KindleLazy - remote control page turns + brightness

[knc1]

Originally Posted by jscris

Sorry. Maybe someone who knows better how to read the log files should do this. I can't be the only one using KL.

But you are the only one reporting that it does not hang the Kindle after it goes through a "suspend to RAM" ... "resume" cycle.

So the first thing to do in finding out why, is to confirm it is going through the full "suspend" ... "resume" cycle.

The search bar command ;dm should dump the entire log as a set of documents in USB storage.

Just archive them and attach to your next post.
The ";dm" command scrubs them of any personal information.
(There are some dump commands that don't, but this one does.)

[jscris]

So I'm probably wrong.

Attached.

[knc1]

Originally Posted by jscris

So I'm probably wrong.

Attached.

Both yes and no.
Also, this may be firmware version specific.
Post (or re-post) your firmware version and if it is still the factory or the customer flavor.

There are two things at work here - -

You have the device running on the secondary battery (the one in the cover, the cover is plugged into the Kindle).
Other Kindle models do not have a secondary battery.

Oops. This is a KV not a KOA.

The suspend code IRQ handling in the adapter module is not correct for your model Kindle, and perhaps not correct for others:

161017:170527 [132000.846573] PM: suspend of devices complete after 90.659 msecs
161017:170527 [132000.846984] PM: late suspend of devices complete after 0.393 msecs
161017:170527 [132000.847312] triggered: wake_irq_isr[0-3]: 0x0, 0x0, 0x200, 0x0
161017:170527 [132000.847323] gpio4 triggered 0x100000 (0xfed58f71/0x104080)
161017:170527 [132000.847989] PM: early resume of devices complete after 0.216 msecs

Note:
left column: YYMMDD:User land time
next column, bracketed: Kernel time with microsecond resolution.

132000.847312 - 132000.846984 is not much of a 'sleep', in fact, it isn't even a short nap on a computer's timescale.

About the only way that can happen, is that:

161017:170527 [132000.847312] triggered: wake_irq_isr[0-3]: 0x0, 0x0, 0x200, 0x0
161017:170527 [132000.847323] gpio4 triggered 0x100000 (0xfed58f71/0x104080)

was a pre-existing condition (not properly masked when going to sleep).

= = = = =

Yes, I can read the log but I can't do anything about the code.
That will have to be up to the author of this add-in.

[jscris]

Originally Posted by knc1

Both yes and no.
Also, this may be firmware version specific.
Post (or re-post) your firmware version and if it is still the factory or the customer flavor.

There are two things at work here - -

You have the device running on the secondary battery (the one in the cover, the cover is plugged into the Kindle).
Other Kindle models do not have a secondary battery.

Oops. This is a KV not a KOA.

The suspend code IRQ handling in the adapter module is not correct for your model Kindle, and perhaps not correct for others:

161017:170527 [132000.846573] PM: suspend of devices complete after 90.659 msecs
161017:170527 [132000.846984] PM: late suspend of devices complete after 0.393 msecs
161017:170527 [132000.847312] triggered: wake_irq_isr[0-3]: 0x0, 0x0, 0x200, 0x0
161017:170527 [132000.847323] gpio4 triggered 0x100000 (0xfed58f71/0x104080)
161017:170527 [132000.847989] PM: early resume of devices complete after 0.216 msecs

Note:
left column: YYMMDD:User land time
next column, bracketed: Kernel time with microsecond resolution.

132000.847312 - 132000.846984 is not much of a 'sleep', in fact, it isn't even a short nap on a computer's timescale.

About the only way that can happen, is that:

161017:170527 [132000.847312] triggered: wake_irq_isr[0-3]: 0x0, 0x0, 0x200, 0x0
161017:170527 [132000.847323] gpio4 triggered 0x100000 (0xfed58f71/0x104080)

was a pre-existing condition (not properly masked when going to sleep).

= = = = =

Yes, I can read the log but I can't do anything about the code.
That will have to be up to the author of this add-in.

Thanks. Maybe that explains the battery usage and sounds like a lot of work for zxczxc. FW 5.8.1 Updated after your wonderful downgrade and BD's jailbreak. That was a really happy day!

[knc1]

Originally Posted by jscris

Thanks. Maybe that explains the battery usage and sounds like a lot of work for zxczxc. FW 5.8.1 Updated after your wonderful downgrade and BD's jailbreak. That was a really happy day!

I don't think he supports 5.8.2 yet - so fixing this could be combined with that work.
It shouldn't be too hard, all that needs to be done is look up that address in a running KV, 5.8.1, /proc symbol table, look up the symbol in the source code, begin reading.

[jscris]

Originally Posted by knc1

I don't think he supports 5.8.2 yet - so fixing this could be combined with that work.
It shouldn't be too hard, all that needs to be done is look up that address in a running KV, 5.8.1, /proc symbol table, look up the symbol in the source code, begin reading.

[zxczxc]

Ok I haven't touched KindleLazy since I last updated it (0.7 I think? -oops, 0.9!)

I will add some form of specifying the input device in the config file.
For example, you will will be able to enter:
/dev/input/event3
if that is where the kindle system designates your usb device.
or, for my kindle+my controller I would enter:
/dev/input/event2
/dev/input/event3
in the config file since for some reason my device spans both these input devices

Maybe also have a button for the KUAL KindleLazy menu that outputs your current input devices to a text file.
So the text file created may look something like this:
/dev/input/event0
/dev/input/event1
/dev/input/event2
/dev/input/event3
/dev/input/event4

And then by a little thinking/guessing/trial and error, you would add the lines that your device is mapped to, in the config file.


I can explain more if needed.


As I have said before, the way to automatically check what 'input' on the kindle to read is confusing and difficult for me - someone can push some code to my on github if they want to help but until then I guess we will do it manually.
Sorry it's not more plug and play!


-
About 5.8.2, I have no idea. I would like to think it works...
But yes I only tested on 5.8.1

[knc1]

And now you have 5.8.5 to deal with.

Its "that time of the year" for Amazon.

[zxczxc]

The change log seems beneficial - "Books that were manually loaded to your device will now appear in both the All and Downloaded library views."

That took them long enough to add back in at least, I'm not sure when it ever left, or how many times it left and was reintroduced!

I will update.

[knc1]

The 5.8.5 firmware is using kernel 3.0.35

[zxczxc]

ok, the same as 5.8.1
uname -r -> "3.0.35-lab126"

so there shouldn't be any problem with the modules for usbhid, mouse etc, at least for pw3.

--
strange, I just let my device update, ota, and it's 5.8.2.1, not 5.8.2...
i'll leave it overnight and see if it updates again.

[knc1]

Originally Posted by zxczxc

ok, the same as 5.8.1
uname -r -> "3.0.35-lab126"

so there shouldn't be any problem with the modules for usbhid, mouse etc, at least for pw3.

--
strange, I just let my device update, ota, and it's 5.8.2.1, not 5.8.2...
i'll leave it overnight and see if it updates again.

You do not have to wait.
In the JB post, in Step #7, I updated the Amazon download links to include 5.8.5 yesterday.

Amazon changed the URL path on s3.amazonaws.com to the downloads, so you can't just edit in the new version number in the old paths.

[zxczxc]

I made a mistake, I meant 5.8.5 instead of 5.8.2.1.

Anyway, what I meant was, I was on 5.8.1, I removed the update blocker file trick and let my kindle update ota, but it brought me to 5.8.2.1. Not 5.8.5.

Why would my kindle download not the newest update?

I understand that I can update manually, but I tried ota instead.

[knc1]

Originally Posted by zxczxc

I made a mistake, I meant 5.8.5 instead of 5.8.2.1.

Anyway, what I meant was, I was on 5.8.1, I removed the update blocker file trick and let my kindle update ota, but it brought me to 5.8.2.1. Not 5.8.5.

Why would my kindle download not the newest update?

I understand that I can update manually, but I tried ota instead.

OTA update jobs are queued on the Kindle in its 'to do' list and it only queues one OTA update at a time.
(You can find that in the messages log. Searchbar ;dm
It is also most likely why the old firmware downloads remain on the download server forever, there is no telling the oldest job queued of all customer Kindles.)

Amazon moved the download locations on s3.amazonaws.com at about the same time as the 5.8.5 release.

Q.E.D: Your Kindle had the old pathname queued, not the new pathname. (or some other screw-up by Amazon/Lab126 in the relocation).

tl;dr: It is working as intended.
Just install it manually, the only difference is the waiting time.

[knc1]

Originally Posted by zxczxc

ok, the same as 5.8.1
uname -r -> "3.0.35-lab126"

so there shouldn't be any problem with the modules for usbhid, mouse etc, at least for pw3.

--
strange, I just let my device update, ota, and it's 5.8.2.1, not 5.8.2...
i'll leave it overnight and see if it updates again.

I was wrong, must have been looking at something else.

KT-3, 5.8.5 -> 3.10.53-lab126

Getting the auto-loader to work (a requirement of hot-plugging) -

Add modules in the appropriate sub-tree of:
/lib/modules/<version>/*
Amazon does not appear to be using /lib/firmware - they have device firmware scattered in the system tree.

For any modules you add, update:
/lib/modules/3.10.53-lab126/modules.builtin (you will want to check this list before building anything - it might already be there).
/lib/modules/3.10.53-lab126/modules.order

And then update:
/opt/factory/rootfs_md5_list
(The system manifest.)