It's not broken, just something else we could be displaying if someone
wants to add support for it.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Document the bits for bmEthernetStatistics if someone wants to do this
in the future. It's not a FIXME, but rather a TODO if someone has a
device they can test this with.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is no such thing as Wireless USB anymore, it's long gone and
obsolete and never will be seen. So remove the FIXME comment and just
state it as such.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In dump_comm_descriptor() we have a FIXME that we aren't going to deal
with here anytime soon. If people show up with a device that has
additional descriptors, we'll be glad to add them. But until that
happens, just comment it as such as we aren't going to change the code
any time soon.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Gets rid of a CodeQL warning that it might be "confusing" to have
commented out code in a file.
Kids these days, surely they must use an editor that shows comments in
different colors so that they notice this isn't really going to do
anything...
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a severe number of false-positive in code scanning caused by
inspecting meson-internal test files like
'build/meson-private/tmpzb46osmq/testfile.c'.
As a workaround, use the 'filter-sarif' action to filter out these
results before uploading the SARIF (Static Analysis Results Interchange
Format).
This PR was inspired by https://github.com/rauc/rauc/pull/1346 and the
example from https://github.com/advanced-security/filter-sarif.
Signed-off-by: Enrico Joerns <ejo@pengutronix.de>
[Emil: port from rauc, use checksums for actions]
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
There's no need to manually create and cd into the build directory,
since meson knows how to handle that.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Make sure we are running in non-interactive mode and that we don't get
blocked on tzdata update.
Signed-off-by: Lucas De Marchi <lucas.de.marchi@gmail.com>
[Emil: port from kmod]
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Fix the following issues with the usb2 device capability descriptor
dump:
Currently, the BESL values claim to be in microseconds, but are actually
the raw, unshifted values from the descriptor. Use a lookup table for
the BESL values, based on Table X-X1 from "Errata for USB 2.0 ECN: Link
Power Management (LPM) - 7/2007" to display the actual microsecond
values.
Currently, the BESL values are displayed at the same indentation level
as bmAttributes, even though they are fields in bmAttributes. Adjust
indentation to align all fields of bmAttributes.
Change "BESL value" to "Baseline BESL value", to reflect the actual
names of the field.
Remove two unnecessary line breaks.
Before change:
USB 2.0 Extension Device Capability:
bLength 7
bDescriptorType 16
bDevCapabilityType 2
bmAttributes 0x0000f41e
BESL Link Power Management (LPM) Supported
BESL value 1024 us
Deep BESL value 61440 us
After change:
USB 2.0 Extension Device Capability:
bLength 7
bDescriptorType 16
bDevCapabilityType 2
bmAttributes 0x0000f41e
BESL Link Power Management (LPM) Supported
Baseline BESL value 400 us
Deep BESL value 10000 us
Signed-off-by: Torleiv Sundre <torleiv@huddly.com>
Commit 008e330759 ("lsusb.py: remove @DATADIR@ instance") removed the
possibility to define the location of the usb.ids file through @DATADIR@
and instead hardcodes the possible paths to the usb.ids file. However
the Debian path is missing, fix that.
Fixes: 008e330759 ("lsusb.py: remove @DATADIR@ instance")
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Some .c files changed, so update the file checksums for them in the
.spdx file to keep things up to date for a release.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Again, code scanners really don't like C string functions, and warn
about the potential for a buffer to possibly be too big for a snprintf()
call as the checking wasn't quite correct. Fix the check for a buffer
overflow up better and handle any potential issues that checking tools
could come up with.
No real functional change as the sysfs path should be just fine. And
really, this is all just userspace code...
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We "know" the sysfs attribute is not going to be bigger than the buffer,
but the thousands of "ooh, a strcpy() call is bad for you!" scanning
tools do not. So to shut them up, use strncpy() just to make things
quiet and for us to stop getting foolish reports.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a new file in the repo, and the checksums have changed in many
others, so update the .spdx file to reflect this.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
That's our defined style and that we've been using in the majority of
the codebase.
Disable clang-format on a couple of sections, where it gets a silly
moment and insists on reformatting everything.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Throw clang into the matrix, so we get a bit of extra coverage. As
evidenced by the disabled warning - it does catch extra issues.
Currently enabled only for 64bit builds, since fiddling with environment
variables and cross-files is more tricky than I can be asked atm x-D
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
The meson build system can create and change into the directory on its
own. Drop the manual steps.
Silly copy/paste left over from another project this file originates.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
I've introduced the container an year or so ago, to avoid the manual
pacman.conf fiddle. Although the last few months it wasn't quite working
as expected due to expired publishing tokens et al.
With that fixed we can use it \o/
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Add support for parsing the Class-specific VC Interrupt Endpoint
Descriptor as described in UVC 1.1 spec 3.8.2.2.
Sample output, including the preceding endpoint descriptor:
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x81 EP 1 IN
bmAttributes 3
Transfer Type Interrupt
Synch Type None
Usage Type Data
wMaxPacketSize 0x0010 1x 16 bytes
bInterval 8
bMaxBurst 0
VideoControl Endpoint Descriptor:
bLength 5
bDescriptorType 37
bDescriptorSubtype 3 (EP_INTERRUPT)
wMaxTransferSize 16
Signed-off-by: Torleiv Sundre <torleiv@huddly.com>
The lists were never being cleaned up when finished, so properly walk
them and free the allocated the memory that was allocated.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The list.h code is REALLY old, from a very old kernel version. Import
the CCAN version instead, and move the code to use it. This is in
anticipation of fixing up the memory leaks in this area.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add Miguel's copyright information as he was the original committer of
the file to Linux. Also remove the unneeded Documentation/ entry as
that makes no sense in this project.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It will help enforce the style defined in .clang-format, only for newly
introduced changes. Pre-existing style issues will not be flagged,
unless the lines are changed with given commit.
If a particular section needs to be ignored, it should be wrapped in
clang-format off/on section.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
We all have things larger than 80 columns these days. If this gets too
out of hand, we can reduce it later.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Most editors support the config format out of the box and with some
patches in the works, it's great to have something in-tree to avoid
(further) mixing the style.
The format is sufficient to handle basic indentation size/style and line
length. More elaborate style tracking can be done with clang-format...
incoming with a later patch.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
The former is clang-only, which has been clean for a while. While the
latter is gone with the hash tables removal as of earlier commits.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
This seems like a premature ahead-of-time optimisation, which actually
hurts both code/binary size while providing no performance benefit.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
This seems like a premature ahead-of-time optimisation, which actually
hurts both code/binary size while providing no performance benefit.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
This seems like a premature ahead-of-time optimisation, which actually
hurts both code/binary size while providing no performance benefit.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Now that the man pages moved around, and we flushed out the remaining
copyright/license file descriptions, update the .spdx file with all of
the latest information.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
