Re: Request for reverting the commit for Samsung HID driver

Re: Request for reverting the commit for Samsung HID driver

[gregkh]

On Wed, Mar 30, 2022 at 03:21:22PM +0900, 조준완 wrote:
>  
> 
>  
> 
> Dear Greg Kroah-Hartman,

<snip>

Please resend in non-html format so it properly makes it to the mailing
lists and I will gladly respond.

thanks,

greg k-h

Re: Request for reverting the commit for Samsung HID driver

[Jiri Kosina]

On Wed, 30 Mar 2022, 조준완 wrote:

> author
> Greg Kroah-Hartman <gregkh@linuxfoundation.org> 2021-12-01 19:35:03 +0100
> 
> committer
> Benjamin Tissoires <benjamin.tissoires@redhat.com> 2021-12-02 15:36:18 +0100
> 
> commit
> 93020953d0fa7035fd036ad87a47ae2b7aa4ae33 (patch)
> 
> tree
> f734e9962e28cedcccfbb109e510d39a18ed6d34 /drivers/hid/hid-samsung.c
> 
> parent
> 720ac467204a70308bd687927ed475afb904e11b (diff)
> 
> download
> linux-93020953d0fa7035fd036ad87a47ae2b7aa4ae33.tar.gz
>  
> HID: check for valid USB device for many HID drivers
> 
> Many HID drivers assume that the HID device assigned to them is a USB 
> device as that was the only way HID devices used to be able to be 
> created in Linux. However, with the additional ways that HID devices can 
> be created for many different bus types, that is no longer true, so 
> properly check that we have a USB device associated with the HID device 
> before allowing a driver that makes this assumption to claim it.
> 
> 
> diff --git a/drivers/hid/hid-samsung.c b/drivers/hid/hid-samsung.c
> index 2e1c31156eca0..cf5992e970940 100644
> --- a/drivers/hid/hid-samsung.c
> +++ b/drivers/hid/hid-samsung.c
> 
> @@ -152,6 +152,9 @@ static int samsung_probe(struct hid_device *hdev,
> 
> int ret;
> 
> unsigned int cmask = HID_CONNECT_DEFAULT;
> 
> 
> + if (!hid_is_usb(hdev))
> 
> + return -EINVAL;
> 
> +
> 
> ret = hid_parse(hdev);
> 
> if (ret) {
> 
> hid_err(hdev, "parse failed\n");
> 
> 
> Bluetooth HID devices like Keyboard and mice don't work at all because 
> of this commit.

Could you please elaborate a little bit more -- which Bluetooth device 
(VID/PID) are you referring to please? hid-samsung as-is in mainline 
should only match against these two:

	static const struct hid_device_id samsung_devices[] = {
	        { HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_IR_REMOTE) },
	        { HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD_MOUSE) },
	        { }
	};
	MODULE_DEVICE_TABLE(hid, samsung_devices);

which are both USB devices, not Bluetooth.

> We Samsung applies several patches to Samsung drivers for Android models 
> based on Linux kernel Samsung Driver. For example, add Samsung 
> accessories or define key codes for special key processing.
> 
> We don't want any changes in Samsung hid driver by others.

That's not how Linux development works though.

> Soorer or later, I have a plan to contribute a modified driver by 
> Samsung that is currently in use on Samsung Android.

That would be very much appreciated, thanks.

Once you do so, you can very well become maintainers of hid-samsung driver 
(which would make a lot of sense for someone from Samsung to actually 
participate in development of that driver), and have much better influence 
on what goes in and in which form.

> Anyway, we sincerely hope you will revert this commit because Samsung 
> driver is not just for USB accessories.

I still would like to understand this part, because as far as I can tell, 
the in-kernel one is.

Thanks,

-- 
Jiri Kosina
SUSE Labs

Re: Request for reverting the commit for Samsung HID driver

[gregkh]

On Wed, Mar 30, 2022 at 04:01:59PM +0900, 조준완 wrote:
>  
> 
> Dear Greg Kroah-Hartman,

<snip>

Again, this was sent in html format, which is rejected by the lists.
Please fix your email client and try again.

thanks,

greg k-h

RE:(2) Request for reverting the commit for Samsung HID driver

[조준완]

Dear Jiri Kosina,

Thank you for your propt reponse.


Please refer to accossories below.


static const struct hid_device_id samsung_devices[] = {
	{ HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_IR_REMOTE) },
	{ HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD_MOUSE) },
    { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD) },
    { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_GAMEPAD) },
    { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_ACTIONMOUSE) },
    { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_UNIVERSAL_KBD) },
    { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_MULTI_HOGP_KBD) },
	{ }
};
MODULE_DEVICE_TABLE(hid, samsung_devices);


Mobile users prefer to use Bluetooth devices instead of USB devices.
The commit should be reverted for Bluetooth accessories.

Due to internal problem, it takes time to upload Samsung's patch.
So, please first revert the commit :)

Thanks

BR.
Junwan.

 
--------- Original Message ---------
Sender : Jiri Kosina <jikos@kernel.org>
Date : 2022-03-30 16:52 (GMT+9)
Title : Re: Request for reverting the commit for Samsung HID driver
 
On Wed, 30 Mar 2022, 조준완 wrote:
 
> author
> Greg Kroah-Hartman <gregkh@linuxfoundation.org> 2021-12-01 19:35:03 +0100
> 
> committer
> Benjamin Tissoires <benjamin.tissoires@redhat.com> 2021-12-02 15:36:18 +0100
> 
> commit
> 93020953d0fa7035fd036ad87a47ae2b7aa4ae33 (patch)
> 
> tree
> f734e9962e28cedcccfbb109e510d39a18ed6d34 /drivers/hid/hid-samsung.c
> 
> parent
> 720ac467204a70308bd687927ed475afb904e11b (diff)
> 
> download
> linux-93020953d0fa7035fd036ad87a47ae2b7aa4ae33.tar.gz
>  
> HID: check for valid USB device for many HID drivers
> 
> Many HID drivers assume that the HID device assigned to them is a USB 
> device as that was the only way HID devices used to be able to be 
> created in Linux. However, with the additional ways that HID devices can 
> be created for many different bus types, that is no longer true, so 
> properly check that we have a USB device associated with the HID device 
> before allowing a driver that makes this assumption to claim it.
> 
> 
> diff --git a/drivers/hid/hid-samsung.c b/drivers/hid/hid-samsung.c
> index 2e1c31156eca0..cf5992e970940 100644
> --- a/drivers/hid/hid-samsung.c
> +++ b/drivers/hid/hid-samsung.c
> 
> @@ -152,6 +152,9 @@ static int samsung_probe(struct hid_device *hdev,
> 
> int ret;
> 
> unsigned int cmask = HID_CONNECT_DEFAULT;
> 
> 
> + if (!hid_is_usb(hdev))
> 
> + return -EINVAL;
> 
> +
> 
> ret = hid_parse(hdev);
> 
> if (ret) {
> 
> hid_err(hdev, "parse failed\n");
> 
> 
> Bluetooth HID devices like Keyboard and mice don't work at all because 
> of this commit.
 
Could you please elaborate a little bit more -- which Bluetooth device 
(VID/PID) are you referring to please? hid-samsung as-is in mainline 
should only match against these two:
 
        static const struct hid_device_id samsung_devices[] = {
                { HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_IR_REMOTE) },
                { HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD_MOUSE) },
                { }
        };
        MODULE_DEVICE_TABLE(hid, samsung_devices);
 
which are both USB devices, not Bluetooth.
 
> We Samsung applies several patches to Samsung drivers for Android models 
> based on Linux kernel Samsung Driver. For example, add Samsung 
> accessories or define key codes for special key processing.
> 
> We don't want any changes in Samsung hid driver by others.
 
That's not how Linux development works though.
 
> Soorer or later, I have a plan to contribute a modified driver by 
> Samsung that is currently in use on Samsung Android.
 
That would be very much appreciated, thanks.
 
Once you do so, you can very well become maintainers of hid-samsung driver 
(which would make a lot of sense for someone from Samsung to actually 
participate in development of that driver), and have much better influence 
on what goes in and in which form.
 
> Anyway, we sincerely hope you will revert this commit because Samsung 
> driver is not just for USB accessories.
 
I still would like to understand this part, because as far as I can tell, 
the in-kernel one is.
 
Thanks,
 
-- 
Jiri Kosina
SUSE Labs
 
 

RE:(2) Request for reverting the commit for Samsung HID driver

[Jiri Kosina]

On Wed, 30 Mar 2022, 조준완 wrote:

> Dear Jiri Kosina,
> 
> Thank you for your propt reponse.
> 
> 
> Please refer to accossories below.
> 
> 
> static const struct hid_device_id samsung_devices[] = {
> 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_IR_REMOTE) },
> 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD_MOUSE) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_GAMEPAD) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_ACTIONMOUSE) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_UNIVERSAL_KBD) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_MULTI_HOGP_KBD) },
> 	{ }
> };
> MODULE_DEVICE_TABLE(hid, samsung_devices);

This is not how the in-kernel driver looks like in Linus' tree though.

> Mobile users prefer to use Bluetooth devices instead of USB devices. The 
> commit should be reverted for Bluetooth accessories.

Yeah, but the driver in kernel tree doesn't support any Bluetooth devices, 
so there is no need to revert that commit there.

Once you submit your patches to support Bluetooth devices, we'd of course 
have to adjust that condition.

Thanks,

-- 
Jiri Kosina
SUSE Labs

Re: (2) Request for reverting the commit for Samsung HID driver

[gregkh]

On Wed, Mar 30, 2022 at 05:09:37PM +0900, 조준완 wrote:
> 
> Dear Jiri Kosina,
> 
> Thank you for your propt reponse.
> 
> 
> Please refer to accossories below.
> 
> 
> static const struct hid_device_id samsung_devices[] = {
> 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_IR_REMOTE) },
> 	{ HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD_MOUSE) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_GAMEPAD) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_ACTIONMOUSE) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_UNIVERSAL_KBD) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_MULTI_HOGP_KBD) },
> 	{ }
> };
> MODULE_DEVICE_TABLE(hid, samsung_devices);
> 
> 
> Mobile users prefer to use Bluetooth devices instead of USB devices.
> The commit should be reverted for Bluetooth accessories.
> 
> Due to internal problem, it takes time to upload Samsung's patch.
> So, please first revert the commit :)

No, not at all.

The commit is correct, as the driver code, as contained in the kernel
tree, requires that check to solve a problem where this driver can be
used to exploit the system.

Any out-of-tree changes you make, you are required to maintain and keep
up to date in order to have them remain working.  That is a requirement
of you, we have no idea what changes anyone else makes, that would be
impossible.

This is the real cost of keeping out-of-tree changes, your management
knows this and plans for it in their budgeting.  There is nothing that
we can do about this.

So the change needs to remain in order for the code to be correct.
Without it, you have a broken and totally insecure system.

If you wish to make the changes you list above to the driver, you now
need to make more changes in order to properly handle the fixes we made
to the code.  Please submit your changes so that we can review them and
accept them if they are correct.

thanks,

greg k-h

RE:(2) (2) Request for reverting the commit for Samsung HID driver

[조준완]

As I said, AP manufacturers don't consider validation of patches.
They just import linux patches into their Board Support Packages without consideration.


So, I can remove the condition that you made once I add Bluetooh accessories to Samsung driver in Linux Mainline?



--------- Original Message ---------
Sender : gregkh@linuxfoundation.org <gregkh@linuxfoundation.org>
Date : 2022-03-30 17:14 (GMT+9)
Title : Re: (2) Request for reverting the commit for Samsung HID driver
 
On Wed, Mar 30, 2022 at 05:09:37PM +0900, 조준완 wrote:
> 
> Dear Jiri Kosina,
> 
> Thank you for your propt reponse.
> 
> 
> Please refer to accossories below.
> 
> 
> static const struct hid_device_id samsung_devices[] = {
>         { HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_IR_REMOTE) },
>         { HID_USB_DEVICE(USB_VENDOR_ID_SAMSUNG, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD_MOUSE) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_KBD) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_GAMEPAD) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_ACTIONMOUSE) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_UNIVERSAL_KBD) },
>     { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_SAMSUNG_ELECTRONICS, USB_DEVICE_ID_SAMSUNG_WIRELESS_MULTI_HOGP_KBD) },
>         { }
> };
> MODULE_DEVICE_TABLE(hid, samsung_devices);
> 
> 
> Mobile users prefer to use Bluetooth devices instead of USB devices.
> The commit should be reverted for Bluetooth accessories.
> 
> Due to internal problem, it takes time to upload Samsung's patch.
> So, please first revert the commit :)
 
No, not at all.
 
The commit is correct, as the driver code, as contained in the kernel
tree, requires that check to solve a problem where this driver can be
used to exploit the system.
 
Any out-of-tree changes you make, you are required to maintain and keep
up to date in order to have them remain working.  That is a requirement
of you, we have no idea what changes anyone else makes, that would be
impossible.
 
This is the real cost of keeping out-of-tree changes, your management
knows this and plans for it in their budgeting.  There is nothing that
we can do about this.
 
So the change needs to remain in order for the code to be correct.
Without it, you have a broken and totally insecure system.
 
If you wish to make the changes you list above to the driver, you now
need to make more changes in order to properly handle the fixes we made
to the code.  Please submit your changes so that we can review them and
accept them if they are correct.
 
thanks,
 
greg k-h
 

RE:(2) (2) Request for reverting the commit for Samsung HID driver

[Jiri Kosina]

On Wed, 30 Mar 2022, 조준완 wrote:

> As I said, AP manufacturers don't consider validation of patches. They 
> just import linux patches into their Board Support Packages without 
> consideration.

We as a Linux kernel community have no insight into the process/setup you 
have with your AP vendors though.

> So, I can remove the condition that you made once I add Bluetooh 
> accessories to Samsung driver in Linux Mainline?

Of course; the code then needs to be adjusted in a way that the USB 
codepath and Bluetooth codepath handle the transport driver intrinsics 
properly.
Which wasn't the case before Greg's commit -- a specially crafted device 
could be bound by hid-samsung with non-HID transport, crashing the kernel 
once it tried to perform USB operations on it.

Thanks,

-- 
Jiri Kosina
SUSE Labs

Re: (2) (2) Request for reverting the commit for Samsung HID driver

[gregkh]

On Wed, Mar 30, 2022 at 05:23:08PM +0900, 조준완 wrote:
> 
> 
> As I said, AP manufacturers don't consider validation of patches.

That is not our issue, but yours.

> They just import linux patches into their Board Support Packages without consideration.

That is good for our changes, but not for yours.

> So, I can remove the condition that you made once I add Bluetooh accessories to Samsung driver in Linux Mainline?

No, not at all unless you wish to have a broken system?

Again, you have the budget and resources to handle the maintenance and
support of your out-of-tree changes.  That is your responsibility to
your customers as per your contracts with them.

As written, the in-kernel code is correct and now fixed.  If you have
modified the driver to do something else, it is your responsibility to
ensure that it remains working as there is nothing that we can do about
that for obvious reasons.

And again, if you remove those lines, you have a known insecure system,
which I doubt you want for your devices.

Please discuss this with your management as they understand the risks
and responsibilities that you have undertaken by having out-of-tree
code.  That is not our responsibility, but yours.

thanks,

greg k-h

RE:(2) (2) (2) Request for reverting the commit for Samsung HID driver

[조준완]

Sorry that I'm not familier with policy of Linux.
That means I cannot add the code to support Bluetoth accessories in Linux Main line?
Samsung's Bluetooth accessories can't be supported forever on a Linux PC?

Could you share any material or link for exploit related with this commit?

 
--------- Original Message ---------
Sender : gregkh@linuxfoundation.org <gregkh@linuxfoundation.org>
Date : 2022-03-30 17:32 (GMT+9)
Title : Re: (2) (2) Request for reverting the commit for Samsung HID driver
 
On Wed, Mar 30, 2022 at 05:23:08PM +0900, 조준완 wrote:
> 
> 
> As I said, AP manufacturers don't consider validation of patches.
 
That is not our issue, but yours.
 
> They just import linux patches into their Board Support Packages without consideration.
 
That is good for our changes, but not for yours.
 
> So, I can remove the condition that you made once I add Bluetooh accessories to Samsung driver in Linux Mainline?
 
No, not at all unless you wish to have a broken system?
 
Again, you have the budget and resources to handle the maintenance and
support of your out-of-tree changes.  That is your responsibility to
your customers as per your contracts with them.
 
As written, the in-kernel code is correct and now fixed.  If you have
modified the driver to do something else, it is your responsibility to
ensure that it remains working as there is nothing that we can do about
that for obvious reasons.
 
And again, if you remove those lines, you have a known insecure system,
which I doubt you want for your devices.
 
Please discuss this with your management as they understand the risks
and responsibilities that you have undertaken by having out-of-tree
code.  That is not our responsibility, but yours.
 
thanks,
 
greg k-h
 

RE:(2) (2) (2) Request for reverting the commit for Samsung HID driver

[Jiri Kosina]

On Wed, 30 Mar 2022, 조준완 wrote:

> Sorry that I'm not familier with policy of Linux.
> That means I cannot add the code to support Bluetoth accessories in Linux Main line?

You of course can (and I'd be actually very happy to take any in-house 
patches you have in order for the hid-samsung driver to support much 
bigger group of devices), but together with that, the hid-samsung driver 
needs to be adjusted so that it treats the BT-specific and USB-speicific 
codepaths properly.

-- 
Jiri Kosina
SUSE Labs

RE:(3) (2) (2) Request for reverting the commit for Samsung HID driver

[조준완]

 
Thank you for good advice.

I'll check security problem when I add in-house patch.
I have no idea about how the commit improves security level.
Was there any security problem?  Could you let us know?

 
--------- Original Message ---------
Sender : Jiri Kosina <jikos@kernel.org>
Date : 2022-03-30 17:49 (GMT+9)
Title : RE:(2) (2) (2) Request for reverting the commit for Samsung HID driver
 
On Wed, 30 Mar 2022, 조준완 wrote:
 
> Sorry that I'm not familier with policy of Linux.
> That means I cannot add the code to support Bluetoth accessories in Linux Main line?
 
You of course can (and I'd be actually very happy to take any in-house 
patches you have in order for the hid-samsung driver to support much 
bigger group of devices), but together with that, the hid-samsung driver 
needs to be adjusted so that it treats the BT-specific and USB-speicific 
codepaths properly.
 
-- 
Jiri Kosina
SUSE Labs
 
 

Re: (2) (2) (2) Request for reverting the commit for Samsung HID driver

[gregkh]

On Wed, Mar 30, 2022 at 05:44:01PM +0900, 조준완 wrote:
> 
> Sorry that I'm not familier with policy of Linux.

There is no "policy" here in the community.  Please contact your Linux
kernel group for how to maintain and handle devices that use the kernel,
they know how to do this quite well.

> That means I cannot add the code to support Bluetoth accessories in Linux Main line?
> Samsung's Bluetooth accessories can't be supported forever on a Linux PC?

Not at all, as Jiri said, please submit your changes.  You just have to
handle the changes that others make as well.

> Could you share any material or link for exploit related with this commit?

No, sorry, I do not think that would be a good idea for obvious reasons.

thanks,

greg k-h

Re: (3) (2) (2) Request for reverting the commit for Samsung HID driver

[gregkh]

On Wed, Mar 30, 2022 at 06:01:50PM +0900, 조준완 wrote:
>  
> Thank you for good advice.
> 
> I'll check security problem when I add in-house patch.
> I have no idea about how the commit improves security level.

Then please do not remove it.

> Was there any security problem?

Yes there was, it is now fixed.

thanks,

greg k-h

RE:(2) (3) (2) (2) Request for reverting the commit for Samsung HID driver

[조준완]

No. If so, I will not apply in house patches.
I'm a Bluetooth person, and I'm only interested in Bluetooth devices.
Codes for Bluetooth devices will never be worked along the commit, anyway, no reason to include them.

Thanks.

--------- Original Message ---------
Sender : gregkh@linuxfoundation.org <gregkh@linuxfoundation.org>
Date : 2022-03-30 18:13 (GMT+9)
Title : Re: (3) (2) (2) Request for reverting the commit for Samsung HID driver
 
On Wed, Mar 30, 2022 at 06:01:50PM +0900, 조준완 wrote:
>  
> Thank you for good advice.
> 
> I'll check security problem when I add in-house patch.
> I have no idea about how the commit improves security level.
 
Then please do not remove it.
 
> Was there any security problem?
 
Yes there was, it is now fixed.
 
thanks,
 
greg k-h
 

Re: (2) (3) (2) (2) Request for reverting the commit for Samsung HID driver

[gregkh]

A: http://en.wikipedia.org/wiki/Top_post
Q: Were do I find info about this thing called top-posting?
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

A: No.
Q: Should I include quotations after my reply?


http://daringfireball.net/2007/07/on_top

On Wed, Mar 30, 2022 at 06:20:58PM +0900, 조준완 wrote:
> 
> No. If so, I will not apply in house patches.

I do not understand what you are referencing here, sorry.  Please
explain.

> I'm a Bluetooth person, and I'm only interested in Bluetooth devices.

That's fine, but this driver, as is in the tree today, does not support
bluetooth devices.  Please submit a patch to resolve that.

> Codes for Bluetooth devices will never be worked along the commit, anyway, no reason to include them.

I do not understand.  Please explain.

thanks,

greg k-h

RE:(2) (2) (3) (2) (2) Request for reverting the commit for Samsung HID driver

[조준완]

Any problmes with my email style?


Anyway, If your commit has to be maintained, the patches I'm going to pu in won't work anyway.
My patches is only for Bluetooth. But your commit restircts usage of Bluetooth devices.

What do you think of making new function like "hid_is_bluetooth"?


 
--------- Original Message ---------
Sender : gregkh@linuxfoundation.org <gregkh@linuxfoundation.org>
Date : 2022-03-30 18:26 (GMT+9)
Title : Re: (2) (3) (2) (2) Request for reverting the commit for Samsung HID driver
 
A: http://en.wikipedia.org/wiki/Top_post
Q: Were do I find info about this thing called top-posting?
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
 
A: No.
Q: Should I include quotations after my reply?
 
 
http://daringfireball.net/2007/07/on_top
 
On Wed, Mar 30, 2022 at 06:20:58PM +0900, 조준완 wrote:
> 
> No. If so, I will not apply in house patches.
 
I do not understand what you are referencing here, sorry.  Please
explain.
 
> I'm a Bluetooth person, and I'm only interested in Bluetooth devices.
 
That's fine, but this driver, as is in the tree today, does not support
bluetooth devices.  Please submit a patch to resolve that.
 
> Codes for Bluetooth devices will never be worked along the commit, anyway, no reason to include them.
 
I do not understand.  Please explain.
 
thanks,
 
greg k-h
 

Re: (2) (2) (3) (2) (2) Request for reverting the commit for Samsung HID driver

[gregkh]

On Wed, Mar 30, 2022 at 06:43:54PM +0900, 조준완 wrote:
> 
> Any problmes with my email style?

Yes, please do not top-post.  Read the links I provided in my last
response.

> Anyway, If your commit has to be maintained, the patches I'm going to pu in won't work anyway.

Then you will need to change the driver to work properly.  This is not a
big problem.

> My patches is only for Bluetooth. But your commit restircts usage of Bluetooth devices.

Yes, so change the driver to support both properly.

> What do you think of making new function like "hid_is_bluetooth"?

I do not think that is necessary here at all.

thanks,

greg k-h

RE:(2) (2) (2) (3) (2) (2) Request for reverting the commit for Samsung HID driver

[조준완]

You are the first person to point this out while exchanging emails.
I don't know how to send in your favorite style in our mail system.
This is the last email, so don't be annoyed.

Anyway because of your commit, Bluetooth accesorries cannot use Samsung driver.
Our driver will be well customized by us.


Thank you for your time.
 
--------- Original Message ---------
Sender : gregkh@linuxfoundation.org <gregkh@linuxfoundation.org>
Date : 2022-03-30 18:49 (GMT+9)
Title : Re: (2) (2) (3) (2) (2) Request for reverting the commit for Samsung HID driver
 
On Wed, Mar 30, 2022 at 06:43:54PM +0900, 조준완 wrote:
> 
> Any problmes with my email style?
 
Yes, please do not top-post.  Read the links I provided in my last
response.
 
> Anyway, If your commit has to be maintained, the patches I'm going to pu in won't work anyway.
 
Then you will need to change the driver to work properly.  This is not a
big problem.
 
> My patches is only for Bluetooth. But your commit restircts usage of Bluetooth devices.
 
Yes, so change the driver to support both properly.
 
> What do you think of making new function like "hid_is_bluetooth"?
 
I do not think that is necessary here at all.
 
thanks,
 
greg k-h
 

Re: Request for reverting the commit for Samsung HID driver

[Krzysztof Kozlowski]

On 30/03/2022 11:58, 조준완 wrote:
> You are the first person to point this out while exchanging emails.
> I don't know how to send in your favorite style in our mail system.
> This is the last email, so don't be annoyed.

Greg is not the first person pointing this out. We all did. :)
Top-posting makes reading and responding much more difficult. It's
inefficient. Proper quoting is the basis of efficient email
communication and well known between tech folks in internet.

Best regards,
Krzysztof

RE:(2) Request for reverting the commit for Samsung HID driver

[조준완]

If bottom posting is natural, why doesn't our mailing system support bottom positing?
It's not clear which method is more efficient. Still a lot of people are arguing which is the most effective way.
I don't want to scroll down to read your reply.
Don't force it on others.

I have simply sent an inquiry email, and you only need to respond to it.

I didn't write a tech thread. This is a kind of business mail. 
I'm sending you a business email about fixes that affect our products.
Do you understand?
Do I have to copy your most recent mail and add the '>' myself to do it the way you like it?
If you tell me how, I'll think about it.

 
 
--------- Original Message ---------
Sender : Krzysztof Kozlowski <krzk@kernel.org>
Date : 2022-03-31 03:43 (GMT+9)
Title : Re: Request for reverting the commit for Samsung HID driver
 
On 30/03/2022 11:58, 조준완 wrote:
> You are the first person to point this out while exchanging emails.
> I don't know how to send in your favorite style in our mail system.
> This is the last email, so don't be annoyed.
 
Greg is not the first person pointing this out. We all did. :)
Top-posting makes reading and responding much more difficult. It's
inefficient. Proper quoting is the basis of efficient email
communication and well known between tech folks in internet.
 
Best regards,
Krzysztof
 
 

Re: (2) Request for reverting the commit for Samsung HID driver

[gregkh]

On Thu, Mar 31, 2022 at 10:51:05AM +0900, 조준완 wrote:
> If bottom posting is natural, why doesn't our mailing system support bottom positing?

Mine does, perhaps you need a better email client?

> It's not clear which method is more efficient. Still a lot of people are arguing which is the most effective way.
> I don't want to scroll down to read your reply.
> Don't force it on others.

Sorry, but that is the Linux kernel development style, and has been for
20+ years.  Again, see these links for why:

	http://en.wikipedia.org/wiki/Top_post
	http://daringfireball.net/2007/07/on_top

> I have simply sent an inquiry email, and you only need to respond to it.

And we did, and asked you nicely to perhaps use a better format to
discuss technical things easier.

> I didn't write a tech thread. This is a kind of business mail. 

This is not a business :)

Good luck!

greg k-h