2.6.0-test7 and HIDBP

2.6.0-test7 and HIDBP

[Norman Diamond]

OK, I already know that I'm half-blind, but now either I'm 75% blind or else
these are facts:

make xconfig has options for HIDBP.
make gconfig doesn't.

Of course I want full HID so this might not matter, but I have memories of
needing HIDBP a few years ago.

Re: 2.6.0-test7 and HIDBP

[Randy.Dunlap]

On Fri, 10 Oct 2003 18:34:45 +0900 "Norman Diamond" <ndiamond@wta.att.ne.jp> wrote:

| OK, I already know that I'm half-blind, but now either I'm 75% blind or else
| these are facts:
| 
| make xconfig has options for HIDBP.
| make gconfig doesn't.
| 
| Of course I want full HID so this might not matter, but I have memories of
| needing HIDBP a few years ago.

I don't know about gconfig, but you don't want or need HIDBP with full HID.

--
~Randy

Re: 2.6.0-test7 and HIDBP

[Norman Diamond]

Randy Dunlap replied to me:

> | OK, I already know that I'm half-blind, but now either I'm 75% blind or
> | else these are facts:
> |    make xconfig has options for HIDBP.
> |    make gconfig doesn't.
> | Of course I want full HID so this might not matter, but I have memories
> | of needing HIDBP a few years ago.
>
> I don't know about gconfig, but you don't want or need HIDBP with full
> HID.

Indeed I don't want or need HIDBP with full HID, which is the reason I said
"so this might not matter".  Also I don't know if anyone still needs HIDBP
or not, which is the reason I said "I have memories of needing HIDBP a few
years ago".

Surely if anyone does still need HIDBP then it should be added to gconfig?
Surely if no one needs HIDBP any more then it should be deleted entirely?

Or maybe I'm 75% blind, maybe I it really is in gconfig and I couldn't see
it.  But no one has said that this is the case.

Re: 2.6.0-test7 and HIDBP

[Sam Ravnborg]

On Sat, Oct 11, 2003 at 07:10:50PM +0900, Norman Diamond wrote:
> 
> Or maybe I'm 75% blind, maybe I it really is in gconfig and I couldn't see
> it.  But no one has said that this is the case.

If there are dofferences between what is shown in xconfig and gconfig, then
there is a bug in one of the front-ends.
They use the same back-end - and what they display is based on the same
input files.

	Sam

Re: 2.6.0-test7 and HIDBP

[Norman Diamond]

Sam Ravnborg replied to me:

> If there are dofferences between what is shown in xconfig and gconfig,
> then there is a bug in one of the front-ends.
> They use the same back-end - and what they display is based on the same
> input files.

Or maybe there are bugs in both of these front-ends.  I finally thought of
looking at two .config files in text editors, one produced by make xconfig
and one produced by make gconfig.  Both are for kernel 2.6.0-test7 though
the machines and original distros are different.

In the .config generated by make xconfig, the following lines exist:

#
# USB Human Interface Devices (HID)
#
CONFIG_USB_HID=m
CONFIG_USB_HIDINPUT=y
# CONFIG_HID_FF is not set
# CONFIG_USB_HIDDEV is not set

#
# USB HID Boot Protocol drivers
#
# CONFIG_USB_KBD is not set
# CONFIG_USB_MOUSE is not set
# CONFIG_USB_AIPTEK is not set
# CONFIG_USB_WACOM is not set
[...]

OK, aside from making me wonder what HIDDEV refers to and why it isn't set,
and making it rather non-obvious which drivers are HID Boot Protocol and
which aren't, the settings look pretty reasonable to me.

In the .config generated by make gconfig, the following lines exist:

#
# USB Human Interface Devices (HID)
#
CONFIG_USB_HID=y
CONFIG_USB_HIDINPUT=y
# CONFIG_HID_FF is not set
CONFIG_USB_HIDDEV=y
# CONFIG_USB_AIPTEK is not set
# CONFIG_USB_WACOM is not set
[...]

With this file in a text editor, searching for string USB_KBD results in the
string not being found.

Re: 2.6.0-test7 and HIDBP

[Randy.Dunlap]

On Sat, 11 Oct 2003 22:15:50 +0900 "Norman Diamond" <ndiamond@wta.att.ne.jp> wrote:

| Sam Ravnborg replied to me:
| 
| > If there are dofferences between what is shown in xconfig and gconfig,
| > then there is a bug in one of the front-ends.
| > They use the same back-end - and what they display is based on the same
| > input files.
| 
| Or maybe there are bugs in both of these front-ends.  I finally thought of
| looking at two .config files in text editors, one produced by make xconfig
| and one produced by make gconfig.  Both are for kernel 2.6.0-test7 though
| the machines and original distros are different.
| 
| In the .config generated by make xconfig, the following lines exist:
| 
| #
| # USB Human Interface Devices (HID)
| #
| CONFIG_USB_HID=m
| CONFIG_USB_HIDINPUT=y
| # CONFIG_HID_FF is not set
| # CONFIG_USB_HIDDEV is not set
| 
| #
| # USB HID Boot Protocol drivers
| #
| # CONFIG_USB_KBD is not set
| # CONFIG_USB_MOUSE is not set
| # CONFIG_USB_AIPTEK is not set
| # CONFIG_USB_WACOM is not set
| [...]
| 
| OK, aside from making me wonder what HIDDEV refers to and why it isn't set,
| and making it rather non-obvious which drivers are HID Boot Protocol and
| which aren't, the settings look pretty reasonable to me.
| 
| In the .config generated by make gconfig, the following lines exist:
| 
| #
| # USB Human Interface Devices (HID)
| #
| CONFIG_USB_HID=y
| CONFIG_USB_HIDINPUT=y
| # CONFIG_HID_FF is not set
| CONFIG_USB_HIDDEV=y
| # CONFIG_USB_AIPTEK is not set
| # CONFIG_USB_WACOM is not set
| [...]
| 
| With this file in a text editor, searching for string USB_KBD results in the
| string not being found.

Let me see if I am understanding this..
Is the problem that some generated .config files contain different
strings for options that are _not_ set, depending on which
front-end is used?
Or are you discussing a bug in the (on-screen) viewable part of a
config front-end?

If so, I'll confirm that different config front-ends generate different
.config files, and it makes using "vi .config && make oldconfig"
not as usable as I'd like it to be, but I don't know that it's a bug.

--
~Randy

Re: 2.6.0-test7 and HIDBP

[Randy.Dunlap]

On Mon, 13 Oct 2003 09:49:43 -0700 "Randy.Dunlap" <rddunlap@osdl.org> wrote:

| On Sat, 11 Oct 2003 22:15:50 +0900 "Norman Diamond" <ndiamond@wta.att.ne.jp> wrote:
| 
| | Or maybe there are bugs in both of these front-ends.  I finally thought of
| | looking at two .config files in text editors, one produced by make xconfig
| | and one produced by make gconfig.  Both are for kernel 2.6.0-test7 though
| | the machines and original distros are different.
| | 
| | In the .config generated by make xconfig, the following lines exist:
| | 
| | #
| | # USB Human Interface Devices (HID)
| | #
| | CONFIG_USB_HID=m
| | CONFIG_USB_HIDINPUT=y
| | # CONFIG_HID_FF is not set
| | # CONFIG_USB_HIDDEV is not set
| | 
| | #
| | # USB HID Boot Protocol drivers
| | #
| | # CONFIG_USB_KBD is not set
| | # CONFIG_USB_MOUSE is not set
| | # CONFIG_USB_AIPTEK is not set
| | # CONFIG_USB_WACOM is not set
| | [...]
| | 
| | OK, aside from making me wonder what HIDDEV refers to and why it isn't set,
| | and making it rather non-obvious which drivers are HID Boot Protocol and
| | which aren't, the settings look pretty reasonable to me.
| | 
| | In the .config generated by make gconfig, the following lines exist:
| | 
| | #
| | # USB Human Interface Devices (HID)
| | #
| | CONFIG_USB_HID=y
| | CONFIG_USB_HIDINPUT=y
| | # CONFIG_HID_FF is not set
| | CONFIG_USB_HIDDEV=y
| | # CONFIG_USB_AIPTEK is not set
| | # CONFIG_USB_WACOM is not set
| | [...]
| | 
| | With this file in a text editor, searching for string USB_KBD results in the
| | string not being found.
| 
| Let me see if I am understanding this..
| Is the problem that some generated .config files contain different
| strings for options that are _not_ set, depending on which
| front-end is used?
| Or are you discussing a bug in the (on-screen) viewable part of a
| config front-end?
| 
| If so, I'll confirm that different config front-ends generate different
| .config files, and it makes using "vi .config && make oldconfig"
| not as usable as I'd like it to be, but I don't know that it's a bug.

oops, I mis-diffed the config files, but I'll keep looking since
I think that I've seen something like this before.

--
~Randy

Re: 2.6.0-test7 and HIDBP

[Roman Zippel]

Hi,

On Sat, 11 Oct 2003, Norman Diamond wrote:

> In the .config generated by make gconfig, the following lines exist:
> 
> #
> # USB Human Interface Devices (HID)
> #
> CONFIG_USB_HID=y
> CONFIG_USB_HIDINPUT=y
> # CONFIG_HID_FF is not set
> CONFIG_USB_HIDDEV=y
> # CONFIG_USB_AIPTEK is not set
> # CONFIG_USB_WACOM is not set
> [...]
> 
> With this file in a text editor, searching for string USB_KBD results in the
> string not being found.

If USB_HIDDEV is set to 'y', then you indeed won't see USB_KBD, as it's 
only selectable if USB_HIDDEV is 'n' or 'm'. It seems to work here.

bye, Roman

Re: 2.6.0-test7 and HIDBP

[Norman Diamond]

Roman Zippel replied to me:

> If USB_HIDDEV is set to 'y', then you indeed won't see USB_KBD, as it's
> only selectable if USB_HIDDEV is 'n' or 'm'. It seems to work here.

You are right.  On the machine with Gnome, I ran "make gconfig" again,
changed one of the HID options from "y" to "m", and then the HIDBP options
became available.

Sorry, I didn't guess that changing a setting between "y" and "m" could
affect the visibility of other options this way.  In this case it seems
reasonable, but I still didn't imagine this kind of thing.  Sorry.

Re: 2.6.0-test7 and HIDBP

[Randy.Dunlap]

Norman Diamond wrote:
| Or maybe there are bugs in both of these front-ends.  I finally thought of
| looking at two .config files in text editors, one produced by make xconfig
| and one produced by make gconfig.  Both are for kernel 2.6.0-test7 though
| the machines and original distros are different.

Hi Norman,

I just tested make menuconfig/xconfig/gconfig and got the same results
from all 3 of them.

| In the .config generated by make xconfig, the following lines exist:
| 
| #
| # USB Human Interface Devices (HID)
| #
| CONFIG_USB_HID=m
| CONFIG_USB_HIDINPUT=y
| # CONFIG_HID_FF is not set
| # CONFIG_USB_HIDDEV is not set
| 
| #
| # USB HID Boot Protocol drivers
| #
| # CONFIG_USB_KBD is not set
| # CONFIG_USB_MOUSE is not set
| # CONFIG_USB_AIPTEK is not set
| # CONFIG_USB_WACOM is not set
| [...]

USB_KBD and USB_MOUSE show up here (above) because USB_HID=m.
Somehow it got changed to USB_HID=y in the config file below,
and that changes the rules for some of the items under it.

| OK, aside from making me wonder what HIDDEV refers to and why it isn't set,
| and making it rather non-obvious which drivers are HID Boot Protocol and
| which aren't, the settings look pretty reasonable to me.
| 
| In the .config generated by make gconfig, the following lines exist:
| 
| #
| # USB Human Interface Devices (HID)
| #
| CONFIG_USB_HID=y
| CONFIG_USB_HIDINPUT=y
| # CONFIG_HID_FF is not set
| CONFIG_USB_HIDDEV=y
| # CONFIG_USB_AIPTEK is not set
| # CONFIG_USB_WACOM is not set
| [...]
| 
| With this file in a text editor, searching for string USB_KBD results in the
| string not being found.

I think that Roman Zippel already replied to this, but you won't
find USB_KBD there because USB_HID=y and they are mutually
exclusive.

In summary, I checked again and still don't see a problem.

--
~Randy