On 22-08-09 23:00:29, Saravana Kannan wrote: > This patch series improves fw_devlink in the following ways: > > 1. It no longer cares about a fwnode having a "compatible" property. It > figures this our more dynamically. The only expectation is that > fwnode that are converted to devices actually get probed by a driver > for the dependencies to be enforced correctly. > > 2. Finer grained dependency tracking. fw_devlink will now create device > links from the consumer to the actual resource's device (if it has one, > Eg: gpio_device) instead of the parent supplier device. This improves > things like async suspend/resume ordering, potentially remove the need > for frameworks to create device links, more parallelized async probing, > and better sync_state() tracking. > > 3. Handle hardware/software quirks where a child firmware node gets > populated as a device before its parent firmware node AND actually > supplies a non-optional resource to the parent firmware node's > device. > > 4. Way more robust at cycle handling (see patch for the insane cases). > > 5. Stops depending on OF_POPULATED to figure out some corner cases. > > 6. Simplifies the work that needs to be done by the firmware specific > code. > > This took way too long to get done due to typo bugs I had in my rewrite or > corner cases I had to find and handle. But it's fairly well tested at this > point and I expect this to work properly. > > Abel & Doug, > > This should fix your cyclic dependency issues with your display. Can you > give it a shot please? Tested the specific case we discussed about here: https://lore.kernel.org/all/CAGETcx8F0wP+RA0KpjOJeZfc=DVG-MbM_=SkRHD4UhD2ReL7Kw@xxxxxxxxxxxxxx/raw Thanks for fixing this. Tested-by: Abel Vesa <abel.vesa@xxxxxxxxxx> > > Alexander, > > This should fix your issue where the power domain device not having a > compatible property. Can you give it a shot please? > > Tony, > > This should handle the odd case of the child being the supplier of the > parent. Can you please give this a shot? I want to make sure the cycle > detection code handles this properly and treats it like it's NOT a cycle. > > Geert, > > Can you test the renesas stuff I changed please? They should continue > working like before. Any other sanity test on other hardware would be > great too. > > Sudeep, > > I don't think there are any unfixed issues you had reported in my other > patches that this series might fix, but it'll be nice if you could give > this a sanity test. > > Guenter, > > I don't think this will fix the issue you reported in the amba patch, but > it's worth a shot because it improves a bunch of corner case handling. So > it might be better at handling whatever corner cases you might have in the > qemu platforms. > > Thanks, > Saravana > > Cc: Abel Vesa <abel.vesa@xxxxxxxxxx> > Cc: Alexander Stein <alexander.stein@xxxxxxxxxxxxxxx> > Cc: Tony Lindgren <tony@xxxxxxxxxxx> > Cc: Sudeep Holla <sudeep.holla@xxxxxxx> > Cc: Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> > Cc: John Stultz <jstultz@xxxxxxxxxx> > Cc: Doug Anderson <dianders@xxxxxxxxxxxx> > Cc: Guenter Roeck <linux@xxxxxxxxxxxx> > > Saravana Kannan (9): > driver core: fw_devlink: Don't purge child fwnode's consumer links > driver core: fw_devlink: Improve check for fwnode with no > device/driver > soc: renesas: Move away from using OF_POPULATED for fw_devlink > gpiolib: Clear the gpio_device's fwnode initialized flag before adding > driver core: fw_devlink: Add DL_FLAG_CYCLE support to device links > driver core: fw_devlink: Allow marking a fwnode link as being part of > a cycle > driver core: fw_devlink: Consolidate device link flag computation > driver core: fw_devlink: Make cycle detection more robust > of: property: Simplify of_link_to_phandle() > > drivers/base/core.c | 437 +++++++++++++++++++++----------- > drivers/gpio/gpiolib.c | 6 + > drivers/of/property.c | 84 +----- > drivers/soc/renesas/rcar-sysc.c | 2 +- > include/linux/device.h | 1 + > include/linux/fwnode.h | 12 +- > 6 files changed, 323 insertions(+), 219 deletions(-) > > -- > 2.37.1.559.g78731f0fdb-goog >
