1. PIRQ table derived from standard bios, but it is still wrong. 2. have a hook in arima/hdama/mainboard.c for CONF_PASS_PRE_BOOT, but it is a no-op 3. targets/arima/hdama/Config.lb now enables SMP and sets max_cpus to 2
ron
ron minnich rminnich@lanl.gov writes:
- PIRQ table derived from standard bios, but it is still wrong.
Of course we change all of the device id, on bus 0. You simply can not cut and paste a binary table any more.
- have a hook in arima/hdama/mainboard.c for CONF_PASS_PRE_BOOT, but it is a no-op
Once these things merge I have a few more comments. But the more I see of it, the enumeration value instead of separate functions looks increasingly like the wrong thing.
- targets/arima/hdama/Config.lb now enables SMP and sets max_cpus to 2
Good. Hopefully I can do some interesting development now and get the static configuration where it needs to be.
Eric
On 5 Aug 2003, Eric W. Biederman wrote:
Once these things merge I have a few more comments. But the more I see of it, the enumeration value instead of separate functions looks increasingly like the wrong thing.
hmm, not sure we agree from here but we're open to discussion. Greg had an interesting point that given the enumeration value, hardwaremain could actually be a fairly straightforward state machine, and lots of current special cases could go away.
Also, inserting a new enumeration value is still lots easier than inserting a new special function all over the place, but as always on this project we can evolve by working this out on the list.
ron
-
ebiederman＠lnxi.com -
ron minnich