Asrock P4V88 and 4 GB barrier – Part 3

Mit dem auf dem DrupalCamp gewonnenen BuchUsing Drupal” hab ich heute endlich mal CCK und Views funktionierend hinbekommen.

Eigentlich ist das gar nicht so schwierig, wenn man die Methode erst einmal verstanden hat, aber so ist das ja meistens. Ich kann das Buch aber schonmal bedenkenlos jedem empfehlen, der einen Einstieg in Drupal sucht und mit Hilfe von Praxisbeispielen die Möglichkeiten von Drupal erkunden möchte.

Published

13 comments

  1. Ingo, the problem is likely that the MTRRs are not set up correctly (another BIOS bug unfortunately) which means that your main memory is likely being marked as uncachable, resulting in, well, no use of your caches.

    The reason Xen sees all your memory is because it forces the use of PAE. If you select the HIGHMEM64G option in a stock kernel config, you'll see the same effect quite likely.

    Please feel free to shoot me an email if you'd like to try the upstream kernel and track getting it fixed there.

    regards, Kyle

  2. No cache could explain the slowness as well… hmmm…

    Of course it would be nice to have >3 GB usuable without being that slow.

    Although I hope as well that Asrock support will fix that problem as well, I've sent you an email… 🙂

  3. Hello
    I have the same (similar, because i am loosing only 29mb ram. Excuse me for my bad english. I have 4×1 gb (2 kits) identically ram memories. My motherboard is asus p4-p800 (latest stable bios), with 4 ram slots, with maximum installable 4gb memory with *. I know for the problem with 4gb limitation of the south(i think it was south) bridge. So:
    # dmesg|more
    BIOS EBDA/lowmem at: 0009ac00/0009ac00
    Linux version 2.6.28.3 (root@kombina.org) (gcc version 4.3.2 (Debian 4.3.2-1.1) ) #1 SMP
    PREEMPT Mon Feb 2 23:21:59 EET 2009
    KERNEL supported cpus:
    Intel GenuineIntel
    AMD AuthenticAMD
    NSC Geode by NSC
    Cyrix CyrixInstead
    Centaur CentaurHauls
    Transmeta GenuineTMx86
    Transmeta TransmetaCPU
    UMC UMC UMC UMC
    BIOS-provided physical RAM map:
    BIOS-e820: 0000000000000000 – 000000000009fc00 (usable)
    BIOS-e820: 000000000009fc00 – 00000000000a0000 (reserved)
    BIOS-e820: 00000000000e8000 – 0000000000100000 (reserved)
    BIOS-e820: 0000000000100000 – 00000000bfdb0000 (usable)
    BIOS-e820: 00000000bfdb0000 – 00000000bfdc0000 (ACPI data)
    BIOS-e820: 00000000bfdc0000 – 00000000bfdf0000 (ACPI NVS)
    BIOS-e820: 00000000bfdf0000 – 00000000bfe00000 (reserved)
    BIOS-e820: 00000000ffb80000 – 0000000100000000 (reserved)
    DMI 2.3 present.
    AMI BIOS detected: BIOS may corrupt low RAM, working it around.
    last_pfn = 0xbfdb0 max_arch_pfn = 0x100000
    x86 PAT enabled: cpu 0, old 0x7040600070406, new 0x7010600070106
    WARNING: BIOS bug: CPU MTRRs don't cover all of memory, losing 29MB of RAM.
    ————[ cut here ]————
    WARNING: at arch/x86/kernel/cpu/mtrr/main.c:1662 mtrr_trim_uncached_memory+0x2ca/0x306()
    Modules linked in:
    Pid: 0, comm: swapper Not tainted 2.6.28.3 #1
    Call Trace:
    [] warn_on_slowpath+0x40/0x61
    [] up+0x9/0x2a
    [] _spin_unlock_irqrestore+0x18/0x22
    [] release_console_sem+0x171/0x19e
    [] e820_update_range_map+0x13a/0x1a0
    [] printk+0xe/0x17
    [] mtrr_trim_uncached_memory+0x2ca/0x306
    [] e820_end_pfn+0x9e/0xa8
    [] setup_arch+0x399/0x46c
    [] reserve_ebda_region+0x68/0x6e
    [] start_kernel+0x54/0x236
    —[ end trace 4eaa2a86a8e2da22 ]—
    update e820 for mtrr
    modified physical RAM map:
    modified: 0000000000000000 – 0000000000010000 (reserved)
    modified: 0000000000010000 – 000000000009fc00 (usable)
    modified: 000000000009fc00 – 00000000000a0000 (reserved)
    modified: 00000000000e8000 – 0000000000100000 (reserved)
    modified: 0000000000100000 – 00000000be000000 (usable)
    modified: 00000000be000000 – 00000000bfdb0000 (reserved)
    modified: 00000000bfdb0000 – 00000000bfdc0000 (ACPI data)
    modified: 00000000bfdc0000 – 00000000bfdf0000 (ACPI NVS)
    modified: 00000000bfdf0000 – 00000000bfe00000 (reserved)
    modified: 00000000ffb80000 – 0000000100000000 (reserved)
    last_pfn = 0xbe000 max_arch_pfn = 0x100000
    kernel direct mapping tables up to 377fe000 @ 10000-16000
    ACPI: RSDP 000FAD80, 0021 (r2 ACPIAM)
    ACPI: XSDT BFDB0100, 003C (r1 A M I OEMXSDT 9000505 MSFT 97)
    ACPI: FACP BFDB0290, 00F4 (r3 A M I OEMFACP 9000505 MSFT 97)
    ACPI: DSDT BFDB03F0, 3833 (r1 A0030 A0030011 11 INTL 2002026)
    ACPI: FACS BFDC0000, 0040
    ACPI: APIC BFDB0390, 005C (r1 A M I OEMAPIC 9000505 MSFT 97)
    ACPI: OEMB BFDC0040, 003F (r1 A M I OEMBIOS 9000505 MSFT 97)
    ACPI: Local APIC address 0xfee00000
    2152MB HIGHMEM available.
    887MB LOWMEM available.
    mapped low ram: 0 – 377fe000
    low ram: 00000000 – 377fe000
    bootmap 00012000 – 00018f00
    (8 early reservations) ==> bootmem [0000000000 – 00377fe000]
    #0 [0000000000 – 0000001000] BIOS data page ==> [0000000000 – 0000001000]
    #1 [0000001000 – 0000002000] EX TRAMPOLINE ==> [0000001000 – 0000002000]
    #2 [0000006000 – 0000007000] TRAMPOLINE ==> [0000006000 – 0000007000]
    #3 [0000200000 – 000055f208] TEXT DATA BSS ==> [0000200000 – 000055f208]
    #4 [0000560000 – 0000563000] INIT_PG_TABLE ==> [0000560000 – 0000563000]
    #5 [000009ac00 – 0000100000] BIOS reserved ==> [000009ac00 – 0000100000]
    #6 [0000010000 – 0000012000] PGTABLE ==> [0000010000 – 0000012000]
    #7 [0000012000 – 0000019000] BOOTMAP ==> [0000012000 – 0000019000]
    Zone PFN ranges:
    DMA 0x00000010 -> 0x00001000
    Normal 0x00001000 -> 0x000377fe
    HighMem 0x000377fe -> 0x000be000
    Movable zone start PFN for each node
    early_node_map[2] active PFN ranges
    0: 0x00000010 -> 0x0000009f
    0: 0x00000100 -> 0x000be000
    On node 0 totalpages: 778127
    free_area_init_node: node 0, pgdat c04bfa00, node_mem_map c1000200
    DMA zone: 32 pages used for memmap
    DMA zone: 0 pages reserved
    DMA zone: 3951 pages, LIFO batch:0
    Normal zone: 1744 pages used for memmap
    Normal zone: 221486 pages, LIFO batch:31
    HighMem zone: 4305 pages used for memmap
    HighMem zone: 546609 pages, LIFO batch:31
    Movable zone: 0 pages used for memmap
    ACPI: PM-Timer IO Port: 0x808
    ACPI: Local APIC address 0xfee00000
    ACPI: LAPIC (acpi_id[0x01] lapic_id[0x00] enabled)
    ACPI: LAPIC (acpi_id[0x02] lapic_id[0x01] enabled)
    ACPI: IOAPIC (id[0x02] address[0xfec00000] gsi_base[0])
    IOAPIC[0]: apic_id 2, version 32, address 0xfec00000, GSI 0-23
    ACPI: INT_SRC_OVR (bus 0 bus_irq 0 global_irq 2 dfl dfl)
    ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 high level)
    ACPI: IRQ0 used by override.
    ACPI: IRQ2 used by override.
    ACPI: IRQ9 used by override.
    Enabling APIC mode: Flat. Using 1 I/O APICs
    SMP: Allowing 2 CPUs, 0 hotplug CPUs
    Allocating PCI resources starting at c0000000 (gap: bfe00000:3fd80000)
    PERCPU: Allocating 32768 bytes of per cpu data
    NR_CPUS: 2, nr_cpu_ids: 2, nr_node_ids 1

  4. Oh, well, I wouldn't care about loosing 29 MB… Kyle, the other commenter, tries to fix the problem in mainline kernel. Maybe it will fix your problem as well? If Kyle and you agree, I would sent you the others email adresses…

  5. Probably way off base with this so I'm just throwing it up for a quick double check.

    When changing my BIOS IDE/SATA setting the kernel was complaining about having to wait for DMA when both IDE channels was set to Auto Detect
    I saw a significant performance gain with my system when I changed my IDE settings to:

    IDE 1 Primary channel
    User Defind
    PO4
    DMA2

    IDE 2 Primary channel
    User Defined
    PO4
    DMA4

    for my two cd-roms attached to the IDE controllers
    both secondary channels are empty and
    hard disks are attached to SATA

    Hope this is helpful…

  6. I have a system with 4G of ram and I have to boot it with
    dom0_mem=256M

    in order for things to work right.

  7. Well, that wouldn't make much sense for me as I would use dom0 as my main desktop system. I intended to use Xen VMs for testing purposes.

    I still hope that things can be sorted out, one way or the other…

  8. Asrock P4V88 and 4 GB barrier – Part 4
    After being able to plug in all four 1G DIMMs, the machine recognizes them as 4x 1GB dual-ranked. Sadly only 3.1 GB are usuable., though. But more annoying are random lockups from the radeon driver: Feb 8 12:01:50 muaddib kernel: [239359.738458] [drm:

  9. I just tried to max out m P4v88 and also ran into problems with my radeon card (9600). They resolved when I set my AGP voltage to “high.”

  10. Asrock P4V88 and 4 GB bar
    Hello. How you got boot Asrock P4V88 with 4 Gb RAM (4×1 Gb PC2700 (DDR333)? Which setting in BIOS (latest 1.80 version)? I’m trying it but system not boot. 🙁

Comments are closed.