From mboxrd@z Thu Jan 1 00:00:00 1970 MIME-Version: 1.0 In-Reply-To: <1955b997f424ec1e9d2021d52e43ece3@mikro.quanstro.net> References: <1955b997f424ec1e9d2021d52e43ece3@mikro.quanstro.net> Date: Sun, 9 Mar 2014 16:42:05 -0400 Message-ID: From: yan cui To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net> Content-Type: multipart/alternative; boundary=089e0160bd82157b6c04f4328597 Subject: Re: [9fans] GSOC 2014 proposal for Plan 9 Topicbox-Message-UUID: c707792c-ead8-11e9-9d60-3106f5b1d025 --089e0160bd82157b6c04f4328597 Content-Type: text/plain; charset=ISO-8859-1 Erik, thanks for your reply! I will read the document carefully and submit an application. Best Wishes! Yan 2014-03-09 16:30 GMT-04:00 erik quanstrom : > On Sun Mar 9 15:32:48 EDT 2014, ccuiyyan@gmail.com wrote: > > > Dear All, > > > > I am a student of Columbia university(Yan Cui) and want to > > contribute to > > plan 9 in google summer code. I am interested in one idea posted on > > the idea list, which is titled "MCS locks (i.e. queueing locks) for > > the kernel with a compatable calling interface to > > *lock*(9)". > > Actually, my research direction is operating system scalability on > > multicore systems and knows kinds of synchronization methods very well > > including Linux kernel ticket lock, MCS lock, CLH lock, etc. For MCS > > lock, I have done many experiments (such as replacing the Linux kernel > > spin lock with MCS lock) and made improvements to it. I think I have > > all backgrounds and experiences required. I believe I can complete > > the task in time. Wonder to know can I join you in the Plan 9 > > development? > > excellent! i think you will find that plan 9 lends itself to this > sort of work. you are certainly welcome at any time to join plan 9 > development. > > the key here is to work out how to keep the existing lock api. > this is described here: http://www.9atom.org/magic/man2html/9/lock > i've prototyped this in user space, but did not get any further. > > as for gsoc, please do submit an application. there does happen to > be another student interested in MCS locks. so we may need to > work out something to avoid having to pick between two good > applications. but i am confident that we can figure this out as > things move forward. > > if you have any more questsion, please ask the list, or contact > me off list via email or g+. > > cheers! > > - erik > > -- Think big; Dream impossible; Make it happen. --089e0160bd82157b6c04f4328597 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable
Erik, thanks for your reply! I will read the doc= ument carefully and submit an application.



Best Wishe= s!
Yan


2014-03-09 16:30 GMT-04:00 erik quanstrom <quanstro@quanstro.net&g= t;:
On Sun Mar =A09 15:32:48 EDT 2014, ccuiyyan@gmail.com wrote:

> Dear All,
>
> =A0 =A0 I am a student of Columbia university(Yan Cui) and want to
> =A0 =A0 contribute to
> plan 9 in google summer code. =A0I am interested in one idea posted on=
> the idea list, which is titled "MCS locks (i.e. =A0queueing locks= ) for
> the kernel with a compatable calling interface to
> *lock*(9)<http://plan9.bell-labs.com/magic/man2html/9/loc= k>".
> Actually, my research direction is operating system sc= alability on
> multicore systems and knows kinds of synchronization methods very well=
> including Linux kernel ticket lock, MCS lock, CLH lock, etc. =A0For MC= S
> lock, I have done many experiments (such as replacing the Linux kernel=
> spin lock with MCS lock) and made improvements to it. =A0I think I hav= e
> all backgrounds and experiences required. =A0I believe I can complete<= br> > the task in time. =A0Wonder to know can I join you in the Plan 9
> development?

excellent! =A0i think you will find that plan 9 lends itself to this<= br> sort of work. =A0you are certainly welcome at any time to join plan 9
development.

the key here is to work out how to keep the existing lock api.
this is described here: http://www.9atom.org/magic/man2html/9/lock
i've prototyped this in user space, but did not get any further.

as for gsoc, please do submit an application. =A0there does happen to
be another student interested in MCS locks. =A0so we may need to
work out something to avoid having to pick between two good
applications. =A0but i am confident that we can figure this out as
things move forward.

if you have any more questsion, please ask the list, or contact
me off list via email or g+.

cheers!

- erik




--
Thin= k big; Dream impossible; Make it happen.=A0=A0
--089e0160bd82157b6c04f4328597--