I normally run OS/2 and haven’t seen any hardware related error messages. I installed a test LEAP 15.1 system and it is reporting a lot of single bit ECC errors. Could Linux be overclocking?
Message from syslogd@localhost at Mar 2 22:05:27 ...
kernel: 1250.276047] [Hardware Error]: Corrected error, no action required.
Message from syslogd@localhost at Mar 2 22:05:27 ...
kernel: 1250.276060] [Hardware Error]: CPU:0 (f:6b:1) MC4_STATUS[Over|CE|MiscV|-|AddrV|CECC]: 0xdc18c100d1080a13
Message from syslogd@localhost at Mar 2 22:05:27 ...
kernel: 1250.276066] [Hardware Error]: Error Addr: 0x000000002a7cb510
Message from syslogd@localhost at Mar 2 22:05:27 ...
kernel: 1250.276069] [Hardware Error]: MC4 Error (node 0): DRAM ECC error detected on the NB.
Message from syslogd@localhost at Mar 2 22:05:27 ...
kernel: 1250.276088] [Hardware Error]: cache level: L3/GEN, mem/io: MEM, mem-tx: RD, part-proc: RES (no timeout)
Message from syslogd@localhost at Mar 2 22:10:39 ...
kernel: 1561.568050] [Hardware Error]: Corrected error, no action required.
Message from syslogd@localhost at Mar 2 22:10:39 ...
kernel: 1561.568089] [Hardware Error]: CPU:0 (f:6b:1) MC4_STATUS[Over|CE|MiscV|-|AddrV|CECC]: 0xdc18c100d1080a13
Message from syslogd@localhost at Mar 2 22:10:39 ...
kernel: 1561.568118] [Hardware Error]: Error Addr: 0x000000002a7cb510
Message from syslogd@localhost at Mar 2 22:10:39 ...
kernel: 1561.568144] [Hardware Error]: MC4 Error (node 0): DRAM ECC error detected on the NB.
Message from syslogd@localhost at Mar 2 22:10:39 ...
kernel: 1561.568185] [Hardware Error]: cache level: L3/GEN, mem/io: MEM, mem-tx: RD, part-proc: RES (no timeout)
Message from syslogd@localhost at Mar 2 22:15:50 ...
kernel: 1872.864039] [Hardware Error]: Corrected error, no action required.
Message from syslogd@localhost at Mar 2 22:15:50 ...
kernel: 1872.864077] [Hardware Error]: CPU:0 (f:6b:1) MC4_STATUS[Over|CE|MiscV|-|AddrV|CECC]: 0xdc18c100d1080a13
Message from syslogd@localhost at Mar 2 22:15:50 ...
kernel: 1872.864106] [Hardware Error]: Error Addr: 0x000000002a7cb510
Message from syslogd@localhost at Mar 2 22:15:50 ...
kernel: 1872.864133] [Hardware Error]: MC4 Error (node 0): DRAM ECC error detected on the NB.
Message from syslogd@localhost at Mar 2 22:15:50 ...
kernel: 1872.864172] [Hardware Error]: cache level: L3/GEN, mem/io: MEM, mem-tx: RD, part-proc: RES (no timeout)
Message from syslogd@localhost at Mar 2 22:21:01 ...
kernel: 2184.160052] [Hardware Error]: Corrected error, no action required.
Message from syslogd@localhost at Mar 2 22:21:01 ...
kernel: 2184.160092] [Hardware Error]: CPU:0 (f:6b:1) MC4_STATUS[Over|CE|MiscV|-|AddrV|CECC]: 0xdc18c100d1080a13
Message from syslogd@localhost at Mar 2 22:21:01 ...
kernel: 2184.160121] [Hardware Error]: Error Addr: 0x000000002a7cb510
Message from syslogd@localhost at Mar 2 22:21:01 ...
kernel: 2184.160147] [Hardware Error]: MC4 Error (node 0): DRAM ECC error detected on the NB.
Message from syslogd@localhost at Mar 2 22:21:01 ...
kernel: 2184.160188] [Hardware Error]: cache level: L3/GEN, mem/io: MEM, mem-tx: RD, part-proc: RES (no timeout)
Message from syslogd@localhost at Mar 2 22:26:12 ...
kernel: 2495.456051] [Hardware Error]: Corrected error, no action required.
Message from syslogd@localhost at Mar 2 22:26:12 ...
kernel: 2495.456097] [Hardware Error]: CPU:0 (f:6b:1) MC4_STATUS[Over|CE|MiscV|-|AddrV|CECC]: 0xdc18c100d1080a13
Message from syslogd@localhost at Mar 2 22:26:12 ...
kernel: 2495.456126] [Hardware Error]: Error Addr: 0x000000002a7cb510
Message from syslogd@localhost at Mar 2 22:26:12 ...
kernel: 2495.456152] [Hardware Error]: MC4 Error (node 0): DRAM ECC error detected on the NB.
Message from syslogd@localhost at Mar 2 22:26:12 ...
kernel: 2495.456192] [Hardware Error]: cache level: L3/GEN, mem/io: MEM, mem-tx: RD, part-proc: RES (no timeout)
Do you still experience apparent slowness as reported in your other thread? If the whole PC is 14 years old, there could be components expired or getting tired. Are you aware of the capacitor plague? 2006 is near the end of that era. Reseating RAM could help, but it may be time to open up for cleaning and inspection of not only the motherboard, but particularly the power supply.
High disk activity can be caused by weak sectors on the drive. Linux will attempt to read until the checksums agree or the heat death of the Universe. Run smartctl to check health of drive.
For cable plugs and pins, use a Contact Spray – there are special products available from Electronics suppliers …
For cards, especially those with gold plate on the connectors, use a non-abrasive pencil eraser – be very careful: more than 50% of the erasers contain abrasive particles!!
For the card slots on the mainboard, use a Contact Spray.
OS/2 (Warp I assume) does not know about the hardware chips in your computer. It was designed for MicroChannel devices from the mid 1980’s.
Linux Mint has almost every chipset know to be interrogated for issues.
That is why Linux logs errors that OS/2 cannot see.
Other than Lotus 123 support - I cannot imagine why you would be running OS/2. Everything OS/2 except Lotus 123 runs better in Windows or Linux.
I run OS/2 24/7 for Paradox for DOS and Quattro Pro for DOS, which were never provided a migration path elsewhere. I run them in ATI proprietary SVGA text mode, 132x43, which has no workable counterpart anywhere other than real DOS on hardware, where disk I/O is painfully slow for large files compared to OS/2.
QPro to me is much like cell phones to others, indispensable.
Did not think about DOS under OS/2 - I would suggest DOSBOX but it does not support any thing but 80x24 screen - I do use it for MS Quickbasic in Linux. Sometime 16 bit apps are still good to have.
Have you seen this site - they have all the old Borland Products for free downloads including new versions. You could get the MS Windows version of both here for free.
There’s no file format conversion or importing from the DOS versions to the Windows versions, unless something came later that I don’t know about. Even if there was, my large macro library would most likely be broken or even useless.
Serendipity. Cleaning the DIMM contacts didn’t help, but there was a leaking capacitor on the graphics card.Replacing the graphics card also didn’t help, but it would have eventually given me grief.
I currently have 2 GiB; if I have to replace the memory I’ll probably spring for 4 GiB. Likewise, if I have to replace the disk I’ll probably spring for a larger SSD.
The confusing thing is that this isn’t affecting OS/2; I’d expect it to get read failures or retries as well. Or did LEAP win the jackpot and get all the bad sectors?
I would run a few passes of memtest on the hardware to see what it sees about your memory. I know that OS/2 was never designed for newer hardware - good motherboards do the ECC fixes in hardware and the support memory chip will set a recovery that can be read (my old Gigabyte ECC motherboard logged the last corrections, my MSI only logged the last 8 - it was a lot older and was the wonderful ramdac memory of the future - it went bad the week after the warranty ended - the memory company went under - so much for the lifetime warranty.)
You can get a bootable USB image of memtest. Gecko Linux image has memtest on the grub menu - it is based on OpenSUSE 15.0
My second choice would boot a USB version of OpenSUSE to see what it sees as the host. I would dd your hard disk from this to see if the drive is the failing part.
Since you have no harddrive errors, you are having soft ECC errors that the Linux kernel chipset supports are seeing - memtest will not see soft ECC errors as it does not look at the support chips - just the L1, L2 and L3 cache and memory for errors.
I don’t know if there is a way to stop logging soft ECC errors in Linux kernels - there was a way in HPUX.
I know with DDR2 ECC memory if the chipset and the memory had different timings (chipset was 7-7-7 and motherboard 8-8-8) we would see lots of soft ECC errors when the same memory chip was hit a lot. Sometime in the DDR3 chipsets the chips would figure out what the memory was set to ( it could read it ) and end the soft ECC errors.