Page 1 of 1
Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 1:19 pm
by br1anstorm
I have been trying to keep an ancient desktop computer operating, essentially as a 'spare' machine on which to experiment.
The computer has a single 160GB hard disk drive, obviously formatted legacy/BIOS. I think it originally had WinXP on it. I have reformatted the hard drive and managed to install the 32-bit version of MX23.5 on it.
The essential
inxi system information is as follows:
Code: Select all
System:
Kernel: 6.1.0-33-686-pae [6.1.133-1] arch: i686 bits: 32 compiler: gcc v: 12.2.0
parameters: BOOT_IMAGE=/boot/vmlinuz-6.1.0-33-686-pae root=UUID=<filter> ro quiet splash
Desktop: Xfce v: 4.20.0 tk: Gtk v: 3.24.38 info: xfce4-panel wm: xfwm v: 4.20.0 vt: 1
dm: LightDM v: 1.32.0 Distro: MX-23.6_386 Libretto Jan 12 2025 base: Debian GNU/Linux 12
(bookworm)
Machine:
Type: Desktop System: DIXONSXP product: N/A v: N/A serial: <superuser required>
Mobo: Foxconn model: P4M800P7MB serial: <superuser required> BIOS: Phoenix v: 6.00 PG
date: 09/21/2006
CPU:
Info: model: Intel Celeron D bits: 64 arch: Netburst Presler built: 2006 process: Intel 65nm
family: 0xF (15) model-id: 6 stepping: 4 microcode: 0x4
Topology: cpus: 1x cores: 1 smt: <unsupported> cache: L1: 16 KiB desc: d-1x16 KiB L2: 512 KiB
desc: 1x512 KiB
Speed (MHz): 3320 min/max: N/A core: 1: 3320 bogomips: 6639
Flags: ht lm nx pae sse sse2 sse3
Vulnerabilities:
Type: gather_data_sampling status: Not affected
Type: itlb_multihit status: KVM: VMX unsupported
Type: l1tf mitigation: PTE Inversion
Type: mds status: Vulnerable: Clear CPU buffers attempted, no microcode; SMT disabled
Type: meltdown mitigation: PTI
Type: mmio_stale_data status: Unknown: No mitigations
Type: reg_file_data_sampling status: Not affected
Type: retbleed status: Not affected
Type: spec_rstack_overflow status: Not affected
Type: spec_store_bypass status: Vulnerable
Type: spectre_v1 mitigation: usercopy/swapgs barriers and __user pointer sanitization
Type: spectre_v2 mitigation: Retpolines; STIBP: disabled; RSB filling; PBRSB-eIBRS: Not
affected; BHI: Not affected
Type: srbds status: Not affected
Type: tsx_async_abort status: Not affected
Graphics:
Device-1: VIA CN700/P4M800 Pro/P4M800 CE/VN800 Graphics [S3 UniChrome Pro] vendor: Foxconn
driver: N/A bus-ID: 01:00.0 chip-ID: 1106:3344 class-ID: 0300
Display: server: X.Org v: 1.21.1.7 compositor: xfwm v: 4.20.0 driver: X: loaded: vesa
dri: swrast gpu: N/A display-ID: :0.0 screens: 1
Screen-1: 0 s-res: 1280x1024 s-dpi: 95 s-size: 342x271mm (13.46x10.67") s-diag: 436mm (17.18")
Monitor-1: default res: 1280x1024 size: N/A modes: N/A
API: OpenGL v: 4.5 Mesa 22.3.6 renderer: llvmpipe (LLVM 15.0.6 128 bits) direct-render: Yes
Audio:
Device-1: VIA VT8233/A/8235/8237 AC97 Audio vendor: Foxconn driver: snd_via82xx v: kernel
bus-ID: 00:11.5 chip-ID: 1106:3059 class-ID: 0401
API: ALSA v: k6.1.0-33-686-pae status: kernel-api tools: alsamixer,amixer
Server-1: PipeWire v: 1.0.0 status: active with: 1: pipewire-pulse status: active
2: wireplumber status: active 3: pipewire-alsa type: plugin 4: pw-jack type: plugin
tools: pactl,pw-cat,pw-cli,wpctl
Network:
Device-1: Realtek RTL-8100/8101L/8139 PCI Fast Ethernet Adapter vendor: Foxconn driver: 8139too
v: 0.9.28 modules: 8139cp port: fc00 bus-ID: 00:09.0 chip-ID: 10ec:8139 class-ID: 0200
IF: eth0 state: unknown speed: 100 Mbps duplex: full mac: <filter>
Drives:
Local Storage: total: 148.99 GiB used: 8.11 GiB (5.4%)
SMART Message: Unable to run smartctl. Root privileges required.
ID-1: /dev/sda maj-min: 8:0 vendor: Western Digital model: WD1600BB-56RDA0 size: 148.99 GiB
block-size: physical: 512 B logical: 512 B speed: <unknown> type: N/A serial: <filter> rev: 0K20
scheme: MBR
Partition:
ID-1: / raw-size: 59.59 GiB size: 58.36 GiB (97.93%) used: 7.98 GiB (13.7%) fs: ext4
dev: /dev/sda1 maj-min: 8:1
ID-2: /home raw-size: 89.39 GiB size: 87.43 GiB (97.81%) used: 127.6 MiB (0.1%) fs: ext4
dev: /dev/sda2 maj-min: 8:2
Swap:
Kernel: swappiness: 15 (default 60) cache-pressure: 100 (default)
ID-1: swap-1 type: file size: 1024 MiB used: 0 KiB (0.0%) priority: -2 file: /swap/swap
Sensors:
Src: lm-sensors+/sys Message: No sensor data found using /sys/class/hwmon or lm-sensors.
Repos:
Packages: pm: dpkg pkgs: 2129 libs: 1060 tools: apt,apt-get,aptitude,nala,synaptic pm: rpm
pkgs: 0 pm: flatpak pkgs: 0
No active apt repos in: /etc/apt/sources.list
Active apt repos in: /etc/apt/sources.list.d/debian-stable-updates.list
1: deb http://deb.debian.org/debian bookworm-updates main contrib non-free non-free-firmware
Active apt repos in: /etc/apt/sources.list.d/debian.list
1: deb http://deb.debian.org/debian bookworm main contrib non-free non-free-firmware
2: deb http://security.debian.org/debian-security bookworm-security main contrib non-free non-free-firmware
Active apt repos in: /etc/apt/sources.list.d/mx.list
1: deb http://mxlinux.mirrors.uk2.net/packages/mx/repo/ bookworm main non-free
Info:
Processes: 176 Uptime: 15m wakeups: 947 Memory: 1.9 GiB used: 1.09 GiB (57.3%) Init: SysVinit
v: 3.06 runlevel: 5 default: graphical tool: systemctl Compilers: gcc: 12.2.0 alt: 12
Client: shell wrapper v: 5.2.15-release inxi: 3.3.26
Boot Mode: BIOS (legacy, CSM, MBR)
There is however a problem over getting it to boot up. I have experimented, and tried three routes. Only one seems to work, and it is a bit complicated. There seems to be a problem with Grub, or its location.
On first powering up, the initial screen display (from Phoenix Technologies) offers three options: click TAB to see Post Screen, DEL to enter Setup, ESC to enter Boot Menu.
Esc> Boot menu then offers a choice, either to boot from the HDD, or from a CDROM/DVD. Here is what happens in each case.
Route 1: boot from HDD - doesn't work
The screen display gives a terminal-style text listing the drives and PCI devices, and a message saying
"Verifying DMI Pool Data....". This is then followed after a few minutes by a message saying
"Boot from CD: disk boot failure, insert system disk and press Enter" and a blinking cursor. Putting a CD/DVD into the tray results in no further action, just a blinking cursor. The only further step is to Ctrl-Alt-Delete, which brings back the initial screen display with TAB/DEL/ESC.
Route 2: boot from Live CD/DVD
Choosing Esc from the initial screen (with an MX23.5 bootable ISO disk in the CD/DVD tray) and selecting to boot from that CD/DVD launches a Live session. The first splash screen is of course familiar, and offers various options.
Route 2(a) : MX23.5 live session. This works normally.... the computer launches into a Live session in the usual way. But of course that doesn't get us very far, and only lasts until shutdown.
Route 2(b): One of the options listed on the initial MX Live splash screen menu is to
'Boot from Hard Disk'. Selecting this option opens a black, terminal-style screen with the words
"Grub loading, Welcome to Grub...". Then after a few moments the message
"error: attempt to read and write outside of disk 'hd0'", followed by a new line saying
"grub rescue>". I have no idea what I could type into the space which follows. So I can only again do Ctrl-Alt-Del to get back to the initial screen display, in order to try the next possibility....
Route 2(c): The last option listed on the MX Live splash screen menu is
"Switch to Grub bootloader". Choosing that launches the MX23.5 Libretto screen [from the installed version of MX23.5 which is actually on the hard drive}. But this is followed by a terminal-style text screen with the message
"ata3: SRST failed (error no =-16)".
This message appears a couple of times. But if left alone, eventually the system boots into a CLI interface, with a series of terminal-type instructions, thus
[clans is the name I gave the computer when I installed MXLinux, so I enter my login name, which is brian]
[so I enter the password selected when I installed MXLinux]
At that line I enter
. I cannot now remember where I picked up that trick (probably a forum post I read years ago). But it works, in the sense that the computer then boots up into the MX23.5 which is actually installed on the desktop computer's hard drive.
And that's where I am now and posting this help request into the forum.
It seems to my non-expert eye that there is something wrong with the Grub. But I don't know what to do about it. Is it corrupted? (it seems not - it does work eventually if I follow route 2(c). Is it in the wrong place on the desktop hard disk drive? I simply don't have the knowledge or skill to work out the significance of the 'grub rescue' message or the reference to 'disk hd0'.
I am of course slightly tempted to throw the entire computer into a junkyard skip on the grounds that the machine is obsolete! But it is frustrating to think that the solution to this current problem might in fact be relatively simple. There is no precious personal data on the system which is at risk of loss: it would just be nice to get this old system to boot up and function normally. Can anyone suggest what to do? I'm ready to provide whatever further descriptive information might help, and/or to try out terminal commands.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 1:34 pm
by oops
... to see first some potential errors and warninge you can try in the terminal :
sudo dmesg -k -l emerg,alert,crit,err,warn ; dmesg | egrep -i --color 'error|critical|failed'
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 1:39 pm
by j2mcgreg
@br1anstorm wrote:
Route 2(b): One of the options listed on the initial MX Live splash screen menu is to 'Boot from Hard Disk'. Selecting this option opens a black, terminal-style screen with the words "Grub loading, Welcome to Grub...". Then after a few moments the message "error: attempt to read and write outside of disk 'hd0'", followed by a new line saying "grub rescue>". I have no idea what I could type into the space which follows. So I can only again do Ctrl-Alt-Del to get back to the initial screen display, in order to try the next possibility....
From here:
https://www.donordrives.com/western-dig ... DYsNFjv1tA
and here:
https://gearspace.com/board/geekzone/86 ... 424lv.html
Your hard drive is nineteen years old and has almost certainly failed. You should still be able to find a replacement on Amazon and while you are at it pick up a ribbon cable too.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 1:56 pm
by br1anstorm
oops wrote: ↑Sat Apr 19, 2025 1:34 pm
... to see first some potential errors and warninge you can try in the terminal :
sudo dmesg -k -l emerg,alert,crit,err,warn ; dmesg | egrep -i --color 'error|critical|failed'
Interesting to try some investigative detective work. This is what came up in the terminal...
Code: Select all
$ sudo dmesg -k -l emerg,alert,crit,err,warn ; dmesg | egrep -i --color 'error|critical|failed'
[sudo] password for brian:
[ 0.457367] acpi PNP0A03:00: fail to add MMCONFIG information, can't access extended PCI configuration space under this bridge.
[ 0.503459] pnp 00:00: disabling [mem 0x000c8e00-0x000cbfff] because it overlaps 0000:01:00.0 BAR 6 [mem 0x000c0000-0x000dffff]
[ 10.666962] ata1: link is slow to respond, please be patient (ready=0)
[ 15.526961] ata1: SRST failed (errno=-16)
[ 20.726961] ata1: link is slow to respond, please be patient (ready=0)
[ 25.586960] ata1: SRST failed (errno=-16)
[ 30.786960] ata1: link is slow to respond, please be patient (ready=0)
[ 233.587426] device-mapper: core: CONFIG_IMA_DISABLE_HTABLE is disabled. Duplicate IMA measurements will not be recorded in the IMA log.
dmesg: read kernel buffer failed: Operation not permitted
I see some familiar information there: the "SRST failed" message is what I got when trying one of the various boot-up routes, as described in my original post. But I'm not sure what further insights the rest of the quoted text reveals.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 2:16 pm
by br1anstorm
j2mcgreg wrote: ↑Sat Apr 19, 2025 1:39 pm
Your hard drive is nineteen years old and has almost certainly failed.
Yup, it is an ancient machine. But I did try to check the condition of the hard drive using the "Disks" utility. The verdict was that the drive was OK, and had just 3 bad sectors. That doesn't seem to me to explain the problem I have described.
Getting a replacement hard drive is perhaps one way to go. But (a) it doesn't actually diagnose the existing problem; and (b) I wasn't planning to get into the business of rebuilding, or replacing parts in, such an old and slow machine - at least not without first pinning down exactly what is causing the current problem. After all, the drive seems to be working all right once it is actually booted up.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 2:59 pm
by j2mcgreg
@br1anstorm wrote:
Yup, it is an ancient machine. But I did try to check the condition of the hard drive using the "Disks" utility. The verdict was that the drive was OK, and had just 3 bad sectors. That doesn't seem to me to explain the problem I have described.
It does if you consider that the drive's motor may no longer be spinning up properly or consistently, the sealants will have dried out and are leaking, and the plastic components will have dried out and shrunk.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 4:06 pm
by i_ri
ram try one stick at a time.
two 1G sticks; try one alone; try other one alone.
place one in slot nearest cpu. it may like just one.
when usb works, dvd works, hdd not work; check hdd cables.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 4:48 pm
by br1anstorm
When troubleshooting problems - whether on a computer, or in areas like cars with which I am more experienced - I like to be methodical, examine the evidence, and try to reach a diagnosis. I'm not normally inclined to the "suck it and see" approach, followed by swapping out parts on the basis of best-guess or assumptions.
That's why I outlined the detailed steps I had taken to try to boot up this system. Of course I could replace the HDD... the CPU ... the RAM ... hell, I could replace the entire computer.
But sorry, guys, that slightly misses the point. I'm hoping, first, that the problem can be more precisely identified on the basis of the evidence I have offered, or further checks.... not just by a strategy that is based on ".... it's an old machine, replace whatever looks the most likely culprit".
Is it not possible to run further checks, or terminal commands, to narrow down the cause/source of the problem?
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 4:56 pm
by CharlesV
MX Boot repair will give you a few options to try, you might start there and see if either reinstalling or repairing grub might help.
This issue sounds to me like something is possibly running the wrong grub? and then eventually figuring it out. If you can wipe the drive, that might be another method to see whats up - wipe the drive and then reinstall.. and personally I would install into one partition / and see if that helps things.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 5:38 pm
by oops
... You can also try for an old PC, a old kernel (ex: 5.10.xx , or 4.19.xx antiX via MXPI)
https://www.kernel.org/category/releases.html
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 5:43 pm
by br1anstorm
CharlesV wrote: ↑Sat Apr 19, 2025 4:56 pm
MX Boot repair will give you a few options to try, you might start there and see if either reinstalling or repairing grub might help.
This issue sounds to me like something is possibly running the wrong grub? and then eventually figuring it out. If you can wipe the drive, that might be another method to see whats up - wipe the drive and then reinstall.. and personally I would install into one partition / and see if that helps things.
Thanks, @CharlesV . Although I didn't say so in my original post, I have tried actions on those lines already, but they don't seem to have made any difference. I used to have both MX and Mint on this old computer, and encountered difficulties with booting up. So in an attempt to simplify matters I wiped and reformatted the hard drive and then installed MX 23.5 alone, using the default settings, as a clean fresh install on the whole disk (ie with just the / (root) and /home partitions). That's how the setup is currently configured.
I assume that this default installation puts the MX Grub into the MBR (which is I think the norm). I noted that MX had a Boot Repair tool, so I tried that. It gives the option of reinstalling the [repaired] Grub into the MBR and/or the partition containing MX 23. I ticked the boxes for both. I may thus have two Grubs (how do I check?) . But neither appears to function automatically when I try to boot up from the hard drive. And I am still mystified by the error message which refers to
"....failed attempts to read or write outside of disk 'hd0'" . What does that message actually mean?
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 6:07 pm
by br1anstorm
That's an approach that is way outside my limited experience. I'm familiar with normal updates which deliver (usually) newer versions of the same kernel. I have no idea how to go about installing an old(er) kernel.
To use the same analogy about cars again, that seems again to be like throwing parts at a problem without knowing the reason. Installing a different - older - kernel seems to me a bit like saying ".. the engine on that classic car won't start or is misfiring: so remove it and put in an engine from an even older vehicle....".
Also, I have to comment that - once I manage to get it booted up - this old desktop seems to run the latest MX23.5, with kernel 6.1.0-33, perfectly reliably, if slowly. How will changing to an older kernel cure issues around booting-up, and access to and operation of Grub?
Having said that, I still have older versions of MXLinux (v16. 1 from 2017, and v21.3 from 2023) as Live CDs. Is there any point in wiping the hard disk, reformatting it, and then installing one of these older versions of MX instead? If it really is worth trying, I'm prepared to experiment.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 6:25 pm
by oops
It is pretty simple via MXPI:
Code: Select all
su-to-root -x -c mx-packageinstaller &
Popular Appli > Kernel > ...
PS: about the ata1: SRST failed (errno=-16) ... maybe it is the jumper setting.
https://www.donordrives.com/media/catal ... 2-0046.jpg
https://forums.opensuse.org/t/solved-at ... rs/62083/7
"...Now I went back to the Western Digital manual. This time I found a fourth jumper option: cable select. Ah, that sounds vaguely familiar! I moved the jumper to the CS position and booted one more time.
This time the boot proceeded at lightning speed, and the “ata1: SRST failed” messages were gone..."
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 19, 2025 6:30 pm
by CharlesV
br1anstorm wrote: ↑Sat Apr 19, 2025 5:43 pm
CharlesV wrote: ↑Sat Apr 19, 2025 4:56 pm
MX Boot repair will give you a few options to try, you might start there and see if either reinstalling or repairing grub might help.
This issue sounds to me like something is possibly running the wrong grub? and then eventually figuring it out. If you can wipe the drive, that might be another method to see whats up - wipe the drive and then reinstall.. and personally I would install into one partition / and see if that helps things.
Thanks, @CharlesV . Although I didn't say so in my original post, I have tried actions on those lines already, but they don't seem to have made any difference. I used to have both MX and Mint on this old computer, and encountered difficulties with booting up. So in an attempt to simplify matters I wiped and reformatted the hard drive and then installed MX 23.5 alone, using the default settings, as a clean fresh install on the whole disk (ie with just the / (root) and /home partitions). That's how the setup is currently configured.
I assume that this default installation puts the MX Grub into the MBR (which is I think the norm). I noted that MX had a Boot Repair tool, so I tried that. It gives the option of reinstalling the [repaired] Grub into the MBR and/or the partition containing MX 23. I ticked the boxes for both. I may thus have two Grubs (how do I check?) . But neither appears to function automatically when I try to boot up from the hard drive. And I am still mystified by the error message which refers to
"....failed attempts to read or write outside of disk 'hd0'" . What does that message actually mean?
SO.. that error message is telling you that grub is attempting to get onto the wrong partition. ( I suspect because you had / have two partitions for boot.)
You might try the reverse of the 'auto install' . Been a bit since I did it, but as I recall it was something like:
new install using the advanced options (don't use the "erase disk and install ") do this instead -
- Create a 1 GB (1024 MB) ext4 partition on the beginning of the disk; mounted in "/boot"
- Create your desired install space in ext4 mounted in "/" MINUS your swap area
- Use remaining space for swap. (ALL partitions will be primary)
In the boot install drop down menu, select your "/boot" partition. Not the defaulted drive root!
do the install.
If that still doesnt work for you, then liveUSB, run gparted and look at the drive and see what is going on. ( look around too... possibly some other partition left over some place ? )
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 20, 2025 6:52 am
by br1anstorm
CharlesV wrote: ↑Sat Apr 19, 2025 6:30 pm
....
You might try the reverse of the 'auto install' .
....
Well, I tried that. Using a Live CD/DVD , I reformatted the entire hard drive using Gparted, created a separate 1GB /boot partition, a 50GB /root partition, and a 2 GB swap space, and left the rest unallocated.
Ran the customised install to put MX21.3 (which has an older kernel) on to the hard drive. When it was done I rebooted. Got the initial screen which listed all the devices etc and the message
"Verifying DMI Pool Data.....". Then after a short while, the same old further message saying
"Boot from CD: Disk boot failure, insert system disk and press enter". So no joy there.
I did Ctrl-Alt-Del to get back to the initial screen, chose Esc to get to the boot menu, and opted to boot from the CD/DVDROM (which I had put back into the tray).
From the launch screen of that Live CD/DVD I selected "Switch to Grub bootloader" . Something else happened which I hadn't seen before. A black screen with the heading
"GNU Grub version 2.06-3~deb11u5", and then the text
"Minimal BASH-like line editing is supported. For the first word, TAB lists possible command completions. Anywhere else TAB lists possible device or file completions." Then on the next line
grub>, with a flashing cursor.
This is way outside my comfort zone. So I got out (Ctrl-Alt-Del), and went back to the Live CD/DVD. This time I chose 'Boot from hard disk' from the list.
Again, didn't work. Up came a black screen with text as follows:
Booting...
GRUB loading
Welcome to GRUB!
error: file '/boot/grub/i386-pc/normal.mod' not found
grub rescue>
...... and a flashing cursor.
Again, nothing I could do but get out (Ctrl-Alt-Del)
So I'm stumped. I had put Grub into its own dedicated boot partition (sda1) when I did this latest install. But the system apparently can't find it. There are no other mystery partitions or leftover bits on the hard drive (as far as I know).
So all I can do is get into a Live session. Is there anything else I can do - in that Live session - either to investigate or to repair whatever is on the hard drive?
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 20, 2025 7:36 am
by br1anstorm
oops wrote: ↑Sat Apr 19, 2025 6:25 pm
PS: about the ata1: SRST failed (errno=-16) ... maybe it is the jumper setting.
Just to follow up on this separate strand of discussion. As regards the kernel, my experiment attempting to install and run MX21.3 (which has kernel 5.10.xxx) failed at the first hurdle. See separate post in reply to
@CharlesV . The issue seems to be with the location and operation of Grub. Which suggests that using a different kernel is unlikely to be relevant or helpful (or am I missing something?).
As for the SRST failure and the idea of changing jumper settings, this is unfamiliar territory. I deduce that it means opening up the casing and physically playing around with wires and cables. As the case has never been opened and no hardware removed, installed or changed, it's a little hard to grasp how a machine which was working normally should now somehow need rewiring or replacing of hardware. Is that what's being suggested? Or are we talking about setting, or resetting, something in the BIOS? (I can get into the BIOS from the initial splash screen by hitting Del. But what BIOS setting should I check or change?)
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 20, 2025 8:37 am
by oops
... For the Hdd jumper, yes, or just see if the actual setting is correct (CS or Master, not Slave).
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 20, 2025 8:55 am
by br1anstorm
oops wrote: ↑Sun Apr 20, 2025 8:37 am
... For the Hdd jumper, yes, or just see if the actual setting is correct (CS or Master, not Slave).
According to the information in the BIOS, the hard disk drive is set as the Primary Master. The CD/DVD is shown as the Secondary Master. No Slave disks are identified.
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 20, 2025 10:55 am
by CharlesV
Well... one of the things I really like about Western Digital drives... is also one of the most problematic things I have ever seen on hard drives - *many* WD's dont just die.. they fail slowly... and typically with spectacular results. (in my experience.)
WDs are know for a "complex failure model" which includes:
Corrupt File Structures - which includes
- Accidental deletion
Accidental format
File corruption
Software "bugs"
File system corruption
And much more
Firmware Failures - including
- Drive powers up, but is not recognized by the computer;
Drive powers up, but is recognized wrongly, sometimes with nonsensical characters;
Drive freezes during booting up;
Bad Platter Symptoms
- Hard disk still accessible but appear to “hang” or “sluggish”;
Constant Cyclic Redundancy (CRC) errors;
Unable to access folders or files which can be seen
I am starting to think that j2mcgreg may be 100% correct on the drive giving out (or a part of it is bad - say the first part of FAT)
If it were me... the next thing I would do is to put in a new / different drive and test with that. ALL of what you are seeing very well could be that of a drive that has several poorly performing / bad sectors in the very worst spot.
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 20, 2025 11:14 am
by oops
br1anstorm wrote: ↑Sun Apr 20, 2025 8:55 am
oops wrote: ↑Sun Apr 20, 2025 8:37 am
... For the Hdd jumper, yes, or just see if the actual setting is correct (CS or Master, not Slave).
According to the information in the BIOS, the hard disk drive is set as the Primary Master. The CD/DVD is shown as the Secondary Master. No Slave disks are identified.
"Secondary Master" means slave for PATA Hdds & CD/DVD. (maybe try to disconnect the CD/DVD ) ...
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 20, 2025 12:03 pm
by br1anstorm
CharlesV wrote: ↑Sun Apr 20, 2025 10:55 am
Well... one of the things I really like about Western Digital drives... is also one of the most problematic things I have ever seen on hard drives - *many* WD's dont just die.. they fail slowly... and typically with spectacular results.
...
I am starting to think that j2mcgreg may be 100% correct on the drive giving out (or a part of it is bad - say the first part of FAT)
If it were me... the next thing I would do is to put in a new / different drive and test with that. ALL of what you are seeing very well could be that of a drive that has several poorly performing / bad sectors in the very worst spot.
It looks as if we are indeed looking at a failing hard drive. Probably not much point in wrestling further with investigative and diagnostic efforts....
So it looks like a simple choice: surgery or the skip. Either open up the case and start swapping out components and cables. Or consign the entire machine to the recycling centre. As the desktop is almost 20 years old, and has a feeble processor and minimal memory by today's standards, perhaps the latter is the more logical option?
I imagine a refurbished HDD might be easy and cheap to find online, so I might just explore that. But otherwise, I'm not sure the gain would be worth the pain (and time) involved.
Thanks to all who have contributed to this discussion. Even if we haven't solved the problem, it's been a voyage of discovery...
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 20, 2025 11:41 pm
by FullScale4Me
New 2.5 SATA SSD 128GB can be had fairly cheap $12 USD (8.97 GBP)
https://www.newegg.com/p/0D9-004U-00093?Item=9SIA1DSKDN8255
Re: Strange boot-up problems on old desktop computer
Posted: Thu Apr 24, 2025 2:32 pm
by br1anstorm
Well, the story isn't over yet... and the search for a solution continues. So for @CharlesV, @j2mcgreg, @oops and others, this is the latest instalment.
As the general consensus seemed to be that the hard drive in the desktop was failing, I got a replacement hard drive: used, wiped and tested, guaranteed clean and good. To ensure compatibility, I got exactly the same make, model and spec as the original: a Western Digital 160GB WD1600BB-56RDA0.
Installed and freshly formatted, I ran a live session of MX21.3 from a CD/DVD. The live session was absolutely fine: everything worked as expected. So I installed MX21.3 on the "new" hard drive. I followed @CharlesV suggestion, and created a dedicated boot partition, as well as a swap partition and a partition for /rootMX. The installation proceeded and completed perfectly normally. So I rebooted.
Sadly, I got the same sequence of problems as described in my original post.
(Route 1) the initial launch screen, as before, lists Tab to view the POST screen, Del to enter the BIOS, and ESC to access the Boot menu. Leaving it to run brings up a black screen filled with text listing the devices and PCI etc, with a message at the bottom saying "Verifying DMI Pool Data ..." and a flashing cursor. After three or four minutes, up comes the dreaded and now familiar error message again: "Boot From CD: Disk Boot Failure, Insert System Disk and Press Enter."
So the first and obvious conclusion is that if the new and good HDD is producing the same result as the original HDD, then there was nothing wrong with that original hard drive.
(Route 2) from the initial splash screen, I chose Esc to get the boot order menu and chose the [MX] CD/DVD. From that MX launch screen menu list I chose the last option listed "Switch to Grub bootloader". As before, the screen displayed "Grub loading... Welcome to Grub"....and then on a new line "grub rescue>" and a flashing cursor. So no joy there.
(Route 3) again from the splash screen I chose again Esc and the MX CD/DVD. And this time from that MX Live session launch screen I chose the third option, "Boot from Hard Disk". Once again, I got the same sequence as described in my earlier posts: a screen with a message, repeated on two lines, consisting of a string of numbers and text saying something like "1234567xx: ata3: SRST failed (errno=-16)". And then after two or three minutes of clicking and whirring from the machine, the MX21.3 'Wildflower' launch menu appeared with the box to for my login password. In other words it had found and launched the MX21.3 which I had installed on the hard disk drive.
So what is the problem? Clearly the hard drive is fine. Clearly the CD/DVD drive works and enables a Live session. Clearly MX21.3 is properly installed, with its Grub, on the [replacement] hard drive.
The problem seems to be that the system is still not finding or recognising the Grub on that hard drive in the way it should, but is only [eventually] seeing it via the Route 3 described above, via the menu in the CD/DVD launch screen for the Live session, and after a couple of SRST (soft reset or system reset?) error pauses.
Is there an issue with the BIOS settings? Or some other glitch in the sequence of invisible actions which the system presumably goes through when booting up? What is triggering the "grub rescue" message? What is prompting the SRST process? Any advice, comments or insights would be welcome!
Re: Strange boot-up problems on old desktop computer
Posted: Thu Apr 24, 2025 2:50 pm
by j2mcgreg
@br1anstorm wrote:
I got a replacement hard drive: used, wiped and tested, guaranteed clean and good. To ensure compatibility, I got exactly the same make, model and spec as the original: a Western Digital 160GB WD1600BB-56RDA0.
I think that your mistake was in duplicating the original model rather than using a different brand. Your method assumes that the defect in the original was unique rather than being series wide. Rather than taking that chance, I would have searched for the newest IDE drive I could find.
Re: Strange boot-up problems on old desktop computer
Posted: Thu Apr 24, 2025 3:33 pm
by br1anstorm
j2mcgreg wrote: ↑Thu Apr 24, 2025 2:50 pm
@br1anstorm wrote:
I got a replacement hard drive: used, wiped and tested, guaranteed clean and good. To ensure compatibility, I got exactly the same make, model and spec as the original: a Western Digital 160GB WD1600BB-56RDA0.
I think that your mistake was in duplicating the original model rather than using a different brand. Your method assumes that the defect in the original was unique rather than being series wide. Rather than taking that chance, I would have searched for the newest IDE drive I could find.
Hmmm - very interesting point. When I browsed online for a replacement drive, there were obviously several makes and models to choose from (and they are not particularly expensive). I did look at another Western Digital model - WD1600BB-
55RDAO, which is very similar. But one aspect had me rather puzzled - and led me to select a direct equivalent.
The specific HDD in my old machine, WD1600BB-
56RDA0, is described as having a PATA interface (which it seems is slightly older and slower than SATA). The
55DRA0 for example is very similar but is listed as having a SATA interface. And various websites which I checked suggested that the two interfaces were different and had different connectors and sockets. Confusingly it turns out however that PATA and IDE are two different terms for the same thing.....! See for example this website
https://askleo.com/whats_the_difference ... _and_ide/.
I therefore had to assume that I could not order any old (or new) 160GB HDD, but would have to get one which had the right connections/sockets to fit the cables and plugs inside the casing of this particular and rather elderly machine. Hence my choice.
Moreover we ought not to overlook the fact that - for the last decade or more, the HDD in this laptop worked totally reliably and well. That of itself rather undermines the assumption that there might be a series-wide flaw with this particular model of HDD.
Do I now go looking for a different brand or model - and if so what? The jungle of model names, acronyms and connection-codes is not exactly simple or clear. And I see little point in a scattergun approach, buying and fitting a random selection of different HDDs of different brands on the basis that some may turn out to have the right connectors or interfaces and maybe one of them might work. Just throwing parts at an old computer without knowing where the fault lies, nor even knowing if they'll fit, seems to me a fool's game.
Re: Strange boot-up problems on old desktop computer
Posted: Thu Apr 24, 2025 5:53 pm
by Gabriel_M
It seems that the BIOS is having trouble finding your hard drive.
Try changing the dip switches (according to what you see on the internet you have them) as Master and as CS cable select.
You can also try adding to the boot line:
Re: Strange boot-up problems on old desktop computer
Posted: Thu Apr 24, 2025 6:32 pm
by br1anstorm
Gabriel_M wrote: ↑Thu Apr 24, 2025 5:53 pm
It seems that the BIOS is having trouble finding your hard drive.
Try changing the dip switches (according to what you see on the internet you have them) as Master and as CS cable select.
You can also try adding to the boot line:
Thanks
@Gabriel_M ...
You may be right that the BIOS is having trouble finding my hard drive.
But I don't completely understand your advice. What and where are "dip switches"? Are you referring to what some people call the "jumper" connectors on the back (or side) of the hard disk drive (between the PATA/IDE connector and the power-supply socket?
A quick internet search and a check on my setup shows that the jumper is in the "cable select" position on the hard drive; and the screen display when i try to boot (when the "Verifiying DMI Pool Data " message appears), gives a list of devices which shows the hard drive as Master, and the CDROM as Slave. I believe that is how the devices are supposed to be configured. What should I change?
And for your other suggestion - what and where is "the boot line"? Can you give step-by-step guidance on exactly what to do and where? Is this changing something in the BIOS setup? Or editing some file? And if so what file, and how can i do this? In a Live session?
Re: Strange boot-up problems on old desktop computer
Posted: Thu Apr 24, 2025 8:49 pm
by Gabriel_M
If the jumper is on CS select cable and the disk is installed as master, set the jumper as Master.
Regarding the boot line (first test the jumper) MX Tools --->MX Boot Options ---> Kernel parameters
If the distribution used does not have these options, edit the GRUB configuration template in /etc/default/grub.
Add the kernel parameter in the GRUB_CMDLINE_LINUX_DEFAULT variable.
Code: Select all
GRUB_CMDLINE_LINUX_DEFAULT=“insmod = ide-generic”.
Then run the following command to regenerate the GRUB configuration file.
Re: Strange boot-up problems on old desktop computer
Posted: Fri Apr 25, 2025 1:32 pm
by br1anstorm
Thanks again @Gabriel_M ... but unfortunately I don't think I have moved anything forward.
The jumper is on CS (cable select), which is shown as the default setting. That is to say, it links pins 1 and 2, according to the diagrams I have found.
As I understand it - and as shown in the initial screen when I try to boot up - the Western Digital hard disk drive is already set as the Master. The CDROM (is that considered a "drive"?) is listed as the Slave.
So what do I need to change? The three other settings possible for the jumper are
- Single: no jumper at all
- Dual (Master): jumper pins 5 and 6
- Dual (Slave): jumper pins 3 and 4
I have left the jumper on the default CS setting.
I booted into a Live session of MX21.3, and went to look at MX Tools>MX Boot Options. The screen shows four headings
The first heading is Options. It shows a menu timeout of 5 secs, and the 'Boot to' box is blank. Below that, Kernel parameters, and in that box the only entry is 'quiet splash'.
The other three headings are Background, Splash, and Messages. At the bottom is a 'Display Log' button. Clicking that gives a message that "Nothing has been logged yet..."
I have no idea how to edit the Grub configuration template, and no idea how or where to add a kernel parameter.
Sorry. I just don't have the knowledge, the instruction manual and the step-by-step guidance that might enable me to change or fix things.
Right now I feel like I am being asked to overhaul a car engine using a knife and fork and with a blindfold on!
Re: Strange boot-up problems on old desktop computer
Posted: Fri Apr 25, 2025 1:47 pm
by br1anstorm
A quick follow up. Without messing with the jumpers, I booted into a Live session, went looking for the file /etc/default/grub, found it, and - I think - successfully added the two new code-entries "sudo nano /etc/default/grub" and "insmod = ide generic". Saved and exited.
Then in the terminal, tried "sudo update-grub". got an error message: "... failed to get canonical path of overlay". That means absolutely nothing to me, and leaves me even more confused.
What else, or what more, can I do? This seems to be going deeper and deeper down the rabbit hole, with no clear route-map to a solution.
Re: Strange boot-up problems on old desktop computer
Posted: Fri Apr 25, 2025 2:27 pm
by j2mcgreg
This would open grub with administrative powers in the nano text editor. Is that what you were trying to do?
Re: Strange boot-up problems on old desktop computer
Posted: Fri Apr 25, 2025 3:32 pm
by br1anstorm
j2mcgreg wrote: ↑Fri Apr 25, 2025 2:27 pm
This would open grub with administrative powers in the nano text editor. Is that what you were trying to do?
The truthful answer,
@j2mcgreg , is that I really don't know! I was simply following the suggestion put forward by
@Gabriel_M , in the hope that this was designed to address and resolve whatever the problem appeared to be with the non-booting of this desktop.
I don't normally like doing things to a computer (or indeed any piece of machinery and technology) without knowing exactly what I am doing, why, how, and what the outcome ought to be. But in this situation (and this thread is now four pages long) all I have been able to do is describe, as accurately as I can, what happens when I try to boot up, and what error messages I see on screen. I have then been trying to carry out whatever remedial actions (replacing parts, editing files, changing connections) have been suggested as a solution.
But it has to be said that at each stage, and with each different attempt to get things going, a different error message seems to appear.
I have been hoping, and assuming, that the problem (or problems) I am experiencing are not unique, and that the various error messages and behaviours I have observed with this old desktop might mean something, and be diagnosed, by others with greater knowledge and experience. Thus far, I'm not sure the problem has been clearly identified (is it a physical hard drive issue, a BIOS setting, a Grub corruption, or what?).
My feeling is that until the cause of the "Disk Boot Failure" is clearly identified, we've just been trying 'suck it and see' attempts to find a solution. Is the situation really beyond rescue or recovery? I just hate admitting defeat and giving up!
Re: Strange boot-up problems on old desktop computer
Posted: Fri Apr 25, 2025 3:43 pm
by j2mcgreg
When you replaced the HDD did you also replace the ribbon cable as I suggested?
Re: Strange boot-up problems on old desktop computer
Posted: Fri Apr 25, 2025 3:57 pm
by br1anstorm
j2mcgreg wrote: ↑Fri Apr 25, 2025 3:43 pm
When you replaced the HDD did you also replace the ribbon cable as I suggested?
Nope.... didn't have (and couldn't find locally) another ribbon cable. So all I did was wiggle the connections, unplug, try to get rid of any dust and dirt etc, and reconnect.
Thinking more generally about the problem(s) - and I try to be logical - I have found it possible (a) to install an operating system on the hard disk, as i did, from a Live session; and (b) get that system to run, albeit by going a somewhat unorthodox route (I described it earlier as 'route 3') to get it to boot up eventually after the SRST error messages. That surely means that the physical components and connections (hard drives old and new, ribbon cable, plugs, sockets, connectors) are sound and operational.
Several of the error messages and unusual screen displays allude in one way or another to Grub. An online search suggests that many of these messages are quite common and familiar: the tricky aspect is that there seem to be multiple causes, and a range of possible explanations...
The issue seems to be over the process or sequence of steps involved in normal boot-up. Surely that indicates that the fault is not with the hardware as such, but with the way it is operating when going through the boot process? A problem of settings rather than physical integrity.
Or is that just too simple and logical?
Re: Strange boot-up problems on old desktop computer
Posted: Fri Apr 25, 2025 4:41 pm
by Gabriel_M
**Try setting the HDD as Master: jumper pins 5 and 6**
Older PATA drives, if not configured correctly for master/slave (master, slave, single, cable-selected), can cause serious problems.
Re: Strange boot-up problems on old desktop computer
Posted: Fri Apr 25, 2025 5:19 pm
by br1anstorm
Gabriel_M wrote: ↑Fri Apr 25, 2025 4:41 pm
**Try setting the HDD as Master: jumper pins 5 and 6**
Older PATA drives, if not configured correctly for master/slave (master, slave, single, cable-selected), can cause serious problems.
I can see how to do this,
@Gabriel_M , and I'm willing to try it.
But what is the "correct" configuration? Surely the default Cable Select (CS) - as at present - with the jumper on pins 1 and 2, is the recommended arrangement? It certainly worked from the time the computer was built back in about 2009, until a few weeks ago.
And just to help me understand this, if I now set the jumper to pins 5 and 6, this is described - on the WD disk itself, which has a little diagram - as Dual (Master). That's fine. But what happens to the CDROM? It is currently listed and connected, apparently, as the Slave, and it works more or less normally and enables Live sessions. Am I, by resetting the jumper to pins 5 and 6, disconnecting the CDROM?
Maybe that doesn't matter if I do this as an experiment. But obviously, at present the CD/DVD is the only way available to boot the computer at all (as described in my earlier posts about 'route 3'). And equally obviously, if I do need to install or reinstall MX21.3 (or any operating system) then the CD/DVD drive has to be connected, properly configured, and operating. Will it do so if the jumper is on pins 5 and 6?
Re: Strange boot-up problems on old desktop computer
Posted: Fri Apr 25, 2025 8:00 pm
by Gabriel_M
It only affects how the HDD is recognized by the BIOS, it has no effect on the CD/DVD, nor on the data contained on the HDD.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 26, 2025 7:48 am
by oops
Gabriel_M wrote: ↑Fri Apr 25, 2025 8:00 pm
It only affects how the HDD is recognized by the BIOS, it has no effect on the CD/DVD, nor on the data contained on the HDD.
If HDD is master, CD/DVD must be slave for PATA.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 26, 2025 8:59 am
by Gabriel_M
brianstorm: As I understand it - and as shown in the initial screen when I try to boot up - the Western Digital hard disk drive is already set as the Master. The CDROM (is that considered a "drive"?) is listed as the Slave.
As the user already reported, the CD/DVD drive is installed as a slave. The HDD is installed as a master drive and therefore requires configuration.
Note: The HDD's ribbon cable must be connected to IDE port 1 blue. The CD/DVD ribbon cable (which has no configuration jumpers) to be a slave must go into the IDE2 port gray.
In short: you should try HDD with dip switch as Master, ribbon cable in IDE1 port blue. CD/DVD , with ribbon cable connected to IDE2 port gray.
-----------------------------------------------
HDD Connectors: PIDE & SIDE
This connectors supports the provided Ultra DMA 133/100/66 IDE hard disk ribbon cable. Connect the cable's blue connector to the primary (recommended)
or secondary IDE connector, then connect the gray connector to the Ultra DMA 133/100/66 slave device (hard disk drive) and the black connector to the Ultra
DMA 133/100/66 master device. If you install two hard disks, you must configure the second drive as a slave device by setting its jumper accordingly. Refer to the
hard disk documentation for the jumper settings.
source:
https://www.manualslib.com/manual/67510 ... =19#manual
-----------------------------------------------
Standard CMOS Features
This sub-menu is used to set up the standard CMOS features, such as the date,
time, HDD model and so on. Use the arrow keys select the item to set up, and
then use the <PgUp> or <PgDn> keys to choose the setting values
IDE Channel 0/1 Master/Slave, IDE Channel 2/3 Master
These categories identify the HDD types of 2 IDE channels installed in the
computer system. There are three choices provided for the Enhanced IDE BIOS:
None, Auto, and Manual. "None" means no HDD is installed or set; "Auto" means
the system can auto-detect the hard disk when booting up; by choosing "Manual"
and changing Access Mode to "CHS", the related information should be entered
manually. Enter the information directly from the keyboard and press < Enter>
source:
https://www.manualslib.com/manual/67510 ... =29#manual
----------------------------------------------
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 26, 2025 5:38 pm
by br1anstorm
Not really a progress report - but an update, and an apology.
I opened up the computer casing again, to access the hard drive. And I successfully moved the jumper from pins 1 and 2 (the default Cable-Select setting), to pins 5 and 6.
Connected it all up again, and powered on to boot up.
Sadly, absolutely nothing different. Computer won't boot up into MX21.3 (which is installed on the hard disk).
Same sequence of screens as before. The first (splash?) screen has the computer logo and the three options: hit TAB for POST, Del to enter Setup, and Esc for the Boot menu.
After a minute or two, the next screen is the one which lists all the hardware and devices and at the bottom has the message, "Verifying DMI Pool Data..." and a flashing cursor, and then - yes, you guessed it - that same old message: "Boot from CD... Disk Boot Failure, Insert System Disk and Press Enter". And nothing else happens.
So, in terms of the experiment, a failure. Changing the jumper to different pins made no difference and changed nothing in terms of the behaviour at start-up.
I should also add that I did not mess with the ribbon cables. I unplugged and then replugged only the ribbon cable connector at the back of the HDD (and the four-pin power connector). I did not remove or mess with the other end of the ribbon, where it connects to the motherboard - although i did manage to see that it was plugged into a blue socket on the motherboard. And I did not tamper at all with the ribbon cable or any other wiring to and from the CD/DVD drive.
However.... there is an aspect here which may be significant and on which I fear I may have given misleading info in an earlier post. Most of the detailed information displayed on the screen listing all the devices etc means very little to me. In my previous posts I have focused on, and reported very precisely, the error messages that appeared.
But in this changing-jumpered pins experiment, I looked much more closely at the list of hardware, devices, and connections after shifting the jumper. As I have reported, there was no change in any of the info listed. I can't take a screenshot. But I reproduce below some of the key information, as it may shed light on this Master/Slave situation.
In the first text box on the screen the list is as follows:
Diskette Drive A: None
Pri. Master Disk: LRG, ATA 100, 160GB
Pri. Slave Disk: None
Sec. Master Disk: DVD-RW, ATA 33
Sec. Slave Disk : None
Over to the right of this list is more information...
Display Type: EGA/VGA
Serial Ports: 3E8, 2F8
Parallel Ports: 3 7 8
DDR SDRAM at Rank 0 1 2 3
Below that is the PCI device list, with information in several columns (Bus No ... Device No ... etc ... IRQ...) for 11 'devices'
And then below that come the various error messages when they appear.
The key point, and the reason for apology, is that I mistakenly or carelessly said earlier that the Hard Disk was listed as the Master and the CD/DVD as the Slave. That was wrong. I had misread, or misremembered, that listing on the screen. As quoted above, the HDD is the Pri(mary) Master Disk and the CD/DVD is the Sec(ondary?) Master Disk. There is no Slave Disk listing.
Both before (with jumper on pins 1 and 2) and after (jumper on pins 5 and 6) the display information on the screen has remained the same. I have checked by reinstating the jumper on pins 1 and 2, and checking with both the 'old' HDD and the 'replacement' HDD. The info listed on the display screen remains the same. In neither configuration does the system boot up, or even show the Grub menu.
I still cannot interpret all these details. But maybe someone else can? Do they reveal something significant about the disk setup and connections? Does something need to be checked, or changed, in the BIOS settings?
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 26, 2025 6:04 pm
by Gabriel_M
See if the information in this manual is helpful:
https://www.manualslib.com/manual/67510 ... =30#manual. Essentially, pages 26 through 30.
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 26, 2025 6:09 pm
by trawglodyte
Your QSI says you have a 64-bit processor. Can you install the 64-bit version of MX Linux or does it have to be 32-bit for some other reason?
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 26, 2025 6:32 pm
by br1anstorm
trawglodyte wrote: ↑Sat Apr 26, 2025 6:09 pm
Your QSI says you have a 64-bit processor. Can you install the 64-bit version of MX Linux or does it have to be 32-bit for some other reason?
Well,
@trawglodyte , I'm not sure how this suggestion relates to the problem of not being able to boot at all. But - for the record - I have in the past installed and run both 32-bit and 64-bit versions of MXLinux, Linux Mint, Solydx and PCLinuxOS on this rather old desktop computer. All worked fairly well, albeit not always very speedily.
I have recently been relying on MXLinux (because I find it user-friendly) and I installed the 32-bit rather than the 64-bit on the assumption that this would be less demanding on the not-very powerful CPU and limited RAM in the machine.
But choice of OS version seems not to be the key issue. Or is it? (if so please explain....) The difficulty is that I cannot get the system to boot normally at all. Not even to get to the Grub screen. And messages about disk boot failure, Grub rescue, and "cannot find the canonical pathway...(??)" all seem to appear before the machine has booted into any operating system.
Do you have any insight into the cause of the problems and the reason for the error messages which I have documented in this thread?
Re: Strange boot-up problems on old desktop computer
Posted: Sat Apr 26, 2025 6:42 pm
by trawglodyte
br1anstorm wrote: ↑Sat Apr 26, 2025 6:32 pm
Do you have any insight into the cause of the problems and the reason for the error messages which I have documented in this thread?
No, I don't. Thanks for answering my question though, I thought perhaps it had something to do with your troubles. Just a spitball, I do hope you get your machine working.
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 27, 2025 12:53 am
by FullScale4Me
Just some ramblings from a retired server tech:
Boot flags are a thing not touched on in this thread. Old PCs like these
*may* need either LBA or Legacy_Boot on their boot volume to spring to life. Boot from live media and use GParted to investigate and initiate experiments.
There is a
*recent* thread here in this forum on MX 23 32-bit version NOT booting to a usable desktop on 64-bit PCs (
32-bit version won't boot on 64-bit machine) so, just saying.....MX 21 is NOT affected by whatever issue is present in MX23.
I've had this happen to me a few times over the years. Does your boot media boot to a usable desktop on another PC? (It may be a bad ISO or boot media.)
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 27, 2025 6:37 am
by br1anstorm
Thanks @FullScale4Me for your ramblings...
Just a quick response on the thoughts you offer.
The subject of boot flags, LBA etc is way above my pay grade. I had a look at the Wikipedia explanation of LBA. It might as well have been written in Chinese or Greek. I couldn't begin to understand it. It follows that I'm hardly equipped to investigate and experiment in GParted: I wouldn't know what to look for, what to do, or what to change.
On the 32-bit v 64-bit debate, in this case I think it's a red herring. I have run Live sessions off DVDs of both MX21.3 and MX23.5, (and other distros) in both 32 and 64-bit versions. When I tried it, MX23.5 32-bit ran in Live session, and I installed it on the HDD. I reverted to MX21.3 32-bit on the assumption that it was well tried and tested.
On the question of "boot media", I have an assortment of distros on CDs or DVDs, and one or two on USB sticks. All verified (MD5 or SHA256). From them I have installed my choice of operating system on various other computers (mostly more recent laptops such as the ThinkPad I'm using to type this). So I'm confident that the ISOs I'm using are not corrupt or faulty.
But - as described at length in this thread - I have all kinds of difficulty getting whatever I have installed on the HDD to boot. I have tried LMDE and Mint, as well as MX. I have wiped and reformatted the HDD using GParted. I have tried installing Grub on the MBR, and I have tried creating a separate /boot partition and putting Grub there. None of these has enabled the system to boot into whatever is installed on the HDD.
We come back to the central issue. I can run a Live session in the old desktop computer from a CD/DVD. I can install MX, Mint or whatever on to the hard disk. I can run SMART checks on the integrity of the HDD and the results say that the disk is good. BUT.... when I try to boot up the computer to run whatever is on the HDD, I get the error messages already described. Disk Boot Failure.... grub rescue ... cannot find the canonical pathway ... SRST .... etc. Running MX Boot Repair or sudo update-grub results in similar error messages.
In my simple layman's mind, this tells me two things. One, the actual physical hardware (HDD - I've tried two; the CPU; the display screen; etc etc) is all OK. Two, whatever OS I install on the hard disk - if and when I can actually launch it via the unorthodox route of choosing the 'boot from hard disk' option from the CD/DVD Live session splash/launch menu - runs perfectly normally.
The problem seems to be that on powering up, as the boot process (BIOS?) goes through whatever steps it has to go through, it cannot see, find or access the hard drive. Why not? Is it a bootloader problem? Is it somehow related to the settings in the BIOS? That remains the unanswered question. And so far, I haven't found an answer. I'm still hoping that someone with greater expertise than me will review all the evidence and suggest possible solutions!
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 27, 2025 3:24 pm
by FullScale4Me
br1anstorm wrote: ↑Sun Apr 27, 2025 6:37 am
Thanks @FullScale4Me for your ramblings...
Just a quick response on the thoughts you offer.
The subject of
boot flags, LBA etc is way above my pay grade. I had a look at the Wikipedia explanation of LBA. It might as well have been written in Chinese or Greek. I couldn't begin to understand it. It follows that I'm hardly equipped to investigate and experiment in GParted: I wouldn't know what to look for, what to do, or what to change.
<snip>
The problem seems to be that on powering up, as the boot process (BIOS?) goes through whatever steps it has to go through, it cannot see, find or access the hard drive. Why not? Is it a bootloader problem? Is it somehow related to the settings in the BIOS? That remains the unanswered question. And so far, I haven't found an answer. I'm still hoping that someone with greater expertise than me will review all the evidence and suggest possible solutions!
I feel you are overanalyzing the flag issue one level beyond what we would normally ask in a support thread. A flag only persists until the next formatting.
GParted manages
flags via a checkbox.
- Left click to highlight the partition.
There's no heavy lifting involved for what's a simple on/off affair. Try it!
What it
may look like below.
While researching something else, I came across the Rufus program (USB burner) having support for older PCs like yours. It has (in addition to using a MBR aka msDOS portion table), under 'Show advanced drive properties', a checkable option 'Add fixes for old BIOSes (extra partition, align, etc.)' checked. Above is the GParted view of the resulting USB.
Re: Strange boot-up problems on old desktop computer
Posted: Sun Apr 27, 2025 5:24 pm
by br1anstorm
Thanks @FullScale4Me , I'm prepared to try almost any uncomplicated experiment to get this old computer functioning normally.
I try to understand what I'm doing, but a lot of this is the proverbial steep learning curve.
However your suggestions on managing flags in GParted look pretty straightforward. This is what I found when I booted into a Live session and launched GParted.
I should start by saying that - following earlier comments from @CharlesV - I currently have the hard drive configured with a dedicated boot partition on which Grub is installed.
The configuration is as follows:
/dev/sda1 .........format ext4.........label boot........size 1GB........flags boot
/dev/sda2 ........linux-swap ..............................size 2GB
/dev/sda3 .......format ext4 ........label root MX21.3 ..........size 30GB
/dev/sda4 .........extended partition ...............................size 116GB, containing
/dev/sda5.........format ext4 ........label DATA...........size 50GB,
with the remaining 66GB in the extended partition unallocated.
I clicked as advised to get to the Manage Flags for /dev/sda1, the boot partition. It already shows the 'boot' flag. The list of flags in the popup window was as shown in your posted screenshot. Only the 'boot' box is ticked. So I selected and clicked the box for 'lba' . A short pause with the rotating pointer icon.... and then no action. The 'lba' box flag cannot be ticked.
Just out of curiosity I then tried to select and tick each of the other flag-boxes separately and in turn. I could tick and activate any of the flags on the list except 'hidden', 'lba' and 'msftres'. These three flags could not be set on /dev/sda1.
Still curious, I looked at the Manage Flags for the MX root partition, dev/sda3. No flags are set. I don't know what any of them represent. I tried to set 'lba' on that partition. No joy: it wouldn't do so. But it would allow me to set the 'boot' flag on that partition as well. Should I do so? Is it acceptable to have a 'boot' flag on more than one partition?
What does that reveal? Anything significant? Suggests that the computer doesn't like, or won't accept, 'lba'. Is that why it won't boot normally?
I haven't yet looked into the Rufus angle. I have used Rufus to create bootable USB drives of Linux ISOs, but not for any other troubleshooting/problem-solving.
Re: Strange boot-up problems on old desktop computer
Posted: Mon Apr 28, 2025 3:35 am
by putih
Maybe MBR have some issue
During install select PBR