[00:11:00] user890104: yes, yes and yes [00:11:12] or technically not a reset but a soft disconnect/reconnect [00:11:23] maybe windows is too slow, try adding some delay... [00:13:02] TheSeven: unplugging my classic after launching umstest and not installing any debug interface drivers, then plugging it again, crashes windows, and it's reproducible [00:13:36] user890104: even with the debug interface disabled on the ipod side? (last parameter changed to false) [00:13:53] i haven't tried that [00:14:46] i'll try it later, when i finish what i'm currently working on, because... you know... :) [02:04:07] *** Joins: clustur (~logger@c-68-53-250-91.hsd1.tn.comcast.net) [02:04:08] *** Quits: clustur (~logger@c-68-53-250-91.hsd1.tn.comcast.net) (Remote host closed the connection) [02:57:13] *** Quits: Rurd2di (~Rurd2di@hotaru.1337.net.nz) (Read error: Operation timed out) [03:12:24] *** Joins: [Saint] (~saint@rockbox/user/saint) [03:16:14] *** Joins: [Saint_] (~saint@rockbox/user/saint) [03:16:50] *** Quits: [Saint] (~saint@rockbox/user/saint) (Disconnected by services) [03:16:59] *** [Saint_] is now known as [Saint] [03:19:56] *** Joins: Rurd2di (~Rurd2di@hotaru.1337.net.nz) [05:17:18] *** Quits: [Saint] (~saint@rockbox/user/saint) (Remote host closed the connection) [05:28:22] *** Joins: [Saint] (~saint@rockbox/user/saint) [05:42:56] *** Quits: [Saint] (~saint@rockbox/user/saint) (Remote host closed the connection) [05:46:29] *** Joins: [Saint] (~saint@rockbox/user/saint) [05:47:41] *** Quits: [Saint] (~saint@rockbox/user/saint) (Remote host closed the connection) [06:11:02] *** Quits: TheSeven (~quassel@rockbox/developer/TheSeven) (Disconnected by services) [06:11:08] *** Joins: [7] (~quassel@rockbox/developer/TheSeven) [08:04:09] *** Joins: clustur (~logger@c-68-53-250-91.hsd1.tn.comcast.net) [08:04:09] *** Quits: clustur (~logger@c-68-53-250-91.hsd1.tn.comcast.net) (Remote host closed the connection) [12:46:46] *** Joins: el3ctrik-away (~el3ctrik@2a00:1c10:5:201:216:3eff:fe98:c8f1) [12:46:54] *** el3ctrik-away is now known as el3ctrik [14:04:08] *** Joins: clustur (~logger@c-68-53-250-91.hsd1.tn.comcast.net) [14:04:08] *** Quits: clustur (~logger@c-68-53-250-91.hsd1.tn.comcast.net) (Remote host closed the connection) [19:51:45] *** el3ctrik is now known as el3ctrik-away [20:04:10] *** Joins: clustur (~logger@c-68-53-250-91.hsd1.tn.comcast.net) [20:04:10] *** Quits: clustur (~logger@c-68-53-250-91.hsd1.tn.comcast.net) (Remote host closed the connection) [21:23:29] *** el3ctrik-away is now known as el3ctrik [21:38:41] [7]: can you check if there's anything wrong in the usb descriptors that are restored after unplugging umstest? or do some packet sniffing? [21:39:16] <[7]> I see no reason why those could be damaged [21:39:58] memory corruption? [21:40:17] i'm just making wild guesses, since i can't properly debug the issue myself [21:40:19] <[7]> it's buried somewhere in the middle of the kernel memory [21:40:28] <[7]> if that would be corrupted, I doubt the ipod would still be running [21:41:23] debugging such OS crashes without a logic analyzer seems impossible to me [21:41:32] <[7]> I also didn't have *any* problems accessing it on linux [21:42:27] it also doesn't boot if there's an ipod in that state, connected to usb [21:42:39] also, it crashes if i hard-reset the ipod, while running umstest [21:43:03] that's without installing the monitor drivers [21:43:21] <[7]> sounds like it doesn't like suddenly disappearing mass storage devices [21:43:46] <[7]> hm, I might be able to try hooking up an ipod to a windows vm to check if that crashes as well [21:44:37] well, it's ok with the device disappearing, but when an emcore debugger appears later, it freaks out [21:44:50] yeah, that would be a good start [21:46:23] for the monitor driver, i'm using libusb-win32 1.2.6.0 (the latest release), which claims to be KMCS, but my OS doesn't think so [21:47:00] <[7]> hm, or it doesn't like a device changing its descriptors, and on top of that is too slow to recognize the soft disconnect in between [21:49:29] is there a way to shut down usb controller's power between the descriptor changes? [21:50:25] i'll get another laptop tomorrow, so i can test if putting some delay fixes the issue [21:50:28] <[7]> I'm telling it to detach, but probably not for long enough [21:53:14] i need to insert the delay in usbmanager_uninstall_custom() that's inside http://websvn.freemyipod.org/filedetails.php?repname=freemyipod&path=%2Femcore%2Ftrunk%2Fusb%2Fusbglue.c before if (usb_connected) usb_init(usb_data); , right? [21:58:05] <[7]> yes, or within that if statement [21:58:23] <[7]> (and ideally also on the install side, after usb_shutdown or usb_exit or whatever it's called) [23:15:24] *** Parts: jy2wong (~jy2wong@corn-syrup.csclub.uwaterloo.ca) ("WeeChat 0.4.1")