Tools Snapshots of NiLuJe's hacks

[NiLuJe]

@7hir7een: Good to know, at least, and it confirms the fact that I've only heard of this kind of thing on the KOA2 . Also, that stuff survives FW 5.9.5.2, which I'm not sure anyone had ever quite said yet .

I have a few crappy ideas left to try, I'll check them tomorrow, to at least confirm that they don't break anything on my PW2...

[NiLuJe]

@7hir7een: Tried some stuff in the latest round of snapshots, and after that, I have one final thing to try before being completely out of ideas :'(.

[7hir7een]

!!!!!

So I was about to report that it didn't work. But I wanted to try everything I could first. And I had poked around on your source and saw your note

This means check_is_plugged_in() would have to be nerfed in the KUAL extension on the KOA2 .

So since you said the protection regarding switching modes while plugged in was moved to KUAL (implying it was not in the usbnet plugin itself), I tried one more thing...

I did the usual turn usbnet on, remote connect to ssh in terminal, throw out a couple commands, exit, and turn usbnet off process -- but this time I used the ;un command in search to maybe bypass that protection and switch modes while plugged in??? And that??? Worked???

So in summary
- yes crash: using kual toggle off as intended (required usb to be removed)
- no crash: using ;un to toggle off (left usb plugged into mac)
- yes crash: using ;un to toggle off while usb unplugged from device

I don't know what that means but maybe you do?

[NiLuJe]

@7hir7een: Hmm, mmmmaybe? .

I was hoping for something of the sort, but it's a bit surprising that it works without a few other code changes...

Which exact revision of the hack is that?

[7hir7een]

2018-06-02 20:29 NiLuJe

* [r15022] src/usbnet/bin/usbnetwork:

Kindle Touch Hacks:
* USBNetwork:
* Jot that down, because goldfish brain.

this is the last thing in your changelog in the version i downloaded

Attaching my logs (should theoretically show a successful toggle right before creation) if you need them, too.

[NiLuJe]

@7hir7een: Yeah, the logs were indeed useful, if only to see what happens on a successful switch .

So, let's try r15029!

Hooopefully it works in every situation, AND it doesn't crash...

A few things to look out for:

If it doesn't crash, it *may* not actually switch to USBMS until the next time the device is unplugged. Wait for ~10s, there *should* be a message asking you to do just that.

If stuff switches properly immediately, it'll instead proceed to shut down telnet & sshd as usual without that extra 6s delay.

If you're up for it, testing both while plugged in and not will be helpful (KUAL will no longer prevent you to do that).

Thanks again .

[7hir7een]

Sorry it takes so long to get back to you, that 10 minute reboot is murder.

[Try 1] Used KUAL to disconnect while plugged in (I figured that was the safer bet). Saw the messages you described. Status printout indicated usbnet was still up. Unplugged, replugged, and it froze (can't recall if that was immediate, or when I pressed the status button in KUAL -- will pay more attention on next attempt)

[Try 2] Used ;un to disconnect while plugged in. Before, this resulted in a near-immediate transition to usbms. Now, received unplug device message. Unplugged. Replugged. The freeze is immediate. Kindle force-restarts itself.

So whatever was magically allowing it to work is no longer doing so in revision 15029

[Try 3] Using KUAL to disconnect while not plugged in is an immediate crash and burn as before

[NiLuJe]

@7hir7een: Huh. That's... not good, I really thought that'd do the trick...

Well, the good news is that stomped on approximately all my current ideas, so, there's that .

A few non-crash inducing questions: How quick were those unplug/replug (in particular, how long did you wait between the two, and did you notice at which point it froze: after the unplug or the replug?).

And while USBNet is still running, does unplugging/replugging (in itself, without toggling the hack) goes 'splodey, too?

EDIT: I guess that, for now, the best I can do is prevent the switch from happening if the device is unplugged, and restore the same codepath as in r15022 when plugged and hope for the best that the timings are not too finicky to still cause a crash in some circumstances...

[7hir7een]

Unplugging and replugging while usbnet is running has never caused a crash for me in and of itself. I did that several times while working on my jars for margin adjustments and had no issues.

To answer your other question:
- waited about 10 seconds after I received the "try to unplug your device" message
- status message at this point reads * USBNetwork: enabled (usbnet, ssh down?) *
- unplugged, waited ~5 seconds
- tapped screen, no response... so, it appears it's on the unplug.
- kindle force restarts itself a few moments after

(trying with it unplugged to begin with, it's immediate. the text printout at the bottom garbles & it freezes -- which has been the case for all versions of usbnet I've ran on the device)

[NiLuJe]

@7hir7een: Okay, thanks. At least the timing makes sense . That the watchdog catches the crash and/or deadlock and reboots the device on its own is also somewhat of a good news, even if it doesn't really help me .

And the rest is good news, it means there's hope for my crappy workaround ^^.

[NiLuJe]

@7hir7een: r15031 ready!

It should *hopefully* behave like r15022 when plugged in (even if I'm afraid that there might be timing issues involved...), and, otherwise, simply abort and warn to avoid the crash.

[7hir7een]

Good news and bad news. Tested the functionality of being able to successfully switch usbnet off while plugged in, and that part is behaving as in r15022 (which is to say that usbms works after the toggle off and usbnet still works after toggling on again).

Unfortunately, when testing to see if your abort failsafe worked when unplugged, it crapped out. This was on the same... not sure what to call it... same boot(?) as when I'd switched it on and off and on already -- so I'll give it one more try on a fresh go.

But it's the same behavior as most of the rest of it: hit the toggle in kual and the readout at the bottom immediately garbles and the whole thing freezes and eventually reboots itself

[NiLuJe]

@7hir7een: Yeah, I was kind of afraid that the second time round would go kablooey, but if that was a test of the "unplugged" safety and it blew up anyway, it might be that I just stupidly messed up the check (or that the method I'm using is unreliable while in USBNet mode on the KOA2, which wouldn't be something unheard of, unfortunately...).

[7hir7een]

It is in fact the same on a fresh boot. So the unplugged toggle-off failsafe part is failing

ETA: to clarify, I'm 90% sure it's letting me toggle back and forth between usbms and usbnet multiple times while plugged in. The issue that crashed it was testing whether it'd successfully yell at me for trying to do so unplugged.

Update: Yep. usbms-usbnet-usbms-usbnet-usbms-usbnet-usbms all worked fine (tested with file transfers) so long as I leave it plugged in

[NiLuJe]

@7hir7een: Yeah, then let's assume I just borked the safety check .

Sooo, over SSH, could you tell me the result of these commands:

cat /sys/class/udc/ci_hdrc.0/subsystem/ci_hdrc.0/connected
cat /sys/class/power_supply/max77796-charger/charging
cat /sys/class/power_supply/max77796-charger/present
cat /sys/class/power_supply/max77796-charger/online

And something completely unrelated, but that's bothering me:

rdev
cat /proc/mounts