Re: [PATCH 0/8] tty: n_gsm: Make mux work as a responder station

[Andrej Krpic]

On 21.02.2016 23:42, One Thousand Gnomes wrote:

On Sun, 21 Feb 2016 22:38:29 +0100
Andrej Krpic  wrote:

When using n_gsm you have to explicitly set it to work as a 
initiator
station. This led me to believe that it can also work as a 
responder.

snip
This looks reasonable to me. It was never intended to work as a 
responder
but it seems clean enough to do so. Have you tested it against some 
other

modems with these changes applied ?


It has been tested against SIM900 (SIMCom) and M66 (Quectel).

(I'm always wary of patches to this going in without testing on 
actual

modems, because the spec is complex and we are not the only ones with
bugs)


While second and third patch don't change anything for initiator mode 
mux,

others certainly do.


Also can you please cc these patches to
xinhuix@intel.com


This address got rejected. Yours and LKML's didn't.


-Andrej

Re: [PATCH 0/8] tty: n_gsm: Make mux work as a responder station

[Andrej Krpic]

On 21.02.2016 23:42, One Thousand Gnomes wrote:

On Sun, 21 Feb 2016 22:38:29 +0100
Andrej Krpic  wrote:

When using n_gsm you have to explicitly set it to work as a 
initiator
station. This led me to believe that it can also work as a 
responder.

snip
This looks reasonable to me. It was never intended to work as a 
responder
but it seems clean enough to do so. Have you tested it against some 
other

modems with these changes applied ?


It has been tested against SIM900 (SIMCom) and M66 (Quectel).

(I'm always wary of patches to this going in without testing on 
actual

modems, because the spec is complex and we are not the only ones with
bugs)


While second and third patch don't change anything for initiator mode 
mux,

others certainly do.


Also can you please cc these patches to
xinhuix@intel.com


This address got rejected. Yours and LKML's didn't.


-Andrej

Re: [PATCH 0/8] tty: n_gsm: Make mux work as a responder station *DUPLICATE*

[Andrej Krpic]

Please ignore this duplicate thread.

Re: [PATCH 0/8] tty: n_gsm: Make mux work as a responder station *DUPLICATE*

[Andrej Krpic]

Please ignore this duplicate thread.

Re: [PATCH 0/8] tty: n_gsm: Make mux work as a responder station

[One Thousand Gnomes]

On Sun, 21 Feb 2016 22:38:29 +0100
Andrej Krpic  wrote:

> When using n_gsm you have to explicitly set it to work as a initiator
> station. This led me to believe that it can also work as a responder.
> 
> In a use case where I needed responder station mux I came to the
> conclusion that it actually does not work. Second and third patch
> fix dealings with frame C/R bit that then enable mux to work as a
> responder station.
> 
> Next patches in the series then fix bugs that were found after two 
> instances of n_gsm connected over null-modem cable were used.
> 
> First patch fixes formatting errors, such as space before comma, and
> most of the warnings reported by the checkpatch script.

This looks reasonable to me. It was never intended to work as a responder
but it seems clean enough to do so. Have you tested it against some other
modems with these changes applied ?

(I'm always wary of patches to this going in without testing on actual
modems, because the spec is complex and we are not the only ones with
bugs)

Also can you please cc these patches to
xinhuix@intel.com

Thanks

Alan

Re: [PATCH 0/8] tty: n_gsm: Make mux work as a responder station

[One Thousand Gnomes]

On Sun, 21 Feb 2016 22:38:29 +0100
Andrej Krpic  wrote:

> When using n_gsm you have to explicitly set it to work as a initiator
> station. This led me to believe that it can also work as a responder.
> 
> In a use case where I needed responder station mux I came to the
> conclusion that it actually does not work. Second and third patch
> fix dealings with frame C/R bit that then enable mux to work as a
> responder station.
> 
> Next patches in the series then fix bugs that were found after two 
> instances of n_gsm connected over null-modem cable were used.
> 
> First patch fixes formatting errors, such as space before comma, and
> most of the warnings reported by the checkpatch script.

This looks reasonable to me. It was never intended to work as a responder
but it seems clean enough to do so. Have you tested it against some other
modems with these changes applied ?

(I'm always wary of patches to this going in without testing on actual
modems, because the spec is complex and we are not the only ones with
bugs)

Also can you please cc these patches to
xinhuix@intel.com

Thanks

Alan