nichcream / 4.9-msm8937 Goto Github PK

We need a working msm iommu v1 driver to replace the unsupported tz-smmu, then hook on other drivers to the new iommu driver instead.
mdss
vidc

Will need to backport 3.18 camera and attach the driver to iommu v1

i2c was unable to register aw2013 at 0x45

microphone near the charging port doesn't work

fpc blob hardcoded sysfs path to /sys/devices/soc but that changed to /sys/devices/platform/soc in 4.9

happened only recently. probably might be fixed after recompilation. Just adding this as reminder

currently broken

We didn't see you for a while. Can u contact me please? Need to discuss something with you
email [email protected]

i have no leads on this.

venus fw failed to load

Now rebased, all that's left is to cleanup the commit history.

Relook:

• 93ac736 power: qpnp-smbcharger: Reverse the sign of CURRENT_NOW
  The last I picked this commit the current_now became negative when charging (which is what the commit claimed to fix).
• 1929776 HACK: media: Restore old entity_id logic
  Rebase corrrectly
• 9bd5527 ARM64: dts: land: optimize the vbat-estimate-diff-mv
  Not needed
• 8cc4596 FIXME: power: reset: msm: Always perform warm reset
  Use dt prop instead

Also in the future, I'd like to keep device specific commits on a separate branch (i.e xm8937). Upstreamed Linux/CAF branches will merge xm8937 branch. This is so I can keep track of the device specific commits easily (rebasing will be easier later).

Just a reminder

Issue seen on both santoni & land after playing videos

[23059.216892] msm_vidc: info: Opening video instance: 0000000000000000, 1
[23059.223866] subsys-restart: __subsystem_get(): Changing subsys fw_name to venus
[23059.228185] subsys-pil-tz 1de0000.qcom,venus: loading /vendor/firmware_mnt/image/venus.mdt failed with error -13
[23059.229848] ueventd: firmware: loading 'venus.mdt' for '/devices/platform/soc/1de0000.qcom,venus/firmware/venus.mdt'
[23059.230406] ueventd: loading /devices/platform/soc/1de0000.qcom,venus/firmware/venus.mdt took 0ms
[23059.245673] subsys-pil-tz 1de0000.qcom,venus: venus: loading from 0x000000008f800000 to 0x000000008fc00000
[23059.260430] subsys-pil-tz 1de0000.qcom,venus: loading /vendor/firmware_mnt/image/venus.b02 failed with error -13
[23059.262291] ueventd: firmware: loading 'venus.b02' for '/devices/platform/soc/1de0000.qcom,venus/firmware/venus.b02'
[23059.264913] ueventd: loading /devices/platform/soc/1de0000.qcom,venus/firmware/venus.b02 took 2ms
[23059.265547] subsys-pil-tz 1de0000.qcom,venus: loading /vendor/firmware_mnt/image/venus.b03 failed with error -13
[23059.266766] ueventd: firmware: loading 'venus.b03' for '/devices/platform/soc/1de0000.qcom,venus/firmware/venus.b03'
[23059.267114] ueventd: loading /devices/platform/soc/1de0000.qcom,venus/firmware/venus.b03 took 0ms
[23059.270446] subsys-pil-tz 1de0000.qcom,venus: loading /vendor/firmware_mnt/image/venus.b04 failed with error -13
[23059.274474] ueventd: firmware: loading 'venus.b04' for '/devices/platform/soc/1de0000.qcom,venus/firmware/venus.b04'
[23059.274989] ueventd: loading /devices/platform/soc/1de0000.qcom,venus/firmware/venus.b04 took 0ms
[23059.296097] subsys-pil-tz 1de0000.qcom,venus: venus: Brought out of reset
[23063.231862] FG: soc_work_fn: adjust_soc: s 46 r 140869 i 513913 v 3746316 t 355
[23064.930258] msm_vidc: info: Opening video instance: 0000000000000000, 1
[23064.930451] msm_vidc: err: Failed to create debugfs for msm_vidc
[23065.787025] msm_vidc: info: Closed video instance: 0000000000000000
[23066.738924] msm_vidc: info: Opening video instance: 0000000000000000, 1
[23067.377830] msm_vidc: info: Opening video instance: 0000000000000000, 1
[23067.378063] msm_vidc: err: Failed to create debugfs for msm_vidc
[23068.082909] msm_vidc: info: Closed video instance: 0000000000000000
[23068.104344] msm_vidc: info: Closed video instance: 0000000000000000
[23069.812740] msm_vidc: info: Opening video instance: 0000000000000000, 1
[23081.022759] msm_vidc: info: Opening video instance: 0000000000000000, 1
[23081.022948] msm_vidc: err: Failed to create debugfs for msm_vidc
[23083.552258] msm_vidc: info: Opening video instance: 0000000000000000, 1
[23083.552503] msm_vidc: err: Failed to create debugfs for msm_vidc
[23083.559632] msm_vidc: err: H/W is overloaded. needed: 378000 max: 360960
[23083.559641] msm_vidc: err: Running instances:
[23083.559645] msm_vidc: err: type| w| h| fps|prop
[23083.559649] msm_vidc: err: 1| 640|1136| 30|
[23083.559652] msm_vidc: err: 1| 640|1136| 30|
[23083.559656] msm_vidc: err: 1| 640|1136| 30|
[23083.559659] msm_vidc: err: 1|1920|1088| 15|
[23083.560530] msm_vidc: warn: msm_comm_generate_session_error
[23083.560536] msm_vidc: warn: Got a response for an inactive session
[23083.560540] msm_vidc: warn: msm_vidc_check_session_supported: Hardware is overloaded
[23083.560543] msm_vidc: err: msm_vdec_s_fmt: session not supported
[23083.562932] msm_vidc: info: Closed video instance: 0000000000000000
[23083.711483] FG: soc_work_fn: adjust_soc: s 46 r 131591 i 706477 v 3720223 t 355
[23084.528062] msm_vidc: info: Opening video instance: 0000000000000000, 1
[23084.528292] msm_vidc: err: Failed to create debugfs for msm_vidc
[23084.531576] msm_vidc: err: H/W is overloaded. needed: 378000 max: 360960
[23084.531581] msm_vidc: err: Running instances:
[23084.531585] msm_vidc: err: type| w| h| fps|prop
[23084.531589] msm_vidc: err: 1| 640|1136| 30|
[23084.531598] msm_vidc: err: 1| 640|1136| 30|
[23084.531604] msm_vidc: err: 1| 640|1136| 30|
[23084.531610] msm_vidc: err: 1|1920|1088| 15|
[23084.532521] msm_vidc: warn: msm_comm_generate_session_error
[23084.532529] msm_vidc: warn: Got a response for an inactive session
[23084.532533] msm_vidc: warn: msm_vidc_check_session_supported: Hardware is overloaded
[23084.532539] msm_vidc: err: msm_vdec_s_fmt: session not supported
[23084.533187] msm_vidc: info: Closed video instance: 0000000000000000

display flicks momentarily after that

Sony's iommu is racy. It causes panics when a device attaches to iommu too early.
Stage 1: subsys init: probe iommu -> bus is registered, ops are now exposed
Stage 2: module_init(): call attach device
Stage 3: call of_xlate() to add ctx banks

Stage 2 and 3 should actually be swoped

that part was real quick.

setting brightness from 1-5 will not change the backlight