NexusHD2 Android Forum

First off I wish to thank kokotas for his efforts with both the extended TWRP and Black cLK bootloader projects. I have been a user of MAGLDR in combo with CWR, having found previous versions of cLK to be lacking in stability. NO more!

I've been using both of the latest versions of kokotas' efforts for a week now (see edit below) and I must say I'm impressed. For the first time I truly feel like my phone is a native Android. It boots quickly and I haven't had a dropped connection of any type yet. Kudos for your work, especially with Black cLK. I've been having all kinds of issues with dieing nand blocks, with most in residing in the user data area (logical since this is where most of the reading and writing takes place). I love how Black cLK when combined with extended TWRP can check for and mark bad blocks (Black cLK) and check system file integrity at boot (extended TWRP). The file checker can be set to launch at "X" number of boots with the default being 16. Since I'm experiencing so many block failures I've adjusted down to every 10 reboots. It takes slightly longer to boot-up when the check is going on, but worth it IMHO. Besides I own no land line anymore so my phone is seldom rebooted (like when updated ROMs).

In fact, I'm now down to only 937megs of usable nand (without using the 24 hidden "extra" megs). So a few days ago I decided to convert to a data on SD-ext4 install. The transition went rather smoothly though it took me a while to figure out that I had to use the included file manager to move my saved CWR recovery image so my backed up data could be read by TWRP.

@kokotas: Since you don't have have a thread for Black cLK I'm going to ask you a couple of questions about it here.
First, what are the advantages or disadvantages to having cLK or internal control of "off-mode" charging and why all of the voltage adjustments?
Second, is there some way to manually increase font size for cLK (or have them enlarged with future versions)? I'm 50 and have been growing far-sighted over the last 6-7 years. Even with my reading glasses on, my arms just aren't long enough to read the menus on cLK or MAGLDR anymore lol. TIA and keep up the great work!!!

Ciao

Edited on 7/27/2013 @7:59pm EDT to comment about extended TWRP's file checking capability.

Kokotas has released a new updated version of his "EXTENDED TWRP |v2.6.1.0". It can be found on this page over at xda http://forum.xda-developers.com/showthr ... ?t=2120297

A special thanks to kokotas for his continuing efforts in keeping this open sourced recovery project alive and up-to-date for us HD2 owners.

Hi kokotas, thank you for your continuing support for our hd2.
Reading the changelog for 2.6.1.0, i noticed "* Proper detection of bootloader(cLK/MAGLDR/haret)", I was wondering how this was achieved. There have been problems in some rom's aroma installer detecting bootloader, perhaps your method might help. I could not think of a way to do this myself
Best regards
Robbie

Edit; I was using 2.6.1.0 SD version to try to install sense5 on NativeSD using magldr. Unfortunately, looking at the log, the bootloader was detected as HaRET Have attached log.
Also, an error appeared in the log about executing binary updater in the zip (before end of log). Does this mean the aroma binaries need updating or something?
Thanks again

Hi Robbie P,

Robbie P wrote:Reading the changelog for 2.6.1.0, i noticed "* Proper detection of bootloader(cLK/MAGLDR/haret)", I was wondering how this was achieved.

The code responsible for the bootloader detection is using the proc/cmdline:

Code: Select all

int DataManager::Detect_BLDR() {
   if (BLDR == -1) {
      string result;
      string cmd = "grep -Fxq \"clk=\" /proc/cmdline";

      if (TWFunc::Exec_Cmd(cmd, result) == 0) {
         BLDR = 1; // cLK
         SetValue("tw_bldr_is_clk", 1);
      } else {
         SetValue("tw_bldr_is_clk", 0);
         cmd = "grep -Fxq \"rel_path=\" /proc/cmdline";
         if (TWFunc::Exec_Cmd(cmd, result) == 0)            
            BLDR = 0; // Haret
         else {
            BLDR = 2; // Magldr
            SetValue("tw_bldr_is_magldr", 1);
         }
      }
   }
   return BLDR;
}

Robbie P wrote:Edit; I was using 2.6.1.0 SD version to try to install sense5 on NativeSD using magldr. Unfortunately, looking at the log, the bootloader was detected as HaRET Have attached log.

Could you also provide the contents of "proc/cmdline" after booting into the 2.6.1.0 SD version?

Robbie P wrote:Also, an error appeared in the log about executing binary updater in the zip (before end of log). Does this mean the aroma binaries need updating or something?

Does the script request the device to reboot at the end?

Thanks for the feedback! It's been a while since anyone reported anything...
Regards!

--off-topic--

Odysseus wrote:First, what are the advantages or disadvantages to having cLK or internal control of "off-mode" charging and why all of the voltage adjustments?
Second, is there some way to manually increase font size for cLK (or have them enlarged with future versions)? I'm 50 and have been growing far-sighted over the last 6-7 years. Even with my reading glasses on, my arms just aren't long enough to read the menus on cLK or MAGLDR anymore lol.

The whole cLK project was/is a learning experience for me. The built-in off-mode-charging was more of a proof of concept hence all those voltage adjustments for experimenting. However it made cLK autonomous regarding the off-mode-charging issue: no need for a recovery that has the correct setup(i.e. someone could use 4EXT with cLK and still be able to charge his device while turned off).
As for the font size, it's not currently possible to change it through cLK itself. We must have a larger font compiled into it.

Hi kokotas, /proc/cmdline has "rel_path=TWRP2" in it after booting up sd version from magldr.
I have tried re-flashing the same zip and I now get a "successful" installation, so no idea what the earlier error was for.I'm pretty sure aroma asked for a reboot which i declined and then found the error on the twerp screen and then saved log, if that helps.
Best regards and thanks again

Robbie P wrote:Hi kokotas, /proc/cmdline has "rel_path=TWRP2" in it after booting up sd version from magldr.
I have tried re-flashing the same zip and I now get a "successful" installation, so no idea what the earlier error was for.I'm pretty sure aroma asked for a reboot which i declined and then found the error on the twerp screen and then saved log, if that helps.
Best regards and thanks again

Robbie I've had similar issues with my setup using Black LK too. It seems that occasionally and for no apparent reason after the aroma installer completes what looks like a successful install TWRP will output the install failed. I have observed this behavior with both 2.6.0.0 and 2.6.1.0 versions. (I've never tried any other versions as I used MAGLDR and CMR before I made the switch a couple of months ago.) In every instance when I flashed the ROM again it did so without errors. Go figure?

On the other hand I've had issues pertaining to partition sizing with Black LK / extended-TWRP that I never had with MAGLDR and CMR. That is, if I make the Boot and or recovery partition(s) too large it will fail to boot-up or be recognized internally by Black LK. I noticed this because I wanted to increase the size of recovery since every iteration of TWRP seems to grow larger than the last and I wanted to get ahead of the curve and "future proof" my install. I was aiming to create a 15MB recovery and a 10MB boot, however, when I did the ROM would appear to install correctly but wouldn't boot up, instead it would reboot straight into Black LK with it reporting KERNEL @ boot (INVALID).

What I discovered is that if I made boot larger than 7MB it fails to be seen by Black LK even though the installation appears to be successful, it fails to boot. The same holds true if I make recovery larger than 12MB. After much experimenting I found that the largest I could make the two partitions and have my device boot-up correctly was 12MB recovery and 7MB boot.

Now on MAGLDR I can set boot to 10MB with no problems. I just find it odd that I can't seem to be able to do so with Black LK.

kokotas wrote:Thanks for the feedback! It's been a while since anyone reported anything...

Like London Buses, you wait for ages, then two come along at the same time

Robbie P wrote:Like London Buses, you wait for ages, then two come along at the same time

The buses are like that in London too? I thought this was only a San Francisco phenomenon. Public transportation is quite good in the Bay Area (unlike here in South Florida), especially so in the city of San Francisco proper, however, if you were in a hurry or stuck somewhere it never failed that after a period of extended delay, two (or sometimes three) would show up within a minute of each other. Also it seems that if no bus was in sight, as soon as you or someone else waiting lights up a cigarette, one appears from out of nowhere not leaving enough time to finish it. Go figure? lol

Robbie P wrote:Hi kokotas, /proc/cmdline has "rel_path=TWRP2" in it after booting up sd version from magldr.

So, afaiu, Magldr automatically adds that arg to the cmdline, right? And all this time I thought only Haret was using that arg... Ok then, maybe there is a way to use the detected partitions.
Since I don't use Magldr, could you help me by running a couple of cmds after booting twrp_sd and posting their outputs?

cat /proc/partitions
df
mount

Robbie P wrote:Like London Buses, you wait for ages, then two come along at the same time

I thought that was a greek-only privilege:) However, Odysseus is right about the cigarette trick:):)

Odysseus wrote:Robbie I've had similar issues with my setup using Black LK too. It seems that occasionally and for no apparent reason after the aroma installer completes what looks like a successful install TWRP will output the install failed. I have observed this behavior with both 2.6.0.0 and 2.6.1.0 versions. (I've never tried any other versions as I used MAGLDR and CMR before I made the switch a couple of months ago.) In every instance when I flashed the ROM again it did so without errors. Go figure?

I had occasionally seen the "failure" message after installing aroma based packages which at the end asked for a reboot. That's why I asked Robbie about it. In any case I have not found the cause of that behavior... But it doesn't seem to affect the installation process!

Regards!

cat /proc/partitions
major minor #blocks name
179___0____1921024_mmcblk0
179___1_____494560_mmcblk0p1
179___2____1426432_mmcblk0p2

df
Filesystem----------------1K-blocks---------used------availab....
tmpfs----------------------210056-------------48---------2100......
/dev/block/mmcblk0p1-493572---------434040---------595......
/dev/block/mmcblk0p1-493572---------434040---------595.....

mount
rootfs on / type footfs (rw)
tmpfs on /dev type tmpfs (rw,nosuid,relatime.mod........
devpts on /dev/pts type devpts (rw,relatime,mode........
proc on /proc type proc (rw, relatime)
sysfs on /sys type sysfs (rw,relatime)
/dev/block/mmcblk0p1 on /sdcard type vfat (rw,re........
/dev/block/mmcblk0p1 on /and-sec type vfat (rw,r........

The dots on the end of the lines are where the lines went offscreen and i couldn't read them. I tried to echo to a txt file but i am a noob, so had to type the above.
BTW in terminal, you have to delete the last line before you can enter a new one.
Thanks
Edit; went to stock wm6.5, now have

cat /proc/partitions
major minor #blocks name
31____0_____524288_mtdblock0
179___0____1921024_mmcblk0
179___1_____494560_mmcblk0p1
179___2____1426432_mmcblk0p2

df
Filesystem----------------1K-blocks---------used------availab....
tmpfs----------------------210056-------------48---------2100......
/dev/block/mmcblk0p1-493572---------434044---------595......
/dev/block/mmcblk0p1-493572---------434044---------595.....

mount
rootfs on / type footfs (rw)
tmpfs on /dev type tmpfs (rw,nosuid,relatime.mod........
devpts on /dev/pts type devpts (rw,relatime,mode........
proc on /proc type proc (rw, relatime)
sysfs on /sys type sysfs (rw,relatime)
/dev/block/mmcblk0p1 on /sdcard type vfat (rw,re........
/dev/block/mmcblk0p1 on /and-sec type vfat (rw,r........
rtname=mixed.errors=remount-ro)

the red colour is the diff, the blue colour is a line i found in wm6.5, but i am sure was also in magldr (just didn't copy it)
so it looks like wm6.5 has an extra partition, i would have thought magldr had one, but there you go.