Re: [rtmk-discuss] regarding migrating threads

[Johan Rydberg]

On 2002.02.08 13:06 Carlos Manuel Duclos Vergara wrote:

: Hello,

Hi Carlos!

I'm CC'ing the list, of archival purposes.

: what about resources?
: i mean, will an activation  hold resources? or will resources belong to
: processes?
: i'm not sure about that point, but i think needs some thoughts.

The concept of "processes" does not exist in rtmk.  Instead there's
threads, tasks and now activations.  The old notion of a "process" is
in rtmk a task, with one thread populating an activation.

Regarding resources.  I have not yet figured that out really.  As it
seems now there's only need for one kernel stack per thread, not
per activation.  If this is possible (which I think it is), it will
reduce the number of kernel stacks.

If you're thinking about "holding resources" as in locking a
resource (using a mutex or whatever) this will be done by the thread,
since the thread holds the scheduling state.  Remember that an
activation is just, in theory, a set of hardware dependent states
(i.e. hardware registers).

: BTW, i'm interested in running this micro kernel on Motorola Coldfire
: processors, but this will take some long time because i'm finishing some
: things before start this port.

Sounds great.  I have tried to make the kernel as portable as
possible, so I guess there should be no real problem with porting
it to the Coldfire architecture.

I just wanna state this; Currently the PPC port is broken.  And I
will probably not work anything on it until I write a good bootloader,
or find one that I can use.  The bootloader must be able to read
file systems and load both kernel and modules.  Much like the GRUB
bootloader for the x86-arch.

brgds
johan


--
Johan Rydberg

$ ON F$ERROR("LANGUAGE","ENGLISH","IN_MESSAGE").GT.F$ERROR("NORMAL") -
             THEN EXCUSE/OBJECT=ME