Lirc 0.9.4.d Misses first packet?

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

Lirc 0.9.4.d Misses first packet?

James A R Brown
HI,

I am working with a HP RC6 MCE remote with -H default -d /dev/lirc0.

Everything works, but if the remote is dormant for say 15 sec, it seems
to need a number of packets to "wake" up for irw to give me feedback.
Mode2 always gets.

I checked the toggle bit and it seems ok, once it is working it gets
every press. I then thought maybe the pulses were badly done by
irrecord, so generated some keys with IrScrutinizer and pronto2lirc,
followed by irrecord -a..... same result.

Then tried to renice the lircd daemon and irw. Same result.

Can anyone advise any ideas, feel I am missing something obvious.

James


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
Reply | Threaded
Open this post in threaded view
|

Re: Lirc 0.9.4.d Misses first packet?

Bengt Martensson-2
On 02/12/17 21:23, James A R Brown wrote:
> HI,
>
> I am working with a HP RC6 MCE remote with -H default -d /dev/lirc0.
>
> Everything works, but if the remote is dormant for say 15 sec, it seems
> to need a number of packets to "wake" up for irw to give me feedback.
> Mode2 always gets.

I do not understand the last sentence. Gets what? Is the program mode2
always responsive? You can use IrScrutinizer (version >= 1.3)) with
capturing device /dev/lirc (/dev/lirc0) to verify the function of
/dev/lirc and the connected hardware. Please do so. That will isolate
the problem to either /dev/lirc + kernel or to Lirc proper.

> I checked the toggle bit and it seems ok, once it is working it gets
> every press. I then thought maybe the pulses were badly done by
> irrecord, so generated some keys with IrScrutinizer and pronto2lirc,
> followed by irrecord -a..... same result.

If the problem is with Lirc proper, start lircd with highest logging
(lircd ---n --logfile=/tmp(lircd.log --loglevel=9 ... ) and try to
understand the log file. Or post it for Alec to analyze :-).


Greetz,

Bengt


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
Reply | Threaded
Open this post in threaded view
|

Re: Lirc 0.9.4.d Misses first packet?

James A R Brown
Please find attached logs as requested.

At first you can see the remote is working. I then echo a PAUSE>>> into the log and waited a bit. then it comes with a failed on header and space expected errors.

But then later you can see where i have quickly sent the button press twice, it starts to work again.

From the brief test on 0.9.0 I could not recreate, which I guess I could use, but the lircd.conf.d directory sure makes things simpler on 0.9.2 and above.

For most remotes with continuous repeat sending I don't think this would be noticeable, but came to light when using a B&O Beo5 remote, which does one transmission per button press only. Although I hasten to add, if I use the original HP MCE remote and ensure I press quickly for 1 transmission the same problem is there.

James

On Sun, Feb 12, 2017 at 9:14 PM, James A R Brown <[hidden email]> wrote:
Hi Bengt,

Sorry I mean mode2 always sees the first packet data / single key press, even after a period of not using, so I don't think its an issue with the receiver / remote.

IRW does not see the first packet data / single key press after a period of not using, but wakes if I keep repressing the key and once awake does not miss a press.

I have tried with IrScrutinizer on my laptop (Fedora25), away from the TV box, 1.3, 1.3dev, 1.1, appimage and bin, and I simply can not get it to grab itself with /dev/lirc0 or mode2. I have checked the lock groups, dialout, installed fedora rxtx, disabled selinux. For /dev/lirc0 I get a hardware exception (i forget the full exception).

In the end I was using mode2 -m and pasting the timings to get the remote key details.

I have just removed lirc0.9.4d on the TV box (Ubuntu 16.04LTS) and installed 0.9.0 ubuntu stock which seems to have fixed the issue.

I will do more testing and advise.

James




On 12/02/17 20:41, Bengt Martensson wrote:
On 02/12/17 21:23, James A R Brown wrote:
HI,

I am working with a HP RC6 MCE remote with -H default -d /dev/lirc0.

Everything works, but if the remote is dormant for say 15 sec, it seems
to need a number of packets to "wake" up for irw to give me feedback.
Mode2 always gets.
I do not understand the last sentence. Gets what? Is the program mode2
always responsive? You can use IrScrutinizer (version >= 1.3)) with
capturing device /dev/lirc (/dev/lirc0) to verify the function of
/dev/lirc and the connected hardware. Please do so. That will isolate
the problem to either /dev/lirc + kernel or to Lirc proper.

I checked the toggle bit and it seems ok, once it is working it gets
every press. I then thought maybe the pulses were badly done by
irrecord, so generated some keys with IrScrutinizer and pronto2lirc,
followed by irrecord -a..... same result.
If the problem is with Lirc proper, start lircd with highest logging
(lircd ---n --logfile=/tmp(lircd.log --loglevel=9 ... ) and try to
understand the log file. Or post it for Alec to analyze :-).


Greetz,

Bengt


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot



------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot

lirclog.txt (258K) Download Attachment
HP_MCE.conf (4K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Lirc 0.9.4.d Misses first packet?

James A R Brown
And here is a mode2 reception after a period of unuse, which I am pretty sure is good IR reception

      2750      800      500      400      450      450
      450      850      500      850     1350      850
      500      400      450      450      450      400
      500      400      450      450      450      450
      450      400      500      400      450      450
      450      400      950      400      450      450
      450      450      450      400      500      850
      500      400      450      400      500      400
      900      900      450      400      500      400
      450      450      450      400      950      400
      500      400      500      350      500      400
      500 16777215



On Mon, Feb 13, 2017 at 8:21 PM, James Brown <[hidden email]> wrote:
Please find attached logs as requested.

At first you can see the remote is working. I then echo a PAUSE>>> into the log and waited a bit. then it comes with a failed on header and space expected errors.

But then later you can see where i have quickly sent the button press twice, it starts to work again.

From the brief test on 0.9.0 I could not recreate, which I guess I could use, but the lircd.conf.d directory sure makes things simpler on 0.9.2 and above.

For most remotes with continuous repeat sending I don't think this would be noticeable, but came to light when using a B&O Beo5 remote, which does one transmission per button press only. Although I hasten to add, if I use the original HP MCE remote and ensure I press quickly for 1 transmission the same problem is there.

James

On Sun, Feb 12, 2017 at 9:14 PM, James A R Brown <[hidden email]> wrote:
Hi Bengt,

Sorry I mean mode2 always sees the first packet data / single key press, even after a period of not using, so I don't think its an issue with the receiver / remote.

IRW does not see the first packet data / single key press after a period of not using, but wakes if I keep repressing the key and once awake does not miss a press.

I have tried with IrScrutinizer on my laptop (Fedora25), away from the TV box, 1.3, 1.3dev, 1.1, appimage and bin, and I simply can not get it to grab itself with /dev/lirc0 or mode2. I have checked the lock groups, dialout, installed fedora rxtx, disabled selinux. For /dev/lirc0 I get a hardware exception (i forget the full exception).

In the end I was using mode2 -m and pasting the timings to get the remote key details.

I have just removed lirc0.9.4d on the TV box (Ubuntu 16.04LTS) and installed 0.9.0 ubuntu stock which seems to have fixed the issue.

I will do more testing and advise.

James




On 12/02/17 20:41, Bengt Martensson wrote:
On 02/12/17 21:23, James A R Brown wrote:
HI,

I am working with a HP RC6 MCE remote with -H default -d /dev/lirc0.

Everything works, but if the remote is dormant for say 15 sec, it seems
to need a number of packets to "wake" up for irw to give me feedback.
Mode2 always gets.
I do not understand the last sentence. Gets what? Is the program mode2
always responsive? You can use IrScrutinizer (version >= 1.3)) with
capturing device /dev/lirc (/dev/lirc0) to verify the function of
/dev/lirc and the connected hardware. Please do so. That will isolate
the problem to either /dev/lirc + kernel or to Lirc proper.

I checked the toggle bit and it seems ok, once it is working it gets
every press. I then thought maybe the pulses were badly done by
irrecord, so generated some keys with IrScrutinizer and pronto2lirc,
followed by irrecord -a..... same result.
If the problem is with Lirc proper, start lircd with highest logging
(lircd ---n --logfile=/tmp(lircd.log --loglevel=9 ... ) and try to
understand the log file. Or post it for Alec to analyze :-).


Greetz,

Bengt


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot




------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
Reply | Threaded
Open this post in threaded view
|

Re: Lirc 0.9.4.d Misses first packet?

Bengt Martensson-2
On 02/14/17 08:30, James Brown wrote:

> And here is a mode2 reception after a period of unuse, which I am pretty
> sure is good IR reception
>
>       2750      800      500      400      450      450
>       450      850      500      850     1350      850
>       500      400      450      450      450      400
>       500      400      450      450      450      450
>       450      400      500      400      450      450
>       450      400      950      400      450      450
>       450      450      450      400      500      850
>       500      400      450      400      500      400
>       900      900      450      400      500      400
>       450      450      450      400      950      400
>       500      400      500      350      500      400
>       500 16777215

This decodes in IrScrutinizer (paste it to "Scrutinize signal" and press
"Scrutinize", may like to select "Options -> ignoring ending silence")
as protocol = MCE, device = 4, subdevice = 15, obc = 31, misc = T=1.

In other words, this is fine.

>
> On Mon, Feb 13, 2017 at 8:21 PM, James Brown <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>     Please find attached logs as requested.
>
>     At first you can see the remote is working. I then echo a PAUSE>>>
>     into the log and waited a bit. then it comes with a failed on header
>     and space expected errors.

It appears as lircd get "out of phase". Possibly it is related to Linux'
non-interlaced spaces/marks (Alec's "favorite"). Please make sure that
you are using the current Git version (or at least the latest release),
and file a bug https://sourceforge.net/p/lirc/tickets/

Alec, any comments?

>         I have tried with IrScrutinizer on my laptop (Fedora25), away
>         from the TV box, 1.3, 1.3dev, 1.1, appimage and bin, and I
>         simply can not get it to grab itself with /dev/lirc0 or mode2. I
>         have checked the lock groups, dialout, installed fedora rxtx,
>         disabled selinux. For /dev/lirc0 I get a hardware exception (i
>         forget the full exception).

Should be fixed in the CI build.

Greetz,

Bengt


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
Reply | Threaded
Open this post in threaded view
|

Re: Lirc 0.9.4.d Misses first packet?

James A R Brown
Thanks Bengt.

Tested on 0.9.3a and that's OK.  Have filed a bug against 0.9.4d

Also tried to build git,  but that drops with config.h not found at first lib compile.  I can see Alec has done some commits on filepath,  so probably just temporary issue. 

Irsqrutinizer also working from ci now. 

On 14 Feb 2017 10:10, "Bengt Martensson" <[hidden email]> wrote:
On 02/14/17 08:30, James Brown wrote:
And here is a mode2 reception after a period of unuse, which I am pretty
sure is good IR reception

      2750      800      500      400      450      450
      450      850      500      850     1350      850
      500      400      450      450      450      400
      500      400      450      450      450      450
      450      400      500      400      450      450
      450      400      950      400      450      450
      450      450      450      400      500      850
      500      400      450      400      500      400
      900      900      450      400      500      400
      450      450      450      400      950      400
      500      400      500      350      500      400
      500 16777215

This decodes in IrScrutinizer (paste it to "Scrutinize signal" and press "Scrutinize", may like to select "Options -> ignoring ending silence") as protocol = MCE, device = 4, subdevice = 15, obc = 31, misc = T=1.

In other words, this is fine.


On Mon, Feb 13, 2017 at 8:21 PM, James Brown <[hidden email]
<mailto:[hidden email]>> wrote:

    Please find attached logs as requested.

    At first you can see the remote is working. I then echo a PAUSE>>>
    into the log and waited a bit. then it comes with a failed on header
    and space expected errors.

It appears as lircd get "out of phase". Possibly it is related to Linux' non-interlaced spaces/marks (Alec's "favorite"). Please make sure that you are using the current Git version (or at least the latest release), and file a bug https://sourceforge.net/p/lirc/tickets/

Alec, any comments?


        I have tried with IrScrutinizer on my laptop (Fedora25), away
        from the TV box, 1.3, 1.3dev, 1.1, appimage and bin, and I
        simply can not get it to grab itself with /dev/lirc0 or mode2. I
        have checked the lock groups, dialout, installed fedora rxtx,
        disabled selinux. For /dev/lirc0 I get a hardware exception (i
        forget the full exception).

Should be fixed in the CI build.

Greetz,

Bengt



------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot