Reverse engineer video BIOS initialization
[email protected] Tue, 26 Jun 2007 10:29:33 +0200 (CEST)
| Newsgroups | gmane.linux.ports.hppa |
|---|---|
| Message-ID | <[email protected]> |
Hi, recently I bought a Geforce PCI FX5200 and some HP j6750. (OK, it was the other way around ;-) Of course, the graphics card do not work with xorg on this machine. In your article "Graphics cards with PA-RISC Linux", you stated, that such a graphics card has to be awaken by its BIOS and that the exact wake-up sequence is yet unknown. It happens, that I have access to some up-to-date logic-analyzers. So I plan to attach some probes to the pci-slots of an old x86 mainboard. The LA then can record all read- and write-accesses on the PCI-bus addressed on the graphics-card. It should be very simple to write a small c-code that will reproduce these reads- and writes. My question now is: Where would be the best place to execute such a wake-up code? As the company I work for is more into IA, I don't know much about PA-RISC architecture. Does PA-RISC machines have a kind of real-mode in which all BIOS-codes must run? Or would it be sufficient to issue the wake-up code at any time on an mmaped memory? Thanx, Gregor Rebel