[9fans] Partition problems

[9fans] Partition problems

[Steve Simon]

Hi,

I run a modified labs kernel with a few bits of 9atom to support my atom
Motherboard.

I have added an ssd to the mirrored disks in my plan9 server.
The initialisation of devfs fails without a helpful error message (something I will sort out)

However the source of the problem seems to be boot/parts.c which fails to instantiate the last partition on my disks. This causes devfs to give up rather than continuing to retry other partitions. My kernel then cannot find its venti partition and dies with a panic.

If I boot using a different partition structure I can bring up the machine, and see that this one partition is indeed missing.

If I run disk/prep on the disk and rewrite the partition table unchanged, the missing partition appears.

Anyone seen problems with the partition table generation code in boot/part.c?
If not I will keep digging but I thought it was worth asking.

-Steve

Re: [9fans] Partition problems

[tlaronde]

Hello,

On Sun, Nov 20, 2016 at 11:07:03PM +0000, Steve Simon wrote:
> Hi,
>
> I run a modified labs kernel with a few bits of 9atom to support my atom
> Motherboard.
>
> I have added an ssd to the mirrored disks in my plan9 server.
> The initialisation of devfs fails without a helpful error message (something I will sort out)
>
> However the source of the problem seems to be boot/parts.c which fails to instantiate the last partition on my disks. This causes devfs to give up rather than continuing to retry other partitions. My kernel then cannot find its venti partition and dies with a panic.
>
> If I boot using a different partition structure I can bring up the machine, and see that this one partition is indeed missing.
>
> If I run disk/prep on the disk and rewrite the partition table unchanged, the missing partition appears.
>
> Anyone seen problems with the partition table generation code in boot/part.c?
> If not I will keep digging but I thought it was worth asking.

I had a problem (and still have) with the plan9 program because the code
rewrites the whole partition entry, recomputing the values of the start
and so on even with partitions "untouched" rendering the machine
unbootable because the starting block was not the correct one.

I have never found neither the time nor the incentive to correct this
part.

I don't know if this has anything to do with your problem at hand.

Best,
--
        Thierry Laronde <tlaronde +AT+ polynum +dot+ com>
                     http://www.kergis.com/
Key fingerprint = 0FF7 E906 FBAF FE95 FD89  250D 52B1 AE95 6006 F40C

Re: [9fans] Partition problems

[Iruatã Souza]

9front solved that 5 years ago.

On Mon, Nov 21, 2016 at 5:20 AM,  <tlaronde@polynum.com> wrote:
> Hello,
>
> On Sun, Nov 20, 2016 at 11:07:03PM +0000, Steve Simon wrote:
>> Hi,
>>
>> I run a modified labs kernel with a few bits of 9atom to support my atom
>> Motherboard.
>>
>> I have added an ssd to the mirrored disks in my plan9 server.
>> The initialisation of devfs fails without a helpful error message (something I will sort out)
>>
>> However the source of the problem seems to be boot/parts.c which fails to instantiate the last partition on my disks. This causes devfs to give up rather than continuing to retry other partitions. My kernel then cannot find its venti partition and dies with a panic.
>>
>> If I boot using a different partition structure I can bring up the machine, and see that this one partition is indeed missing.
>>
>> If I run disk/prep on the disk and rewrite the partition table unchanged, the missing partition appears.
>>
>> Anyone seen problems with the partition table generation code in boot/part.c?
>> If not I will keep digging but I thought it was worth asking.
>
> I had a problem (and still have) with the plan9 program because the code
> rewrites the whole partition entry, recomputing the values of the start
> and so on even with partitions "untouched" rendering the machine
> unbootable because the starting block was not the correct one.
>
> I have never found neither the time nor the incentive to correct this
> part.
>
> I don't know if this has anything to do with your problem at hand.
>
> Best,
> --
>         Thierry Laronde <tlaronde +AT+ polynum +dot+ com>
>                      http://www.kergis.com/
> Key fingerprint = 0FF7 E906 FBAF FE95 FD89  250D 52B1 AE95 6006 F40C
>

Re: [9fans] Partition problems

[Sigrid Haflinadóttir]

[-- Attachment #1: Type: text/plain, Size: 1987 bytes --]

He didn't ask about 9front.

On Wed, Nov 23, 2016 at 4:13 PM, Iruatã Souza <iru.muzgo@gmail.com> wrote:

> 9front solved that 5 years ago.
>
> On Mon, Nov 21, 2016 at 5:20 AM,  <tlaronde@polynum.com> wrote:
> > Hello,
> >
> > On Sun, Nov 20, 2016 at 11:07:03PM +0000, Steve Simon wrote:
> >> Hi,
> >>
> >> I run a modified labs kernel with a few bits of 9atom to support my atom
> >> Motherboard.
> >>
> >> I have added an ssd to the mirrored disks in my plan9 server.
> >> The initialisation of devfs fails without a helpful error message
> (something I will sort out)
> >>
> >> However the source of the problem seems to be boot/parts.c which fails
> to instantiate the last partition on my disks. This causes devfs to give up
> rather than continuing to retry other partitions. My kernel then cannot
> find its venti partition and dies with a panic.
> >>
> >> If I boot using a different partition structure I can bring up the
> machine, and see that this one partition is indeed missing.
> >>
> >> If I run disk/prep on the disk and rewrite the partition table
> unchanged, the missing partition appears.
> >>
> >> Anyone seen problems with the partition table generation code in
> boot/part.c?
> >> If not I will keep digging but I thought it was worth asking.
> >
> > I had a problem (and still have) with the plan9 program because the code
> > rewrites the whole partition entry, recomputing the values of the start
> > and so on even with partitions "untouched" rendering the machine
> > unbootable because the starting block was not the correct one.
> >
> > I have never found neither the time nor the incentive to correct this
> > part.
> >
> > I don't know if this has anything to do with your problem at hand.
> >
> > Best,
> > --
> >         Thierry Laronde <tlaronde +AT+ polynum +dot+ com>
> >                      http://www.kergis.com/
> > Key fingerprint = 0FF7 E906 FBAF FE95 FD89  250D 52B1 AE95 6006 F40C
> >
>
>

[-- Attachment #2: Type: text/html, Size: 2699 bytes --]

Re: [9fans] Partition problems

[Iruatã Souza]

[-- Attachment #1: Type: text/plain, Size: 2195 bytes --]

I forgot only mac os should be mentioned, sorry.

Em 23/11/2016 4:00 PM, "Sigrid Haflinadóttir" <
sigrid.haflinadottir@gmail.com> escreveu:

> He didn't ask about 9front.
>
> On Wed, Nov 23, 2016 at 4:13 PM, Iruatã Souza <iru.muzgo@gmail.com> wrote:
>
>> 9front solved that 5 years ago.
>>
>> On Mon, Nov 21, 2016 at 5:20 AM,  <tlaronde@polynum.com> wrote:
>> > Hello,
>> >
>> > On Sun, Nov 20, 2016 at 11:07:03PM +0000, Steve Simon wrote:
>> >> Hi,
>> >>
>> >> I run a modified labs kernel with a few bits of 9atom to support my
>> atom
>> >> Motherboard.
>> >>
>> >> I have added an ssd to the mirrored disks in my plan9 server.
>> >> The initialisation of devfs fails without a helpful error message
>> (something I will sort out)
>> >>
>> >> However the source of the problem seems to be boot/parts.c which fails
>> to instantiate the last partition on my disks. This causes devfs to give up
>> rather than continuing to retry other partitions. My kernel then cannot
>> find its venti partition and dies with a panic.
>> >>
>> >> If I boot using a different partition structure I can bring up the
>> machine, and see that this one partition is indeed missing.
>> >>
>> >> If I run disk/prep on the disk and rewrite the partition table
>> unchanged, the missing partition appears.
>> >>
>> >> Anyone seen problems with the partition table generation code in
>> boot/part.c?
>> >> If not I will keep digging but I thought it was worth asking.
>> >
>> > I had a problem (and still have) with the plan9 program because the code
>> > rewrites the whole partition entry, recomputing the values of the start
>> > and so on even with partitions "untouched" rendering the machine
>> > unbootable because the starting block was not the correct one.
>> >
>> > I have never found neither the time nor the incentive to correct this
>> > part.
>> >
>> > I don't know if this has anything to do with your problem at hand.
>> >
>> > Best,
>> > --
>> >         Thierry Laronde <tlaronde +AT+ polynum +dot+ com>
>> >                      http://www.kergis.com/
>> > Key fingerprint = 0FF7 E906 FBAF FE95 FD89  250D 52B1 AE95 6006 F40C
>> >
>>
>>
>

[-- Attachment #2: Type: text/html, Size: 3187 bytes --]