Re: [gentoo-user] Re: Externel drive should be /dev/sda1, but /dev/sda1 does not exist [SOLVED somehow]

2010-01-02 Thread Robert Bridge
On Sat, Jan 2, 2010 at 2:37 PM, Michael Sullivan  wrote:
> I guess I should have checked before I sent that last post because
> now /dev/sda is there.  Now, how do I make it be there when I first boot
> into Linux?

If your / filesystem is on the USB drive, you need to insert a delay
before mounting it to allow the USB system to stabilise. This is
either done using an initrd or by passing a command line option to the
kernel in the boot loader. Essentially, there is a race condition
between USB initialisation and mounting /.

Appending "rootdelay=10" to the kernel line in your bootloader should solve it.

http://www.reactivated.net/weblog/archives/2005/11/booting-linux-userland-from-an-external-usb-flash-disk/
has more details of the problem.



Re: [gentoo-user] Re: Externel drive should be /dev/sda1, but /dev/sda1 does not exist [SOLVED somehow]

2010-01-02 Thread Michael Sullivan
I guess I should have checked before I sent that last post because
now /dev/sda is there.  Now, how do I make it be there when I first boot
into Linux?

On Sat, 2010-01-02 at 08:35 -0600, Michael Sullivan wrote:
> On Fri, 2010-01-01 at 11:37 -0800, walt wrote:
> > On 01/01/2010 05:48 AM, Michael Sullivan wrote:
> > > Hello
> > >
> > > My wife's computer is pretty slow, so I've attached and old hard drive
> > > into a hard drive enclosure and hooked it into her USB port for
> > > additional swap space.  It used to work.  The swap space is supposed to
> > > be /dev/sda1.  The problem is that for some reason when I rebooted this
> > > morning with a new kernel, /dev/sda does not exist anymore...
> > 
> > Hm.  So the only thing you changed was the new kernel?  Might help to
> > know why you built the new kernel.  What problem were you solving by
> > doing it?
> > 
> OK here goes.  I built the new kernel because with the old one /dev/sda
> didn't seem to exist when I know it should.  And besides, there was a
> newer kernel marked as stable and aren't we always supposed to use the
> newest stable kernel for security reasons, unless there's something that
> we need done that the new kernel doesn't do?
> 
> > I would try booting the machine without the USB swap disk and then
> > hotplug it when the machine is already running.  What does dmesg say
> > then?  Can be simpler to interpret when you know exactly which lines
> > were printed in response to the newly connected drive.
> 
> 
> I rebooted the machine with the drive unplugged and ran dmesg:
> 
> catherine ~ # cat dmesg.txt
> Initializing cgroup subsys cpuset
> Linux version 2.6.31-gentoo-r6 (r...@catherine) (gcc version 4.3.4
> (Gentoo 4.3.4 p1.0, pie-10.1.5) ) #1 SMP Thu Dec 31 19:13:07 CST 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:  - 0009f800 (usable)
>  BIOS-e820: 0009f800 - 000a (reserved)
>  BIOS-e820: 000f - 0010 (reserved)
>  BIOS-e820: 0010 - 0dff (usable)
>  BIOS-e820: 0dff - 0dff3000 (ACPI NVS)
>  BIOS-e820: 0dff3000 - 0e00 (ACPI data)
>  BIOS-e820: fec0 - fec01000 (reserved)
>  BIOS-e820: fee0 - fee01000 (reserved)
>  BIOS-e820:  - 0001 (reserved)
> DMI 2.2 present.
> Phoenix BIOS detected: BIOS may corrupt low RAM, working around it.
> e820 update range:  - 0001 (usable) ==>
> (reserved)
> last_pfn = 0xdff0 max_arch_pfn = 0x10
> MTRR default type: uncachable
> MTRR fixed ranges enabled:
>   0-9 write-back
>   A-B uncachable
>   C-C7FFF write-protect
>   C8000-E uncachable
>   F-F write-protect
> MTRR variable ranges enabled:
>   0 base 0 mask FF000 write-back
>   1 base 00E00 mask FFE00 uncachable
>   2 base 0D000 mask FF800 write-combining
>   3 disabled
>   4 disabled
>   5 disabled
>   6 disabled
>   7 disabled
> initial memory mapped : 0 - 0080
> init_memory_mapping: -0dff
>  00 - 40 page 4k
>  40 - 000dc0 page 2M
>  000dc0 - 000dff page 4k
> kernel direct mapping tables up to dff @ 1-15000
> RAMDISK: 0dcc5000 - 0dfdf8aa
> ACPI: RSDP 000f7170 00014 (v00 AWARD )
> ACPI: RSDT 0dff3000 0002C (v01 AWARD  AWRDACPI 42302E31 AWRD )
> ACPI: FACP 0dff3040 00074 (v01 AWARD  AWRDACPI 42302E31 AWRD )
> ACPI: DSDT 0dff30c0 038FA (v01 AWARD  AWRDACPI 1000 MSFT 010E)
> ACPI: FACS 0dff 00040
> ACPI: APIC 0dff69c0 0005A (v01 AWARD  AWRDACPI 42302E31 AWRD )
> ACPI: Local APIC address 0xfee0
> 0MB HIGHMEM available.
> 223MB LOWMEM available.
>   mapped low ram: 0 - 0dff
>   low ram: 0 - 0dff
>   node 0 low ram:  - 0dff
>   node 0 bootmap 00011000 - 00012c00
> (9 early reservations) ==> bootmem [00 - 000dff]
>   #0 [00 - 001000]   BIOS data page ==> [00 -
> 001000]
>   #1 [001000 - 002000]EX TRAMPOLINE ==> [001000 -
> 002000]
>   #2 [006000 - 007000]   TRAMPOLINE ==> [006000 -
> 007000]
>   #3 [10 - 637f00]TEXT DATA BSS ==> [10 -
> 637f00]
>   #4 [000dcc5000 - 000dfdf8aa]  RAMDISK ==> [000dcc5000 -
> 000dfdf8aa]
>   #5 [09f800 - 10]BIOS reserved ==> [09f800 -
> 10]
>   #6 [638000 - 63a06a]  BRK ==> [638000 -
> 63a06a]
>   #7 [01 - 011000]  PGTABLE ==> [01 -
> 011000]
>   #8 [011000 - 013000]  BOOTMAP ==> [011000 -
> 013000]
> found SMP MP-table at [c00f57a0] f57a0
> Zone PFN ranges:
>   DMA  0x001