🚀 go-pugleaf

In article <slrn5381sv.mr.davep@hagbard.demon.co.uk>, Dave Pearson
<davep@hagbard.demon.co.uk> writes
>
>I've just upgraded my Linux system this weekend, and one of the things
>I added was an IDE CDROM. Everything seems to be working fine except
>for one very annoying problem that I can't even begin to work out the
>cause of.
>
>I can read "small" files from the CD with no problems, but "large"
>files seem to cause a problem. I'm stil trying to work out the line
>between "small" and "large".
>
>
>
>Has anyone seen this kind of problem before, or do you know what the
>cause would be?
>
>I'm currently running Slackware 3.0 1.2.13 kernel.

I had a similar problem which was solved by adding the line:
  append = "hda=serialize"
to my /etc/lilo.conf file.
It tells you about this in the kernel source code directory for (e)ide
drivers. I believe that the problem does not exist with 2.0 kernels.

--------------------------------------------------------- 
Malcolm R Peacock              |      class need
malcolm@mpeacock.demon.co.uk   |      {  
                               |         friend indeed;
                               |      };

Dave Pearson (davep@hagbard.demon.co.uk) wrote:

: I've just upgraded my Linux system this weekend, and one of the things
: I added was an IDE CDROM. Everything seems to be working fine except
: for one very annoying problem that I can't even begin to work out the
: cause of.

: I can read "small" files from the CD with no problems, but "large"
: files seem to cause a problem. I'm stil trying to work out the line

At a guess you've got a problem with automatic text file translation,
where dos style CRLF's are mangled to unix standards.
maybe have a close look at mount manpage?
It may also be compile time option

: between "small" and "large".

: The problem is this. If I copy a "large" tgz file, and then try to
: (t)est or or e(x)tract the contents, it will get so far and then come
: up with "tar: Skipping to next file header...". The file I've copied
: is the correct length, but it seems as if the file has possibly become
: corrupted during the read and/or the copy. As I say, I can read a
: large file with tar on the CD, or I can copy it and the same will
: happen.

: Has anyone seen this kind of problem before, or do you know what the
: cause would be?

: IDE devices (as found during bootup) in the machine are:

: hda: Conner Peripherals 420MB - CFS420A, 406MB w/64KB Cache,
:                         CHS=826/16/63, MaxMult=64
: hdb: Conner Peripherals 540MB - CFS540A, 516MB w/64KB Cache, LBA,
:                         CHS=1050/16/63, MaxMult=16
: hdc: FUJITSU M1636TAU, 1226MB w/128KB Cache, LBA, CHS=2490/16/63, MaxMult=16
: hdd: TOSHIBA CD-ROM XM-5602B, ATAPI, CDROM drive

: I'm currently running Slackware 3.0 1.2.13 kernel. Part of the reason
: for this upgrade (including the third hard drive which is also new)
: was so I could play with one of the new 2.0.x distributions (either
: RedHat or Debian probably) and then "move over". However, with a CDROM
: drive that only seems to be able to read "small" files, that's gonna
: be tricky.

: Thanks in advance for any help or assistance you can provide.

: --
: Take a look in Hagbard's World: |     w3ng - The WWW Norton Guide reader.
: http://www.acemake.com/hagbard  |  ng2html - The NG to HTML converter.
: Resist UK Internet Censorship:  |       eg - Norton Guide reader for OS/2.
: http://babylon.ivision.co.uk    |   dgscan - DGROUP scanner for Clipper.



--
Ian Stirling.
Homepage: http://www.mauve.demon.co.uk/
Very under construction, Await eagerly homepage that sings it's contents
to you via reelawdioo, using OfFkeyOtron Text to song converter.

Get off a shot FAST, this upsets him long enbough to let yo make your
second shot perfect.
Robert A Heinlein.

Malcolm R Peacock <malcolm@mpeacock.demon.co.uk> wrote:

> >I can read "small" files from the CD with no problems, but "large"
> >files seem to cause a problem. I'm stil trying to work out the line
> >between "small" and "large".
> 
> I had a similar problem which was solved by adding the line:
>   append = "hda=serialize"
> to my /etc/lilo.conf file.
> It tells you about this in the kernel source code directory for (e)ide
> drivers. I believe that the problem does not exist with 2.0 kernels.

Cheers, I must have missed that bit, I'll check it out. 

Also, I've found I can make the problem go away if I mount the CDROM
with blocksize set to 1024. I'm not quite sure why this would be, I'm
still digging around so see if I can find out why. But, I'm also happy
that I can now use the CD with no problems. :-)

-- 
Take a look in Hagbard's World: |     w3ng - The WWW Norton Guide reader.
http://www.acemake.com/hagbard  |  ng2html - The NG to HTML converter.
Resist UK Internet Censorship:  |       eg - Norton Guide reader for OS/2.
http://babylon.ivision.co.uk    |   dgscan - DGROUP scanner for Clipper.

Ian Stirling <root@mauve.demon.co.uk> wrote:

> : I can read "small" files from the CD with no problems, but "large"
> : files seem to cause a problem. I'm stil trying to work out the line
> 
> At a guess you've got a problem with automatic text file
> translation, where dos style CRLF's are mangled to unix standards.
> maybe have a close look at mount manpage?  It may also be compile
> time option

Hmm, possible, but I would assume that this would not differ from CD
drive to CD drive (the settings I'm using for the new IDE drive are
exactly the same as I was using for my old CDU31A, well, apart from
the drive names <g>)?

However, thanks for the idea, I'll consider this. Also, as I've posted
elsewhere, I've now found that I can make it work if I mount with the
blocksize option set to 1024, I'm still trying to work out how/why
this makes a difference.

Again, thanks for your input.

-- 
Take a look in Hagbard's World: |     w3ng - The WWW Norton Guide reader.
http://www.acemake.com/hagbard  |  ng2html - The NG to HTML converter.
Resist UK Internet Censorship:  |       eg - Norton Guide reader for OS/2.
http://babylon.ivision.co.uk    |   dgscan - DGROUP scanner for Clipper.