6 Apr
2005
6 Apr
'05
10:10 p.m.
Phillip Susi wrote:
So... is any of this massive rewrite in a branch in
the repository or
is it all still in his working copy?
Thomas will have a diff of my Pushlock patch and collaborate on
finishing the missing function.
The Kernel re-write will be in a branch in May as soon as it's bootable.
I am against branching unbootable versions of ROS. Buggy, yes, but not
unbootable. As soon as it's bootable (even if it crashes every 5 seconds
and doesn't run anything anymore), it will be branched. Also, my design
isn't finished yet, so it's another reason why it's not ready for
primetime (for example, I haven't even added NPX saving for user-mode
context switch), and I've added no SMP support yet.
The same applies for the ob re-write. It is incomplete, doesn't even
compile (I would write it on the bus in notepad without even checking if
it compiles..it's probably full of typos). As soon as it's bootable,
I'll branch it for comments.
As for Spinlocks/IRQL, it's also incomplete... I've written only 1/4th
of the necessry HAL routines, and this too is not yet ready for review
(useless, imo, to review something that isn't even complete).
I'm quite interested in the changes so even if it's not ripe yet I'd
like to svn switch to the branch and take a look at the changes so far.
If you really really want to see some of the new functions in the
scheduler I can send them to you on a case-by-case basis. This is not
yet possible for the OB stuff because it's much less ripe. It's still a
flower, not a fruit :)
I can also send you the basic IRQL stuff I've written (only 4 functions
so far) if you really want to see it as well.
Best regards,
Alex Ionescu