2021-06-17 16:34:12

by Ruslan Bilovol

[permalink] [raw]
Subject: [PATCH] usb: gadget: f_hid: fix endianness issue with descriptors

Running sparse checker it shows warning message about
incorrect endianness used for descriptor initialization:

| f_hid.c:91:43: warning: incorrect type in initializer (different base types)
| f_hid.c:91:43: expected restricted __le16 [usertype] bcdHID
| f_hid.c:91:43: got int

Fixing issue with cpu_to_le16() macro

Fixes: 71adf1189469 ("USB: gadget: add HID gadget driver")
Cc: Fabien Chouteau <[email protected]>
Cc: Segiy Stetsyuk <[email protected]>
Cc: [email protected]
Signed-off-by: Ruslan Bilovol <[email protected]>
---
drivers/usb/gadget/function/f_hid.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/usb/gadget/function/f_hid.c b/drivers/usb/gadget/function/f_hid.c
index 70774d8cb14e..02683ac0719d 100644
--- a/drivers/usb/gadget/function/f_hid.c
+++ b/drivers/usb/gadget/function/f_hid.c
@@ -88,7 +88,7 @@ static struct usb_interface_descriptor hidg_interface_desc = {
static struct hid_descriptor hidg_desc = {
.bLength = sizeof hidg_desc,
.bDescriptorType = HID_DT_HID,
- .bcdHID = 0x0101,
+ .bcdHID = cpu_to_le16(0x0101),
.bCountryCode = 0x00,
.bNumDescriptors = 0x1,
/*.desc[0].bDescriptorType = DYNAMIC */
--
2.17.1


2021-06-17 19:22:56

by Greg Kroah-Hartman

[permalink] [raw]
Subject: Re: [PATCH] usb: gadget: f_hid: fix endianness issue with descriptors

On Thu, Jun 17, 2021 at 07:27:55PM +0300, Ruslan Bilovol wrote:
> Running sparse checker it shows warning message about
> incorrect endianness used for descriptor initialization:
>
> | f_hid.c:91:43: warning: incorrect type in initializer (different base types)
> | f_hid.c:91:43: expected restricted __le16 [usertype] bcdHID
> | f_hid.c:91:43: got int
>
> Fixing issue with cpu_to_le16() macro
>
> Fixes: 71adf1189469 ("USB: gadget: add HID gadget driver")
> Cc: Fabien Chouteau <[email protected]>
> Cc: Segiy Stetsyuk <[email protected]>
> Cc: [email protected]
> Signed-off-by: Ruslan Bilovol <[email protected]>
> ---
> drivers/usb/gadget/function/f_hid.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/usb/gadget/function/f_hid.c b/drivers/usb/gadget/function/f_hid.c
> index 70774d8cb14e..02683ac0719d 100644
> --- a/drivers/usb/gadget/function/f_hid.c
> +++ b/drivers/usb/gadget/function/f_hid.c
> @@ -88,7 +88,7 @@ static struct usb_interface_descriptor hidg_interface_desc = {
> static struct hid_descriptor hidg_desc = {
> .bLength = sizeof hidg_desc,
> .bDescriptorType = HID_DT_HID,
> - .bcdHID = 0x0101,
> + .bcdHID = cpu_to_le16(0x0101),

This is a BCD value, not a little-endian value, are you sure this
conversion is correct?

thanks,

greg k-h

2021-06-18 02:58:43

by Alan Stern

[permalink] [raw]
Subject: Re: [PATCH] usb: gadget: f_hid: fix endianness issue with descriptors

On Thu, Jun 17, 2021 at 06:52:54PM +0200, Greg Kroah-Hartman wrote:
> On Thu, Jun 17, 2021 at 07:27:55PM +0300, Ruslan Bilovol wrote:
> > Running sparse checker it shows warning message about
> > incorrect endianness used for descriptor initialization:
> >
> > | f_hid.c:91:43: warning: incorrect type in initializer (different base types)
> > | f_hid.c:91:43: expected restricted __le16 [usertype] bcdHID
> > | f_hid.c:91:43: got int
> >
> > Fixing issue with cpu_to_le16() macro
> >
> > Fixes: 71adf1189469 ("USB: gadget: add HID gadget driver")
> > Cc: Fabien Chouteau <[email protected]>
> > Cc: Segiy Stetsyuk <[email protected]>
> > Cc: [email protected]
> > Signed-off-by: Ruslan Bilovol <[email protected]>
> > ---
> > drivers/usb/gadget/function/f_hid.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/usb/gadget/function/f_hid.c b/drivers/usb/gadget/function/f_hid.c
> > index 70774d8cb14e..02683ac0719d 100644
> > --- a/drivers/usb/gadget/function/f_hid.c
> > +++ b/drivers/usb/gadget/function/f_hid.c
> > @@ -88,7 +88,7 @@ static struct usb_interface_descriptor hidg_interface_desc = {
> > static struct hid_descriptor hidg_desc = {
> > .bLength = sizeof hidg_desc,
> > .bDescriptorType = HID_DT_HID,
> > - .bcdHID = 0x0101,
> > + .bcdHID = cpu_to_le16(0x0101),
>
> This is a BCD value, not a little-endian value, are you sure this
> conversion is correct?

It's a BCD value, but the storage format is little endian. So yes, the
conversion is correct.

But even more, the conversion is correct because 0x0101 yields exactly
the same sequence of bytes in little-endian and big-endian orders!
Either way, it is two bytes each containing 0x01.

Alan Stern

2021-06-21 09:24:10

by Greg Kroah-Hartman

[permalink] [raw]
Subject: Re: [PATCH] usb: gadget: f_hid: fix endianness issue with descriptors

On Thu, Jun 17, 2021 at 07:44:21PM -0400, Alan Stern wrote:
> On Thu, Jun 17, 2021 at 06:52:54PM +0200, Greg Kroah-Hartman wrote:
> > On Thu, Jun 17, 2021 at 07:27:55PM +0300, Ruslan Bilovol wrote:
> > > Running sparse checker it shows warning message about
> > > incorrect endianness used for descriptor initialization:
> > >
> > > | f_hid.c:91:43: warning: incorrect type in initializer (different base types)
> > > | f_hid.c:91:43: expected restricted __le16 [usertype] bcdHID
> > > | f_hid.c:91:43: got int
> > >
> > > Fixing issue with cpu_to_le16() macro
> > >
> > > Fixes: 71adf1189469 ("USB: gadget: add HID gadget driver")
> > > Cc: Fabien Chouteau <[email protected]>
> > > Cc: Segiy Stetsyuk <[email protected]>
> > > Cc: [email protected]
> > > Signed-off-by: Ruslan Bilovol <[email protected]>
> > > ---
> > > drivers/usb/gadget/function/f_hid.c | 2 +-
> > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/usb/gadget/function/f_hid.c b/drivers/usb/gadget/function/f_hid.c
> > > index 70774d8cb14e..02683ac0719d 100644
> > > --- a/drivers/usb/gadget/function/f_hid.c
> > > +++ b/drivers/usb/gadget/function/f_hid.c
> > > @@ -88,7 +88,7 @@ static struct usb_interface_descriptor hidg_interface_desc = {
> > > static struct hid_descriptor hidg_desc = {
> > > .bLength = sizeof hidg_desc,
> > > .bDescriptorType = HID_DT_HID,
> > > - .bcdHID = 0x0101,
> > > + .bcdHID = cpu_to_le16(0x0101),
> >
> > This is a BCD value, not a little-endian value, are you sure this
> > conversion is correct?
>
> It's a BCD value, but the storage format is little endian. So yes, the
> conversion is correct.
>
> But even more, the conversion is correct because 0x0101 yields exactly
> the same sequence of bytes in little-endian and big-endian orders!
> Either way, it is two bytes each containing 0x01.

Ah, which is why no one has noticed this yet :)

I'll go apply this just to be "correct".

thanks,

greg k-h