| rinigus | @rfasciani: your porter ( Thaodan ) should make sure it is supported. what's the status of waydroid on xperia 5, I don't know. then act as described at https://github.com/sailfishos-open/waydroid | 05:18 |
|---|---|---|
| rinigus | @b100dian: looking at https://github.com/sonyxperiadev/bug_tracker/issues/834 . maybe they shift focus now to aosp14 ... no idea what is sony's asap; could be similar to jolla's soon | 06:34 |
| T42 | <edp_17> @b100dian : strace /vendor/bin/hw/android.hardware.biometrics.fingerprint@2.1-fpcservice: https://paste.ubuntu.com/p/8dzrJxs5dd/ | 07:40 |
| T42 | <edp_17> There are a few (Permission denied) like: openat(AT_FDCWD, "/dev/qseecom", O_RDWR) = -1 EACCES (Permission denied) | 07:41 |
| T42 | <edp_17> and a Error::Failed to open /dev/qseec | 07:41 |
| T42 | <edp_17> this ^^ was when I run the strace as defaultuser | 07:42 |
| T42 | <edp_17> However, when I ran it as root, those errors are gone: https://paste.ubuntu.com/p/YtDKnhXG9H/ | 07:42 |
| T42 | <edp_17> Both errors ("fpc_hal_open failed\0" and "sensor_command, Failed to send command") are there but for me there is no details about the root cause. | 07:45 |
| T42 | <edp_17> @elros34 : Do you know how can I expand that suppressed errors? Like: "chatty : uid=1000(system) /vendor/bin/hw/android.hardware.biometrics.fingerprint@2.1-fpcservice expire 30 lines" | 07:47 |
| *** amccarthy is now known as Guest6030 | 09:40 | |
| *** amccarthy_ is now known as amccarthy | 09:40 | |
| T42 | <edp_17> Since I reflashed sfos with formatted /data, the /sys/devices/platform/soc/6019000.jtagmm is gone. Now I do have all sort of 6*.cti folders in /sys/devices/platform/soc/, but 6019000.jtagmm is no longer there. | 10:12 |
| T42 | <edp_17> However, the fingerprint service is still complaining about Unable to open /sys/devices/platform/soc/6*.cti/modalias: https://paste.ubuntu.com/p/JJym9DBsdW/ | 10:18 |
| T42 | <edp_17> And this modalias do exist in some folders in there: https://paste.ubuntu.com/p/8JjXgR8rh2/ | 10:19 |
| T42 | <edp_17> Mostly in /sys/devices/platform/soc/6*.jtagmm/ folders. | 10:19 |
| T42 | <edp_17> Is it possible to amend the service to look for that modalias in these 6*.jtagmm folders instead of 6*.cti ? | 10:20 |
| T42 | <edp_17> I think those "Unable to open '/sys/devices/platform/soc/6*.cti/modalias" messages are red herring because strace shows those "openat(AT_FDCWD, "/sys/devices/platform/soc/6*.jtagmm/modalias". | 10:50 |
| T42 | <edp_17> Should I see something fingerprint related in binder-list? https://paste.ubuntu.com/p/GnpCGBGWtR/ | 12:47 |
| T42 | <edp_17> Looks like ServiceManagemet is waiting for Could not find 'android.hardware.biometrics.fingerprint@2.1::IBiometricsFingerprint/default then moves on. Line 2488 : https://paste.ubuntu.com/p/GCvQRPP2Dx/ | 12:49 |
| T42 | <edp_17> @b100dian ^^ | 12:50 |
| T42 | <edp_17> I've seen mal has had issues with fingerprint on fp4 and he said: "I have the service in /vendor/odm/etc/vintf/manifest.xml but still I get the manifest error in logcat and then "Unable to get FP service"" | 12:51 |
| T42 | <edp_17> On my device, I don't have /vendor/odm/etc/vintf/manifest.xml. Could it be the issue? | 12:51 |
| T42 | <edp_17> In device on my host, I've found "vendor/etc/vintf/manifest/vendor.qti.gnss@3.0-service.xml" | 12:52 |
| T42 | <edp_17> Wait, no. That is only listed in device/motorola/sdm632-common/proprietary-files.txt. | 12:54 |
| T42 | <edp_17> I cannot find this vendor.qti.gnss@3.0-service.xml in dev environment. | 12:57 |
| mal | hmm, forgot about that fp4 issues, need to test in my latest build | 13:32 |
| T42 | <b100dian> rinigus: indeed, that means that we migh want to wait. You mentioned custom ROMs for nagara, did you find any? | 13:32 |
| T42 | <b100dian> @edp_17 I'm reading now your messages. | 13:32 |
| T42 | <b100dian> From the root strace: ioctl commands sent to fd 6 are to /dev/qseecom, when fpc_hal_open failed is logged. This means that you could have dmesg -w messages from kernel as I believe that device should have a kernel driver. | 13:41 |
| T42 | <b100dian> If it's not a kernel driver, you have qseecom process I guess? | 13:42 |
| rinigus | @b100dian: no, I haven't found any when briefly looking around at xda | 13:42 |
| rinigus | let's see when they publish new blobs | 13:43 |
| T42 | <b100dian> rinigus: yeah. Unfortunately the builds I've tried and the last 3 oem SW versions + the 5.10 on did not fare better | 14:06 |
| T42 | <b100dian> @edp_17 this is my binder list https://paste.ubuntu.com/p/FFj9cwWSwk/ - of course you won't have xiaomi or inscreen | 14:07 |
| rinigus | @b100dian: thank you very much for trying! | 14:08 |
| T42 | <edp_17> @b100dian , Thanks. I did a grep and find in kernel tree for qseecom. It seems it is not in defconfig: https://paste.ubuntu.com/p/g4RKKSWCmP/ | 15:00 |
| T42 | <b100dian> I think you have it https://github.com/edp17/android_kernel_motorola_sdm632/blob/lineage-17.1/arch/arm64/configs/ocean_defconfig#L1312 | 15:12 |
| T42 | <b100dian> So what does dmesg say when you try to start that service you straced earlier? | 15:13 |
| T42 | <elros34> I think it would be better to fix /dev/qseecom permission and strace this fingerprint exe with correct user instead as root or defaultuser Also use strace -f argument | 16:34 |
| T42 | <elros34> -s 256 would be nice, at least to see generated logs if you don't include fresh logcat | 16:39 |
| T42 | <b100dian> rinigus: I see XperiFirm can be run under mono:) I have downloaded a 64.0 FW, but it's all .sin files (which seem to be e.g. oem*.sin is a .tar with three oem.001|002|cms files.. Do you have experience with switching the base FW on Sony? | 16:41 |
| T42 | <b100dian> ah nevermind, probably Emma is needed for flashing :-S | 16:49 |
| rinigus | @b100dian: emma is safer bet, I think. | 16:56 |
| rinigus | there is also Newflasher. that's the one I used long time ago as it runs on linux | 16:58 |
| rinigus | but note that this bricking issue while flashing of xperia was reported with that | 16:59 |
| mal | what bricking issue? | 17:13 |
| mal | you mean because vendor_boot cannot be flashed from normal fastboot | 17:13 |
| mal | that is annoying | 17:13 |
| rinigus | mal: no, its about https://github.com/sonyxperiadev/bug_tracker/issues/824 | 17:14 |
| rinigus | basically boot partition was different probably due to use of Newflasher | 17:15 |
| rinigus | what about vendor_boot? what is "normal" fastboot? | 17:15 |
| rinigus | the fact that you have to do "fastboot reboot fastboot​" | 17:17 |
| rinigus | ? | 17:17 |
| mal | ok | 17:21 |
| mal | yes, userspace fastboot i.e. that "fastboot reboot fastboot​" is needed on newer sonys to flash vendor_boot | 17:21 |
| T42 | <edp_17> @elros34 : what should be the correct permission of /dev/qseecom? Its current system : drmrpc | 17:29 |
| rinigus | mal: it was already needed for tama. but at that time it was attributed to some bug, at least I thought so | 17:32 |
| T42 | <elros34> @edp_17 ah right maybe it's correct but stracing as user was incorrect | 18:12 |
| T42 | <edp_17> Thanks, will try that you suggested. | 18:15 |
| *** deathmist1 is now known as deathmist | 20:40 | |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!