goes like this : |
---|
| 10 | |
---|
| 11 | - make a RTEMS executable, |
---|
| 12 | - put is as data section in the bootloder binary, |
---|
| 13 | - relink the loader (see make-exe macros for details), |
---|
| 14 | |
---|
| 15 | I would like to thank Gabriel for his support and his code. |
---|
| 16 | The original code can be found in form of a patch to official linux |
---|
| 17 | kernel at (I insist not vger ppc kernel or Imac ppc kernels!!) : |
---|
| 18 | |
---|
| 19 | <ftp://vlab1.iram.es/pub/linux-2.2/> |
---|
| 20 | |
---|
| 21 | After applying the patch, the code is located in a new directory |
---|
| 22 | called prepboot. |
---|
| 23 | |
---|
| 24 | (NB : note use ftp not netscape...) |
---|
| 25 | |
---|
| 26 | Note that the actual code differs a lot since Gabriel choose to use |
---|
| 27 | a CHRP compliant mapping instead of a Prep Mapping to save |
---|
| 28 | BATs. I had no time to upgrade the code to its new one allthough |
---|
| 29 | I agree it should be done... |
---|
| 30 | |
---|
| 31 | I have also splitted the original code to have a more modular |
---|
| 32 | design enabling to reuse code between the loader and RTEMS |
---|
| 33 | initialization (e.g printk, ...). |
---|
| 34 | |
---|
| 35 | Eric Valette (valette@crf.canon.fr) |
---|
| 36 | |
---|
| 37 | |
---|
[3a3e0b0e] | 38 | ************************************************** |
---|
| 39 | 2003/5/7, Greg Menke, gregory.menke@gsfc.nasa.gov |
---|
| 40 | |
---|
| 41 | Reworked the pci bus 0 initialization a little and added support for |
---|
| 42 | configuring an arbitrary number of other busses & their respective |
---|
| 43 | bridges. Also added support for configuring IO ranges below 0x10000, |
---|
| 44 | which I think is reasonable given this is a PowerPC bsp. |
---|
| 45 | |
---|