Closed Bug 994846 Opened 11 years ago Closed 9 years ago

[QRD] Device is sluggish, CPU use is constantly high, logcat is non-functional

Categories

(Firefox OS Graveyard :: Performance, defect, P2)

Product:
Firefox OS Graveyard
Component:
Performance
Platform:
ARM
Gonk (Firefox OS)
Type:
defect

Tracking

(Not tracked)

Status:
RESOLVED WONTFIX

People

(Reporter: mikeh, Unassigned)

References

Details

(Keywords: perf, Whiteboard: [c=power p= s= u=] [Power])

While attempting to investigate Camera issues on the QRD, I have observed the following: 1. CPU use is consistently high: 130|root@msm8226:/ # top -m 10 -d 1 -s cpu -t User 16%, System 39%, IOW 0%, IRQ 0% User 39 + Nice 1 + Sys 99 + Idle 110 + IOW 0 + IRQ 0 + SIRQ 0 = 249 PID TID PR CPU% S VSS RSS PCY UID Thread Proc 2786 2804 0 26% R 218012K 69416K root Gecko_IOThread /system/b2g/b2g 2786 2786 0 24% R 218012K 69416K root b2g /system/b2g/b2g 2101 2101 1 4% R 1372K 620K root top top 2786 2813 1 4% S 218012K 69416K root BgHangManager /system/b2g/b2g 289 289 0 0% S 4592K 232K root adbd /sbin/adbd 1020 1034 0 0% S 6228K 580K root mpdecision /system/bin/mpdecision 21 21 1 0% S 0K 0K root khelper 22 22 1 0% S 0K 0K root netns 23 23 0 0% S 0K 0K root kworker/0:1 27 27 0 0% S 0K 0K root kworker/0:1H 2. logcat is either swamped with RIL and IO errors: 04-09 14:41:49.920 201 201 I Gonk : RIL[0]: OnConnectError 04-09 14:41:49.920 201 201 I Gonk : RIL[0]: OnDisconnect 04-09 14:41:49.920 201 489 I I/O : connect failed with error 2 (No such file or directory) 3. or logcat output is not available at all: 130|root@msm8226:/ # logcat -v threadtime --------- beginning of /dev/log/system 01-01 11:15:19.500 193 193 I Vold : Vold 2.1 (the revenge) firing up 01-01 11:15:19.500 193 193 D Vold : Volume sdcard state changing -1 (Initializing) -> 0 (No-Media) 01-01 11:15:19.580 193 218 D Vold : Volume sdcard state changing 0 (No-Media) -> 2 (Pending) 01-01 11:15:19.580 193 218 D Vold : Volume sdcard state changing 2 (Pending) -> 1 (Idle-Unmounted) 01-01 11:15:20.290 193 230 I Vold : /dev/block/vold/179:65 being considered for volume sdcard 01-01 11:15:20.290 193 230 D Vold : Volume sdcard state changing 1 (Idle-Unmounted) -> 3 (Checking) 01-01 11:15:20.450 633 633 D fsck_msdos: sectors Per cluster :64 01-01 11:15:20.450 633 633 D fsck_msdos: boot->ClusterOffset :8064 01-01 11:15:20.450 633 633 I fsck_msdos: FAT32 Filesystem 01-01 11:15:20.450 633 633 D fsck_msdos: Using cluster_chain_descriptor 01-01 11:15:20.450 633 633 I fsck_msdos: Total clusters 118914 01-01 11:15:20.510 633 633 I fsck_msdos: Begin to compare FAT 01-01 11:15:20.510 633 633 I fsck_msdos: ** Phase 2 - Check Cluster Chains 01-01 11:15:20.510 633 633 I fsck_msdos: Begin to handle the cluster chain 01-01 11:15:20.520 633 633 D fsck_msdos: rootDir->head :2 01-01 11:15:20.680 633 633 D fsck_msdos: Verify Filesystem information 01-01 11:15:20.680 633 633 W fsck_msdos: 372 files, 3052096 free (95378 clusters) 01-01 11:15:20.680 633 633 I fsck_msdos: free_rb_tree 01-01 11:15:20.680 633 633 I fsck_msdos: free_fragment_tree 01-01 11:15:20.680 633 633 I fsck_msdos: free_fragment_tree 01-01 11:15:20.680 633 633 I fsck_msdos: free_fragment_tree: rb_tree is empty 01-01 11:15:20.690 193 230 I Vold : Filesystem check completed OK Killed 137|root@msm8226:/ #
Inder, Bhargav -- please help with this! It is proving difficult to make progress on reproducing and fixing QRD only bugs. However, Mike is trying though on his QRD device...
Flags: needinfo?(ikumar)
Flags: needinfo?(bhargavg1)
Flags: needinfo?(bhargavg1)
few things, did we load the moz ril and see if still see the logcat spam... for capturing logs try this, adb reboot && adb wait-for-device && adb root && adb wait-for-device && adb logcat | tee log1.txt
Flags: needinfo?(ikumar)
In the logcat, I can see that between getting the CameraManager and creating the CameraControl instance, over 9 seconds pass: 04-10 13:05:10.170 1114 1114 I Gecko : nsDOMCameraManager::nsDOMCameraManager(nsPIDOMWindow*):56 : this=0xb3fd81f0, windowId=4 04-10 13:05:10.170 1114 1114 I Gecko : getListOfCameras : getNumberOfCameras() returned 2 04-10 13:05:10.170 1114 1114 I Gecko : GetCameraName : getNumberOfCameras() returned 2 04-10 13:05:10.180 1114 1114 I Gecko : GetCameraName : getNumberOfCameras() returned 2 04-10 13:05:13.340 206 206 E libsensor1: sensor1_open: Error in connect() errno=111 Connection refused 04-10 13:05:13.340 206 206 E qcom_sensors_hal: _hal_sensors_get_sensors_list: sensor1 init failed 04-10 13:05:16.350 206 206 E libsensor1: sensor1_open: Error in connect() errno=111 Connection refused 04-10 13:05:16.350 206 206 E qcom_sensors_hal: _hal_sensors_get_sensors_list: sensor1 init failed 04-10 13:05:19.370 206 206 E libsensor1: sensor1_open: Error in connect() errno=111 Connection refused 04-10 13:05:19.370 206 206 E qcom_sensors_hal: _hal_sensors_get_sensors_list: sensor1 init failed 04-10 13:05:19.430 206 206 I Gonk : Setting nice for pid 1522 to 18 04-10 13:05:19.430 206 206 I Gonk : Changed nice for pid 1522 from 0 to 18. 04-10 13:05:19.490 206 1540 E Profiler: BPUnw: [9 total] thread_register_for_profiling(me=0x186c958, stacktop=0xa86d4cc3) 04-10 13:05:19.580 1114 1114 I Gecko : void nsDOMCameraManager::GetCamera(const nsAString_internal&, const mozilla::dom::CameraConfiguration&, mozilla::dom::GetCameraCallback&, const mozilla::dom::Optional<mozilla::dom::OwningNonNull<mozilla::dom::CameraErrorCallback> >&, mozilla::ErrorResult&):115 Then it's another 3 seconds before the hardware is reported open: 04-10 13:05:22.800 1114 1114 I Gecko : DOM OnHardwareStateChange(0) And another ~190ms to for the preview to be started: 04-10 13:05:22.990 1114 1548 I Gecko : Preview started
(In reply to Mike Habicher [:mikeh] from comment #3) > In the logcat, I can see that between getting the CameraManager and creating > the CameraControl instance, over 9 seconds pass: > > Then it's another 3 seconds before the hardware is reported open: > > 04-10 13:05:22.800 1114 1114 I Gecko : DOM OnHardwareStateChange(0) > > And another ~190ms to for the preview to be started: > > 04-10 13:05:22.990 1114 1548 I Gecko : Preview started this should help you to overcome delays, https://bugzilla.mozilla.org/show_bug.cgi?id=962697#c1
Keywords: perf
Priority: -- → P1
Whiteboard: [c=power p= s= u=]
Mike - do you have a QRD device? We (the performance team) do not, nor do we expect to get one in the future, so are you planning on looking at this in the future?
Priority: P1 → P2
Flags: needinfo?(mhabicher)
(In reply to Jon Hylands [:jhylands] from comment #5) > > Mike - do you have a QRD device? We (the performance team) do not, nor do we > expect to get one in the future, so are you planning on looking at this in > the future? I have one, but will wait for bug 1014485 to land before investigating further.
Depends on: 1014485
Flags: needinfo?(mhabicher)
Whiteboard: [c=power p= s= u=] → [c=power p= s= u=] [Power]
gsvelto says: "IIRC QRD devices are not available anymore, I returned mine months ago and I think they were sent back to the vendor, I suggest closing this bug."
Status: NEW → RESOLVED
Closed: 9 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.