Wacom digitizer and antix boot – [solved]

Forum Forums Official Releases antiX-17 “Heather Heyer” Wacom digitizer and antix boot – [solved]

This topic contains 54 replies, has 8 voices, and was last updated by stevesr0 Dec 3-4:34 pm.

Viewing 15 posts - 1 through 15 (of 55 total)
  • Author
    Posts
  • #10634
    Member
    Avatar
    stevesr0

    I have liveUSBs of antix 16 and antix17.
    (See also my thread about a slow boot with anti17 on my Fujitsu laptop.)

    Antix 16 doesn’t appear to “recognize” the Wacom digitizer on my Fujitsu T4220 laptop, but antix17 does.

    Antix16 boots in 24 seconds and 17 boots in 5 minutes.

    Is the delayed boot caused by the software “trying” to setup the digitizer? In the /var/log Xorg.log file, Wacom entries are listed at 68 seconds and then at the next entry at 318 seconds.

    Is there a way to disable the setting up of the digitizer to test this?

    Is there a way to speed up the setup of the digitizer?

    • This topic was modified 11 months, 3 weeks ago by rokytnji.
    • This topic was modified 10 months, 2 weeks ago by anticapitalista.
    #10462
    Member
    Avatar
    stevesr0

    I didn’t see this question discussed, so I am posting it in hopes of personal enlightenment.

    I made two live-USBs of the same downloaded copy of 17.1, using the same program (Rufus).

    One USB (Sandisk) will boot both machines, but the boot screen is somewhat different. On one machine (a Fuitsu dual core with 4 gig of RAM) it shows function keys, but no text option. On the other machine (Lenovo with a quad core celeron and 8 gig of RAM), it shows a “full” option, a customize option and a text option without any function keys.

    The other USB (Lexar) only boots on the Lenovo.

    The boot time on the Fujitsu is prolonged by a 5 minnute “gap” with the screen blank, compared to about a 5 second gap with the Lenovo.

    Final difference: When the antiX screen launches and a the amount of RAM and /disk size is displayed, the Lenovo shows 7.68 gig of RAM vs 3.85 gig in the Fujitsu shows 3.85 and 3.05 respectively. While the RAM differences match the actual RAM on the two computers, I am unclear why the /disk is different in size.

    Once launched, antiX works fine on both machines. On the older Fuitsu, the somewhat larger screen makes it easier to see.

    **Last point — the Fuitsu is a BIOS/MBR system, while the Lenovo is a UEFI/GPT system.

    • This reply was modified 12 months ago by stevesr0.
    • This reply was modified 11 months, 3 weeks ago by rokytnji.
    #10466
    Forum Admin
    anticapitalista
    anticapitalista

    The last point gives you the answer

    Philosophers have interpreted the world in many ways; the point is to change it.

    #10470
    Member
    Avatar
    stevesr0

    Hi anticapitalista,

    Thanks for response.

    So, all of those differences are attributable to BIOS vs UEFI and/or MBR vs GPT. I was hoping there were specific adjustments I could make in a persistent liveUSB that would at least reduce the 5 minute blank period, just before the desktop appears.

    Do you have any thoughts on whether one of those differences (BIOS/UEFI or MBR/GPT) might be mainly responsible.

    Rufus formats the USB so that it is (allegedly) compatible with either BIOS or UEFI system.

    I am planning to use GPT partitioning on this system after I empty the drive. That will be a BIOS/GPT combination, (unless that is likely to produce a problem installing or running antiX.

    I will look for a site that explains how the boot can be so affected by these differences.

    Steve

    #10471
    Forum Admin
    dolphin_oracle
    dolphin_oracle

    the difference in the boot menus is because on the live-USB the bios/mbr boot uses syslinux and the UEFI boot uses grub2. An installed system will always use grub2. the “customize” option replaces the function keys on the mbr/bios boot. They functionally do the same thing.

    As to why you have a 5 minute delay on the fujitisu, that is a good question. perhaps running with the bootchart option enabled will shed some light. I think there is a live boot log under /var/log as well that might be beneficial in this case.

    As long as your PC’s bios can boot a gpt partition table, you should be OK with that combination, although its not 100% and its hardware dependent. The mbr still needs written to the disk on a bios boot, but gpt is supposed to have a “protected mbr” area, that while not a real mbr, will serve the function. I’ve got a least one PC that will not boot bios/mbr on a gpt disk, although that PC is very very old.

    #10474
    Member
    Avatar
    stevesr0

    Thanks, dolphin_oracle, That is enlightening <g>.

    I will try running the bootchart option, the next time I run the liveUSB.

    Since GRUB2 is the bootloader for installs and for UEFI/GPT platforms, why is syslinux used in the BIOS/MBR liveUSB?

    I guess syslinux would be used to boot a BIOS/GPT system with a liveUSB? Perhaps bootchart will tell me.

    I will report back if I have anything I think might be useful.

    Steve

    P.S. I have/am enjoyed/enjoying your videos.

    #10475
    Forum Admin
    dolphin_oracle
    dolphin_oracle

    syslinux would be used for live-boot on MBR systems, that is correct. the partition table doesn’t matter a whole lot as long as your system can boot from a gpt’s protected MBR.

    I don’t know the exact history of why syslinux is used, but its used on a lot of distros for iso and live-USB boots. the antiX graphical boot menus were developed before grub2/UEFI came along, so those are kept where they work, and grub2 was tacked on later for UEFI boots.

    #10476
    Member
    Avatar
    stevesr0

    Thanks for the additional information.

    The system I am going to wipe off has become encrusted with several bootloaders chained together over the years for different distros.

    I have grub4dos, syslinux and grub2 (and perhaps one other one that I can’t bring to mind at the moment). Don’t currently have LILO on this system.

    I am looking forward to a simpler setup.

    Steve

    #10487
    Member
    Avatar
    stevesr0

    I used the bootchart option and have two image files (one for each computer). According to bootchart, the older BIOS computer took 319 seconds and the newer Lenovo took 55 seconds.

    After looking over the two files, I realize I don’t know how to use them to figure out where the delay is (or the multiple points of delay), and how to “fix” it (them) if possible.

    So, I am attaching them to this post and would appreciate anyone’s comments.

    Thanks in advance.

    Steve

    #10529
    Member
    Avatar
    stevesr0

    Well, after looking back and forth at the two charts for several days, it appears to me that avahi-daemon is involved in launching the desktop-session (along with slim) on my Lenovo which boots in 54 seconds, but terminates well before the launch of antix on my Fuitsu (319 seconds).

    I am not sure if this suggests a fix (or a bug when booting BIOS/MBR based systems).

    I will check if it is possible to stop this from running during boot.

    Appreciate any comments…

    Thanks in advance.

    Steve

    • This reply was modified 11 months, 4 weeks ago by stevesr0.
    • This reply was modified 11 months, 3 weeks ago by rokytnji.
    #10535
    Member
    Avatar
    crestiah

    the 5 minute blank screen would drive me barmy.

    questions
    what usb rating are the usb sticks and the ports they are plugged into? usb1, usb2, usb3, usb3.1, given the read, write, transfer rates this could make a bit of a difference especially if you are coupling them with a usb port rated for usb1 as an example.

    1 computer is a dual core but what is the cpu speed, the other a quad core whats its cpu speed? surely this would add a difference to the boot up time.

    i dont know much about the uefi/gpt thing, but the impression i get from this thing is that it acts more like a save state. the equivalent on a bios/mbr would be hibernate/sleep mode. at least this is how i look at it.

    i dont know anything about grub2 so not sure how that operates, grub4dos had some quite interesting ways of doing things though.

    i wouldnt think an “if then do” statement (ie if bios/mbr then boot syslinux but if gpt then boot uefi) would take upto 5 minutes to process, but i suppose it might if grub2 is geared more towards a uefi boot than a bios boot, maybe its taking 5 minutes to reprocess the commmands to be acceptible to bios boot?

    • This reply was modified 11 months, 4 weeks ago by crestiah.
    • This reply was modified 11 months, 3 weeks ago by rokytnji.
    #10537
    Member
    Avatar
    stevesr0

    Hi crestiah,

    Thanks for the comments. I copied your questions below, with my answers are preceded by **.

    Steve

    “what usb rating are the usb sticks and the ports they are plugged into? ”
    **My ports are all USB 2.0 and the USB sticks are USB 3.0.

    “1 computer is a dual core but what is the cpu speed.”
    **2 GHz.
    “the other a quad core whats its cpu speed?”
    **1.83 GHz.

    “i wouldnt think an “if then do” statement (ie if bios/mbr then boot syslinux but if gpt then boot uefi) would take upto 5 minutes to process, but i suppose it might if grub2 is geared more towards a uefi boot than a bios boot, maybe its taking 5 minutes to reprocess the commmands to be acceptible to bios boot?”
    **I doubt that, but an answer would have to come from someone familiar with the code.

    #10538
    Member
    Avatar
    stevesr0

    Just for kicks, I replaced 17 with 16.1 on one of the USB sticks. It boots in about 20 seconds on the Fujitsu machine that takes more than 5 minutes with 17.1.

    Does that suggest that the delay is due to something in particular that changed between the two releases?

    Hoping for some simple fix. I did a bootchart.png of that one too, but didn’t save it. If anyone wants to analyze it <g> I will generate a new one and upload it.

    Comments and suggestions welcomed!!

    Steve

    #10539
    Member
    Avatar
    crestiah

    hope this makes sense.
    Does that suggest that the delay is due to something in particular that changed between the two releases?

    based on what i read here
    https://stackoverflow.com/questions/10897087/multi-threaded-bios
    https://stackoverflow.com/questions/27949362/create-a-thread-in-uefi
    and a thread i read about the difference between dual core and quad core PC’s (sorry closed tab so cant point to it).

    but basically im going to make an assumption or 2:
    antix 16.1 probably uses uefi boot with single thread in mind. so in theory there would be next to no difference in the boot up process.( ie the quad core is probably only using 1-2 cores like the dual core in the inital boot process cores 3 and 4 sitting idle)

    antix 17 uefi boot possibly includes MpService protocol (for lack of a better term “task delegator or multi threader” utilizes or makes use of the extra cores more efficently) “probably what gives me the saved state affect/feeling”.

    so in the instance of the quad core
    core 1 = boot processs with multi thread delegation
    core 2 = start up hd/usb initialization
    core 3 = start up video display
    core 4 = any other components
    all run in parallel arriving at user-space quicker.

    in the case of the dual core
    core 1 = boot process with single thread delegation that may be split to run on core 2.
    core 2 = either idle or running a split of the single thread (Core 1 has to wait till core 2 finishes to move to next stage).

    if grub2 uefi is setup for multi thread delegation as its preffered first option then its probably trying to pipe multithread exec code through a single thread pipe, this could be done at the same speed, except the dual core processor would need to be at least twice the speed of the quad core.
    ie quad core 1.83ghz dual core 4ghz.

    “pipe multithread exec code through a single thread pipe” by this im thinking uefi(multithread)/gpt piped through syslinux bios/mbr.

    i would be curious to know if the syslinux component is booting with boot=quiet (usually means to me blank screen)

    #10540
    Member
    Avatar
    736b69646f6f

    pressing tab (or whatever is instructed on boot screen, idunno don’t have efi here) should let you see the full bootline parameters, to check whether quiet is specified

    The liveboot efi is probably blackbox, unchanged by antiX. We get whatever the upstream provides and debian packages.

    okay put on your sunglasses, here come a few bright ideas:
    Possibly an updated firmware is available from lenovo which would remedy the slow booting.
    Visit F2 bios setup and check whether you can manually choose device boot order, to “usb first, if present” or whatever option is available.
    Check /var/log/ or /var/log/live/ to see what error messages are being spewed to logfile during the 5 minute boot stall.
    Bootchart indicates “udev” is running for minutes. Struggling to find a driver? Bad/intermittent hardware errors?

    Something in the bootchart about set-console-width…. and same shown at right side of bootchart.
    It’s not a solid line like that process was running the whole time. It gets interrupted, suspended, while udev it doing something else???
    Far out possibility: typed a custom boot parameter, intended to type “conwidth=140” but typo, or key repeat, wound up with conwidth=1400 or something, and the init is struggling trying to satisfy requested out-of-range width value?

    If the boot device is liveusb, I don’t understand how grub2 enters the picture. AFAIK, syslinux provides the liveboot efi bootloader and I see nothing on the syslinux.org site mentioning ability to tweak the behavior for “threads”.

Viewing 15 posts - 1 through 15 (of 55 total)

You must be logged in to reply to this topic.