nvidia-tegra-modules.git/drivers/platform/tegra/tegra-camera-rtcpu.c, branch rtss22-ae NVIDIA's kernel modules to support tegra chips (used in Jetson boards) tegra: camera: Wait on empty interrupt 2021-01-19T12:39:22+00:00 Aniket Bahadarpurkar aniketb@nvidia.com 2021-01-13T10:55:45+00:00 e7565d131437f83b629436dfdaa6eeb6d067b6f5 Polling empty irq status increases boot time, hence use empty irq interrupt instead. Bug 200671014 Change-Id: I5c263df5ac0f2f7990ae5bf821290439b3d64fa3 Signed-off-by: Aniket Bahadarpurkar <aniketb@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2469939 (cherry picked from commit 6e10f43c6fcdd45910d79ec07a492f73ea910463) Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2469944 Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> Reviewed-by: Automatic_Commit_Validation_User Reviewed-by: Ken Chang <kenc@nvidia.com> Reviewed-by: Nitin Pai <npai@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> GVS: Gerrit_Virtual_Submit 
Polling empty irq status increases boot time, hence
use empty irq interrupt instead.

Bug 200671014

Change-Id: I5c263df5ac0f2f7990ae5bf821290439b3d64fa3
Signed-off-by: Aniket Bahadarpurkar <aniketb@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2469939
(cherry picked from commit 6e10f43c6fcdd45910d79ec07a492f73ea910463)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2469944
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Ken Chang <kenc@nvidia.com>
Reviewed-by: Nitin Pai <npai@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
tegra: rtcpu: use polling in tegra_camrtc_wait_for_empty() 2020-12-03T21:24:49+00:00 Ken Chang kenc@nvidia.com 2020-11-24T03:36:29+00:00 2c43dace7abd54fed82c08b8e24ce809bfe52963 The current implementation of tegra_camrtc_wait_for_empty() is to test mabilbox empty by both interrupt and polling. This is causing __might_sleep() being called after prepare_to_wait_event() sets the current process in !TASK_RUNNING state, that's why we are seeing below warning: [ 6.790029] do not call blocking ops when !TASK_RUNNING; state=2 set at [<ffffff8008116378>] prepare_to_wait_event+0x70/0x120 [ 6.790054] ------------[ cut here ]------------ [ 6.790061] WARNING: CPU: 2 PID: 1 at __might_sleep+0x84/0x90 [ 6.790068] Modules linked in: ... [ 6.790240] Call trace: [ 6.790246] [<ffffff80080e1ccc>] __might_sleep+0x84/0x90 [ 6.790252] [<ffffff80081306dc>] synchronize_irq+0x54/0xc8 [ 6.790259] [<ffffff8008be610c>] tegra_hsp_sm_pair_is_empty+0x24/0x80 [ 6.790267] [<ffffff8008bb9638>] tegra_camrtc_mbox_exchange+0xd8/0x298 [ 6.790272] [<ffffff8008bb9850>] tegra_camrtc_command+0x58/0x80 [ 6.790276] [<ffffff8008bba620>] tegra_cam_rtcpu_probe+0x2b0/0x638 [ 6.790283] [<ffffff8008880220>] platform_drv_probe+0x60/0xc0 [ 6.790287] [<ffffff800887d360>] really_probe+0x160/0x378 [ 6.790291] [<ffffff800887d66c>] driver_probe_device+0x4c/0xc8 [ 6.790295] [<ffffff800887d7e8>] __driver_attach+0x100/0x128 [ 6.790307] [<ffffff800887ab30>] bus_for_each_dev+0x90/0xe0 [ 6.790311] [<ffffff800887cb90>] driver_attach+0x30/0x40 [ 6.790315] [<ffffff800887c4f4>] bus_add_driver+0x16c/0x270 [ 6.790319] [<ffffff800887eae8>] driver_register+0x68/0x100 [ 6.790323] [<ffffff8008880164>] __platform_driver_register+0x5c/0x68 [ 6.790331] [<ffffff8009832ef8>] tegra_cam_rtcpu_driver_init+0x18/0x20 [ 6.790337] [<ffffff80080835c0>] do_one_initcall+0x58/0x160 [ 6.790343] [<ffffff80097e0dec>] kernel_init_freeable+0x1e4/0x278 [ 6.790349] [<ffffff8008ef0ee0>] kernel_init+0x18/0x100 [ 6.790353] [<ffffff8008083390>] ret_from_fork+0x10/0x40 Simplify the code to use polling only to avoid above problem during probing. Bug 3175120 Change-Id: Ifc49e02af84ea1cf5bc4c747fb2d1cd139c9d9ed Signed-off-by: Ken Chang <kenc@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2449955 Reviewed-by: automaticguardword <automaticguardword@nvidia.com> Reviewed-by: Pekka Pessi <ppessi@nvidia.com> Reviewed-by: Bibek Basu <bbasu@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> GVS: Gerrit_Virtual_Submit Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> 
The current implementation of tegra_camrtc_wait_for_empty() is to test
mabilbox empty by both interrupt and polling.
This is causing __might_sleep() being called after prepare_to_wait_event()
sets the current process in !TASK_RUNNING state, that's why we are seeing
below warning:

[ 6.790029] do not call blocking ops when !TASK_RUNNING; state=2 set at [<ffffff8008116378>] prepare_to_wait_event+0x70/0x120
[ 6.790054] ------------[ cut here ]------------
[ 6.790061] WARNING: CPU: 2 PID: 1 at __might_sleep+0x84/0x90
[ 6.790068] Modules linked in:
...
[ 6.790240] Call trace:
[ 6.790246] [<ffffff80080e1ccc>] __might_sleep+0x84/0x90
[ 6.790252] [<ffffff80081306dc>] synchronize_irq+0x54/0xc8
[ 6.790259] [<ffffff8008be610c>] tegra_hsp_sm_pair_is_empty+0x24/0x80
[ 6.790267] [<ffffff8008bb9638>] tegra_camrtc_mbox_exchange+0xd8/0x298
[ 6.790272] [<ffffff8008bb9850>] tegra_camrtc_command+0x58/0x80
[ 6.790276] [<ffffff8008bba620>] tegra_cam_rtcpu_probe+0x2b0/0x638
[ 6.790283] [<ffffff8008880220>] platform_drv_probe+0x60/0xc0
[ 6.790287] [<ffffff800887d360>] really_probe+0x160/0x378
[ 6.790291] [<ffffff800887d66c>] driver_probe_device+0x4c/0xc8
[ 6.790295] [<ffffff800887d7e8>] __driver_attach+0x100/0x128
[ 6.790307] [<ffffff800887ab30>] bus_for_each_dev+0x90/0xe0
[ 6.790311] [<ffffff800887cb90>] driver_attach+0x30/0x40
[ 6.790315] [<ffffff800887c4f4>] bus_add_driver+0x16c/0x270
[ 6.790319] [<ffffff800887eae8>] driver_register+0x68/0x100
[ 6.790323] [<ffffff8008880164>] __platform_driver_register+0x5c/0x68
[ 6.790331] [<ffffff8009832ef8>] tegra_cam_rtcpu_driver_init+0x18/0x20
[ 6.790337] [<ffffff80080835c0>] do_one_initcall+0x58/0x160
[ 6.790343] [<ffffff80097e0dec>] kernel_init_freeable+0x1e4/0x278
[ 6.790349] [<ffffff8008ef0ee0>] kernel_init+0x18/0x100
[ 6.790353] [<ffffff8008083390>] ret_from_fork+0x10/0x40

Simplify the code to use polling only to avoid above problem during probing.

Bug 3175120

Change-Id: Ifc49e02af84ea1cf5bc4c747fb2d1cd139c9d9ed
Signed-off-by: Ken Chang <kenc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2449955
Reviewed-by: automaticguardword <automaticguardword@nvidia.com>
Reviewed-by: Pekka Pessi <ppessi@nvidia.com>
Reviewed-by: Bibek Basu <bbasu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
tegra: rtcpu: inverse SCE reset assert sequence 2019-06-22T09:35:57+00:00 Timo Alho talho@nvidia.com 2019-03-11T09:23:53+00:00 4aef73562676eb7350a2c855ffda90f4ad345009 Reset the SCE R5 core before the APB peripherals to avoid situation where R5 makes accesses to the APB devices under devices and potentially locks-up the APB bus. Bug 200477166 Change-Id: I53e090f592bd175f63172beffa527e093025dcd7 Signed-off-by: Timo Alho <talho@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/2069626 (cherry picked from commit 50d55ff555560b623faa5a1209d3b4b8a242e9be) Reviewed-on: https://git-master.nvidia.com/r/2138038 GVS: Gerrit_Virtual_Submit Reviewed-by: Pekka Pessi <ppessi@nvidia.com> Tested-by: Leon Yu <leoyu@nvidia.com> Reviewed-by: Bibek Basu <bbasu@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> 
Reset the SCE R5 core before the APB peripherals to avoid situation
where R5 makes accesses to the APB devices under devices and potentially
locks-up the APB bus.

Bug 200477166

Change-Id: I53e090f592bd175f63172beffa527e093025dcd7
Signed-off-by: Timo Alho <talho@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2069626
(cherry picked from commit 50d55ff555560b623faa5a1209d3b4b8a242e9be)
Reviewed-on: https://git-master.nvidia.com/r/2138038
GVS: Gerrit_Virtual_Submit
Reviewed-by: Pekka Pessi <ppessi@nvidia.com>
Tested-by: Leon Yu <leoyu@nvidia.com>
Reviewed-by: Bibek Basu <bbasu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
tegra: rtcpu: clear FWLOADDONE bit on suspend 2019-02-16T10:42:51+00:00 Pekka Pessi ppessi@nvidia.com 2018-12-17T17:04:47+00:00 d0bfb9996b84f6737e9526ac7c6dc6608ddac1fc Clear the FWLOADDONE bit when suspending SCE/RCE. Bug 200469610 Change-Id: Ibdb32868c1dd1b678ccd94fb4c2537cf041a845d Signed-off-by: Pekka Pessi <ppessi@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1974615 (cherry picked from commit d232d1edcb29ef8282688eb7bd29faa4f2ea2469) Reviewed-on: https://git-master.nvidia.com/r/2019788 GVS: Gerrit_Virtual_Submit Tested-by: Leon Yu <leoyu@nvidia.com> Reviewed-by: Bibek Basu <bbasu@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> 
Clear the FWLOADDONE bit when suspending SCE/RCE.

Bug 200469610

Change-Id: Ibdb32868c1dd1b678ccd94fb4c2537cf041a845d
Signed-off-by: Pekka Pessi <ppessi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1974615
(cherry picked from commit d232d1edcb29ef8282688eb7bd29faa4f2ea2469)
Reviewed-on: https://git-master.nvidia.com/r/2019788
GVS: Gerrit_Virtual_Submit
Tested-by: Leon Yu <leoyu@nvidia.com>
Reviewed-by: Bibek Basu <bbasu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
tegra: camera: rtcpu: tone down logging 2018-09-13T00:47:33+00:00 Pekka Pessi ppessi@nvidia.com 2018-08-20T21:52:23+00:00 a277218541aa81b31f1eceb0f6e79cac8cf10603 Reduce kernel log spam. Bug 2305627 Change-Id: I9e7c34f631e5e4c5a7aba348664aa8c8cbe6e881 Signed-off-by: Pekka Pessi <ppessi@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1803896 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> 
Reduce kernel log spam.

Bug 2305627

Change-Id: I9e7c34f631e5e4c5a7aba348664aa8c8cbe6e881
Signed-off-by: Pekka Pessi <ppessi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1803896
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
tegra: camera: rtcpu: fix boot timeouts 2018-09-13T00:47:24+00:00 Pekka Pessi ppessi@nvidia.com 2018-08-20T17:25:22+00:00 a16f79eeff8f61dccc679dd84a0ecb1f138c506e The t186-sce boot time sometimes exceeds 2000 milliseconds. The undue boot time is because the memory bandwidth problems. During the boot the DRAM image is read twice (once when copying, once when calculating SHA1). Calculate time from deasserting resets to the first mailbox handshake with camrtc. Add a dedicated bandwidth manager client for camrtc. Request for extra memory bandwidth during boot time. Retry camrtc boot if it fails. Report failed boot handshake correctly to the platform, avoid corruption. Bug 2305627 Change-Id: Ia96e369ee1b09d6298268f7bd309db1c8f326564 Signed-off-by: Pekka Pessi <ppessi@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1803895 Reviewed-by: Automatic_Commit_Validation_User Reviewed-by: Semi Malinen <smalinen@nvidia.com> Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com> GVS: Gerrit_Virtual_Submit Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> 
The t186-sce boot time sometimes exceeds 2000 milliseconds. The undue
boot time is because the memory bandwidth problems. During the boot
the DRAM image is read twice (once when copying, once when calculating
SHA1).

Calculate time from deasserting resets to the first mailbox handshake
with camrtc.

Add a dedicated bandwidth manager client for camrtc. Request for extra
memory bandwidth during boot time.

Retry camrtc boot if it fails.

Report failed boot handshake correctly to the platform, avoid
corruption.

Bug 2305627

Change-Id: Ia96e369ee1b09d6298268f7bd309db1c8f326564
Signed-off-by: Pekka Pessi <ppessi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1803895
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Semi Malinen <smalinen@nvidia.com>
Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
tegra: rtcpu: access camera devices by name 2018-08-14T19:10:47+00:00 Pekka Pessi ppessi@nvidia.com 2018-07-25T10:37:43+00:00 c07b8b92509c0d169e643f2b752f525e4964f508 Some configurations may not expose all the camera devices. Jira CAMASIL-184 Bug 2274108 Change-Id: Ic39da21649615872c7aebd88c4024e4ea2feb6b9 Signed-off-by: Pekka Pessi <ppessi@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1795993 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Jon Mayo <jmayo@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> 
Some configurations may not expose all the camera devices.

Jira CAMASIL-184
Bug 2274108

Change-Id: Ic39da21649615872c7aebd88c4024e4ea2feb6b9
Signed-off-by: Pekka Pessi <ppessi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1795993
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Jon Mayo <jmayo@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
tegra: camrtc: slow clock after resume 2018-06-11T10:03:52+00:00 Pekka Pessi ppessi@nvidia.com 2018-06-06T11:40:22+00:00 d6c049615324f2d42a6cc4cef2e5699075293854 Do not adjust clocks on full speed after powering on Camera RTCPU when resuming from SC7 suspend. Bug 200418877 Change-Id: I0626c07d76e444bed38bec1687e38352b39bafe3 Signed-off-by: Pekka Pessi <ppessi@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1741562 Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> 
Do not adjust clocks on full speed after powering on Camera RTCPU when
resuming from SC7 suspend.

Bug 200418877

Change-Id: I0626c07d76e444bed38bec1687e38352b39bafe3
Signed-off-by: Pekka Pessi <ppessi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1741562
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
tegra: camrtc: allow disabling runtime pm 2018-06-06T13:23:26+00:00 Pekka Pessi ppessi@nvidia.com 2018-06-04T14:14:59+00:00 3ded185377a4d5b7c7e5535da52f39bd12cb57f4 Add dt support for disabling runtime suspend of CAMRTC after boot. Bug 200364940 Change-Id: Ib28c85f5eb96d7eb540be7383eefe9eb75812abe Signed-off-by: Pekka Pessi <ppessi@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1739261 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Vishruth Jain <vishruthj@nvidia.com> Tested-by: Mantravadi Karthik <mkarthik@nvidia.com> Reviewed-by: Sudhir Vyas <svyas@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> 
Add dt support for disabling runtime suspend of CAMRTC after boot.

Bug 200364940

Change-Id: Ib28c85f5eb96d7eb540be7383eefe9eb75812abe
Signed-off-by: Pekka Pessi <ppessi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1739261
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Vishruth Jain <vishruthj@nvidia.com>
Tested-by: Mantravadi Karthik <mkarthik@nvidia.com>
Reviewed-by: Sudhir Vyas <svyas@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
video:host:nvcsi load prod settings at poweron 2018-03-09T12:29:09+00:00 David Bang dbang@nvidia.com 2018-02-23T22:47:10+00:00 fe74f70da73e95cdda450a037022a28baaaa72ac Load the prod settings for nvcsi during finalize poweron Bug 200385222 Change-Id: Ieee4d9fdfd6997ba70edc2ceaabdfcf5b03b120d Signed-off-by: David Bang <dbang@nvidia.com> Reviewed-on: https://git-master.nvidia.com/r/1663271 Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com> Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Bhanu Murthy V <bmurthyv@nvidia.com> Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com> Tested-by: mobile promotions <svcmobile_promotions@nvidia.com> 
Load the prod settings for nvcsi during finalize poweron

Bug 200385222

Change-Id: Ieee4d9fdfd6997ba70edc2ceaabdfcf5b03b120d
Signed-off-by: David Bang <dbang@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1663271
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Bhanu Murthy V <bmurthyv@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>