[Letux-kernel] letux v5.4 + suspend

[H. Nikolaus Schaller]

Hi,

> Am 01.12.2019 um 19:33 schrieb Andreas Kemnade <andreas at kemnade.info>:
> 
> On Wed, 27 Nov 2019 22:05:12 +0100
> Andreas Kemnade <andreas at kemnade.info> wrote:
> 
>> On Wed, 27 Nov 2019 21:49:16 +0100
>> "H. Nikolaus Schaller" <hns at goldelico.com> wrote:
>> 
>>>> Am 27.11.2019 um 21:38 schrieb Andreas Kemnade <andreas at kemnade.info>:
>>>> 
>>>> Hi,
>>>> 
>>>> On Tue, 26 Nov 2019 23:01:08 +0100
>>>> Andreas Kemnade <andreas at kemnade.info> wrote:
>>>> 
>>>>> Hi,
>>>>> 
>>>>> while I get down to the typical 26mA idle current on GTA04A5, I do not
>>>>> wake up from suspend there. 
>>>>> Not investigated yet.
>>>>> 
>>>> Played around on the letux3704:
>>>> 
>>>> init=/bin/bash
>>>> mounting sysfs, proc
>>>> /etc/rcS.d/udev start
>>>> 
>>>> and then doing rtcwake -s 20 -m mem
>>>> does not wake up
>>>> 
>>>> rm -rf /lib/modules/5.4.0-letux+/kernel/drivers/gpu/drm/pvrsgx
>>>> 
>>>> fixes the issue.    
>>> 
>>> Important finding. So this means the SGX driver stops wakeup from suspend?
>>> I have added Tony and the openpvrsgx mailing list.
>>> 
>>> For more dynamic tests you could
>>> * boot
>>> * lsmod | grep pvrsgx
>>> * echo blacklist pvrsrvkm_driver_name >/etc/modprobe.d/pvr.conf
>>> * reboot
>>> * modprobe pvrsrvkm_driver_name to see what it does influence
>>> 
>> well, no_console_suspend gives more interesting results:
>> [  296.323822] Freezing user space processes ... (elapsed 0.001 seconds) done.
>> [  296.342559] OOM killer disabled.
>> [  296.350738] Freezing remaining freezable tasks ... (elapsed 0.001 seconds) done.
>> [  296.374755] dmtimer posted=0
>> [  296.456512] wwan_on_off_suspend: WWAN suspend
>> [  296.473388] ------------[ cut here ]------------
>> [  296.478881] WARNING: CPU: 0 PID: 1410 at drivers/gpu/drm/drm_modeset_lock.c:266 modeset_lock+0xc8/0xf4 [drm]
>> [  296.489410] Modules linked in: omapdrm libertas_sdio libertas cfg80211 snd_soc_simple_card snd_soc_simple_card_utils snd_soc_omap_twl4030 wwan_on_off panel_simple encoder_opa362 pvrsrvkmy
>> [  296.575866] CPU: 0 PID: 1410 Comm: rtcwake Not tainted 5.4.0-letux+ #3
>> [  296.582794] Hardware name: Generic OMAP36xx (Flattened Device Tree)
>> [  296.589508] [<c010fe8c>] (unwind_backtrace) from [<c010b28c>] (show_stack+0x10/0x14)
>> [  296.597717] [<c010b28c>] (show_stack) from [<c07ca8d0>] (dump_stack+0x88/0xa8)
>> [  296.605407] [<c07ca8d0>] (dump_stack) from [<c013130c>] (__warn+0xcc/0xe8)
>> [  296.612701] [<c013130c>] (__warn) from [<c01316c8>] (warn_slowpath_fmt+0x74/0xa0)
>> [  296.621002] [<c01316c8>] (warn_slowpath_fmt) from [<bf03d2b0>] (modeset_lock+0xc8/0xf4 [drm])
> 
> shouln't we initialie the mutex there by calling drm_mode_config_init()?
> Seems we do not do that and then using an unitialized mutex here.

Could at least be a reason for the suspend problem.

> 
>> [  296.630706] [<bf03d2b0>] (modeset_lock [drm]) from [<bf03d370>] (drm_modeset_lock_all_ctx+0x14/0xb4 [drm])
>> [  296.641540] [<bf03d370>] (drm_modeset_lock_all_ctx [drm]) from [<bf280d54>] (drm_atomic_helper_suspend+0x34/0xc8 [drm_kms_helper])
>> [  296.654357] [<bf280d54>] (drm_atomic_helper_suspend [drm_kms_helper]) from [<bf28406c>] (drm_mode_config_helper_suspend+0x28/0x5c [drm_kms_helper])
>> [  296.669006] [<bf28406c>] (drm_mode_config_helper_suspend [drm_kms_helper]) from [<bf2d4a50>] (pvr_suspend+0x18/0x48 [pvrsrvkm_omap3630_sgx530_125])
>> [  296.683349] [<bf2d4a50>] (pvr_suspend [pvrsrvkm_omap3630_sgx530_125]) from [<c0539380>] (dpm_run_callback+0x28/0x54)
> [...]

I have added Tony and the openpvrsgx mailing list for better discussion. The more eyes look on the issue the easier we can solve it.

BR,
Nikolaus