This is effectively a lot of dead code.
* The minimum requirements for this feature are EC 00.00.00.23 and Hub2 1.42.
* "A00" docks shipped with EC 01.00.00.00 and Hub2 1.47
It is far too easy to forget to set FWUPD_DEVICE_FLAG_NO_GUID_MATCHING for new
plugins, and without it it all works really well *until* a user has two devices
of the same type installed at the same time and then one 'disappears' for hard
to explain reasons. Typically we only need it for replug anyway!
Explicitly opt-in to this rarely-required behaviour, with the default to just
use the physical and logical IDs. Also document the update behavior for each
plugin to explain why the flag is being used.
This allows you to have two identical Unifying plugged in without one of them
being hidden from the user, at the same time allowing a HIDRAW<->USB transition
when going to and from bootloader and runtime modes.
This removes the workaround added in 99eb3f06b6.
Fixes https://github.com/fwupd/fwupd/issues/2915
This is typically a Linux sysfs path or USB platform ID and is used in a
different way to the physical ID. The physical ID is only set for some devices
after setup() and depends on the subsystem list, and this would not be defined
for devices that do not match a plugin.
This also fixes an regression where the FuDeviceList fails to match the new
FuUdevDevice device in fu_device_list_get_by_guids_removed() and instead
silently gets 'fixed up' only if FWUPD_DEVICE_FLAG_NO_GUID_MATCHING is not set.
This also allows us to move the various backends device caches to FuBackend as
we now have a suitable ID that is for just the backend to use.
If the user explicitly specifies the device-id then we do not do the front-end
filtering. If the device detaches without being able to attach then the hardware
could be left in the bootloader state.
See https://github.com/fwupd/fwupd/issues/2926#issuecomment-784234180
References to opt_variant and val_variant are sunk in a subsequent
variant creation and consumed by g_dbus_proxy_call_sync. They don't need
to be freed in this context.
Although glibc calls this when memory gets low automatically, we can invoke it
manually when we've loaded all plugins and coldplugged all devices.
This has no effect when using proper heap benchmarks like massif, but random
Reddit users just looking at the RSS column in an 'easy-to-use memory profiler'
will see a much lower numbers.
Where this feature might actually be useful is after a libxmlb metadata rebuild
where we dirtied a lot of heap just parsing XML.
The current approach of adding SBAT metadata after linking is creating
an image that is badly formed in 2 ways:
* The SBAT section's file offset and size are not a multiple of the
file alignment.
* The SBAT section has a virtual address of zero. EDK2 loads the header
here, and so it gets rejected.
This changes the approach to match shim, where an object file is
created with a .sbat section and then the linker takes care of placing
the section at a more appropriate virtual address.
See https://github.com/vathpela/gnu-efi/pull/14 for the section addition.
This is typically when the OEM is using the reference hardware design.
Prevent updates, as there might be a new bug introduced in the reference
firmware that only manifests on one OEM's product. It's up to the OEM to do the
testing and validation.
We need something to tie it back to a physical device model if it's using a
reference firmware and we want to update it.
Also split out the firmware parsing to an object so we can check the firmware
using firmware-parse and also fuzz it.
See also: https://github.com/fwupd/fwupd/issues/1665
This allows a device subclass to call the parent method after doing an initial
action, or even deliberately not call the *generic* parent method at all.
It also simplifies the plugins; you no longer have to remember what the plugin
is deriving from and accidentally clobber the wrong superclass method.
