Re: 2.6.20-rc6 libata PATA ATAPI CDROM is not working

[Lennart Sorensen]

On Sat, Feb 17, 2007 at 11:33:53AM +0000, Joel Soete wrote:
> Hello Lennart,
> 
> Lennart Sorensen wrote:
> >On Tue, Feb 13, 2007 at 05:35:32PM +0000, Joel Soete wrote:
> >>A small update:
> >>your patch also works against 2.6.20
> >>
> >>but seems that open the door to numerous other pb:
> >>1/ pb to burn cd:
> >># md5sum cd060213.iso
> >>6a1248783a21722816b972aa9bae9d5e  cd060213.iso
> >>
> >># ll cd060213.iso
> >>-rwxr-xr-x 1 root root 3213312 Feb 13  2006 cd060213.iso
> >>
> >># dd if=/dev/sr0 bs=1 count=3213312 | md5sum
> >>dd: reading `/dev/sr0': Input/output error
> >>0337e9846d17779945c5c252d4f897f0  -
> >>3129344+0 records in
> >>3129344+0 records out
> >>3129344 bytes (3.1 MB) copied, 36.6963 seconds, 85.3 kB/s
> >>
> >>eventhought cdrecord seems to be successfull???
> >
> >Has that ever worked by any method?
> 
> Yes here was some test made some time ago (not so far):
> >> On Sun January 8 2006 09:28, you wrote:
> > cdrecord: No write mode specified.
> > cdrecord: Asuming -tao mode.
> > cdrecord: Future versions of cdrecord may have different drive dependent
> > defaults.
> > cdrecord: Continuing in 5 seconds...
> > cdrecord: Warning: Running on Linux-2.6.15-686
> > cdrecord: There are unsettled issues with Linux-2.5 and newer.
> > cdrecord: If you have unexpected problems, please try Linux-2.4 or 
> Solaris.
> > scsidev: '/dev/hdb'
> > devname: '/dev/hdb'
> > scsibus: -2 target: -2 lun: -2
> > Warning: Open by 'devname' is unintentional and not supported.
> > Linux sg driver version: 3.5.27
> > cdrecord: Warning: using inofficial version of libscg (ubuntu-0.8ubuntu1
> > '@(#)scsitransp.c      1.91 04/06/17 Copyright 1988,1995,2000-2004 J.
> > Schilling').
> > SCSI buffer size: 64512
> > cdrecord: Asked for SCSI I/O buffer size 64512 bytes, could only get 
> 20480.
> > Cdrecord-Clone 2.01.01a03 (i686-pc-linux-gnu) Copyright (C) 1995-2005
> > Joerg Schilling
> > NOTE: this version of cdrecord is an inofficial (modified) release of
> > cdrecord
> >       and thus may have bugs that are not present in the original version.
> >       Please send bug reports and support requests to
> > <cdrtools@xxxxxxxxxxxxxxxxxxx>.
> >       The original author should not be bothered with problems of this
> > version.
> >
> > TOC Type: 1 = CD-ROM
> > Using libscg version 'ubuntu-0.8ubuntu1'.
> > atapi: 1
> > Device type    : Removable CD-ROM
> > Version        : 0
> > Response Format: 1
> > Vendor_info    : 'PHILIPS '
> > Identifikation : 'CDD3610 CD-R/RW '
> > Revision       : '3.09'
> > Device seems to be: Generic mmc CD-RW.
> > Using generic SCSI-3/mmc   CD-R/CD-RW driver (mmc_cdr).
> > Driver flags   : MMC SWABAUDIO
> > Supported modes: TAO PACKET RAW/R16
> > Drive buf size : 786432 = 768 KB
> > FIFO size      : 4194304 = 4096 KB
> > Track 01: data   546 MB
> > Total size:      627 MB (62:11.78) = 279884 sectors
> > Lout start:      628 MB (62:13/59) = 279884 sectors
> > Current Secsize: 2048
> >   ATIP start of lead in:  -11637 (97:26/63)
> >   ATIP start of lead out: 337350 (75:00/00)
> > Disk type:    Phase change
> > Manuf. index: 3
> > Manufacturer: CMC Magnetics Corporation
> > Blocks total: 337350 Blocks current: 337350 Blocks remaining: 57466
> > Starting to write CD/DVD at speed 2 in real TAO mode for single session.
> > Waiting for reader process to fill input buffer ... input buffer ready.
> > Performing OPC...
> > Starting new track at sector: 0
> > Track 01:  546 of  546 MB written (fifo 100%) [buf  94%]   2.0x.
> > Track 01: Total bytes read/written: 573198336/573198336 (279882 sectors).
> > Writing  time: 1878.179s
> > Average write speed   2.0x.
> > Min drive buffer fill was 94%
> > Fixating...
> > Fixating time:  167.622s
> > cdrecord: fifo had 27989 puts and 27989 gets.
> > cdrecord: fifo was 0 times empty and 27757 times full, min fill was 98%.
> >
> [snip]
> >  # readcd dev=/dev/hdb f=- | md5sum
> > Read  speed:  1059 kB/s (CD   6x, DVD  0x).
> > Write speed:   353 kB/s (CD   2x, DVD  0x).
> > Capacity: 279884 Blocks = 559768 kBytes = 546 MBytes = 573 prMB
> > Sectorsize: 2048 Bytes
> > Copy from SCSI (0,1,0) disk to file '-'
> > end:    279884
> > readcd: Success. read_g1: scsi sendcmd: no error
> > CDB:  28 00 00 04 45 48 00 00 04 00
> > status: 0x2 (CHECK CONDITION)
> > Sense Bytes:
> > Sense Key: 0xFFFFFFFF [], Segment 0
> > Sense Code: 0x00 Qual 0x00 (no additional sense information) Fru 0x0
> > Sense flags: Blk 0 (not valid)
> > cmd finished after 0.005s timeout 40s
> > readcd: Success. Cannot read source disk
> > readcd: Retrying from sector 279880.
> > ....~~-~~~+~~~-~~~+~~~-~~~+~

Hmm, readcd was trying to read 279884 blocks, while cdrecord said it
wrote 279882 blocks.

> [snip]
> > # ll
> > total 1150848
> > [...]
> > -rw-r----- 1 root root   604082176 Jan 15 12:22 hppa-cvs-20060115.iso
> >
> > i.e. Block_Number = 604082176 / 2048 = 294962
> >
> >
> > # md5sum hppa-cvs-20060115.iso
> > 1141489a8b914daff5cca790882fe277  hppa-cvs-20060115.iso
> >
> >  # dd bs=2048 count=294962 if=/dev/hdb  | md5sum
> > 294962+0 records in
> > 294962+0 records out
> > 604082176 bytes (604 MB) copied, 676.972 seconds, 892 kB/s
> > 1141489a8b914daff5cca790882fe277  -
> >
> > ;<)
> more same method here:
> # dd if=/dev/hdd bs=2048 count=1569 | md5sum
> dd: reading `/dev/hdd': Input/output error
> 1528+0 records in
> 1528+0 records out
> 3129344 bytes (3.1 MB) copied, 10.1961 seconds, 307 kB/s
> 0337e9846d17779945c5c252d4f897f0  -

There always seems to have been any issue with reading blocks from a cd
getting broken by the kernel doing readahead and getting failures by
trying to read past the end of the disc even though the user never asked
it to.  Sometimes your size just happens to hit a safe block size so the
readahead doesn't break.

> is the same wrong results.
> 
> mmm I always used successfully dd method at my office with scsi cdrom drive 
> (even for bootable disk).

Well it seems some disks can be read that way, others can not.

> It came from an age (see >> brw-rw---- 1 root disk 8, 32 Dec  1  2001 
> /dev/sdc)
> when 15Gb disk was enough to test severall Linux distro ;-)
> At the same time I was also working with a distro runing on a x486 100Mhz, 
> a few ram and 128Mb of disk ;-)

Well that makes some sense.  I always though testing a distribution
meant installing it and using it a bit, andif it wasn't right, you
installed the next one.  I have never had more than one installed at a
time myself.

--
Len Sorensen
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/