Skip to contentSkip to conversation

Dropping WUSB/UWB from my patch queue

DISCUSSION Greg KH 3 participants 11 messages Jun 30, 2008 – Jul 16, 2008

lore
Share & tools
lore

Greg KH

author
#0 lore raw

Hi David,

In looking further at the WUSB/UWB code, it doesn't look like it is going to make it for the 2.6.27 kernel tree due to the reliance on some contriversial core USB changes as well as a total lack of documentation for the sysfs apis.

Because of this, I've dropped all of the patches from my USB queue, so they will no longer show up in the -next releases anymore.

But I have moved them to the -staging tree, as they are good patches and some people are finding them useful. They will show up in those releases for people to use and develop off of if needed.

So please rebase your local trees off of the -staging tree, and continue to send me updates from there as needed.

thanks,

greg k-h

#1 lore raw
Quoted 6 lines · Greg KH

Hi David,

In looking further at the WUSB/UWB code, it doesn't look like it is going to make it for the 2.6.27 kernel tree due to the reliance on some contriversial core USB changes as well as a total lack of documentation for the sysfs apis.

I assume you are referring to the usb_dev_reset_delayed() change here. This is only required by the wire adapter code and should not prevent the majority of the UWB and WUSB stack from being merged.

You should postpone:

usb-add-usb_dev_reset_delayed.patch wusb-add-the-wire-adapter-core.patch wusb-add-hwa-hc-wireless-host-controller-driver.patch wusb-wa-abort-rpipe-request-type-fix.patch

I will correct the lack of sysfs API documentation this week. Please advise on where the documentation should go and its style and format. Be aware that some of the API is experimental and subject to change. I will ensure the documentation is clear on this and that the Kconfig entries depend on EXPERIMENTAL.

Would this be sufficient to reinstate the majority of the UWB and WUSB stacks for 2.6.27?

Quoted 2 lines

Because of this, I've dropped all of the patches from my USB queue, so they will no longer show up in the -next releases anymore.

In future, it would be appreciated if any issues with UWB patches are discussed rather than summarily dropping them.

David

Signature · 3 lines
David Vrabel, Senior Software Engineer, Drivers
CSR, Churchill House, Cambridge Business Park,  Tel: +44 (0)1223 692562
Cowley Road, Cambridge, CB4 0WZ                 http://www.csr.com/
author
#2 lore raw
Quoted 2 lines · David Vrabel
Quoted 1 line · Greg KH

Hi David,

Sorry for the delay, was on vacation all last week without email access, and am on a business trip this week with very limited email access :(

Quoted 8 lines
Quoted 4 lines

In looking further at the WUSB/UWB code, it doesn't look like it is going to make it for the 2.6.27 kernel tree due to the reliance on some contriversial core USB changes as well as a total lack of documentation for the sysfs apis.

I assume you are referring to the usb_dev_reset_delayed() change here. This is only required by the wire adapter code and should not prevent the majority of the UWB and WUSB stack from being merged.

Ok, thanks for letting me know, I did not realize this.

Quoted 6 lines

You should postpone:

usb-add-usb_dev_reset_delayed.patch wusb-add-the-wire-adapter-core.patch wusb-add-hwa-hc-wireless-host-controller-driver.patch wusb-wa-abort-rpipe-request-type-fix.patch

Will the code still work properly for users with these patches removed? If so, I'll reconsider sending this for 2.6.27.

Quoted 8 lines

I will correct the lack of sysfs API documentation this week. Please advise on where the documentation should go and its style and format. Be aware that some of the API is experimental and subject to change. I will ensure the documentation is clear on this and that the Kconfig entries depend on EXPERIMENTAL.

Would this be sufficient to reinstate the majority of the UWB and WUSB stacks for 2.6.27?

Yes. I'll take what you sent me, and try to reorder things to make it so that we can get the majority of the code into .27, I don't want this to live outside the tree for any longer either, that's why I started working on getting this mess cleaned up in the first place :)

Quoted 5 lines
Quoted 2 lines

Because of this, I've dropped all of the patches from my USB queue, so they will no longer show up in the -next releases anymore.

In future, it would be appreciated if any issues with UWB patches are discussed rather than summarily dropping them.

Hey, they were not "dropped" in the format that they suddenly disappeared, they are still around and useful :)

thanks,

greg k-h

#3 lore raw
Quoted 8 lines · Greg KH
Quoted 7 lines · David Vrabel
Quoted 6 lines · Greg KH

Hi David,

In looking further at the WUSB/UWB code, it doesn't look like it is going to make it for the 2.6.27 kernel tree due to the reliance on some contriversial core USB changes as well as a total lack of documentation for the sysfs apis.

Quoted 12 lines
Quoted 3 lines

I assume you are referring to the usb_dev_reset_delayed() change here. This is only required by the wire adapter code and should not prevent the majority of the UWB and WUSB stack from being merged.

Ok, thanks for letting me know, I did not realize this.

Quoted 6 lines

You should postpone:

usb-add-usb_dev_reset_delayed.patch wusb-add-the-wire-adapter-core.patch wusb-add-hwa-hc-wireless-host-controller-driver.patch wusb-wa-abort-rpipe-request-type-fix.patch

I took another look at where usb_dev_reset_delayed() is used and there are more places. You will also need to postpone:

uwb/uwb-add-hwa-rc-radio-controller-driver.patch uwb/uwb-i1480-driver.patch uwb/uwb-i1480-wlp-driver.patch uwb/uwb-disable-command-event-filtering-for-DUB-1210.patch uwb/uwb-add-intel-i1480-hwa-to-the-uwb-rc-quirk-table.patch

And since there are now no drivers for WLP hardware you may also postpone:

uwb/uwb-add-wimedia-llc-protocol-core.patch uwb/uwb-wlp-messages.patch
uwb/uwb-wlp-wss.patch
uwb/uwb-wlp-build-system.patch uwb/wusb-drivers-uwb-wlp-sysfs.c-move-misplaced-debug-statement.patch

Quoted 2 lines

Will the code still work properly for users with these patches removed? If so, I'll reconsider sending this for 2.6.27.

Yes, users of WHCI WUSB host controllers are unaffected.

Quoted 6 lines
Quoted 6 lines

I will correct the lack of sysfs API documentation this week. Please advise on where the documentation should go and its style and format. Be aware that some of the API is experimental and subject to change. I will ensure the documentation is clear on this and that the Kconfig entries depend on EXPERIMENTAL.

Quoted 7 lines
Quoted 2 lines

Would this be sufficient to reinstate the majority of the UWB and WUSB stacks for 2.6.27?

Yes. I'll take what you sent me, and try to reorder things to make it so that we can get the majority of the code into .27, I don't want this to live outside the tree for any longer either, that's why I started working on getting this mess cleaned up in the first place :)

Okay. Good.

David

Signature · 3 lines
David Vrabel, Senior Software Engineer, Drivers
CSR, Churchill House, Cambridge Business Park,  Tel: +44 (0)1223 692562
Cowley Road, Cambridge, CB4 0WZ                 http://www.csr.com/
#4 lore raw
Quoted 3 lines

From: David Vrabel [mailto:[email protected]]

And since there are now no drivers for WLP hardware you may also

postpone:

Quoted 6 lines

uwb/uwb-add-wimedia-llc-protocol-core.patch uwb/uwb-wlp-messages.patch
uwb/uwb-wlp-wss.patch
uwb/uwb-wlp-build-system.patch uwb/wusb-drivers-uwb-wlp-sysfs.c-move-misplaced-debug-statement.patch

What happened to the 1480-wlp code?

author
#5 lore raw
In reply to David Vrabel
Quoted 31 lines · David Vrabel
Quoted 8 lines · Greg KH
Quoted 7 lines · David Vrabel
Quoted 6 lines · Greg KH

Hi David,

In looking further at the WUSB/UWB code, it doesn't look like it is going to make it for the 2.6.27 kernel tree due to the reliance on some contriversial core USB changes as well as a total lack of documentation for the sysfs apis.

Quoted 12 lines
Quoted 3 lines

I assume you are referring to the usb_dev_reset_delayed() change here. This is only required by the wire adapter code and should not prevent the majority of the UWB and WUSB stack from being merged.

Ok, thanks for letting me know, I did not realize this.

Quoted 6 lines

You should postpone:

usb-add-usb_dev_reset_delayed.patch wusb-add-the-wire-adapter-core.patch wusb-add-hwa-hc-wireless-host-controller-driver.patch wusb-wa-abort-rpipe-request-type-fix.patch

I took another look at where usb_dev_reset_delayed() is used and there are more places. You will also need to postpone:

uwb/uwb-add-hwa-rc-radio-controller-driver.patch uwb/uwb-i1480-driver.patch uwb/uwb-i1480-wlp-driver.patch uwb/uwb-disable-command-event-filtering-for-DUB-1210.patch uwb/uwb-add-intel-i1480-hwa-to-the-uwb-rc-quirk-table.patch

Quoted 7 lines

And since there are now no drivers for WLP hardware you may also postpone:

uwb/uwb-add-wimedia-llc-protocol-core.patch uwb/uwb-wlp-messages.patch
uwb/uwb-wlp-wss.patch
uwb/uwb-wlp-build-system.patch uwb/wusb-drivers-uwb-wlp-sysfs.c-move-misplaced-debug-statement.patch

Ugh, so, what will end up getting merged? Anything that people can actually use? :)

I'll poke at this on Friday and see what falls out.

thanks,

greg k-h

#6 lore raw
Quoted 3 lines · Greg KH

Ugh, so, what will end up getting merged? Anything that people can actually use? :)

To summarize what is planned to be merged and what is being postponed.

Merged:

* core UWB stack.
* WHCI radio controller driver
* core WUSB support.
* WHCI host controller driver
* CBAF driver

i.e., a functioning WUSB stack for WHCI (PCI-based) cards.

Postponed:

* HWA radio controller driver
* Wire Adapter core (for HWAs and DWAs).
* HWA host controller
* WLP support
* i1480 firmware download driver.
* i1480 WLP driver.

David

Signature · 3 lines
David Vrabel, Senior Software Engineer, Drivers
CSR, Churchill House, Cambridge Business Park,  Tel: +44 (0)1223 692562
Cowley Road, Cambridge, CB4 0WZ                 http://www.csr.com/
#7 lore raw
Quoted 10 lines

From: David Vrabel [mailto:[email protected]]

Postponed:

* HWA radio controller driver
* Wire Adapter core (for HWAs and DWAs).
* HWA host controller
* WLP support
* i1480 firmware download driver.
* i1480 WLP driver.

If I got you a fix for the USB reset stuff, would you be able to merge these guys?

author
#8 lore raw
Quoted 11 lines · Inaky
Quoted 10 lines

From: David Vrabel [mailto:[email protected]]

Postponed:

* HWA radio controller driver
* Wire Adapter core (for HWAs and DWAs).
* HWA host controller
* WLP support
* i1480 firmware download driver.
* i1480 WLP driver.

Quoted 2 lines

If I got you a fix for the USB reset stuff, would you be able to merge these guys?

I don't see why not.

thanks,

greg k-h

#9 lore raw
In reply to Perez-Gonzalez, Inaky
Quoted 11 lines · Inaky
Quoted 10 lines

From: David Vrabel [mailto:[email protected]]

Postponed:

* HWA radio controller driver
* Wire Adapter core (for HWAs and DWAs).
* HWA host controller
* WLP support
* i1480 firmware download driver.
* i1480 WLP driver.

Quoted 2 lines

If I got you a fix for the USB reset stuff, would you be able to merge these guys?

That would be appreciated. I just don't have the time to tackle tricky problems with wired USB devices.

You will also want to add some documentation for any WLP sysfs files.

David

Signature · 3 lines
David Vrabel, Senior Software Engineer, Drivers
CSR, Churchill House, Cambridge Business Park,  Tel: +44 (0)1223 692562
Cowley Road, Cambridge, CB4 0WZ                 http://www.csr.com/
#10 lore raw
Quoted 15 lines

From: David Vrabel [mailto:[email protected]]

Quoted 12 lines · Inaky
Quoted 10 lines

From: David Vrabel [mailto:[email protected]]

Postponed:

* HWA radio controller driver
* Wire Adapter core (for HWAs and DWAs).
* HWA host controller
* WLP support
* i1480 firmware download driver.
* i1480 WLP driver.

If I got you a fix for the USB reset stuff, would you be able to

merge

Quoted 4 lines
Quoted 1 line

these guys?

That would be appreciated. I just don't have the time to tackle tricky problems with wired USB devices.

I'll see what I can do -- the trip to OLS can be useful for having a straight set of ours for patching the kernel, away from the distractions of work.