What to do with all of the USB UHCI drivers in the kernel?
Ok, now that 2.5.16 is out, we have a total of 4 different USB UHCI controller drivers in the kernel! That's about 3 too many for me :)
So what to do? I propose the following:
From now until July 1, I want everyone to test out both the uhci-hcd and usb-uhci-hcd drivers on just about every piece of hardware they can find. This includes SMP, UP, preempt kernels, big and little endian machines, and loads of different types of USB devices.
(Note, for those who don't realize it, the uhci-hcd driver is based off of the uhci.c driver, and the usb-uhci-hcd driver is based off of the usb-uhci.c driver. Both of these drivers now use the USB HCD interface, hopefully reducing some code complexity and size because of this. So this means that the uhci.c and usb-uhci.c drivers are going to go away, just like the usb-ohci.c driver did in 2.5.16.)
Let me (and the linux-usb-devel list) know about any thoughts you have pertaining to liking one of the drivers over the other one. Speed tests, size tests, code pretty tests, comment spelling tests, documentation tests, you name it, I want to know about it. If you don't want your comments to be public, send them to me directly and I will not let anyone else know what you said, but will use the info to try to pick which one should stay.
I will be doing the same thing (running speed tests, and hardware tests) and will be publishing those results on the linux-usb-devel list for others to verify.
Then, the week of July 1, I will be taking everyone's comments and making a decision about which driver to keep.
thanks,
greg k-h
Quoted 2 lines · [email protected] said:
Ok, now that 2.5.16 is out, we have a total of 4 different USB UHCI controller drivers in the kernel! That's about 3 too many for me :)
Stupid follow-up question for you. Is it possible for me to stay on 2.4.18, but test the 2.5.16 USB subsystem? I really don't want to move my whole system to 2.5.16, but would be willing to experiment with the USB parts. Is this possible? Can I just rebuild the usb modules - I'm guessing this wouldn't work, but thought I would ask anyway. Or to ask another way - any way us 2.4 folks can help?
Thanks much for your (all the USB developers) work on USB - it is great to have USB working under Linux, I do appreciate it.
-----
Timothy Jedlicka, [email protected], 1-630-713-4436, AOL-IM=bonzowork Network Entomologist, Lucent Technologies, Testers For Hire
Quoted 10 lines · Timothy E. Jedlicka - wrk
Quoted 2 lines · [email protected] said:
Ok, now that 2.5.16 is out, we have a total of 4 different USB UHCI controller drivers in the kernel! That's about 3 too many for me :)
Stupid follow-up question for you. Is it possible for me to stay on 2.4.18, but test the 2.5.16 USB subsystem? I really don't want to move my whole system to 2.5.16, but would be willing to experiment with the USB parts. Is this possible? Can I just rebuild the usb modules - I'm guessing this wouldn't work, but thought I would ask anyway. Or to ask another way - any way us 2.4 folks can help?
You _might_ want to try dropping the drivers/usb tree from 2.5.x into 2.4 (and the usb files in include/linux/usb*.h too) and see what happens.
Odds are the build process will not work, and I think some of the usbfs changes will also not work. But it would be interesting to see what you found :)
So in short, I don't think that you could test out the 2.5 code on 2.4. Is there some reason you can't run a 2.5 kernel? Personally I run it on lots of different machines (laptops, desktops, servers, etc.) successfully.
If you can't run a 2.5 kernel, you can just take a look through the code, and tell me what you think of that. Readability is one of the criteria I'm going to be using.
thanks,
greg k-h
----- Original Message -----
From: "Greg KH" <[email protected]>
To: <[email protected]>
Cc: <[email protected]>; <[email protected]>
Sent: Monday, May 20, 2002 6:31 PM
Subject: What to do with all of the USB UHCI drivers in the kernel?
Quoted 10 lines
Ok, now that 2.5.16 is out, we have a total of 4 different USB UHCI controller drivers in the kernel! That's about 3 too many for me :)
So what to do? I propose the following:
From now until July 1, I want everyone to test out both the uhci-hcd and usb-uhci-hcd drivers on just about every piece of hardware they can find. This includes SMP, UP, preempt kernels, big and little endian machines, and loads of different types of USB devices.
The UHCI driver never recognizes my hardware. The OHCI driver (in the 2.4.18 kernel) does however. My Asus A7M266-D doesn't have an onboard USB but they ship an add-on card with the motherboard (made by Asus).
I attached the relevant syslogs and kernel logs for a boot-up with the UHCI driver. The USB startup is towards the end.
List of USB devices attached:
1) AsusTek USB enhanced Host Controller
2) Logitech USB Camera (QuickCam Web)
3) NEC PCI to USB Open Host Controller
4) NEC PCI to USB Open Host Controller
5) USB Root Hub x2 (Must belong to the USB hub on my monitor, ViewSonic G773-2)
6) USB Root Hub x2 (Must belong to the USB hub on my monitor, ViewSonic G773-2)
7) Microsoft IntelliMouse Optical (imPs2)
Thanks!
Quoted 7 lines
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
*****************************************************
André Bonin
Computer Engineering Technologist
Ottawa (Ontario)
Canada
*****************************************************
May 21 20:36:20 Cherlyn kernel: Linux version 2.5.17 (root@Cherlyn) (gcc version 2.95.4 20011002 (Debian prerelease)) #4 SMP Tue May 21 20:26:15 EST 2002 ... May 21 20:36:20 Cherlyn kernel: uhci.c: USB Universal Host Controller Interface driver v1.1 May 21 20:36:20 Cherlyn kernel: usb.c: registered new driver usblp May 21 20:36:20 Cherlyn kernel: printer.c: v0.12: USB Printer Device Class driver May 21 20:36:20 Cherlyn kernel: Initializing USB Mass Storage driver... May 21 20:36:20 Cherlyn kernel: usb.c: registered new driver usb-storage May 21 20:36:20 Cherlyn kernel: USB Mass Storage support registered. May 21 20:36:20 Cherlyn kernel: usb.c: registered new driver hiddev May 21 20:36:20 Cherlyn kernel: usb.c: registered new driver hid May 21 20:36:20 Cherlyn kernel: hid-core.c: v1.31:USB HID core driver May 21 20:36:20 Cherlyn kernel: usb.c: registered new driver usbscanner May 21 20:36:20 Cherlyn kernel: scanner.c: 0.4.6:USB Scanner Driver May 21 20:36:20 Cherlyn kernel: mice: PS/2 mouse device common for all mice
Quoted 15 lines · André Bonin
----- Original Message -----
From: "Greg KH" <[email protected]>
To: <[email protected]>
Cc: <[email protected]>; <[email protected]>
Sent: Monday, May 20, 2002 6:31 PM
Subject: What to do with all of the USB UHCI drivers in the kernel?Quoted 6 lines
Ok, now that 2.5.16 is out, we have a total of 4 different USB UHCI controller drivers in the kernel! That's about 3 too many for me :)
So what to do? I propose the following:
Quoted 8 lines
Quoted 4 lines
From now until July 1, I want everyone to test out both the uhci-hcd and usb-uhci-hcd drivers on just about every piece of hardware they can find. This includes SMP, UP, preempt kernels, big and little endian machines, and loads of different types of USB devices.
The UHCI driver never recognizes my hardware. The OHCI driver (in the 2.4.18 kernel) does however. My Asus A7M266-D doesn't have an onboard USB but they ship an add-on card with the motherboard (made by Asus).
This is probably because you have an OHCI hardware device, not a UHCI device. What does 'lspci -v' say for your machine?
And how does 2.5.17 work for you?
thanks,
greg k-h
Quoted 21 lines · Greg KH
Quoted 19 lines · André Bonin
----- Original Message -----
From: "Greg KH" <[email protected]>
To: <[email protected]>
Cc: <[email protected]>; <[email protected]>
Sent: Monday, May 20, 2002 6:31 PM
Subject: What to do with all of the USB UHCI drivers in the kernel?Quoted 10 lines
Ok, now that 2.5.16 is out, we have a total of 4 different USB UHCI controller drivers in the kernel! That's about 3 too many for me :)
So what to do? I propose the following:
From now until July 1, I want everyone to test out both the uhci-hcd and usb-uhci-hcd drivers on just about every piece of hardware they can find. This includes SMP, UP, preempt kernels, big and little endian machines, and loads of different types of USB devices.
Quoted 7 lines
Quoted 4 lines
The UHCI driver never recognizes my hardware. The OHCI driver (in the 2.4.18 kernel) does however. My Asus A7M266-D doesn't have an onboard USB but they ship an add-on card with the motherboard (made by Asus).
This is probably because you have an OHCI hardware device, not a UHCI device. What does 'lspci -v' say for your machine?
Sorry, i'me not too familiar with the USB architecture. Anyway here is the relevant lspci entries (note: I did this under my working 2.4.18)
02:08.0 USB Controller: NEC Corporation USB (rev 41) (prog-if 10 [OHCI])
Subsystem: Unknown device 807d:0035
Flags: bus master, medium devsel, latency 32, IRQ 19 Memory at cd000000 (32-bit, non-prefetchable) [size=4K]
Capabilities: [40] Power Management version 2
02:08.1 USB Controller: NEC Corporation USB (rev 41) (prog-if 10 [OHCI])
Subsystem: Unknown device 807d:0035
Flags: bus master, medium devsel, latency 32, IRQ 16 Memory at cc800000 (32-bit, non-prefetchable) [size=4K]
Capabilities: [40] Power Management version 2
02:08.2 USB Controller: NEC Corporation USB 2.0 (rev 02) (prog-if 20 [EHCI])
Subsystem: Unknown device 807d:1043
Flags: bus master, medium devsel, latency 32, IRQ 17 Memory at cc000000 (32-bit, non-prefetchable) [size=256]
Capabilities: [40] Power Management version 2
Quoted 1 line
And how does 2.5.17 work for you?
Not too good beacuse I don't have the option of enabling OHCI :) Are we still keeping it?
Quoted 1 line
thanks,
I thank you sir!
Quoted 10 lines
greg k-h
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Quoted 18 lines · Andre Bonin
Quoted 2 lines
This is probably because you have an OHCI hardware device, not a UHCI device. What does 'lspci -v' say for your machine?
Sorry, i'me not too familiar with the USB architecture. Anyway here is the relevant lspci entries (note: I did this under my working 2.4.18)
02:08.0 USB Controller: NEC Corporation USB (rev 41) (prog-if 10 [OHCI])
Subsystem: Unknown device 807d:0035
Flags: bus master, medium devsel, latency 32, IRQ 19 Memory atcd000000(32-bit, non-prefetchable) [size=4K]
Capabilities: [40] Power Management version 202:08.1 USB Controller: NEC Corporation USB (rev 41) (prog-if 10 [OHCI])
Subsystem: Unknown device 807d:0035
Flags: bus master, medium devsel, latency 32, IRQ 16 Memory atcc800000(32-bit, non-prefetchable) [size=4K]
Capabilities: [40] Power Management version 2
Quoted 5 lines
02:08.2 USB Controller: NEC Corporation USB 2.0 (rev 02) (prog-if 20 [EHCI])
Subsystem: Unknown device 807d:1043
Flags: bus master, medium devsel, latency 32, IRQ 17 Memory atcc000000(32-bit, non-prefetchable) [size=256]
Capabilities: [40] Power Management version 2
You only have EHCI and OHCI hardware. No wonder the UHCI drivers do not work :)
Quoted 4 lines
Quoted 1 line
And how does 2.5.17 work for you?
Not too good beacuse I don't have the option of enabling OHCI :) Are we still keeping it?
Yes, use the ohci-hcd driver. Also you can use the ehci-hcd driver if you have any USB 2.0 devices, as it looks like you have a USB 2.0 controller.
thanks,
greg k-h
Uz.ytkownik Greg KH napisa?:
Quoted 30 lines · Andre Bonin
Quoted 18 lines
Quoted 3 lines
This is probably because you have an OHCI hardware device, not a UHCI device. What does 'lspci -v' say for your machine?
Sorry, i'me not too familiar with the USB architecture. Anyway here is the relevant lspci entries (note: I did this under my working 2.4.18)
02:08.0 USB Controller: NEC Corporation USB (rev 41) (prog-if 10 [OHCI])
Subsystem: Unknown device 807d:0035
Flags: bus master, medium devsel, latency 32, IRQ 19 Memory atcd000000(32-bit, non-prefetchable) [size=4K]
Capabilities: [40] Power Management version 202:08.1 USB Controller: NEC Corporation USB (rev 41) (prog-if 10 [OHCI])
Subsystem: Unknown device 807d:0035
Flags: bus master, medium devsel, latency 32, IRQ 16 Memory atcc800000(32-bit, non-prefetchable) [size=4K]
Capabilities: [40] Power Management version 2Quoted 5 lines
02:08.2 USB Controller: NEC Corporation USB 2.0 (rev 02) (prog-if 20 [EHCI])
Subsystem: Unknown device 807d:1043
Flags: bus master, medium devsel, latency 32, IRQ 17 Memory atcc000000(32-bit, non-prefetchable) [size=256]
Capabilities: [40] Power Management version 2You only have EHCI and OHCI hardware. No wonder the UHCI drivers do not work :)
Quoted 8 lines
Quoted 3 lines
Quoted 1 line
And how does 2.5.17 work for you?
Not too good beacuse I don't have the option of enabling OHCI :) Are we still keeping it?
Yes, use the ohci-hcd driver. Also you can use the ehci-hcd driver if you have any USB 2.0 devices, as it looks like you have a USB 2.0 controller.
Could you please just do me a small favour and drop something in to linux/Documentation. Becouse I'm right now already confused about which driver to use and which alias to put in /etc/modules.conf so kudzu stops hollering about not knowing what to do
if I out of a sudden reboot in to 2.5.xx kernel.
Many thank's in advance.
PS. I could of course figure it out of my self, but since I don't attach anything to USB on my box *that* frequently.
Hi Martin,
What do you actually want to know? That an EHCI controller should
use the ehci-hcd driver and that the OHCI controller should use the ohci-hcd controller? Or that the uhci-* drivers can't drive a EHCI or OHCI controller? Or something else?
rgeards,
Stephen.
Quoted 66 lines · Martin Dalecki
Uz.ytkownik Greg KH napisa?:
Quoted 33 lines · Andre Bonin
Quoted 19 lines
Quoted 3 lines
This is probably because you have an OHCI hardware device, not a UHCI device. What does 'lspci -v' say for your machine?
Sorry, i'me not too familiar with the USB architecture. Anyway here is the relevant lspci entries (note: I did this under my working 2.4.18)
02:08.0 USB Controller: NEC Corporation USB (rev 41) (prog-if 10 [OHCI])
Subsystem: Unknown device 807d:0035
Flags: bus master, medium devsel, latency 32, IRQ 19 Memory atcd000000(32-bit, non-prefetchable) [size=4K]
Capabilities: [40] Power Management version 202:08.1 USB Controller: NEC Corporation USB (rev 41) (prog-if 10 [OHCI])
Subsystem: Unknown device 807d:0035
Flags: bus master, medium devsel, latency 32, IRQ 16 Memory atcc800000(32-bit, non-prefetchable) [size=4K]
Capabilities: [40] Power Management version 2Quoted 5 lines
02:08.2 USB Controller: NEC Corporation USB 2.0 (rev 02) (prog-if 20 [EHCI])
Subsystem: Unknown device 807d:1043
Flags: bus master, medium devsel, latency 32, IRQ 17 Memory atcc000000(32-bit, non-prefetchable) [size=256]
Capabilities: [40] Power Management version 2You only have EHCI and OHCI hardware. No wonder the UHCI drivers do not work :)
Quoted 2 lines
Quoted 1 line
And how does 2.5.17 work for you?
Quoted 7 lines
Quoted 2 lines
Not too good beacuse I don't have the option of enabling OHCI :) Are we still keeping it?
Yes, use the ohci-hcd driver. Also you can use the ehci-hcd driver if you have any USB 2.0 devices, as it looks like you have a USB 2.0 controller.
Could you please just do me a small favour and drop something in to linux/Documentation. Becouse I'm right now already confused about which driver to use and which alias to put in
/etc/modules.confso kudzu stops hollering about not knowing what to do
if I out of a sudden reboot in to 2.5.xx kernel.Many thank's in advance.
PS. I could of course figure it out of my self, but since I don't attach anything to USB on my box *that* frequently.
_______________________________________________________________
Don't miss the 2002 Sprint PCS Application Developer's Conference August 25-28 in Las Vegas -- http://devcon.sprintpcs.com/adp/index.cfm
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-users
Signature · 6 lines
/------------------------------------+-------------------------\ |Stephen J. Gowdy | SLAC, MailStop 34, | |http://www.slac.stanford.edu/~gowdy/ | 2575 Sand Hill Road, | |http://calendar.yahoo.com/gowdy | Menlo Park CA 94025, USA | |EMail: [email protected] | Tel: +1 650 926 3144 | \------------------------------------+-------------------------/
Uz.ytkownik Stephen J. Gowdy napisa?:
Quoted 3 lines
Hi Martin,
What do you actually want to know? That an EHCI controller should
use the ehci-hcd driver and that the OHCI controller should use the ohci-hcd controller? Or that the uhci-* drivers can't drive a EHCI or OHCI controller? Or something else?
Well what I know is the following: Kudzu did set up my system to use usb-ohci driver similar. And now you have apparently started to obsolete some of them. I know that there are still two drivers supported. So I would really really wellcome it if you would just write up some documentation about the *whole* transition from using some old driver to the one which is supposed to be the driver used in the future.
Most likely this would be of course module aliasing entires which could be used in /etc/modules.conf to cheat around kudzu - this way it would be for example possible to have a nice dual boot configuration.
(I can adjust my boot scripts to set up an apriopriate symlink from /etc/modules.conf to modules-2.5.18.confg and modules-2.4.18.conf debpending what I'm booting anyway...)
In short - if you change something about the configuration and usage - just document it please.
Thank's in advance.
BTW> one of the reasons I never bothered myself with kbuild-2.5 is for example that nio matter how frequently Keith
is advertising it - every time I go there to have a look at it at sf what I find is a scatter heap of .tar.gz. The documentation about how to install it makes me nervous, since I would rather just expect a diff and a README how to use it, so I never look after it.
kb25 manual [was Re: [Linux-usb-users] Re: What to do with all of the USB UHCI drivers in the kernel?]
In reply to Stephen J. GowdyQuoted 7 lines
BTW> one of the reasons I never bothered myself with kbuild-2.5 is for example that nio matter how frequently Keith
is advertising it - every time I go there to have a look at it at sf what I find is a scatter heap of .tar.gz. The documentation about how to install it makes me nervous, since I would rather just expect a diff and a README how to use it, so I never look after it.
Duh... All you need is the core package (diff #1), the architecture independent modifications package (diff #2) and finally diff #3 that adds the arch-specific stuff. Everything's in a single list on sf.
then just do s/t like
cd /usr/src/linux-2.4.19-pre8 && zcat \ ../kbuild-2.5-core-14.gz \ ../kbuild-2.5-common-2.4.19-pre8-1.gz \ ../kbuild-2.5-i386-2.4.19-pre8-1.gz \ | patch -sp1
and read the comprehensive manual in Documentation/kbuild/
What's there to be nervous about?
T.
Re: kb25 manual [was Re: [Linux-usb-users] Re: What to do with all of the USB UHCI drivers in the kernel?]
On Thu, 23 May 2002, Tomas Szepe wrote:
| > BTW> one of the reasons I never bothered myself with kbuild-2.5 | > is for example that nio matter how frequently Keith | > is advertising it - every time I go there to have a look at it | > at sf what I find is a scatter heap of .tar.gz. The documentation | > about how to install it makes me nervous, since I would | > rather just expect a diff and a README how to use it, so I never | > look after it. | | Duh... All you need is the core package (diff #1), the architecture | independent modifications package (diff #2) and finally diff #3 that | adds the arch-specific stuff. Everything's in a single list on sf. | | then just do s/t like | cd /usr/src/linux-2.4.19-pre8 && zcat \ | ../kbuild-2.5-core-14.gz \ | ../kbuild-2.5-common-2.4.19-pre8-1.gz \ | ../kbuild-2.5-i386-2.4.19-pre8-1.gz \ | | patch -sp1 | | and read the comprehensive manual in Documentation/kbuild/ | | What's there to be nervous about?
so there should just be a short, simple readme about how to install and use it -- not one that is 2400 lines long. Other than that, I like it.
Signature · 1 line
~Randy
Re: kb25 manual [was Re: [Linux-usb-users] Re: What to do with all of the USB UHCI drivers in the kernel?]
In reply to Tomas SzepeQuoted 23 lines
| > BTW> one of the reasons I never bothered myself with kbuild-2.5 | > is for example that nio matter how frequently Keith | > is advertising it - every time I go there to have a look at it | > at sf what I find is a scatter heap of .tar.gz. The documentation | > about how to install it makes me nervous, since I would | > rather just expect a diff and a README how to use it, so I never | > look after it. | | Duh... All you need is the core package (diff #1), the architecture | independent modifications package (diff #2) and finally diff #3 that | adds the arch-specific stuff. Everything's in a single list on sf. | | then just do s/t like | cd /usr/src/linux-2.4.19-pre8 && zcat \ | ../kbuild-2.5-core-14.gz \ | ../kbuild-2.5-common-2.4.19-pre8-1.gz \ | ../kbuild-2.5-i386-2.4.19-pre8-1.gz \ | | patch -sp1 | | and read the comprehensive manual in Documentation/kbuild/ | | What's there to be nervous about?
Quoted 3 lines
so there should just be a short, simple readme about how to install and use it -- not one that is 2400 lines long. Other than that, I like it.
How about submitting one to Keith yourself then? :)
No really, I don't think there's a problem with the manual. If you just want to learn how to compile a kernel, you can skim thru the first section of the file in under 30 seconds.
Quoted 1 line
Other than that, I like it.
Who doesn't. I mean, apart from Wayne Brown who hates to have to remember to issue simpler commands. ;)
T.
Re: kb25 manual [was Re: [Linux-usb-users] Re: What to do with all of the USB UHCI drivers in the kernel?]
In reply to Tomas SzepeUz.ytkownik Tomas Szepe napisa?:
Quoted 14 lines
Quoted 14 lines
| > BTW> one of the reasons I never bothered myself with kbuild-2.5 | > is for example that nio matter how frequently Keith | > is advertising it - every time I go there to have a look at it | > at sf what I find is a scatter heap of .tar.gz. The documentation | > about how to install it makes me nervous, since I would | > rather just expect a diff and a README how to use it, so I never | > look after it. | | Duh... All you need is the core package (diff #1), the architecture | independent modifications package (diff #2) and finally diff #3 that | adds the arch-specific stuff. Everything's in a single list on sf. | | then just do s/t like | cd /usr/src/linux-2.4.19-pre8 && zcat \
Quoted 8 lines
Quoted 8 lines
| ../kbuild-2.5-core-14.gz \ | ../kbuild-2.5-common-2.4.19-pre8-1.gz \ | ../kbuild-2.5-i386-2.4.19-pre8-1.gz \ | | patch -sp1 | | and read the comprehensive manual in Documentation/kbuild/ | | What's there to be nervous about?
Well perhaps lazy would be more adequate to express my feelings. First why the hell three different diff files?
I don't give a damn witt about what the internal architecure of it is. And I don't wan't to miss any non i386 build. (I have ARM for example as well.)
I don't wan't to care whatever this is all complete. And I just expect the documentation about how to use it right at the top in README or INSTALL. I'm a human and humans tend to love to stick to habits. And I already got in to the habit of:
- One thing one patch.
- Something to compile look after README or INSTALL first there.
Even the kbuild name isn't intuitive
new-kernel-build would be more clear.
I really just wan't to see how it works first and I wan't to see that it indeed work's better then what I have already before I look at how it is done
Call me arrogant (and most propably it would be justifyed in this case), but this aggressive splitup gives me prejudictions about the whole thing simply beeing overdesigned...
Re: kb25 manual [was Re: [Linux-usb-users] Re: What to do with all of the USB UHCI drivers in the kernel?]
In reply to Tomas SzepeQuoted 29 lines
Uz.ytkownik Tomas Szepe napisa?:
Quoted 14 lines
Quoted 14 lines
| > BTW> one of the reasons I never bothered myself with kbuild-2.5 | > is for example that nio matter how frequently Keith | > is advertising it - every time I go there to have a look at it | > at sf what I find is a scatter heap of .tar.gz. The documentation | > about how to install it makes me nervous, since I would | > rather just expect a diff and a README how to use it, so I never | > look after it. | | Duh... All you need is the core package (diff #1), the architecture | independent modifications package (diff #2) and finally diff #3 that | adds the arch-specific stuff. Everything's in a single list on sf. | | then just do s/t like | cd /usr/src/linux-2.4.19-pre8 && zcat \Quoted 8 lines
Quoted 8 lines
| ../kbuild-2.5-core-14.gz \ | ../kbuild-2.5-common-2.4.19-pre8-1.gz \ | ../kbuild-2.5-i386-2.4.19-pre8-1.gz \ | | patch -sp1 | | and read the comprehensive manual in Documentation/kbuild/ | | What's there to be nervous about?Well perhaps lazy would be more adequate to express my feelings. First why the hell three different diff files?
I don't give a damn witt about what the internal architecure of it is.
Quoted 5 lines
And I don't wan't to miss any non i386 build. (I have ARM for example as well.) I don't wan't to care whatever this is all complete. And I just expect the documentation about how to use it right at the top in README or INSTALL. I'm a human and humans tend to love to stick to habits. And I already got in to the habit of:
Well, if you took the time to look at the code, you'd realize it's actually very practical. The main package is really just the very core of the system, something that you needn't touch at all if going from one kernel revision to another. Obviously, the "common" patch restructures the Makefile skeleton and other stuff that is shared across all archs. No need to explain the arch-specific diff I guess.
So, when upgrading to another kernel release, you end up only having to "patch -R + patch" the "build recipes" (a couple K), not the system core (a 1MB diff) that scarcely ever changes.
Quoted 2 lines
- One thing one patch.
- Something to compile look after README or INSTALL first there.
Well, I guess Keith could add an INSTALL file to the d/l list on sf. If it's to only read "hey, you grab these 3 files and feed them to patch," it's not much work and apparently there are people who will find such info helpful.
Quoted 2 lines
Even the kbuild name isn't intuitive
new-kernel-build would be more clear.
Who cares about the name, man :)
I mean, as far as I'm concerned, if it's called hrmops55 and does the thing right, I'm not saying a word against it.
Quoted 3 lines
I really just wan't to see how it works first and I wan't to see that it indeed work's better then what I have already before I look at how it is done
-> go -> read -> the -> manual -> thankyou :)
Quoted 3 lines
Call me arrogant (and most propably it would be justifyed in this case), but this aggressive splitup gives me prejudictions about the whole thing simply beeing overdesigned...
It's really not.
T.
Quoted 8 lines · Martin Dalecki
Well what I know is the following: Kudzu did set up my system to use usb-ohci driver similar. And now you have apparently started to obsolete some of them. I know that there are still two drivers supported. So I would really really wellcome it if you would just write up some documentation about the *whole* transition from using some old driver to the one which is supposed to be the driver used in the future.
Hm, Kudzu shouldn't care, you just should not be able to load the usb-ohci module on startup. At least that's what happens to me on my boxes.
Anyway, here's the documentation that you need:
The module usb-ohci is now gone. Use ohci-hcd instead.
The people with UHCI controllers have a big more documentation to read:
The module uhci is now gone. If you used this module, use uhci-hcd instead. The module usb-uhci is now gone. If you used this module, use usb-uhci-hcd instead. If you have a preference over which UHCI module works better for you, please email [email protected] your comments, as one of these modules will be going away in the near future.
Quoted 4 lines
Most likely this would be of course module aliasing entires which could be used in
/etc/modules.confto cheat around kudzu - this way it would be for example possible to have a nice dual boot configuration.
Heh, I don't mess around with module aliasing, so I don't know the first thing of what to do in this situation. Someone reading the above documentation should be able to figure it all out.
Quoted 6 lines
(I can adjust my boot scripts to set up an apriopriate symlink from
/etc/modules.confto modules-2.5.18.confg and modules-2.4.18.conf debpending what I'm booting anyway...)In short - if you change something about the configuration and usage - just document it please.
We really don't use CHANGES anymore. Where do you propose things like the above few sentences go to where everyone will know to look?
thanks,
greg k-h
Re: [Linux-usb-users] Re: What to do with all of the USB UHCI driversin the kernel?
In reply to Stephen J. GowdyQuoted 7 lines · Martin Dalecki
Uz.ytkownik Stephen J. Gowdy napisa?:
Quoted 5 lines
Hi Martin,
What do you actually want to know? That an EHCI controller should
use the ehci-hcd driver and that the OHCI controller should use the ohci-hcd controller? Or that the uhci-* drivers can't drive a EHCI or OHCI controller? Or something else?
Rename the old driver to *-old.
Rename the new drivers to the "normal" names.
There is no reason to confurce the userbase to fiddle with their configurations. You said the old drivers could seamlessly be replacedby s.th. better, so this policy will implement this at it's best.
You will immediately get a lot of testing !!!
Regards, Gunther
Uz.ytkownik Greg KH napisa?:
Quoted 10 lines
Anyway, here's the documentation that you need:
The module usb-ohci is now gone. Use ohci-hcd instead.The people with UHCI controllers have a big more documentation to read:
The module uhci is now gone. If you used this module, use uhci-hcd instead. The module usb-uhci is now gone. If you used this module, use usb-uhci-hcd instead. If you have a preference over which UHCI module works better for you, please email [email protected] your comments, as one of these modules will be going away in the near future.
Thank's that is explaining it.
But I would have loved it if it appeared with + in front in the patch somewhere. That's the only true problem I had. OK?
BTW.> usb-ohci seems to be a more reasonable name, since it tells me directly - hey buddy I'm USB the -hcd doen't tell me anything in addition and is entierly redundant, or is there a ohci.o module there?
And why not just doing the following.
1. Rename usb-ohci to usb-ohci-old
2. Rename ohci-hcd to usb-ohci
Much less grief and guessing what happens :-).
Just a suggestion.
Re: [Linux-usb-users] Re: What to do with all of the USB UHCI drivers
In reply to Stephen J. GowdyQuoted 26 lines · Martin Dalecki
Uz.ytkownik Greg KH napisa?:
Quoted 10 lines
Anyway, here's the documentation that you need:
The module usb-ohci is now gone. Use ohci-hcd instead.The people with UHCI controllers have a big more documentation to read:
The module uhci is now gone. If you used this module, use uhci-hcd instead. The module usb-uhci is now gone. If you used this module, use usb-uhci-hcd instead. If you have a preference over which UHCI module works better for you, please email [email protected] your comments, as one of these modules will be going away in the near future.Thank's that is explaining it.
But I would have loved it if it appeared with + in front in the patch somewhere. That's the only true problem I had. OK?BTW.> usb-ohci seems to be a more reasonable name, since it tells me directly - hey buddy I'm USB the -hcd doen't tell me anything in addition and is entierly redundant, or is there a ohci.o module there?
And why not just doing the following.
Quoted 8 lines
1. Rename usb-ohci to usb-ohci-old
2. Rename ohci-hcd to usb-ohci
Much less grief and guessing what happens :-).
Just a suggestion.
You do not understand the _cause_ for the rename.
hcd stands for Host Controller Device. The names mark them as that. Now you can argue, that all the SCSI controller modules do not have HBC (host bus controller) in their name - and then look at the different naming of the other scsi modules:
scsi_mod.o sd_mod.o sg.o sr_mod.o st.o
Also not very intuitiv. We could put them in different subdirs...
Re: [Linux-usb-users] Re: What to do with all of the USB UHCI drivers
In reply to Stephen J. GowdyQuoted 25 lines · Peter Wächtler
Quoted 24 lines · Martin Dalecki
Uz.ytkownik Greg KH napisa?:
Quoted 10 lines
Anyway, here's the documentation that you need:
The module usb-ohci is now gone. Use ohci-hcd instead.The people with UHCI controllers have a big more documentation to read:
The module uhci is now gone. If you used this module, use uhci-hcd instead. The module usb-uhci is now gone. If you used this module, use usb-uhci-hcd instead. If you have a preference over which UHCI module works better for you, please email [email protected] your comments, as one of these modules will be going away in the near future.Thank's that is explaining it.
But I would have loved it if it appeared with + in front in the patch somewhere. That's the only true problem I had. OK?BTW.> usb-ohci seems to be a more reasonable name, since it tells me directly - hey buddy I'm USB the -hcd doen't tell me anything in addition and is entierly redundant, or is there a ohci.o module there?
Quoted 7 lines
Quoted 7 lines
And why not just doing the following.
1. Rename usb-ohci to usb-ohci-old
2. Rename ohci-hcd to usb-ohci
Much less grief and guessing what happens :-).
Second this (in fact "first this").
Elegant solution which would not only habe prevented this thread but help developers and users no end.
Quoted 6 lines
Quoted 4 lines
Just a suggestion.
You do not understand the _cause_ for the rename.
Martin and me understand what hcd means, and we think it is non-untuitive and user-hostile.
Quoted 10 lines
hcd stands for Host Controller Device. The names mark them as that. Now you can argue, that all the SCSI controller modules do not have HBC (host bus controller) in their name - and then look at the different naming of the other scsi modules:
scsi_mod.o sd_mod.o sg.o sr_mod.o st.o
Also not very intuitiv. We could put them in different subdirs...
Obviously you cannot balance between intuitive and user-friendly.
Re: [Linux-usb-users] Re: What to do with all of the USB UHCI drivers
I don't have a problem with the new names. Users shouldn't care, and when they do they want to report problems. It would help to know what they are actually talking about without need to remember which kernel the rename happened in (and that distributions may backport the changes to earlier kernels would only complicate this more).
Quoted 74 lines · Gunther Mayer
Quoted 25 lines · Peter Wächtler
Quoted 24 lines · Martin Dalecki
Uz.ytkownik Greg KH napisa?:
Quoted 10 lines
Anyway, here's the documentation that you need:
The module usb-ohci is now gone. Use ohci-hcd instead.The people with UHCI controllers have a big more documentation to read:
The module uhci is now gone. If you used this module, use uhci-hcd instead. The module usb-uhci is now gone. If you used this module, use usb-uhci-hcd instead. If you have a preference over which UHCI module works better for you, please email [email protected] your comments, as one of these modules will be going away in the near future.Thank's that is explaining it.
But I would have loved it if it appeared with + in front in the patch somewhere. That's the only true problem I had. OK?BTW.> usb-ohci seems to be a more reasonable name, since it tells me directly - hey buddy I'm USB the -hcd doen't tell me anything in addition and is entierly redundant, or is there a ohci.o module there?
Quoted 7 lines
Quoted 7 lines
And why not just doing the following.
1. Rename usb-ohci to usb-ohci-old
2. Rename ohci-hcd to usb-ohci
Much less grief and guessing what happens :-).
Second this (in fact "first this").
Elegant solution which would not only habe prevented this thread but help developers and users no end.Quoted 6 lines
Quoted 4 lines
Just a suggestion.
You do not understand the _cause_ for the rename.
Martin and me understand what hcd means, and we think it is non-untuitive and user-hostile.
Quoted 10 lines
hcd stands for Host Controller Device. The names mark them as that. Now you can argue, that all the SCSI controller modules do not have HBC (host bus controller) in their name - and then look at the different naming of the other scsi modules:
scsi_mod.o sd_mod.o sg.o sr_mod.o st.o
Also not very intuitiv. We could put them in different subdirs...
Obviously you cannot balance between intuitive and user-friendly.
_______________________________________________________________
Don't miss the 2002 Sprint PCS Application Developer's Conference August 25-28 in Las Vegas -- http://devcon.sprintpcs.com/adp/index.cfm
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-users
Signature · 6 lines
/------------------------------------+-------------------------\ |Stephen J. Gowdy | SLAC, MailStop 34, | |http://www.slac.stanford.edu/~gowdy/ | 2575 Sand Hill Road, | |http://calendar.yahoo.com/gowdy | Menlo Park CA 94025, USA | |EMail: [email protected] | Tel: +1 650 926 3144 | \------------------------------------+-------------------------/
Quoted 5 lines
Could you please just do me a small favour and drop something in to linux/Documentation. Becouse I'm right now already confused about which driver to use and which alias to put in
/etc/modules.confso kudzu stops hollering about not knowing what to do
if I out of a sudden reboot in to 2.5.xx kernel.
I don't think that belongs in Linux/Documentation since it's specific to RedHat. Apart from the "which UHCI" confusion, you could use the /etc/rc.d/init.d/hotplug script to start up USB. (That's not part of RedHat though; they use kudzu to promote what seems like more of a static config model.)
I have an RH 7.3 system that has this ... though I had to tweak it by hand since their setup stored multiple entries for some drivers, and didn't put EHCI first (so devices on USB 2.0 adapters would start out with one driver, disconnect, then reconnect using the "right" host controller):
alias usb-controller ehci-hcd
alias usb-controller1 ohci-hcd
alias usb-controller2 uhci-hcd
alias usb-controller3 usb-ohci
alias usb-controller4 usb-uhci
Note that the "uhci-hcd" would be "usb-uhci-hcd" if you're trying to experiment with that version on the 2.5 kernel, if you're not doing manual rmmod/modprobe (which I usually end up doing).
That causes silly boot messages for the hardware or drivers you may not have, but works reliably/correctly for all current kernels.
- Dave
Quoted 3 lines · Greg KH
From now until July 1, I want everyone to test out both the uhci-hcd and usb-uhci-hcd drivers on just about every piece of hardware they can find.
I need this patch to convince the Makefile to build the new drivers.
drivers/usb/Makefile renamed from drivers/usb/Makefile.orig +2 -0
@@ -12,6 +12,8 @@ subdir-$(CONFIG_USB_EHCI_HCD) += host subdir-$(CONFIG_USB_OHCI_HCD) += host subdir-$(CONFIG_USB_OHCI) += host +subdir-$(CONFIG_USB_UHCI_HCD) += host +subdir-$(CONFIG_USB_UHCI_HCD_ALT) += host subdir-$(CONFIG_USB_UHCI_ALT) += host subdir-$(CONFIG_USB_UHCI) += host
Signature · 2 lines
Peter Osterlund - [email protected] http://w1.894.telia.com/~u89404340
Quoted 7 lines · Peter Osterlund
Quoted 3 lines · Greg KH
From now until July 1, I want everyone to test out both the uhci-hcd and usb-uhci-hcd drivers on just about every piece of hardware they can find.
I need this patch to convince the
Makefileto build the new drivers.
Already in Linus' tree, and will show up in 2.5.18.
thanks,
greg k-h
Quoted 7 lines · Greg KH
Let me (and the linux-usb-devel list) know about any thoughts you have pertaining to liking one of the drivers over the other one. Speed tests, size tests, code pretty tests, comment spelling tests, documentation tests, you name it, I want to know about it. If you don't want your comments to be public, send them to me directly and I will not let anyone else know what you said, but will use the info to try to pick which one should stay.
I did a simple test reading a bunch of files from a Freecom CDRW drive.
usb-uhci-hcd 17840 0 (unused)
pengo:/cdrw$ time wc petero/mp3/madonna/*
...
188462 1091976 50386286 total real 1m24.930s
user 0m10.440s
sys 0m1.360suhci-hcd 24608 0 (unused)
pengo:/cdrw$ time wc petero/mp3/madonna/*
...
188462 1091976 50386286 total real 1m33.095s
user 0m12.670s
sys 0m11.870sSo the usb-uhci-hcd driver is 27% smaller, gives a 10% higher transfer rate and produces less system load during the data transfers. (About 70% idle time versus 50% idle time on my 233MHz MMX system.)
Signature · 2 lines
Peter Osterlund - [email protected] http://w1.894.telia.com/~u89404340
Quoted 4 lines · Greg KH
From now until July 1, I want everyone to test out both the uhci-hcd and usb-uhci-hcd drivers on just about every piece of hardware they can find. This includes SMP, UP, preempt kernels, big and little endian machines, and loads of different types of USB devices.
2.5.17, USB onboard Abit BE6, modem ADSL Speedtouch USB :
1) uhci without loading the modem driver :
# modprobe uhci # rmmod uhci OOPS (not logged) kernel: kmem_cache_destroy: Can't free all objects cff416f8 kernel: uhci: not all urb_priv's were freed
2) usb-uhci uhci-hcd usb-uhci-hcd produce no visible difference in the operation of the modem (62 kbytes/s for 608 kbits/s connexion speed).
Pierre
Signature · 3 lines
------------------------------------------------ Pierre Rousselet <[email protected]> ------------------------------------------------