Re: [PATCH 4/4] quectel: use own cmux implementation over kernel line discipline

2019-09-30 Thread Martin Hundebøll
Hi Denis, On 30/09/2019 21.55, Denis Kenzior wrote: Hi Martin, So I wonder whether instead of fully reverting back to CMUX (only to put in n_gsm in later) you might want to just keep both around and use a /etc/quectel.conf or OFONO_QUECTEL_MUX=n_gsm|internal instead? I'd prefer a single

Re: [PATCH 4/4] quectel: use own cmux implementation over kernel line discipline

2019-09-30 Thread Denis Kenzior
Hi Martin, So I wonder whether instead of fully reverting back to CMUX (only to put in n_gsm in later) you might want to just keep both around and use a /etc/quectel.conf or OFONO_QUECTEL_MUX=n_gsm|internal instead? I'd prefer a single method to keep the needed testing to a minimum. I

Re: [PATCH 4/4] quectel: use own cmux implementation over kernel line discipline

2019-09-30 Thread Martin Hundebøll
Hi Denis, On 27/09/2019 03.00, Denis Kenzior wrote: On 9/26/19 2:27 PM, Martin Hundebøll wrote: The in-kernel implementation of gsm0710 causes deadlocks in the kernel[1], so switch back to the user-space implementation in ofono. [1]

Re: [PATCH 4/4] quectel: use own cmux implementation over kernel line discipline

2019-09-26 Thread Denis Kenzior
Hi Martin, On 9/26/19 2:27 PM, Martin Hundebøll wrote: The in-kernel implementation of gsm0710 causes deadlocks in the kernel[1], so switch back to the user-space implementation in ofono. [1] https://lore.kernel.org/lkml/4b2455c0-25ba-0187-6df6-c63b4ccc6...@geanix.com/ --- plugins/quectel.c

[PATCH 4/4] quectel: use own cmux implementation over kernel line discipline

2019-09-26 Thread Martin Hundebøll
The in-kernel implementation of gsm0710 causes deadlocks in the kernel[1], so switch back to the user-space implementation in ofono. [1] https://lore.kernel.org/lkml/4b2455c0-25ba-0187-6df6-c63b4ccc6...@geanix.com/ --- plugins/quectel.c | 250 -- 1