Tools Snapshots of NiLuJe's hacks

[radamo]

Originally Posted by NiLuJe

@radamo: The dedicated thread (which is also linked to in the first post of this thread for everything listed there) should cover (ha!) the basics. Reply in there if you're encountering an issue that doesn't appear to be specific to the snapshots package (i.e., anything, really, now that you've got it installed ^^).

Will do!!! Thanks so much for the support!!!

[radamo]

Originally Posted by NiLuJe

@radamo: The dedicated thread (which is also linked to in the first post of this thread for everything listed there) should cover (ha!) the basics. Reply in there if you're encountering an issue that doesn't appear to be specific to the snapshots package (i.e., anything, really, now that you've got it installed ^^).

Thanks again for all of your help. Found the answer clearly listed in the Screensaver README.

[7hir7een]

I do have a bit of a strange issue with not being able to toggle usbnet off.

I can turn it on just fine with * Toggle USBNetwork * and it functions great.

But when I need to load books onto the device and press it again in an attempt to return to usbms mode, the KOA2 freezes. Nothing is selectable, pressing the power button won't send it to standby, etc.

IIRC, this was the case with the v0.21 koa2-nomax package as well.

I have to force-restart it to get anywhere with it (a painful process -- with the number of books I've loaded on the device, restarting can take 10 minutes )

How would I go about getting you the relevant logs you might need to look into this?

[knc1]

Check the thread on example "runme" scripts - I think I added one for the purpose of turning USB networking on/off on the KOA2. I have to do a lot of turning USBnetworking on/off while working on the KOA2 stuff.
(It is a simple thing, it just calls the existing USB networking script)

[NiLuJe]

@7hir7een: Yeah, this is somewhat of a known issue.

It's a bit tricky to investigate without the device, but it's potentially trivial to fix, if my hunch is right.

Which FW are you running, so that I take a look at the right rootfs to double-check this?

EDIT: Ah, if @knc1 got something running, then it's even better, I'll try to dig that one up, thanks!

@knc1: Does that apply to the PW3 on FW 5.9.x, too?

EDIT²: Can't seem to find anything USBNet related there?

EDIT³: @7hir7een: In the meantime, you can try to delete line 222 in usbnet/bin/usbnetwork (with a text editor that won't screwup *nix line endings), see if that helps?

[7hir7een]

@NiLuJe Unfortunately, deleting that line did not affect the problem. I'm running 5.9.5.2

I'd be perfectly happy with a runme script (anything to avoid force rebooting, honestly ) but I can't seem to find the script referenced by knc1. Still poking around search results in different posts, but no luck yet.

[NiLuJe]

@7hir7een: Until @knc1 digs up his script, which FW version are you running? I'll take a look at it tomorrow .

EDIT: Also, looking at logs might prove useful, provided nothing gets eaten by an unflushed cache... Next time it crashes, as soon as you can after the reboot, enter ;dumpMessages in the searchbar, and send me the system logs that'll dump in documents/ .

[NiLuJe]

@7hir7een: I slightly tweaked the USBNet -> USBMS process in the latest round of snapshots, this *may* have a positive effect on your issue...

[7hir7een]

No luck with the updated snapshot.

I have attached the logs you requested. I trimmed the results to the relevant timestamps to hopefully reduce the mess [see next post]

I know very little about the usbnetworking process, so I don't know if this is relevant or something that is obvious... but the second time I fetched the logs to send you, I noticed something about the crashing.

It's only triggered after an actual connection has been made. Like, if I toggle usbnetworking on in kual, verify its status, then toggle it off again -- no crash. I can do this any number of times without triggering a freeze. However, if I connect via ssh, exit the connection, and toggle usbnetworking off in kual it crashes every single time.

[NiLuJe]

@7hir7een: That's interesting... Does it mean that it only crashes when it's still plugged to the computer, or is that irrelevant? (meaning it still crashes even if you unplug after your USBNet session).

EDIT: As for the logs, those are indeed trimmed a little bit too much, I'm only seeing the screensavers & fonts hack in there .

[7hir7een]

My process for (attempting) to switch between usbnet and usbms is:
- "exit" in terminal window for remote connection
- unplug from computer
- toggle usb networking in kual

I've never attempted to switch it while still plugged in. I was under the impression that was a no-no haha


ETA: oops! well here are the full ones from the second attempt but they are a mess

ETA2: when I was browsing through those logs, I noticed sections where the timestamps went completely wonky. It that something... expected? (ex: it jumps from 180601:221502 to 180602:021750 and back to 180601:221752?) Because it seems to be occurring around the time everything crashes and burns . Or is that just evidence of the force restart?

[NiLuJe]

@7hir7een: It vaguely rung a bell, and, looking at the pre-FW 5.x version of USBNetwork, I indeed prevented toggling while plugged in because all kinds of weird-ass things might happen.

It's never been an issue on FW 5.x., though, but it was worth checking . I'd still recommend doing the toggle while unplugged because gremlins, but, still .

EDIT: Yeah, nothing in the full version either. It's quite possibly a side-effect of the crashes, though, since I'm seeing the mangled kernel logs replay during the reboot, but the system logs appear to even loop and/or go back in time at some points. It may also be lab126's dumping script going loopy, either in and of itself, either because of the crash, or an unfortunate log rotation.

EDIT²: On the off-chance that leads to better result, you can also try, in a terminal over SSH:

showlog | ag kh_msg

Or digging into the archived logs manually:

for file in /var/local/log/messages_*.gz ; do [ -f "${file}" ] && zcat "${file}" | ag "kh_msg" ; done

That should leave you with only the bits from my stuff.

[7hir7een]

Spoiler:

Welcome to Kindle!

root@192.168.15.244's password:
#################################################
# N O T I C E * N O T I C E * N O T I C E #
#################################################
Rootfs is mounted read-only. Invoke mntroot rw to
switch back to a writable rootfs.
#################################################
[root@kindle root]# for file in /var/local/log/messages_*.gz ; do [ -f "${file}"
] && zcat "${file}" | ag "kh_msg" ; done
180601:210942 system: I linkfonts:kh_msg::restoring fontconfig cache for /mnt/us/linkfonts/fonts . . .
180601:210951 system: I linkss:kh_msg::checking screensavers...
180601:210951 system: I linkss:kh_msg::end of screensavers parsing
180601:210951 system: I linkss:kh_msg::mounting custom screensavers
180601:222556 system: I linkfonts:kh_msg::restoring fontconfig cache for /mnt/us/linkfonts/fonts . . .
180601:222613 system: I linkss:kh_msg::checking screensavers...
180601:222613 system: I linkss:kh_msg::end of screensavers parsing
180601:222613 system: I linkss:kh_msg::mounting custom screensavers
180601:231313 system: W usbnet:kh_msg::will not toggle usbnet while plugged in, unplug your kindle
180601:231326 system: I usbnet:kh_msg::Toggle USBNetwork
180601:231326 system: I usbnet:kh_msg::switching from usbms to usbnet
180601:231326 system: I usbnet:kh_msg::setting volumd useUsbForNetwork 1 . . .
180601:231326 system: I usbnet:kh_msg::sending an usbUnconfigured event
180601:231326 system: I usbnet:kh_msg::sending an usbPlugout event
180601:231326 system: I usbnet:kh_msg::waiting for volumd . . .
180601:231328 system: I usbnet:kh_msg::reconfiguring usb0 interface
180601:231328 system: W usbnet:kh_msg::removing stale telnetd pidfile
180601:231328 system: I usbnet:kh_msg::starting telnetd
180601:231328 system: I usbnet:kh_msg::stopping sshd
180601:231328 system: E usbnet:kh_msg::failed to stop sshd
180601:231330 system: W usbnet:kh_msg::removing stale sshd pidfile
180601:231330 system: I usbnet:kh_msg::starting sshd
180601:231331 system: I usbnet:kh_msg::currently in usbnet mode, sshd should be up
180601:231331 system: I usbnet:kh_msg::* USBNetwork: enabled (usbnet, sshd up) *
[root@kindle root]#

It doesn't appear like that resulted in anything from around the times of the freeze-ups, other than the fontconfig and screensavers line showing up after the force-restarts

ETA: It seems like the logs are missing things, actually. Because I toggled usbnetworking on and off several times, status printing in between around 22:15, and none of that is displaying. The freeze-up (after disconnecting from ssh and trying to toggle off in kual) occurred at 22:17 and the reboot process got to the dim backlight with progress bar stage at 22:25, which is when the fontconfig/screensaver lines do show up. Strange that none of the stuff in between is showing

[NiLuJe]

@7hir7een: Ah, that's at least a bit less weird .

And it confirms the fact that I left the toggle while plugged in protection in, it just moved to the KUAL extension, which is why I'd forgotten about it ^^. So, that's definitely out as a possible source of the issue .


But to recap, yeah, bits of logging probably went poof because of the crash, that's definitely possible. Annoying, but not that unexpected .

(NOTE: I took the liberty of removing your logs from the post, because I'm too lazy to potentially sanitize them for public consumption. Better safe than sorry, these things tend to be very verbose ).

[7hir7een]

FWIW: I don't have my old Voyage updated all the way (I've stopped using it since getting the KOA2 fully set up) -- but it is on 5.9.5.1. And it does not have this issue.

So at the very least, it is not a /universal/ issue with the 5.9.5.x firmware.