| 00:03.12 | *** mode/#edev [+o prpplague] by ChanServ |
| 01:58.42 | *** join/#edev dijenerate (~dijenerat@173.225.251.175) |
| 02:00.16 | *** join/#edev kergoth (~kergoth@ip24-251-173-232.ph.ph.cox.net) |
| 03:21.42 | *** join/#edev ZeZu (~null@c-98-227-57-88.hsd1.in.comcast.net) |
| 03:43.44 | *** join/#edev kergoth_ (~kergoth@ip24-251-173-232.ph.ph.cox.net) |
| 04:51.10 | *** join/#edev mbuf (~user@nat/redhat/x-pvoiducvuquwgobx) |
| 05:07.18 | *** join/#edev avinashhm (~avinash-h@192.94.92.14) |
| 06:17.45 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 06:40.34 | *** join/#edev kirabo (~kirabo@LCaen-156-54-47-233.w80-11.abo.wanadoo.fr) |
| 06:50.36 | *** join/#edev dijenerate (~dijenerat@204.212.240.159) |
| 07:02.40 | *** join/#edev kirabo (~kirabo@LCaen-156-54-47-233.w80-11.abo.wanadoo.fr) |
| 08:23.17 | *** join/#edev avinashhm (~avinash-h@192.91.66.189) |
| 08:55.47 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 10:36.55 | *** join/#edev avinashhm (~avinash-h@192.94.92.14) |
| 11:35.01 | *** join/#edev florian (~fuchs@node-vlb.camp.ccc.de) |
| 11:35.01 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 12:56.01 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 13:41.55 | *** join/#edev orbarron (~orb@nat/ti/x-reypxkibrjbedixj) |
| 13:54.58 | *** join/#edev kergoth_ (~kergoth@ip24-251-173-232.ph.ph.cox.net) |
| 14:23.34 | *** join/#edev CosmicPenguin (~nobody@soa.codeaurora.org) |
| 14:23.34 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 14:28.27 | *** join/#edev sndcrb (~sndcrb@dsl-hkibrasgw2-fef8de00-126.dhcp.inet.fi) |
| 15:04.12 | *** join/#edev dijenerate (~dijenerat@204.212.240.159) |
| 15:21.23 | *** join/#edev prpplague (~danders@nat/ti/x-hvylmgahstdyznlu) |
| 15:25.27 | *** mode/#edev [+o prpplague] by ChanServ |
| 15:51.38 | *** join/#edev jayabharath (~jayabhara@nat/ti/x-nppbiymrohurbwpv) |
| 16:16.11 | *** join/#edev ccole (~cole@coledd.com) |
| 16:32.24 | *** join/#edev dijenerate (~dijenerat@204.212.240.159) |
| 16:33.48 | *** join/#edev gustavoz (~gustavoz@host4.201-252-50.telecom.net.ar) |
| 16:42.31 | *** join/#edev flo_lap (~fuchs@clarke.visitor.camp.ccc.de) |
| 16:42.31 | *** join/#edev flo_lap (~fuchs@Maemo/community/contributor/florian) |
| 17:02.28 | *** join/#edev flo_lap (~fuchs@Maemo/community/contributor/florian) |
| 17:13.28 | *** join/#edev flo_lap (~fuchs@Maemo/community/contributor/florian) |
| 17:20.09 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 17:29.47 | landley | This is crazy. |
| 17:29.59 | *** join/#edev MonMotha (~monmotha@2002:6c47:d23b:1:224:7eff:fedc:893e) |
| 17:30.08 | landley | Yesterday we had something fixed. Today, the same hardware with the same software is showing the bug again. |
| 17:30.54 | landley | Yesterday we plugged and unplugged a USB 1.1 device with a 15 foot cable 38 times in a row, and it worked fine each time. Today, it fails on the first plug. |
| 17:30.58 | landley | Every time. |
| 17:31.13 | landley | Nothing changed except the symptom. There's a variable here and I'm not seeing it. |
| 17:37.01 | SpeedEvil | RFI>? |
| 17:37.19 | SpeedEvil | ESD |
| 17:40.22 | landley | It's the ehci-fsl.c controller. Freescale ehci for powerpc. |
| 17:40.51 | CosmicPenguin | Wasn't there a solar flare the other day? |
| 17:41.23 | prpplague | CosmicPenguin: no, kergoth just had some tex-mex |
| 17:41.31 | landley | Works fine with 2.0 devices, works fine with MOST 1.1 devices, but if you use a long cable and it's shouting down the wire it can listen to its own signals and go into a J/K loop that doesn't even stop when you unplug the device. |
| 17:41.43 | landley | Sounds like improper termination, right? |
| 17:41.54 | landley | I found some unnecessary initialization the local copy of the driver was doing that upstream wasn't. |
| 17:42.07 | landley | It's 2.6.33.3 plus local changes. |
| 17:42.28 | landley | I did a log from 2.6.33 to master to see if anybody else had hit this: they hadn't. |
| 17:42.37 | landley | A dozen changes went in but none related to this. |
| 17:42.50 | landley | So I removed the unnecessary local initialization... and it worked. |
| 17:42.55 | landley | And now it's not working. |
| 17:43.15 | landley | There was a STATE CHANGE when I changed kernels, but the kernel didn't change BACK and it went back to the old behavior. |
| 17:43.30 | landley | I do not understand what's going on. |
| 17:43.49 | landley | Parker's setting up a scope so we can make sure it's back in the J/K loop, but... DUDE. |
| 17:43.53 | landley | I mean... DUDE. |
| 17:45.15 | landley | My _theory_ was that the port write plus 500ms write was freezing the state machine in the PHY so it wasn't adjusting the termination between 2.0 and 1.1 modes. Which it otherwise is supposed to automatically adjust, in the hardware. |
| 17:45.54 | landley | The comments aid it was resetting the hardware, which obviously isn't the case because it was right at the start of the driver probe for a static driver that comes up from powered up hardware. If the state is different after we do this, it's NOT the same state we power on into, thus not a reset. |
| 17:46.07 | landley | Either this reset was a NOP or it was broken. Either way, it goes. |
| 17:46.29 | landley | Especially since ~2 years of upstream releases didn't have it, and nobody else was seeing this problem out in the field. |
| 17:47.02 | landley | I am _stumped_. Why did it work and then stop working? What's different? |
| 18:14.57 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 18:25.51 | *** join/#edev landley (~landley@140.242.26.81) |
| 18:25.51 | *** mode/#edev [+o landley] by ChanServ |
| 18:52.01 | *** join/#edev kergoth__ (~kergoth@ip24-251-173-232.ph.ph.cox.net) |
| 19:00.34 | *** join/#edev dijenerate (~dijenerat@66.54.118.89) |
| 19:07.45 | *** join/#edev florian (~fuchs@81.162.148.243) |
| 19:07.45 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 20:11.46 | *** join/#edev florian (~fuchs@clarke.visitor.camp.ccc.de) |
| 20:11.46 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 20:22.33 | *** join/#edev jnd (~jenda@unaffiliated/jnd) |
| 20:35.33 | *** join/#edev dijenerate (~dijenerat@66.54.118.89) |
| 20:59.24 | *** join/#edev ZeZu (null@c-98-227-57-88.hsd1.in.comcast.net) |