[cgl_discussion] FW: Kick-off the Review of the Device Driver
Hardening Design Spe c
rob.rhoads at intel.com
Thu Sep 19 13:22:11 PDT 2002
Exposure of the hardened drivers project has always been my biggest concern.
As I stated previously in my original e-mail, I will in fact announce it on
the lkml. I wanted to shoot it out to CGL first and gauge reaction, before
spamming it out on the lkml. I completely agree, that without further
feedback/exposure to the Linux kernel developer community it will never take
off and will never make it into the mainstream kernel sources. If it doesn't
make it into the kernel mainstream, I would go so far as to say that we are
just wasting our time.
Stay tuned, you will see it out on lkml very soon.
> -----Original Message-----
> From: Randy.Dunlap [mailto:rddunlap at osdl.org]
> Sent: Thursday, September 19, 2002 11:58 AM
> To: Rhoads, Rob
> Cc: 'cgl_discussion at osdl.org'
> Subject: Re: [cgl_discussion] FW: Kick-off the Review of the Device
> Driver Hardening Design Spe c
> Hi Rob,
> I have serious doubts that announcing and reviewing the
> driver hardening spec and drivers _only_ on the hardeneddrivers
> mailing list is sufficient.
> I even doubt that doing so on the cgl_discussion list is
> sufficient. cgl_discussion so far just isn't being used
> (or at least replied to) by people outside of the CGL
> member companies, and I believe that we need more
> involvement than just the CGL member companies for this
> (driver hardening and other parts of CGL) to succeed.
> [BTW, succeed in this case means being integrated into
> a mainstream kernel release.]
> Do you have other plans to announce the hardeneddrivers
> project and ask for feedback? such as to lkml or more
> focused mailing lists? I think that's what needs to
> happen here.
> I see that the hardened drivers projects is already
> listed under http://linuxkernel.foundries.sourceforge.net,
> and that's good, but I still doubt that location is
> enough either.
> I know that there is a concern about spamming lkml etc.,
> but without adequate exposure and feedback, patches like
> this won't make progress, won't be integrated into a
> mainstream kernel, and will eventually just bitrot.
> This is not an issue only for driver hardening and I'm not
> trying to pick on you, but driver hardening is
> (a) leading the way on this and (b) a very critical
> general infrastructure that needs wide exposure.
> Bottom line to me is that it will need to be discussed
> on lkml and linux-scsi at least, until newer lists like
> cgl_discussion have enough usage to stand alone.
> And we shouldn't discourage such discussion on cgl_discussion
> (i.e., we shouldn't limit it to the hardeneddrivers mailing
> list e.g.).
> On Wed, 18 Sep 2002, Rhoads, Rob wrote:
> | I sent the original message below on the hardened driver
> mailing list. Since
> | I've received only a few responses and only from engineers
> within Intel I'm
> | forwarding it here.
> | If you plan on participating on the hardened drivers
> project, please join
> | the mailing and send me your SourceForge user account name. Thanks.
> | -RobR
> | -----Original Message-----
> | From: Rhoads, Rob
> | Sent: Tuesday, September 17, 2002 9:00 AM
> | To: 'hardeneddrivers-discuss at lists.sourceforge.net'
> | Subject: Kick-off the Review of the Device Driver Hardening
> Design Spec
> | All,
> | As promised, I'd like to kick-off the review of the current
> | "Device Driver Hardening Design Spec". I'd like to do this
> | using this mailing list and the SourceForge bug tracker
> | database.
> | Under the "Tracker" section of the Hardened Drivers project
> | (http://sourceforge.net/projects/hardeneddrivers/), I've
> | created a new tracker database called "Review Design Spec".
> | That is where I'd liked to track all review comments and
> | feedback for the design spec.
> | IMPORTANT: Before you can add your review comments, I need
> | to add your SourceForge account to the project.
> | Send me your SourceForge user account name.
> | What I need from you (Action Items for you):
> | -------------------------------------------
> | 1. If you don't already have a SourceForge developer account,
> | get one (see http://sourceforge.net/account/register.php).
> | Send me your account name, so I can add you to the project.
> | Include a brief intro of who you are, past experience, and
> | your interests on device driver hardening. This last part
> | is just so I can get to know you.
> | 2. Download the spec, read and review it.
> | 3. Submit your review comments to the tracker "Review Design Spec"
> | database. Be sure to include the exact reference in the doc
> | (i.e. section numbers). I'd prefer that you use one tracker entry
> | per comment/suggestion. This is more work, but I want to be able
> | to track each item separately. We can always close, combine, or
> | open new entries if needed. Just use your best judgment.
> | DUE DATE: Submit all feedback by the end of the week
> | Sept. 20, 2002.
> | 4. Send your questions or items you'd like to further discuss,
> | to this mailing list.
> | What I will do (Action Items for me):
> | ------------------------------------
> | 1. I will review each of the tracker entries and resolve,
> | or disposition the ones that I can (e.g. accept, reject,
> | etc.). I will periodically publish the results of the
> | dispositions to this mailing list.
> | Ground Rules for tracker disposition:
> | ------------------------------------
> | Before rejecting an entry outright, I will work with the
> | submitter to resolve any conflicts. The submitter or
> | myself always have the option of opening up a discussion
> | on a particular item to this mailing list for wider
> | review. When in doubt, I will error on sending it out
> | to the mailing list for greater discussion.
> | As the project maintainer, I reserve the right to make the final
> | decision.
> | 2. Next week I will put together a list of items that I think
> | need further discussion, this will include all open or unresolved
> | review comments from the tracker entries.
> | 3. Consolidate all resolved review comments into a new
> | revision of the Device Driver Hardening Spec and post
> | it up on the project homepage.
> | 4. Put together a list of all required design/code changes to
> | the existing project source code, that result from your
> | feedback. From this list we can start assigning owners
> | for any needed code changes.
> | +=+=+
> | Rob Rhoads mailto:rob.rhoads at intel.com
> | Staff Software Engineer office: 503-677-5498
> | Intel Corporation
> | This email message solely contains my own personal views, and not
> | necessarily those of my employer.
> | _______________________________________________
More information about the cgl_discussion