Hi Benson,
> On Sat, Mar 16, 2013 at 1:00 PM, Henrik Rydberg <rydberg@xxxxxxxxxxx> wrote:
> >> Userspace can initiate the firmware update procedure by copying cyapa.bin
> >> to /lib/firmware, and then writing anything to the device sysfs attribute:
> >> /sys/bus/i2c/devices/7-0067/update_fw
> >
> > Why do you need to trigger this from userland via sysfs?
> >
>
> It's a confluence of factors that drives the decision trigger from userland.
> - As a part of the cold-boot probe steps, the trackpad's bootloader
> will load an operational firmware from its flash memory. In the
> average case, the trackpad driver does not need to update the firmware
> because it's current. (the non-average cases are cases where on device
> firmware is corrupted or out of date)
> - The cypress firmware payloads we are working with here do not
> contain an indication of firmware version in the payload itself, so
> the driver can't tell at probe time whether what it would get if it
> request_firmware is newer, older, or the same version as what's
> already loaded on the trackpad.
> - On a 400kHz i2c bus, it takes 13 seconds to send the 31K payload to
> the device and to wait for it to jump into the operational mode
> firmware. This is too long to block probe every time.
> - Because it takes so long to do a firmware update, our user space
> would like to put up a warning message anyhow to tell the user that
> the touchpad is being updated, and not to disturb the system.
>
> >> + if (sysfs_create_group(&client->dev.kobj, &cyapa_sysfs_group))
> >> + dev_warn(dev, "error creating sysfs entries.\n");
> >> +
> >
> > Would it not be neat to invoke the firmware update automatically instead?
>
> The one situation where I could see an automatic firmware update being
> useful would be cases where the driver, during probe, discovered that
> the trackpad is in Bootloader mode because no valid firmware was found
> (a case of corruption of some kind due to the system powering down
> during a previous fw update, or example). In that case, the driver can
> request firmware and update it without user space having to trigger it
> later.
>
> Another driver in input does it the same way. atmel_mxt_ts.c, which
> our team is actively working on as well, also has an update_fw sysfs
> for a lot of the same reasons.
Thanks for the rationale, it looks compelling enough for me.
The cyapa_firmware() function has loops that looks like they could be
simplified. With that and the previous comments said,
Acked-by: Henrik Rydberg <rydberg@xxxxxxxxxxx>
Henrik
--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
