Jump to content

Tom G

  • Content Count

  • Joined

  • Last visited

Community Reputation

41 Excellent

About Tom G

  • Rank

Contact Methods

  • Website URL
  • ICQ

Profile Information

  • Location
    Melbourne Australia
  • Your Current Device(s)
    2X Blade U8220 V9 Vega X10
  1. That could be difficult to fix. I doubt dmesg will show any error. It looks like the new 2.6.32 source that ZTE released yesterday has support for s5k5acgx on blade, so maybe I can find something in there.
  2. No, I haven't looked at any of the camera problems yet and I don't know when I will.
  3. Those are to be expected on your hardware. <4>[01-01 00:00:01.429999] [1: swapper]gpio 31 request is error! <4>[01-01 00:00:01.429999] [1: swapper]cypress_touch: probe of 0-000a failed with error -16 Cypress touch screen did not respond. <3>[01-01 00:00:01.690000] [1: swapper]msm_i2c msm_i2c.0: (007c) Error during data xfer (-5) <3>[01-01 00:00:01.690000] [1: swapper]ft5x0x_ts 0-003e: Fts_i2c_read FAILED: read of register 166 <6>[01-01 00:00:01.690000] [1: swapper]wly: Fts_i2c_read, FT5X0X_REG_FIRMID = 166. <3>[01-01 00:00:01.710000] [1: swapper]msm_i2c msm_i2c.0: (007c) Error during data xfer (-5) <3>[01-01 00:00:01.710000] [1: swapper]ft5x0x_ts 0-003e: Fts_i2c_read FAILED: read of register 166 <6>[01-01 00:00:01.710000] [1: swapper]wly: Fts_i2c_read, FT5X0X_REG_FIRMID = 166. Focaltech touch screen did not respond. You have a synaptics touch screen, so you don't have the hardware these two are looking for. <3>[11-01 02:11:45.520000] [67: mt9p111_wq]msm_i2c msm_i2c.0: (007a) Error during data xfer (-5) <3>[11-01 02:11:45.540000] [67: mt9p111_wq]mt9p111_workqueue: rc != -ENOINIT <3>[11-01 02:11:46.630000] [68: mt9t11x_wq]msm_i2c msm_i2c.0: (007a) Error during data xfer (-5) <3>[11-01 02:11:46.630000] [68: mt9t11x_wq]mt9t11x_sensor_dev_probe: rc=0xfffffffb,model_id = 0x0 You have a Samsung s5k5acgx camera sensor, not one of the mt9's. There may also be errors for the ov5642 in there somewhere. This is just hardware not present in the device failing to initialise. The touch screen problem isn't really fixed, it is just handled better. When the problem occurs you should still get a few of the error below. <3>[01-01 00:00:02.000000] [1: swapper]synaptics-rmi-ts 0-0022: synaptics_i2c_read FAILED: read of register 120 <4>[01-01 00:00:02.000000] [1: swapper]wly: synaptics_i2c_read, x, �,f,�,, Then you should see "synaptics_i2c_read failed. Falling back to polling mode."
  4. That log shows it working normally. Try recreating the condition that was causing the problem before (fully powering off the device, then get a log after first boot). If the problem occurs you should see a few i2c read errors from the synaptics driver in dmesg.
  5. I just realised the link for the new test rom was incorrect, it was linking to the old test rom which explains users getting the same results. Try this one. Link
  6. That is actually normal. Its probably not how it should work, but that is how the kernel driver has always handled generating events. Most apps (such as the phone app) probably see any decrease in distance as a near event and any increase as far (most apps only consider near/far and don't care about distance) The version on z-device test I'm using shows resolution simply as 5.0 (no +/-), it is an older version so the way it is displayed in the app has probably changed. The resolution displayed has no real meaning, its just a label from the sensors lib. I could set it to 1 and the lib would still return values of 0-5. I've never seen anything use the resolution value for anything meaningful, it is just information. Mine is also an original UK OLED and with calibration is around 300-400 (using the kernel currently in CM7). I've changed some values in the kernel to get it back similar to what was in the old kernel, but at the moment using the old defaults (which always worked fine on my devices) I'm only getting 1-5 so it will need a little more adjustment before I submit it. When calibrated I'm now getting 5000-7000 which is higher than I got with the old kernel but closer to what I would expect.
  7. I'm more confused now than I was before. It looks like it should be completely failed when your getting the touch point moved. Still its good progress, I'll look at it more tomorrow. If you had compcache enabled previously it would not have actually worked. It is fixed in the test rom so that may be the speed increase you are seeing. Compcache isn't needed for the 512MB devices, but should have some benefit for the 256MB devices.
  8. Please test this and provide logs (dmesg is all thats really needed). Link Thanks.
  9. Sej's builds include patches that have not yet passed review. They may have serious bugs.
  10. The default values in 2.6.32 where 8000-8500, but calibrated values varied depending on device. Both of mine when calibrated in 2.6.32 were around 3200-4000. Now in 2.6.35 they are both around 300-400, so the 2.6.32 values may be very different to the 2.6.35 values. Edit. I've made some changes to the kernel which will hopefully fix the calibration issues as well as making the calibration values similar to what they were in 2.6.32. A new kernel will be submitted for review a little later.
  11. Please also take note of the values returned from the calibration tool and the values that work. For most users calibration is fine, but for a small percentage it is not working. I worked through this with a user on IRC a couple of days ago, and while working on it after many attempts the sensor calibrated correctly using the calibration tool, but I have no idea why it did not work initially with calibration, but later did. The values that did work were fairly close to the ones that didn't so I'm interested in how close the calibration values are to what they should be. Hopefully a small adjustment to the calibration program can fix the problems, but I need users to provide information (and only users that have the calibration issues). So what I need is what does the calibration tool report which doesn't work? What does work? These values vary largely between different devices, mine calibrate to around 300-400 but the users having trouble mostly seem to be over 1000. Burstlam - It would be great if you could get some information from the Chinese users. I think this may effect them worse than others. BTW, the values burstlam has suggested probably won't work for most users, but play with different values and see what does work. Use a sensor testing app such as z-device test. When calibrated correctly the sensor should return a range of values from 0 to 5cm (if for example you only get values from 0-3 it will work since most apps only looks for 0 or >0 but isn't calibrated correctly).
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.