| View previous topic :: View next topic |
| Author |
Message |
pjp
Administrator
Joined: 16 Apr 2002
Posts: 20529
|
 Posted: Thu Dec 26, 2024 3:10 am Post subject: [solved] Using memmap to reserve multiple memory ranges |
 |
|
After running memtest, I appear to have two memory ranges that are bad. Presuming I'm not mistaken, the subsequent question is how to mask both ranges.
Transcribed: | Code: | Tst Pass Failing Address Good Bad Err-Bits Count CPU
--- ---- ------------------------ -------- -------- -------- ----- ----
7 1 002a0a1d670 - 10,762.8MB dfffffff dffffffd 00000002 1 0
7 1 002a0a1d8b0 - 10,762.8MB ffffffdf ffffffd2 00000002 2 0
7 1 002a0a1d9b0 - 10,762.8MB fffffdff fffffdfd 00000002 3 0
7 1 002e0a1d650 - 11,786.8MB dfffffff dffffffd 00000002 4 0
7 1 002e0a1d890 - 11,786.8MB ffffffdf ffffffdd 00000002 5 0
7 1 002e0a1d990 - 11,786.8MB fffffdff fffffdfd 00000002 6 0 |
Each "segment" appears separated by about 1bn bytes, so two ranges seem appropriate.
The first three: 002a0a1d670, 002a0a1d8b0, 002a0a1d9b0 are in the 4k block: 002A0A1D000 - 002A0A1E000
While the second three: 002e0a1d650, 002e0a1d890, 002e0a1d990 are in the 4k block: 002E0A1D000 - 002E0A1E000
For memmap= it seems that multiple ranges require multiple memmap= declarations.
So, without an actual root= line available (memtest86 is still running), this should be the solution: | Code: | | root=UUID=... memmap=4K\$0x002A0A1D000 memmap=4K\$0x002E0A1D000 |
It seems grub doesn't like $, and escaping is required. Exactly how isn't clear, but I think I came across two references to \\\$.
Did I calculate the memory ranges incorrectly? Does the root= parameter seem OK? Anything else?
_________________
Quis separabit? Quo animo? |
|
| Back to top |
|
 |
pjp
Administrator
Joined: 16 Apr 2002
Posts: 20529
|
| Posted: Thu Dec 26, 2024 5:21 am Post subject: |
|
|
I had an error in the memmap= entries. I forgot the leading 0x.
I had tried \\\$0x2... and that failed. There were some messages I didn't note carefully. As it was non-responsive to keyboard input, I considered it to be "locked up."
I updated the above post with the entries that seem to have worked. To verify: | Code: | $ dmesg |grep -i reserv |grep user: |grep -i 02[ae]0a1d00
[ 0.000000] user: [mem 0x00000002a0a1d000-0x00000002a0a1dfff] reserved
[ 0.000000] user: [mem 0x00000002e0a1d000-0x00000002e0a1dfff] reserved |
I ought to see if there's a chance re-seating the memory would help, or moving the 3rd 4GB stick to another slot, but not today. 22hrs+ of memtest only found the referenced errors on the first pass. After six passes, no new errors were found.
Memory is fairly cheap though, so if it is RAM, maybe an upgrade from 4x4GB to 2x16GB.
I'm also revisiting a Zen build. Cases are mostly awful. :(
_________________
Quis separabit? Quo animo? |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
Kernel & Hardware
