is this new instruction for desktop cpu lineups too? so I would have to buy a new cpu finally after years?
Announcement
Collapse
No announcement yet.
Intel Posts Initial Code For x86 User Interrupts On Linux - Shows Great Performance Potential
Collapse
X
-
Originally posted by sinepgib View Post
You mean Fuchsia, the one referred by ermo in the post you quoted?
What i mean is if Alphabet, the parent company of Google, is going to pay anyone to do anything they're going to pay people to work on their own projects. I hope that clears things up for you.
Comment
-
This could be quite handy for an embedded system and would have been useful for some custom hardware projects I've been involved with. When I need this sort of thing I have had to e.g. write my own driver and set up ioctl() calls that wait on a condition variable so that the driver's interrupt handler can signal the condition variable and allow the ioctl() to return. The exact details are messier but that is the basic idea. The big question for me is the execution context of the interrupt callback. If its e.g. a callback from some kernel thread then there are user space actions that shouldn't be done in that context and something like a condition variable handoff to a user thread would still be needed. Even so, it looks like it would be good to have either way.
Comment
-
Originally posted by Paul Frederick View Post
Do I?
What i mean is if Alphabet, the parent company of Google, is going to pay anyone to do anything they're going to pay people to work on their own projects. I hope that clears things up for you.
All ermo said is, if wishful, consistent with that, except maybe the Intel whiteboxes. Writing userspace drivers for Linux that are compatible with Fuchsia is useful for them to allow a more gradual deploy and easier marketing, for example. Having the source code of the firmware they run available suits them as well (the exception being they can very well keep it to themselves and still reap the benefits).
- Likes 1
Comment
Comment