Staging tree status for the .33 kernel merge
Hi all,
Here's a summary of the state of the drivers/staging/ tree, basically what will be coming in the 2.6.33 merge, and what the status of the different drivers are so far.
Sorry it's late (after the merge), but hey, better late than never.
Again, drivers/staging/ is NOT a dumping ground for dead code. If no one steps up to maintain and work to get the code merged into the main portion of the kernel, the drivers will be removed.
Also, drivers can now be merged from mainling into the staging directory, providing a path out of the kernel for some obsolete and/or broken drivers.
So, here's some drivers that will be removed in the 2.6.33 kernel:
- android drivers. Google and no one else stepped up to maintain them, so they will be dropped. So sad...
- dst. The developer isn't working on this anymore and recommended that it be removed as no one is using it.
Here is some new drivers that will show up in .33:
- arlan, netwave, strip, wavelan - wireless drivers that are on their way out of the kernel. If anyone is actually using this old, obsolete hardware, speak up soon, otherwise they will be removed in a few kernel releases.
- ramzswap - a compressed ram driver
- rtl8192u - yet another wireless driver
- samsung-laptop - laptop for the N128 Samsung laptop
- batman-adv - a network protocol
- dt3155 - a frame grabber driver
- sm7xx - another frame buffer driver
Here's the list of drivers that have had work done on them that will show up in the .33 release:
- comedi - lots of development effort happened here, mostly all cleanups, but there are some logic changes. More is needed, and it's moving along nicely.
- line6 - lots of work happening, very nice to see
- rt* - loads of cleanups and other merges. Will be obsolete soon due to a "real" wireless driver being worked on, but it's still nice to have these be a working alternative until then.
- rtl* - more wireless driver work, horrible code, but it seems to work for the users. Hopefully more development time can be spent here in the new year.
- dream - here's the platform specific code for the Android G1 platform. This might be the way the android code sneaks back into the kernel, as there is developers trying to get this to work. Of course, it's all happening without Google's help. {sigh}
- et131x - loads of cleanups, more left to do. Good solid progress happening here.
- iio - a new driver added to this subsystem, along with other fixes and cleanups. Looking nice.
- poch - still some work happening, nice to see it pick back up.
- panel - minor cleanups.
- vme - cleanups and minor tweaks, still alive and kicking
- vt66* - more wireless drivers, will be obsoleted by a "real" driver again.
- wlags49 - more cleanup work as well.
Here's some drivers not listed so far, that have had work done recently, after the 2.6.33-rc1 merge happened, so it has to wait for the .34 kernel release:
- wlan-ng
- slicoss
- mimo
- asus_oled
- udlfb
- w35und
Hm, so, what's up with all of the other staging drivers, and why have they not had any development? What is the status of them? They are now on the short list to be deleted in 2 kernel releases, unless some _real_ development happens on them.
This means, unless someone steps up and starts doing real work (not trivial spelling fixes) on the following drivers, they will be removed in the future kernel releases.
- arlan, netwave, strip, wavelan - wireless drivers mentioned above that are on the way out. Slated for removal in 2.6.35
- hv - Microsoft Hyper V drivers. The developers again seem to have disappeared, this is getting old. Slated for removal in 2.6.35
- p9auth - this will be removed in .34 unless someone steps up.
- frontier - slated for removal in .35. Will be easy for someone to pick up if they want to (hint, hint, hint)
- altpciechdma - this will be removed in .34 unless someone steps up.
- b3dfg - this will be removed in .34 unless someone steps up.
- pohmelfs - filesystem under development out of tree, would be nice to get the patches merged back into here
- quatech_usb2 and serqt_usb2 - usb to serial drivers that need to get merged into mainline.
- rar and sep - Intel needs to step up here and get this code cleaned up properly, or it too will be removed.
Again, if someone is looking for some kernel development work to do, picking any of the above drivers up to get them merged into mainline, would be a great thing to do.
And, on a final, and sadder note, I'd like to announce the failure of the driver project to complete a driver for a company that requested it. This was totally my fault, and I would publically like to apologize about my lack of getting a SCSI driver written in time for a company that had asked for it. So, while the Linux driver project is doing great work for a large number of companies, every once in a while we do fail, we are only human.
thanks,
greg k-h
Quoted 3 lines · * Greg KH | 2009-12-22 16:54:47 [-0800]:
Here's a summary of the state of the
drivers/staging/tree, basically what will be coming in the 2.6.33 merge, and what the status of the different drivers are so far.
I can't find usbip in your summary. Is there nothing to say or did you just forgot about it?
Quoted 3 lines
thanks,
greg k-h
Sebastian
Quoted 8 lines · Sebastian Andrzej Siewior
Quoted 3 lines · * Greg KH | 2009-12-22 16:54:47 [-0800]:
Here's a summary of the state of the
drivers/staging/tree, basically what will be coming in the 2.6.33 merge, and what the status of the different drivers are so far.I can't find usbip in your summary. Is there nothing to say or did you just forgot about it?
Oops, I forgot about it. It's slated to be removed in .35 as well, because no one is working on it.
thanks,
greg k-h
Quoting Greg KH ([email protected]):
...
Quoted 9 lines
This means, unless someone steps up and starts doing real work (not trivial spelling fixes) on the following drivers, they will be removed in the future kernel releases.
- arlan, netwave, strip, wavelan - wireless drivers mentioned above that are on the way out. Slated for removal in 2.6.35
- hv - Microsoft Hyper V drivers. The developers again seem to have disappeared, this is getting old. Slated for removal in 2.6.35
- p9auth - this will be removed in .34 unless someone steps up.
I think I've decided to try to push it. I'm working with some patches at git://git.kernel.org/pub/scm/linux/kernel/git/sergeh/linux-cr.git
(branch p9auth.jan3.4 is latest). I'll send patches as I feel they are ready - so far they pass testcases, but are too new for me to feel I should push them today.
Ashwin, I'm curious whether you'd think the last patch
(http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/sergeh/linux-cr.git;a=commitdiff;h=1662ba777140a39c21a9b647459d2deab8ffe1ca)
would be a problem with any userspace - but I assume there is no legacy userspace to really worry about?
Apart from plenty more cleanups, another more fundamental issue to address is how to stop unused caphash entries from piling up in memory. Put a timeout on them? Let privileged userspace list and occasionally delete them? Associate a target task with each entry, where either the task or its decendent can use the capability, but if the task dies we free the caphash entry?
-serge
Quoted 12 lines · Serge E. Hallyn
Quoting Greg KH ([email protected]):
...Quoted 9 lines
This means, unless someone steps up and starts doing real work (not trivial spelling fixes) on the following drivers, they will be removed in the future kernel releases.
- arlan, netwave, strip, wavelan - wireless drivers mentioned above that are on the way out. Slated for removal in 2.6.35
- hv - Microsoft Hyper V drivers. The developers again seem to have disappeared, this is getting old. Slated for removal in 2.6.35
- p9auth - this will be removed in .34 unless someone steps up.
Quoted 5 lines
I think I've decided to try to push it. I'm working with some patches at
git://git.kernel.org/pub/scm/linux/kernel/git/sergeh/linux-cr.git
(branch p9auth.jan3.4 is latest). I'll send patches as I feel they are ready - so far they pass testcases, but are too new for me to feel I should push them today.
Ok, send them on to me when you feel they are good enough.
thanks,
greg k-h
Quoted 12 lines · Serge E. Hallyn
Quoting Greg KH ([email protected]):
...Quoted 9 lines
This means, unless someone steps up and starts doing real work (not trivial spelling fixes) on the following drivers, they will be removed in the future kernel releases.
- arlan, netwave, strip, wavelan - wireless drivers mentioned above that are on the way out. Slated for removal in 2.6.35
- hv - Microsoft Hyper V drivers. The developers again seem to have disappeared, this is getting old. Slated for removal in 2.6.35
- p9auth - this will be removed in .34 unless someone steps up.
Quoted 5 lines
I think I've decided to try to push it. I'm working with some patches at
git://git.kernel.org/pub/scm/linux/kernel/git/sergeh/linux-cr.git
(branch p9auth.jan3.4 is latest). I'll send patches as I feel they are ready - so far they pass testcases, but are too new for me to feel I should push them today.
Thanks Serge!
It is useful to continue to have this driver in the tree as there a few other people as well who have shown interest in using this. I have been recently contacted by guys at Glendix (http://www.glendix.org/)
who have started looking at using this driver.
Quoted 5 lines
Ashwin, I'm curious whether you'd think the last patch
(http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/sergeh/linux-cr.git;a=commitdiff;h=1662ba777140a39c21a9b647459d2deab8ffe1ca)
would be a problem with any userspace - but I assume there is no legacy userspace to really worry about?
There is no legacy user space support yet for Linux. This should be fine in that sense.
I still need to look at the patches in detail though but what is the motivation for this change?
Please also cc [email protected] and [email protected] as well when you send out these patches for review.
Quoted 7 lines
Apart from plenty more cleanups, another more fundamental issue to address is how to stop unused caphash entries from piling up in memory. Put a timeout on them? Let privileged userspace list and occasionally delete them? Associate a target task with each entry, where either the task or its decendent can use the capability, but if the task dies we free the caphash entry?
So, there are a couple of options here (I favor the second approach):
1. We can add a timer to expire the capabilities.
2. Add a creation time stamp to every capability. Whenever a capability is used (i.e. written to /dev/caphash) we can go through the list in the kernel and reap the ones whose time stamp has expired. We can optimize the data structure later to make this faster.
Ashwin.
Quoted 32 lines · Ashwin Ganti
Quoted 12 lines · Serge E. Hallyn
Quoting Greg KH ([email protected]):
...Quoted 9 lines
This means, unless someone steps up and starts doing real work (not trivial spelling fixes) on the following drivers, they will be removed in the future kernel releases.
?- arlan, netwave, strip, wavelan - wireless drivers mentioned above ? ?that are on the way out. ?Slated for removal in 2.6.35 ?- hv - Microsoft Hyper V drivers. ?The developers again seem to have ? ?disappeared, this is getting old. Slated for removal in 2.6.35 ?- p9auth - this will be removed in .34 unless someone steps up.
Quoted 5 lines
I think I've decided to try to push it. ?I'm working with some patches at
git://git.kernel.org/pub/scm/linux/kernel/git/sergeh/linux-cr.git
(branch p9auth.jan3.4 is latest). ?I'll send patches as I feel they are ready - so far they pass testcases, but are too new for me to feel I should push them today.Thanks Serge!
It is useful to continue to have this driver in the tree as there a few other people as well who have shown interest in using this. I have been recently contacted by guys at Glendix (http://www.glendix.org/)
who have started looking at using this driver.Quoted 5 lines
Ashwin, I'm curious whether you'd think the last patch
(http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/sergeh/linux-cr.git;a=commitdiff;h=1662ba777140a39c21a9b647459d2deab8ffe1ca)
would be a problem with any userspace - but I assume there is no legacy userspace to really worry about?
Quoted 4 lines
There is no legacy user space support yet for Linux. This should be fine in that sense.
I still need to look at the patches in detail though but what is the motivation for this change?
Well, to have a login daemon be unprivileged, it needs some way to set groups without CAP_SETGID. I was playing with some toy frontend and backend (i.e. login and factotum) code and this seemed a nice simple way to do it.
Quoted 2 lines
Please also cc [email protected] and [email protected] as well when you send out these patches for review.
Will do, thanks.
Quoted 7 lines
Quoted 6 lines
Apart from plenty more cleanups, another more fundamental issue to address is how to stop unused caphash entries from piling up in memory. ?Put a timeout on them? ?Let privileged userspace list and occasionally delete them? ?Associate a target task with each entry, where either the task or its decendent can use the capability, but if the task dies we free the caphash entry?
Quoted 6 lines
So, there are a couple of options here (I favor the second approach):
1. We can add a timer to expire the capabilities.
2. Add a creation time stamp to every capability. Whenever a capability is used (i.e. written to/dev/caphash) we can go through the list in the kernel and reap the ones whose time stamp has expired. We can optimize the data structure later to make this faster.
Ok, I'll do that when I get a chance, do some more testing, and send out in the next two weeks.
thanks,
-serge