Re: [ANN] Userspace M-on-N threading model implementation. Alpharelease.

[Arjan van de Ven]

> > On the other side, the overhead you need to add for every single syscall
> > that might block for the M:N threads and the associated complications
> > which make it far harder to conform to POSIX IMHO far outweight the costs
> > of going into the kernel for a context switch.
> 
> That really wasn't my question, Arjan said that switching real threads 
> wasn't a context switch in the hardware sense, and I was asking if I 
> missed something.

a hardware context switch is basically a CR3 change with associated tlb
flush. That is the part that is the most expensive of a context switch.
Just going into the kernel and getting out with a different EIP/ESP is
really cheap, in the order of "a few hundred cycles"; not a heck of a
lot more expensive than a simple getpid or other simple system call.


>  It may be cheap, but it would seem to be a context 
> switch none-the-less.

it includes a privilege level switch, not so much a full context
switch...

-- 
if you want to mail me at work (you don't), use arjan (at) linux.intel.com
Test the interaction between Linux and your BIOS via http://www.linuxfirmwarekit.org

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/