Re: Fast User Space Mutexes (aka Futexes)

[Hiten Pandya <hmp@xxxxxxxxxxxxx>]

Bill Huey (hui) wrote:

: On Sun, Oct 26, 2003 at 09:29:42AM +0000, Hiten Pandya wrote:
:
:>Bill Huey (hui) wrote:
:>
:>>http://people.redhat.com/drepper/futex.pdf
:>>
:>>Better documentation on how it's used and some implementation issues with
:>>glibc. The problem with futex's in the context of dfBSD is that it's
:>>used for their 1:1 threading model, which is different than the M:N
:>>stuff that's being done recently. I think it's essential to think about
:>
:>	I recall Rusty Russell added the Asynchronous interface to the
:>	Futex interface, yes, I am referring to the FUTEX_AWAIT and
:>	friends.  It was needed for the NGPT work, since it is M:N
:>	effectively.
:
: I didn't see that interface mentioned in the above docs. Should I look
: at the most recent kernel sources ?

	It always helps to have a Linux 2.5/2.6 kernel around for
	getting enlightened.  Yes, I am not exactly sure when the
	change was made, but I recall Rusty posted a patch on the
	Linux kernel mailing list, later intergrated into the
	2.5 kernel.

	The AWAIT (async) interface gave birth to the whole /dev/futex
	thing, and friends, such as futex_poll(), futex_open().  Some
	code with an example userland library implimentation was
	attached in Rusty's mail.

	Checkout for more goodies:

		www.kernel.org/pub/linux/kernel/people/rusty/

	Regards,

--
Hiten Pandya
hmp@xxxxxxxxxxxxx