Hi all, Thanks for sending in data so fast. It is helpful. At least it pinpoints to kernel side issue. So the evtest output is easier to talk to.
I see in all the samples sent in (both xorg.0.log and evtest), that we sometimes get a packet with X/Y coordinates at [0,0] at times that pressure is non-zero. Event: time 1304228221.687840, -------------- Report Sync ------------ Event: time 1304228221.711826, type 3 (Absolute), code 58 (?), value 64 Event: time 1304228221.711832, type 3 (Absolute), code 53 (?), value 0 Event: time 1304228221.711832, type 3 (Absolute), code 54 (?), value 0 Event: time 1304228221.711836, type 3 (Absolute), code 0 (X), value 0 Event: time 1304228221.711836, type 3 (Absolute), code 1 (Y), value 0 Event: time 1304228221.711837, type 3 (Absolute), code 24 (Pressure), value 64 That will cause your jumps big time. Those "(?)" are multi-touch versions of X/Y/Pressure. I guess your repo has older evtest that doesn't know those names yet; but thats ok. Also, I notice in everyone's evtest and Xorg.0.log files, that the pressure seems to be reporting oddly in general. I see what look like magic values of 24 and 88 being sent alot. Probably, Wacom didn't want to increase the packet length on these newer Bamboo's and needed to encode the 4 (or more?) fingers by sending multiple packets with something in pressure to tell which finger report it is. We will need someone that doesn't mind hacking on kernel drivers to figure this out. I'm not sure I can do much more without access to hardware. I know Ping is not available for a while. Perhaps Jason can find some info on any possible packet format changes? Barring that, we'll just need to reverse engineer the packet data ourselves. Chris 2011/5/1 Catalin Balan <laserbeam...@gmail.com>: > Sure: > > xsetwacom set 18 TabletDebugLevel 12 > xsetwacom set 18 ToolDebugLevel 12 > > 18 was touch of course, but I am not sure if those 2 were the debug > levels you asked for. > I attached the Xorg.0.log file. The two commands there are found in > the log file at these entries: > > [ 1323.247] (II) Wacom BambooPT 2FG 4x5 Finger touch (10:wcmSetProperty): > [ 1323.247] (II) Wacom BambooPT 2FG 4x5 Finger touch (10:wcmSetProperty): > > I did about 5-7 jumps + some short drags while logging, hope it's not too > much. > > > > evtest: > > I: Bus=0003 Vendor=056a Product=00d6 Version=0111 > N: Name="Wacom BambooPT 2FG 4x5 Finger" > P: Phys= > S: > Sysfs=/devices/pci0000:00/0000:00:1d.0/usb2/2-1/2-1.2/2-1.2:1.1/input/input14 > U: Uniq= > H: Handlers=mouse2 event13 > B: PROP=0 > B: EV=b > B: KEY=2420 630000 0 0 0 0 > B: ABS=660800001000003 > > I hope this is right so I used /dev/input/event13 there. Again, 5-7 > random presses and short flicks. > > It kept showing me some Absolute values in that log, and I tried > setting touch in Relative mode, but xsetwacom kept showing me "Failed > to open Display" whatever I told it to do. I couldn't really see what > was wrong there while in the TTY1. > > I attached 2 files for evtest in this email, I believe the first one > is done before using the init script, the second one after (as far as > I know, the init script doesn't load automatically). PS, the init > script seemed to work only using X, it gave me "Failed to open > Display" several times in TTY1. > > I am not sure here, but is that "Failed to open Display" an error > because xsetwacom tries to modify X, and X is not loaded in TTY1? > > On Sun, May 1, 2011 at 02:13, Chris Bagwell <ch...@cnpbagwell.com> wrote: >> Can someone that has one of these newer Bamboo's log some data and >> send it to the list or directly to me if its to big? I'll try to >> understand whats going wrong. Sounds like it must be a kernel issue >> so far. >> >> What I would like is for someone to set both "Wacom Debug Levels" to >> 12, do the smallest amount of work that produces then cursor jump and >> then email the Xorg.0.log file. It will give an idea of what kernel >> is sending. >> >> Also, it would be even more helpful if you can install the "evtest" >> app and run it on your touch device and redirect the output to a file. >> Reproduce the issue and send this log file in as well. This one is a >> little tricky because you'll be on a console window and won't see the >> cursor jump. Also, I think you have to hit control-c to quite and >> will lose a little bit of data in log file... So do a little extra >> movement to make sure you don't loss the important parts to send. >> >> http://sourceforge.net/apps/mediawiki/linuxwacom/index.php?title=Analysing_kernel_events >> >> Chris >> >> On Sat, Apr 30, 2011 at 1:50 PM, Catalin Balan <laserbeam...@gmail.com> >> wrote: >>> Yup, right you were right here as well... >>> >>> lsusb shows this: >>> ID 056a:00d6 Wacom Co., Ltd >>> >>> >>> On Sat, Apr 30, 2011 at 20:05, roku <r...@arcor.de> wrote: >>>> >>>> Favux, >>>> >>>> Yes, my Bamboo P&T shows 0xd8 as you have correctly supposed ... >>>> >>>> Roland >>>> >>>> >>>> Favux ... wrote: >>>>> >>>>> Sorry, didn't reply to all. >>>>> >>>>> Thanks Guenther, that helps. I went back over the BambooPT HOW TO >>>>> thread, about the last 18 pages, starting around 3-4-11 which is when >>>>> the first complaint about the pointer jumping appeared. Of the 3 to 4 >>>>> folks who complained two have the 0x0d6 and one the 0x0d8. Fennec had >>>>> a 0x0d6 but once he got it working (with help from you Alexander?) he >>>>> didn't mention a jumpy pointer or random left clicks with touch. >>>>> >>>>> So that brings us up to 3 0xd6's and 2 0xd8's (if Roland has one) and >>>>> maybe 1 unknown complaining about pointer jumpy. All new models. >>>>> >>>>> Favux >>>>> >>>>> On Sat, Apr 30, 2011 at 9:40 AM, Favux ... <favux...@gmail.com> wrote: >>>>>> Hi Alexander, >>>>>> >>>>>> OK, we may have a pattern if we expand it to include the 5 new models. >>>>>> Catalin has the 00d6 like you have and Roland has the 00d8 I think, >>>>>> which he can check with 'lsusb'. >>>>>> >>>>>> The thing is you'd expect the problem to be with the two Special >>>>>> Editions (00dA & 000dB) because they are the ones with 4 finger touch >>>>>> and the code only supports 2 finger touch. So to speculate, is the >>>>>> hardware on the 00d6,00d7, and 00d8 also capable of 4 FGT and has been >>>>>> "crippled" by Wacom? If we find Special Editions reporting the >>>>>> pointer jump that might tend to lend support to the speculation. >>>>>> >>>>>> Which would lead one to suspect the 4FGT capable devices do have >>>>>> protocol differences v.s. the 5 original 2FGT tablets that need to be >>>>>> addressed in the code. Kernel side or X side or both? >>>>>> >>>>>> Chris or Ping, where are you on enabling 4FGT MT in the kernel? I >>>>>> recall some preparatory work at least. >>>>>> >>>>>> Favux >>>>>> >>>>> >>>>> ------------------------------------------------------------------------------ >>>>> WhatsUp Gold - Download Free Network Management Software >>>>> The most intuitive, comprehensive, and cost-effective network >>>>> management toolset available today. Delivers lowest initial >>>>> acquisition cost and overall TCO of any competing solution. >>>>> http://p.sf.net/sfu/whatsupgold-sd >>>>> _______________________________________________ >>>>> Linuxwacom-discuss mailing list >>>>> linuxwacom-disc...@lists.sourceforge.net >>>>> https://lists.sourceforge.net/lists/listinfo/linuxwacom-discuss >>>>> >>>>> >>>> >>>> -- >>>> View this message in context: >>>> http://old.nabble.com/Bamboo-P-T-CTH-661---Cursor-jumps-using-finger-touch-function-tp31503490p31513173.html >>>> Sent from the linuxwacom-discuss mailing list archive at Nabble.com. >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> WhatsUp Gold - Download Free Network Management Software >>>> The most intuitive, comprehensive, and cost-effective network >>>> management toolset available today. Delivers lowest initial >>>> acquisition cost and overall TCO of any competing solution. >>>> http://p.sf.net/sfu/whatsupgold-sd >>>> _______________________________________________ >>>> Linuxwacom-discuss mailing list >>>> linuxwacom-disc...@lists.sourceforge.net >>>> https://lists.sourceforge.net/lists/listinfo/linuxwacom-discuss >>>> >>> >>> >>> >>> -- >>> Cătălin Bălan >>> http://thescrambledbit.wordpress.com >>> I do know everything, just not all at once. It's a virtual memory problem. >>> >>> ------------------------------------------------------------------------------ >>> WhatsUp Gold - Download Free Network Management Software >>> The most intuitive, comprehensive, and cost-effective network >>> management toolset available today. Delivers lowest initial >>> acquisition cost and overall TCO of any competing solution. >>> http://p.sf.net/sfu/whatsupgold-sd >>> _______________________________________________ >>> Linuxwacom-discuss mailing list >>> linuxwacom-disc...@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/linuxwacom-discuss >>> >> > > > > -- > Cătălin Bălan > http://thescrambledbit.wordpress.com > I do know everything, just not all at once. It's a virtual memory problem. > ------------------------------------------------------------------------------ WhatsUp Gold - Download Free Network Management Software The most intuitive, comprehensive, and cost-effective network management toolset available today. Delivers lowest initial acquisition cost and overall TCO of any competing solution. http://p.sf.net/sfu/whatsupgold-sd _______________________________________________ Linuxwacom-devel mailing list Linuxwacom-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxwacom-devel
