Firebox LCD Driver for LCDProc
-
Oh yes, sorry. We only tried the driver on 3 XTM5 boxes… My friend tried to compile in 32bits. You will have to test it. :)
https://transfer.sh/10V6e3/sdeclcd.so
Please let us know if it is working.
Thank you my friend! is working with this file!! ;)
thanks again!
-
Hi,
A friend recompiled the driver. Here is the translated procedure from french.Instruction mod lcdproc :
Stop LCDproc service
Replace sdeclcd.so in /usr/local/lib/lcdproc
sdeclcd.so here : https://transfer.sh/t1Cjm/sdeclcd.so
I made mv /usr/local/lib/lcdproc/sdeclcd.so /usr/local/lib/lcdproc/sdeclcd.so.bak
then curl -O “https://transfer.sh/t1Cjm/sdeclcd.so”
and chmod +x the new sdeclcd.so file.
Then you restard LCDproc service and it runs with backlight without the timeoutWe tried on our XTM5 but it should work with other boxes.
Hey, was wondering if you could reupload the modified driver, it looks like the link may have went stale. Hoping to get my backlight to stay on for my XTM 5 running 2.3.3 x64 Thanks!
-
Hey, was wondering if you could reupload the modified driver, it looks like the link may have went stale. Hoping to get my backlight to stay on for my XTM 5 running 2.3.3 x64 Thanks!
I believe that file you reference is for 32bit..
Non the less the new LCDproc package allows the option in "Settings" to keep the back light on.
-
Hi,
A friend recompiled the driver. Here is the translated procedure from french.Instruction mod lcdproc :
Stop LCDproc service
Replace sdeclcd.so in /usr/local/lib/lcdproc
sdeclcd.so here : https://transfer.sh/t1Cjm/sdeclcd.so
I made mv /usr/local/lib/lcdproc/sdeclcd.so /usr/local/lib/lcdproc/sdeclcd.so.bak
then curl -O “https://transfer.sh/t1Cjm/sdeclcd.so”
and chmod +x the new sdeclcd.so file.
Then you restard LCDproc service and it runs with backlight without the timeoutWe tried on our XTM5 but it should work with other boxes.
Hey, was wondering if you could reupload the modified driver, it looks like the link may have went stale. Hoping to get my backlight to stay on for my XTM 5 running 2.3.3 x64 Thanks!
yes I too would really like this… LCDproc has an option to keep the backlight on but this driver ignores it... only the ancient SDEC LCD displays used EL backlights, there is no reason to turn the backlight off on a modern watchguard like an XTM that has an LED backlight... factory firmware keeps it on 24/7
-
there is no reason to turn the backlight off on a modern watchguard like an XTM that has an LED backlight… factory firmware keeps it on 24/7
If have converted many XTM 5 series , and i have never see one of then that keeps the LED backlight on with the factory firmware (BIOS).
Grtz
DeLorean -
I manage a ton of XTMs with actual watchguard software I assure you the backlight stays on 24/7
it's lcdproc that turns it off in the driver
-
That's true, it's the driver that turns off the back light and not the BIOS.
It would not be that hard to recompile the driver with the timer edited out if you really need that. I personally prefer it going out. If I see the LCD is lit that means the driver has crashed out for some reason.
Steve
-
Greetings all,
I've got an XTM5 box (2.3.4 amd64) and I would like to keep the LED display backlight permanently ON.
If one of you has the modified sdeclcd.so file that removes the backlight timeout code, please post the file or post a link to the file.
Thank you!
-
Has someone gathered the DATA on this issue? I know the original spec I coded the driver from mentioned a low number of hours for the life of the backlight. Some of you seem to suggest that the technology has evolved over time, and that automatic time out should no longer apply. Maybe different hardware revisions of these Watchguard/Lanner boxes used backlights that could stay on on a continuous basis. Maybe the problem is only with the earliest boxes, and few people use those anymore. I am willing to provide a code fix (and work it upstream through lcdproc.org) to implement a parameter in the config file, if applicable, but I would like to know the hard facts. Is it a matter of opening the case and identifying the LCD manufacturer, and, from there, locate a suitable spec sheet?
I coded the driver as per this SDEC spec sheet: http://www.ktechs.net/pvt_filz/lcdproc/LMC-S2D20-01.pdf
Let me know…
-
I manage a ton of XTMs with actual watchguard software I assure you the backlight stays on 24/7
it's lcdproc that turns it off in the driver
That's true, it's the driver that turns off the back light and not the BIOS.
It would not be that hard to recompile the driver with the timer edited out if you really need that. I personally prefer it going out. If I see the LCD is lit that means the driver has crashed out for some reason.
Steve
You guys where 100% right.
I never used these boxes with the Watchguard software, only for a powerup test after receiving it for verifying
if the listed modelnr corresponds with the seller information.
Besides that, these boxes where immediately converted to pfSense, and that's why i never saw the LCD screens keeps on. ::)Grtz
DeLorean -
OK, so… if it's the driver... is there a way to include a command to NOT turn it off or to change the timer value? Perhaps being able to set it to zero to keep it on, or to a number of seconds before it shuts off.
-
It would be nice if the option to keep on the LCD screen under settings was working.
So that you can decide if the LCD screen backlight must be kept on or off.Grtz
DeLorean -
That's true. Since there are options set in the settings interface for LCDProc, I wonder if there isn't some way for the file to disregard it's internal clock for killing the backlight and just allow the settings to override it.
Where is the latest code source for the program? I don't know if I can properly write the changes, but I'll try.
-
Go to the source! ;D
http://lcdproc.cvs.sourceforge.net/viewvc/lcdproc/lcdproc/server/drivers/sdeclcd.c?view=markup
But yeah a code revision to enable or disable the backlight timer from the driver config would be appreciated by many. There can;t be that many people running the original X-Core models now.
Steve
-
It's here, now: https://github.com/lcdproc/lcdproc
So, this backlight timeout is only useful for the old X-Core boxes? What about the X-Core-e (x550e, x750e)? X-Peak?
-
The X-Peak may have a limited life but the display is different hardware. My X-Peak died of bad caps a while back anyway.
The X-e and newer models are all led backlight I believe so no problems.
I would not want to have it on all the time personally but a choice to do so seems reasonable.
Steve
-
–EDITED POST--
This might be the wrong way to disable the backlight off toggle because I am not a driver coder, but I have edited/compiled a sdeclcd.so that has "no timeout" on FreeBSD 10.3 (32 bit and 64 bit) and FreeBSD 11.1 (64 bit) virtual machines
So far the 32 bit works great on my X750e and the 64 bit works on XTM5 according to networkBob
Here's how I proceeded (from scratch on a new install of FreeBSD)
pkg install automake autoconf gcc git nano cd ~ git clone http://github.com/lcdproc/lcdproc cd lcdproc nano server/drivers/sdeclcd.c ----- change BACKLIGHT_OFF to BACKLIGHT_ON on lines 650 and 676 then save CTRL-X, Y, ENTER sh ./autogen.sh ./configure --enable-drivers=sdeclcd make serverThis will generate ~/lcdproc/server/drivers/sdeclcd.so which you can copy somewhere in order to download it to your pfSense box by FTP or whatever method you prefer.
If you don't want to compile it yourself, you can grab the ones I have compiled at the following links or as an attachment to this post
pfSense 2.3 (FreeBSD 10.3)
- 32 bit version : http://www.vizi0n.com/watchguard/sdeclcd.so.32bit.zip
- 64 bit version : http://www.vizi0n.com/watchguard/sdeclcd.so.64bit.zip
pfSense 2.4 (FreeBSD 11.1)
- 64 bit version : http://www.vizi0n.com/watchguard/sdeclcd.so.freebsd11.zip
- 32 bit version : none, pfSense has no 32 bit release since 2.4
You need to extract the zip file's content and move it it /usr/local/lib/lcdproc/
Then make sure that you run "chmod +x sdeclcd.so" to make it executableFeel free to edit my post if the method I haved used is not the proper way to bypass that timeout in the driver.
sdeclcd.so.64bit.zip
sdeclcd.so.32bit.zip
sdeclcd.so.64bit.freebsd11.zip -
I am not a driver coder, this might be the wrong way, but I have compiled a sdeclcd.so that has "no timeout" on a i386 FreeBSD 10.3 virtual machine.
So far it works great on my x750e.
All I did was edit "sdeclcd.c" and replace "BACKLIGHT_OFF" to "BACKLIGHT_ON" on lines 650 and 676
Hope this works for you
You can grab it here : http://www.vizi0n.com/watchguard/sdeclcd.so or as an attachment to this post
Feel free to edit my post if the method I haved used is not the proper way to remove that timeout.
That made LCDProc fail on startup. If I went into the services status, it showed LCDproc not started, and when I tried to start, it just failed.
Error in the system log: Could not open driver module /usr/local/lib/lcdproc/sdeclcd.so: /usr/local/lib/lcdproc/sdeclcd.so: unsupported file layout
But I'm happy that you tried! Thank you for that.
-
On 32bit? ;)
-
I am not a driver coder, this might be the wrong way, but I have compiled a sdeclcd.so that has "no timeout" on a i386 FreeBSD 10.3 virtual machine.
So far it works great on my x750e.
All I did was edit "sdeclcd.c" and replace "BACKLIGHT_OFF" to "BACKLIGHT_ON" on lines 650 and 676
Hope this works for you
You can grab it here : http://www.vizi0n.com/watchguard/sdeclcd.so or as an attachment to this post
Feel free to edit my post if the method I haved used is not the proper way to remove that timeout.
Thank you so much for the effort, vizi0n. However, this did not work on my 64-bit XTM515 box.
As noted by dhoffman98, LCDproc failed to start after replacing the release sdeclcd.so file with your modified version. Maybe it has to do with the 64-bit architecture vs 32-bit. I do not possess the required knowledge to know.
I tried messing around with online C compilers but got super frustrated trying to find a way to compile to a .so file. It seems to me that eliminating the timeout would be more complicated than, say, increasing the timeout from 30 seconds to 300,000 seconds, although that would add a tiny amount of processor overhead.
