[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Libreboot] Libreboot on 15.4" T60 - Some questions on inverters.

From: Diagon
Subject: Re: [Libreboot] Libreboot on 15.4" T60 - Some questions on inverters.
Date: Sun, 7 Feb 2016 17:25:20 -0800
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.5.1

On 01/29/2016 06:04 AM, Diagon wrote:
> So I have the 15.4" version of an intel integrated graphics T60 that I'd
> like to see if I can get libreboot working on.  I gather that the screen
> it comes with:
> Samsung LTN154X3-L0A (15.4" 1280x800)
> will not work, and from the comments on the compatibility list (along
> with an IRC chat with @pehjota), I gather that the problem is with the
> LCD inverter.  That the machine will work with libreboot and even with
> an external monitor, but that initialization of the LCD will fail due to
> some problem initializing the inverter.  (Please correct me if I'm wrong
> here.)

It does appear that the inverter is the issue - hopefully the only one.
 In a discussion with @icon on #coreboot IRC, the suggestion has been
that the problem may be in the Pulse Width Modulation settings.  He said
that this is very old code that people haven't paid much attention to
for some time.  These PWM values are set in the coreboot source at
src/northbridge/intel/i945/gma.c (search for BLC_PWM_CTL in that file).
 The actual number is give in src/mainboard/lenovo/t60/devicetree.cb
(search for gpu_backlight), and has value:


It is similar in the code for the X60 and for the Macbook 2,1, but the
value is 0x01290128, and that 16th bit is a flag.  Also, the X60 started
out with the same value as the T60, but was changed by a later commit.
So it appears the T60 value may be an error.

In any case, by reading the value of BLC_PWM_CTL using the inteltool
(also in the source tree), that value being located at address 0x061254,
I find that my 15.4" T60 has value:


(Notice that 16th bit is set.)  So it may be that by using this value
for BLC_PWM_CTL, I may be able to have a working libreboot.  If there is
anyone with a 14.1" or 15.1" T60 running the original BIOS who would be
willing to run "inteltool -f", I would very much like to know if their
BLC_PWM_CTL is indeed 0x01290128.  Do note that inteltool crashed when I
tried it.  I had to get into the code (which is very short) and change
the loop "for (i=1; i <= MMIO_SIZE; ...)" to "for (i= 0x61254; i ==
0x611254; ...)" so that only one register was read.  For some reason the
machine didn't like all of them being read.

> Also listed on the website are 4 screens that are "untested" and whose
> status is unknown/untested:
> Samsung LTN154P2-L05 (42X4641 42T0329) (15.4" 1680x1050)
> LG-Philips LP154W02-TL10 (13N7020 42T0423) (15.4" 1680x1050)
> LG-Philips LP154WU1-TLB1 (42T0361) (15.4" 1920x1200)
> Samsung LTN154U2-L05 (42T0408 42T0574) (15.4" 1920x1200)
>    (for T61p but it might work in T60. Unknown!)
>    [Both of these are now known, see next...]
> Since the inverter appears to be the issue, I spent some time on
> talking to an old timer to ask if any of the
> available 15.4" screens might have different inverters, in the hope that
> one might work.  The thread is here:
> The upshot is that: (1) any of these screens can be used in that T60.
> Indeed, screens can be freely switched between any of the T6x/p's.  (2)
> All 15.4" screens use the same inverter made by Foxconn, FRU #41W1338.
> So that sounds dismal, but there is one ray of hope (maybe??).  There
> are three other inverters that will also work for all of these screens
> (including the original Samsung LTN154X3).  These are:
> 13R1014/13R1015 made by Sumida.
> 42T0078/42T0079 made by Hitachi, also has marking INVC753.
> 41W1012/39T5658 made by Hitachi, also has marking INVC753.
> I have found the first two of these on ebay (use the FRU to search).
> So my questions:
> (a) Does anyone have a clue whether any of these 3 inverters might work?
> (b) What exactly is the problem the inverter introduces?  The EDID is on
> the screen, as I understand, so what logic resides on the inverter?
> I do have one more question.  If an alternate inverter turns out not to
> be a choice, would coreboot (as opposed to libreboot) work on this
> machine?  I know that it will work on a T60p, but that's with an ATI
> blob, and I gather that blob initializes the LCD, including the
> inverter.  With this Intel T60, there's no blob, so I'm guessing there's
> no way to get coreboot to run (?)
> Thanks!
> /D
> PS. I am pretty confident of my informant at, as you may
> also be if you look here:

reply via email to

[Prev in Thread] Current Thread [Next in Thread]