Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[Nico Schottelius]

[-- Attachment #1: Type: text/plain, Size: 867 bytes --]

Hello!

Just wanted to report of the following problems:

Compile Problems when selecting the following:
- Selected PCMCIA-SCSI
- Selected Framebuffer -> Aty128fb
- Sound / trident.c

Other bugs:
- devfs init is still missing -> /dev/vc/0 is the only console.
- floppy driver still broken
- ntfs sometimes crashes the system: working on a loopback file caused
  ntfs to report corruptions in the file system and this hangs system
  completly.

If you need more informations, just tell me. Currently I've some time
to debug parts of the kernel.

Nico

-- 
Changing mail address: please forget all known @pcsystems.de addresses.

Please send your messages pgp-signed and/or pgp-encrypted (don't encrypt mails
to mailing list!). If you don't know what pgp is visit www.gnupg.org.
(public pgp key: ftp.schottelius.org/pub/familiy/nico/pgp-key)

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[Anton Altaparmakov]

At 18:57 31/07/02, Nico Schottelius wrote:
>Just wanted to report of the following problems:
>
>Other bugs:
>- ntfs sometimes crashes the system: working on a loopback file caused
>   ntfs to report corruptions in the file system and this hangs system
>   completly.
>
>If you need more informations, just tell me. Currently I've some time
>to debug parts of the kernel.

I am interested in this ntfs report. Which way round was the loopback file? 
I.e. did you mount: mount -t ntfs -o loop somefile_on_a_non_ntfs_partition 
or did you mount: mount -t some_file_system -o loop 
somefile_on_an_ntfs_partion?

Can you send me the errors produced? If there is an oops, please decode and 
send it, too.

Also it may be useful to have the debug output from ntfs (depending on what 
the errors/oops say - they may be sufficient to pinpoint the problem), i.e. 
enable debugging when configuring the kernel, and then as root do: echo 1 > 
/proc/sys/fs/ntfs-debug. Note this will absolutely flood you with debug 
output so the system will run slow as hell... So it is best to only enable 
debug messages just before the error occurs if that is possible.

Thanks,

         Anton


-- 
   "I've not lost my mind. It's backed up on tape somewhere." - Unknown
-- 
Anton Altaparmakov <aia21 at cantab.net> (replace at with @)
Linux NTFS Maintainer / IRC: #ntfs on irc.openprojects.net
WWW: http://linux-ntfs.sf.net/ & http://www-stu.christs.cam.ac.uk/~aia21/

Re: Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[Nico Schottelius]

[-- Attachment #1: Type: text/plain, Size: 2509 bytes --]

Anton Altaparmakov [Tue, Jul 30, 2002 at 05:28:05PM +0100]:
> At 18:57 31/07/02, Nico Schottelius wrote:
> >Just wanted to report of the following problems:
> >
> >Other bugs:
> >- ntfs sometimes crashes the system: working on a loopback file caused
> >  ntfs to report corruptions in the file system and this hangs system
> >  completly.
> >
> >If you need more informations, just tell me. Currently I've some time
> >to debug parts of the kernel.
> 
> I am interested in this ntfs report. Which way round was the loopback file? 
> I.e. did you mount: mount -t ntfs -o loop somefile_on_a_non_ntfs_partition 
> or did you mount: mount -t some_file_system -o loop 
> somefile_on_an_ntfs_partion?

mount -t ntfs -o loop file.sav-on-ext2-or-on-xfs[when using 2.4.18] /mnt

> Can you send me the errors produced? If there is an oops, please decode and 
> send it, too.

The test I did was the following [may I call that test ?]:

cd /mnt; mkdir /ntfs_on_ext3; cp -r * /ntfs_on_ext3
While copying, with or without debug, the system hangs, but top only reports
7 % cpu load.

Copying the files results in a input / output error.

It has never been an oops and actually 2.5.29 does _not_ hangup anymore!
Still it stops to copy the files and aborts.
I am currently retrying with debug enabled...

> Also it may be useful to have the debug output from ntfs (depending on what 
> the errors/oops say - they may be sufficient to pinpoint the problem), i.e. 
> enable debugging when configuring the kernel, and then as root do: echo 1 > 
> /proc/sys/fs/ntfs-debug. Note this will absolutely flood you with debug 
> output so the system will run slow as hell... So it is best to only enable 
> debug messages just before the error occurs if that is possible.

oops. forget that above. Oh yes, ntfs is really reporting much.
You can find the output at ftp.schottelius.org:/pub/tmp, it's about
600k compressed.

I am really happy that this time the cp did not hald my system!

Nico

p.s.: what was the maximal file size on ext3 ? I just gunzipped a 4gb
      file (the ntfs image the whole story is about), which could not
      be transfered through scp/ftp in this size...

-- 
Changing mail address: please forget all known @pcsystems.de addresses.

Please send your messages pgp-signed and/or pgp-encrypted (don't encrypt mails
to mailing list!). If you don't know what pgp is visit www.gnupg.org.
(public pgp key: ftp.schottelius.org/pub/familiy/nico/pgp-key)

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[Anton Altaparmakov]

At 12:50 01/08/02, Nico Schottelius wrote:
>Anton Altaparmakov [Tue, Jul 30, 2002 at 05:28:05PM +0100]:
> > I am interested in this ntfs report. Which way round was the loopback 
> file?
> > I.e. did you mount: mount -t ntfs -o loop somefile_on_a_non_ntfs_partition
> > or did you mount: mount -t some_file_system -o loop
> > somefile_on_an_ntfs_partion?
>
>mount -t ntfs -o loop file.sav-on-ext2-or-on-xfs[when using 2.4.18] /mnt
>
> > Can you send me the errors produced? If there is an oops, please decode 
> and
> > send it, too.
>
>The test I did was the following [may I call that test ?]:
>
>cd /mnt; mkdir /ntfs_on_ext3; cp -r * /ntfs_on_ext3
>While copying, with or without debug, the system hangs, but top only reports
>7 % cpu load.
>
>Copying the files results in a input / output error.

Interesting.

>It has never been an oops and actually 2.5.29 does _not_ hangup anymore!
>Still it stops to copy the files and aborts.
>I am currently retrying with debug enabled...
>
> > Also it may be useful to have the debug output from ntfs (depending on 
> what
> > the errors/oops say - they may be sufficient to pinpoint the problem), 
> i.e.
> > enable debugging when configuring the kernel, and then as root do: echo 
> 1 >
> > /proc/sys/fs/ntfs-debug. Note this will absolutely flood you with debug
> > output so the system will run slow as hell... So it is best to only enable
> > debug messages just before the error occurs if that is possible.
>
>oops. forget that above. Oh yes, ntfs is really reporting much.
>You can find the output at ftp.schottelius.org:/pub/tmp, it's about
>600k compressed.

Where is it? It doesn't appear - I just looked...

>I am really happy that this time the cp did not hald my system!
>
>p.s.: what was the maximal file size on ext3 ? I just gunzipped a 4gb
>       file (the ntfs image the whole story is about), which could not
>       be transfered through scp/ftp in this size...

Sorry not sure. I think it is dependent on the fs block size you use...

Anton


-- 
   "I've not lost my mind. It's backed up on tape somewhere." - Unknown
-- 
Anton Altaparmakov <aia21 at cantab.net> (replace at with @)
Linux NTFS Maintainer / IRC: #ntfs on irc.openprojects.net
WWW: http://linux-ntfs.sf.net/ & http://www-stu.christs.cam.ac.uk/~aia21/

Re: Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[Nico Schottelius]

[-- Attachment #1.1: Type: text/plain, Size: 1438 bytes --]

Anton Altaparmakov [Thu, Aug 01, 2002 at 02:02:00PM +0100]:
> >It has never been an oops and actually 2.5.29 does _not_ hangup anymore!
> >Still it stops to copy the files and aborts.
> >I am currently retrying with debug enabled...
> >
> >> Also it may be useful to have the debug output from ntfs (depending on 
> >what
> >> the errors/oops say - they may be sufficient to pinpoint the problem), 
> >i.e.
> >> enable debugging when configuring the kernel, and then as root do: echo 
> >1 >
> >> /proc/sys/fs/ntfs-debug. Note this will absolutely flood you with debug
> >> output so the system will run slow as hell... So it is best to only 
> >enable
> >> debug messages just before the error occurs if that is possible.
> >
> >oops. forget that above. Oh yes, ntfs is really reporting much.
> >You can find the output at ftp.schottelius.org:/pub/tmp, it's about
> >600k compressed.
> 
> Where is it? It doesn't appear - I just looked...

I though i just cp'ed it...but I forget to move it from my home..
now you'll find it.


Nico

Btw, ntfs does not compile in 2.5.30 anymore. Neither with or without debug.

-- 
Changing mail address: please forget all known @pcsystems.de addresses.

Please send your messages pgp-signed and/or pgp-encrypted (don't encrypt mails
to mailing list!). If you don't know what pgp is visit www.gnupg.org.
(public pgp key: ftp.schottelius.org/pub/familiy/nico/pgp-key)

[-- Attachment #1.2: 2.5.30-compile-errors --]
[-- Type: text/plain, Size: 561 bytes --]

<M> NTFS file system support (read only) 
   [*]   NTFS debugging support 

make[3]: Entering directory `/usr/src/linux-2.5.30/fs/partitions'
gcc -Wp,-MD,./.check.o.d -D__KERNEL__ -I/usr/src/linux-2.5.30/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fomit-frame-pointer -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -march=i686 -nostdinc -iwithprefix include    -DKBUILD_BASENAME=check -DEXPORT_SYMTAB  -c -o check.o check.c
check.c: In function `devfs_register_partitions':
check.c:470: array subscript is not an integer


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[Nico Schottelius]

[-- Attachment #1: Type: text/plain, Size: 584 bytes --]

Btw, it was not a NTFS problem, just watched wrong!
It seems to be a general problem in the fs tree with devfs.

I am currently applying the console patch, but this doesn't fix the
error I posted in the last mail.

If I have some more time, I will try to fix it myself...

Nico

-- 
Changing mail address: please forget all known @pcsystems.de addresses.

Please send your messages pgp-signed and/or pgp-encrypted (don't encrypt mails
to mailing list!). If you don't know what pgp is visit www.gnupg.org.
(public pgp key: ftp.schottelius.org/pub/familiy/nico/pgp-key)

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[Anton Altaparmakov]

At 12:50 01/08/02, Nico Schottelius wrote:
>Anton Altaparmakov [Tue, Jul 30, 2002 at 05:28:05PM +0100]:
> > At 18:57 31/07/02, Nico Schottelius wrote:
> > >Just wanted to report of the following problems:
> > >
> > >Other bugs:
> > >- ntfs sometimes crashes the system: working on a loopback file caused
> > >  ntfs to report corruptions in the file system and this hangs system
> > >  completly.

I don't think the hangs had anything to do with ntfs. The io errors, I was 
able to reproduce on 2.4.19+ntfs but only when accessing compressed files. 
I have fixed this in the latest ntfs, at least the errors I saw are gone.

Could you give it a try and let me know if it solves your problem, too?

You can download by pulling from bk://linux-ntfs.bkbits.net/ntfs-tng-2.5 or 
by applying the patch I sent to LKML last night. I can forward it to you if 
you need it...

btw. I think it was compiler problem, so out of interest, which compiler 
are you using? I was using gcc-2.96-RH7.3-latest...

Thanks,

         Anton


-- 
   "I've not lost my mind. It's backed up on tape somewhere." - Unknown
-- 
Anton Altaparmakov <aia21 at cantab.net> (replace at with @)
Linux NTFS Maintainer / IRC: #ntfs on irc.openprojects.net
WWW: http://linux-ntfs.sf.net/ & http://www-stu.christs.cam.ac.uk/~aia21/

Re: Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[James Simmons]

> Just wanted to report of the following problems:
>
> Compile Problems when selecting the following:
> - Selected PCMCIA-SCSI

Been broken. A new driver is being worked on.

> - Selected Framebuffer -> Aty128fb

> Other bugs:
> - devfs init is still missing -> /dev/vc/0 is the only console.

Ug. That is partially fixed. I did get the other vc/X but only root can
access them. I have to talk to linus about the best solution here.

Re: Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[Nico Schottelius]

[-- Attachment #1: Type: text/plain, Size: 1174 bytes --]

James Simmons [Wed, Jul 31, 2002 at 10:22:05AM -0700]:
> > Just wanted to report of the following problems:
> >
> > Compile Problems when selecting the following:
> > - Selected PCMCIA-SCSI
> 
> Been broken. A new driver is being worked on.

good to know. I don't use it anyways, just wanted to report.

> > - Selected Framebuffer -> Aty128fb

btw, this fb driver seems to be really experimental or beta, as it does
nothing when loading on a mobility chip.

> > Other bugs:
> > - devfs init is still missing -> /dev/vc/0 is the only console.
> 
> Ug. That is partially fixed. I did get the other vc/X but only root can
> access them. I have to talk to linus about the best solution here.

What about the patch with con_init_devfs(); ? Isn't that simple and stupid
enough to use ? [this is the only way I can work with those kernels right
now..]

Nico

-- 
Changing mail address: please forget all known @pcsystems.de addresses.

Please send your messages pgp-signed and/or pgp-encrypted (don't encrypt mails
to mailing list!). If you don't know what pgp is visit www.gnupg.org.
(public pgp key: ftp.schottelius.org/pub/familiy/nico/pgp-key)

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Bugs in 2.5.28 [scsi/framebuffer/devfs/floppy/ntfs/trident]

[James Simmons]

> > Ug. That is partially fixed. I did get the other vc/X but only root can
> > access them. I have to talk to linus about the best solution here.
>
> What about the patch with con_init_devfs(); ? Isn't that simple and stupid
> enough to use ? [this is the only way I can work with those kernels right
> now..]

Try my patch now and tell me how it works.

http://www.transvirtual.com/~jsimmons/console.diff.gz

Against Linus current BK tree.