LinuxLists.cc - Can't burn DVD under 2.5.59 with ide-cd

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

>From: Gregoire Favre <[email protected]>

>after reporting this problem, someone pointed me that I should try
>without DAO, I have tried this:

>mkisofs -dvd-video -V $1 $2 | cdrecord-prodvd driveropts=burnfree -dummy -v dev=/dev/hdc fs=64m speed=1 -eject tsize={$SIZE}s -

>And got this:

>Cdrecord-ProDVD-Clone 2.0 (i586-pc-linux-gnu) Copyright (C) 1995-2002 J?rg Schilling
>Unlocked features: ProDVD Clone
>Limited features: speed
>This copy of cdrecord is licensed for: private/research/educational_non-commercial_use
>TOC Type: 1 = CD-ROM
>scsidev: '/dev/hdc'
>devname: '/dev/hdc'
>scsibus: -2 target: -2 lun: -2
>Warning: Open by 'devname' is unintentional and not supported.
>Linux sg driver version: 3.5.27
>Using libscg version 'schily-0.7'
>Driveropts: 'burnfree'
>atapi: 1
>Device type : Removable CD-ROM
>Version : 2
>Response Format: 2
>Capabilities :
>Vendor_info : 'SONY '
>Identifikation : 'DVD RW DRU-500A '
>Revision : '1.0f'
>Device seems to be: Generic mmc2 DVD-R/DVD-RW.
>Using generic SCSI-3/mmc-2 DVD-R/DVD-RW driver (mmc_dvd).
>Driver flags : DVD SWABAUDIO BURNFREE
>Supported modes: TAO PACKET SAO SAO/R96R RAW/R96R
>Drive buf size : 8126464 = 7936 KB
>FIFO size : 67108864 = 65536 KB
>Track 01: data 4001 MB
>Total size: 4001 MB = 2048512 sectors
>Current Secsize: 2048
>Blocks total: 2298496 Blocks current: 2298496 Blocks remaining: 249984
>Starting to write CD/DVD at speed 1 in dummy TAO mode for single session.
>Last chance to quit, starting dummy write in 9 seconds. 0.24% done, estimate finish Tue Jan 21 23:09:02 2003
> 8 seconds. 0.49% done, estimate finish Tue Jan 21 23:09:03 2003
> 7 seconds. 0.73% done, estimate finish Tue Jan 21 23:06:46 2003
> 6 seconds. 0.98% done, estimate finish Tue Jan 21 23:07:21 2003
> 5 seconds. 1.22% done, estimate finish Tue Jan 21 23:07:41 2003
> 4 seconds. 1.46% done, estimate finish Tue Jan 21 23:07:55 2003
> 0 seconds. Operation starts.
>Waiting for reader process to fill input buffer ... input buffer ready.
>BURN-Free is ON.
>Starting new track at sector: 0
>Track 01: 4 of 4001 MB written (fifo 96%) 16.1x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
>CDB: 2A 00 00 00 08 B8 00 00 1F 00
>status: 0x1 (GOOD STATUS)
>resid: 63488
>cmd finished after 0.008s timeout 100s

I can't tell you what happened because the kernel is broken :-(

If you fix the kernel, you will get a readble error message,

J?rg

EMail:[email protected] (home) J?rg Schilling D-13353 Berlin
[email protected] (uni) If you don't have iso-8859-1
[email protected] (work) chars I am J"org Schilling
URL: http://www.fokus.fhg.de/usr/schilling ftp://ftp.berlios.de/pub/schily

2003-01-22 07:57:08

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Wed, Jan 22 2003, Joerg Schilling wrote:
> >From: Gregoire Favre <[email protected]>
>
> >after reporting this problem, someone pointed me that I should try
> >without DAO, I have tried this:
>
> >mkisofs -dvd-video -V $1 $2 | cdrecord-prodvd driveropts=burnfree -dummy -v dev=/dev/hdc fs=64m speed=1 -eject tsize={$SIZE}s -
>
> >And got this:
>
> >Cdrecord-ProDVD-Clone 2.0 (i586-pc-linux-gnu) Copyright (C) 1995-2002 J?rg Schilling
> >Unlocked features: ProDVD Clone
> >Limited features: speed
> >This copy of cdrecord is licensed for: private/research/educational_non-commercial_use
> >TOC Type: 1 = CD-ROM
> >scsidev: '/dev/hdc'
> >devname: '/dev/hdc'
> >scsibus: -2 target: -2 lun: -2
> >Warning: Open by 'devname' is unintentional and not supported.
> >Linux sg driver version: 3.5.27
> >Using libscg version 'schily-0.7'
> >Driveropts: 'burnfree'
> >atapi: 1
> >Device type : Removable CD-ROM
> >Version : 2
> >Response Format: 2
> >Capabilities :
> >Vendor_info : 'SONY '
> >Identifikation : 'DVD RW DRU-500A '
> >Revision : '1.0f'
> >Device seems to be: Generic mmc2 DVD-R/DVD-RW.
> >Using generic SCSI-3/mmc-2 DVD-R/DVD-RW driver (mmc_dvd).
> >Driver flags : DVD SWABAUDIO BURNFREE
> >Supported modes: TAO PACKET SAO SAO/R96R RAW/R96R
> >Drive buf size : 8126464 = 7936 KB
> >FIFO size : 67108864 = 65536 KB
> >Track 01: data 4001 MB
> >Total size: 4001 MB = 2048512 sectors
> >Current Secsize: 2048
> >Blocks total: 2298496 Blocks current: 2298496 Blocks remaining: 249984
> >Starting to write CD/DVD at speed 1 in dummy TAO mode for single session.
> >Last chance to quit, starting dummy write in 9 seconds. 0.24% done, estimate finish Tue Jan 21 23:09:02 2003
> > 8 seconds. 0.49% done, estimate finish Tue Jan 21 23:09:03 2003
> > 7 seconds. 0.73% done, estimate finish Tue Jan 21 23:06:46 2003
> > 6 seconds. 0.98% done, estimate finish Tue Jan 21 23:07:21 2003
> > 5 seconds. 1.22% done, estimate finish Tue Jan 21 23:07:41 2003
> > 4 seconds. 1.46% done, estimate finish Tue Jan 21 23:07:55 2003
> > 0 seconds. Operation starts.
> >Waiting for reader process to fill input buffer ... input buffer ready.
> >BURN-Free is ON.
> >Starting new track at sector: 0
> >Track 01: 4 of 4001 MB written (fifo 96%) 16.1x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
> >CDB: 2A 00 00 00 08 B8 00 00 1F 00
> >status: 0x1 (GOOD STATUS)
> >resid: 63488
> >cmd finished after 0.008s timeout 100s
>
> I can't tell you what happened because the kernel is broken :-(
>
> If you fix the kernel, you will get a readble error message,

How helpful. How about saying what's broken instead and I'd be happy to
fix it.

--
Jens Axboe

2003-01-22 08:15:31

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

>From [email protected] Wed Jan 22 09:06:12 2003

>> >BURN-Free is ON.
>> >Starting new track at sector: 0
>> >Track 01: 4 of 4001 MB written (fifo 96%) 16.1x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
>> >CDB: 2A 00 00 00 08 B8 00 00 1F 00
>> >status: 0x1 (GOOD STATUS)
>> >resid: 63488
>> >cmd finished after 0.008s timeout 100s
>>
>> I can't tell you what happened because the kernel is broken :-(
>>
>> If you fix the kernel, you will get a readble error message,

>How helpful. How about saying what's broken instead and I'd be happy to
>fix it.

I thought it's obvious: It is most likely a problem caused by the broken
bit #defines in the Linux kernel for the SCSI status byte. I assume that
status should be 0x02 instead of 0x01. In addition, I would guess that
for the same reason, a kernel instance did not fetch the sense data as
libscg should try to work around these Linux bugs if at least the first
sense byte is != 0.

J?rg

EMail:[email protected] (home) J?rg Schilling D-13353 Berlin
[email protected] (uni) If you don't have iso-8859-1
[email protected] (work) chars I am J"org Schilling
URL: http://www.fokus.fhg.de/usr/schilling ftp://ftp.berlios.de/pub/schily

2003-01-22 08:26:33

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Wed, Jan 22 2003, Joerg Schilling wrote:
> >From [email protected] Wed Jan 22 09:06:12 2003
>
> >> >BURN-Free is ON.
> >> >Starting new track at sector: 0
> >> >Track 01: 4 of 4001 MB written (fifo 96%) 16.1x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
> >> >CDB: 2A 00 00 00 08 B8 00 00 1F 00
> >> >status: 0x1 (GOOD STATUS)
> >> >resid: 63488
> >> >cmd finished after 0.008s timeout 100s
> >>
> >> I can't tell you what happened because the kernel is broken :-(
> >>
> >> If you fix the kernel, you will get a readble error message,
>
> >How helpful. How about saying what's broken instead and I'd be happy to
> >fix it.
>
> I thought it's obvious: It is most likely a problem caused by the broken
> bit #defines in the Linux kernel for the SCSI status byte. I assume that
> status should be 0x02 instead of 0x01. In addition, I would guess that

Sounds plausible. Patch attached. Anyone care to expand on _why_ these
status bytes are shifted one bit?

===== drivers/ide/ide-cd.c 1.35 vs edited =====
--- 1.35/drivers/ide/ide-cd.c Thu Nov 21 22:56:59 2002
+++ edited/drivers/ide/ide-cd.c Wed Jan 22 09:34:28 2003
@@ -706,7 +706,7 @@
* scsi status byte
*/
if ((rq->flags & REQ_BLOCK_PC) && !rq->errors)
- rq->errors = CHECK_CONDITION;
+ rq->errors = CHECK_CONDITION << 1;

/* Check for tray open. */
if (sense_key == NOT_READY) {

--
Jens Axboe

2003-01-22 09:38:23

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

2003-01-22 11:55:46

by Andre Hedrick

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

Joerg,

Then you are aware of the push/pull by one bit as linux does not set a
reserved place holder for bit0 in the status byte. I have to agree with
you that it stinks and make decoding status codes messy when supporting
N-interfaces of the protocol.

Cheers,

Andre Hedrick
LAD Storage Consulting Group

On Wed, 22 Jan 2003, Joerg Schilling wrote:

> >From [email protected] Wed Jan 22 09:06:12 2003
>
> >> >BURN-Free is ON.
> >> >Starting new track at sector: 0
> >> >Track 01: 4 of 4001 MB written (fifo 96%) 16.1x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
> >> >CDB: 2A 00 00 00 08 B8 00 00 1F 00
> >> >status: 0x1 (GOOD STATUS)
> >> >resid: 63488
> >> >cmd finished after 0.008s timeout 100s
> >>
> >> I can't tell you what happened because the kernel is broken :-(
> >>
> >> If you fix the kernel, you will get a readble error message,
>
> >How helpful. How about saying what's broken instead and I'd be happy to
> >fix it.
>
> I thought it's obvious: It is most likely a problem caused by the broken
> bit #defines in the Linux kernel for the SCSI status byte. I assume that
> status should be 0x02 instead of 0x01. In addition, I would guess that
> for the same reason, a kernel instance did not fetch the sense data as
> libscg should try to work around these Linux bugs if at least the first
> sense byte is != 0.
>
> J?rg
>
> EMail:[email protected] (home) J?rg Schilling D-13353 Berlin
> [email protected] (uni) If you don't have iso-8859-1
> [email protected] (work) chars I am J"org Schilling
> URL: http://www.fokus.fhg.de/usr/schilling ftp://ftp.berlios.de/pub/schily
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>

2003-01-22 12:48:08

by Bob Tracy

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

Jens Axboe wrote:
> On Wed, Jan 22 2003, Joerg Schilling wrote:
> > I thought it's obvious: It is most likely a problem caused by the broken
> > bit #defines in the Linux kernel for the SCSI status byte. I assume that
> > status should be 0x02 instead of 0x01. In addition, I would guess that
>
> Sounds plausible. Patch attached. Anyone care to expand on _why_ these
> status bytes are shifted one bit?

Possibly related... I was using SCSI tape to back up my 2.5.59 system
the other day. A media error was encountered, but the application (cpio
in this case) never saw the error and kept on trying to write to the bad
tape. Fortunately, I caught it after only 40 MB of associated syslog
entries had been made.

Thanks for the patch. I can now go back to "unattended" backup mode
(except for having to change tapes periodically) :-).

--
-----------------------------------------------------------------------
Bob Tracy WTO + WIPO = DMCA? http://www.anti-dmca.org
[email protected]
-----------------------------------------------------------------------

2003-01-22 15:30:29

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

>From [email protected] Wed Jan 22 09:35:39 2003

>> >How helpful. How about saying what's broken instead and I'd be happy to
>> >fix it.
>>
>> I thought it's obvious: It is most likely a problem caused by the broken
>> bit #defines in the Linux kernel for the SCSI status byte. I assume that
>> status should be 0x02 instead of 0x01. In addition, I would guess that

>Sounds plausible. Patch attached. Anyone care to expand on _why_ these
>status bytes are shifted one bit?

I have no idea... About 3-4 years ago, I tried to convice some of the kernel
people to change this but as you see, it did not happen.

>===== drivers/ide/ide-cd.c 1.35 vs edited =====
>--- 1.35/drivers/ide/ide-cd.c Thu Nov 21 22:56:59 2002
>+++ edited/drivers/ide/ide-cd.c Wed Jan 22 09:34:28 2003
>@@ -706,7 +706,7 @@
> * scsi status byte
> */
> if ((rq->flags & REQ_BLOCK_PC) && !rq->errors)
>- rq->errors = CHECK_CONDITION;
>+ rq->errors = CHECK_CONDITION << 1;
>
> /* Check for tray open. */
> if (sense_key == NOT_READY) {

J?rg

EMail:[email protected] (home) J?rg Schilling D-13353 Berlin
[email protected] (uni) If you don't have iso-8859-1
[email protected] (work) chars I am J"org Schilling
URL: http://www.fokus.fhg.de/usr/schilling ftp://ftp.berlios.de/pub/schily

2003-01-22 18:04:04

by William Lee Irwin III

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Wed, Jan 22, 2003 at 09:23:16AM +0100, Joerg Schilling wrote:
> I thought it's obvious: It is most likely a problem caused by the broken
> bit #defines in the Linux kernel for the SCSI status byte. I assume that
> status should be 0x02 instead of 0x01. In addition, I would guess that
> for the same reason, a kernel instance did not fetch the sense data as
> libscg should try to work around these Linux bugs if at least the first
> sense byte is != 0.

I should try to cut SCSI CD's with the fixes for this stuff; sounds like
it'd affect generic SCSI I/O.

Thanks.

-- wli

2003-01-23 17:44:40

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

>From [email protected] Thu Jan 23 18:51:42 2003
> 7 seconds. 0.98% done, estimate finish Thu Jan 23 18:51:05 2003
> 6 seconds. 1.22% done, estimate finish Thu Jan 23 18:51:46 2003
> 5 seconds. 1.46% done, estimate finish Thu Jan 23 18:52:14 2003
> 0 seconds. Operation starts.
>Waiting for reader process to fill input buffer ... input buffer ready.
>BURN-Free is ON.
>Starting new track at sector: 0
>Track 01: 4 of 4001 MB written (fifo 96%) 16.1x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
>CDB: 2A 00 00 00 08 B8 00 00 1F 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)
>resid: 63488
>cmd finished after 0.007s timeout 100s

In one of my mails, I decribed why there are 2 bugs in the kernel.
Only one of them so far has been fixed. The sense data is still missing.

J?rg

EMail:[email protected] (home) J?rg Schilling D-13353 Berlin
[email protected] (uni) If you don't have iso-8859-1
[email protected] (work) chars I am J"org Schilling
URL: http://www.fokus.fhg.de/usr/schilling ftp://ftp.berlios.de/pub/schily

2003-01-23 17:42:35

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Wed, Jan 22, 2003 at 09:35:30AM +0100, Jens Axboe wrote:

> Sounds plausible. Patch attached. Anyone care to expand on _why_ these
> status bytes are shifted one bit?
>
> ===== drivers/ide/ide-cd.c 1.35 vs edited =====
> --- 1.35/drivers/ide/ide-cd.c Thu Nov 21 22:56:59 2002
> +++ edited/drivers/ide/ide-cd.c Wed Jan 22 09:34:28 2003
> @@ -706,7 +706,7 @@
> * scsi status byte
> */
> if ((rq->flags & REQ_BLOCK_PC) && !rq->errors)
> - rq->errors = CHECK_CONDITION;
> + rq->errors = CHECK_CONDITION << 1;
>
> /* Check for tray open. */
> if (sense_key == NOT_READY) {

Well, it's changed, but not solved the problem:

Cdrecord-ProDVD-Clone 2.0 (i586-pc-linux-gnu) Copyright (C) 1995-2002 J?rg Schilling
Unlocked features: ProDVD Clone
Limited features: speed
This copy of cdrecord is licensed for: private/research/educational_non-commercial_use
TOC Type: 1 = CD-ROM
scsidev: '/dev/hdc'
devname: '/dev/hdc'
scsibus: -2 target: -2 lun: -2
Warning: Open by 'devname' is unintentional and not supported.
Linux sg driver version: 3.5.27
Using libscg version 'schily-0.7'
Driveropts: 'burnfree'
atapi: 1
Device type : Removable CD-ROM
Version : 2
Response Format: 2
Capabilities :
Vendor_info : 'SONY '
Identifikation : 'DVD RW DRU-500A '
Revision : '1.0f'
Device seems to be: Generic mmc2 DVD-R/DVD-RW.
Using generic SCSI-3/mmc-2 DVD-R/DVD-RW driver (mmc_dvd).
Driver flags : DVD SWABAUDIO BURNFREE
Supported modes: TAO PACKET SAO SAO/R96R RAW/R96R
Drive buf size : 8126464 = 7936 KB
FIFO size : 67108864 = 65536 KB
Track 01: data 4001 MB
Total size: 4001 MB = 2048512 sectors
Current Secsize: 2048
Blocks total: 2298496 Blocks current: 2298496 Blocks remaining: 249984
0.24% done, estimate finish Thu Jan 23 18:47:41 2003
Starting to write CD/DVD at speed 1 in dummy TAO mode for single session.
Last chance to quit, starting dummy write in 9 seconds. 0.49% done, estimate finish Thu Jan 23 18:51:05 2003
8 seconds. 0.73% done, estimate finish Thu Jan 23 18:52:13 2003
7 seconds. 0.98% done, estimate finish Thu Jan 23 18:51:05 2003
6 seconds. 1.22% done, estimate finish Thu Jan 23 18:51:46 2003
5 seconds. 1.46% done, estimate finish Thu Jan 23 18:52:14 2003
0 seconds. Operation starts.
Waiting for reader process to fill input buffer ... input buffer ready.
BURN-Free is ON.
Starting new track at sector: 0
Track 01: 4 of 4001 MB written (fifo 96%) 16.1x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
CDB: 2A 00 00 00 08 B8 00 00 1F 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)
resid: 63488
cmd finished after 0.007s timeout 100s

write track data: error after 4571136 bytes
Sense Bytes: 70 00 00 00 00 00 00 12 00 00 00 00 00 00 00 00 00 00
Writing time: 5.334s
Average write speed 571.3x.
Fixating...
Fixating time: 77.465s
cdrecord-prodvd: fifo had 1095 puts and 73 gets.
cdrecord-prodvd: fifo was 0 times empty and 2 times full, min fill was 96%.

Thank you very much,

Gr?goire
________________________________________________________________
http://ulima.unil.ch/greg ICQ:16624071 mailto:[email protected]

2003-01-23 17:53:33

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23 2003, Joerg Schilling wrote:
> >From [email protected] Thu Jan 23 18:51:42 2003
> > 7 seconds. 0.98% done, estimate finish Thu Jan 23 18:51:05 2003
> > 6 seconds. 1.22% done, estimate finish Thu Jan 23 18:51:46 2003
> > 5 seconds. 1.46% done, estimate finish Thu Jan 23 18:52:14 2003
> > 0 seconds. Operation starts.
> >Waiting for reader process to fill input buffer ... input buffer ready.
> >BURN-Free is ON.
> >Starting new track at sector: 0
> >Track 01: 4 of 4001 MB written (fifo 96%) 16.1x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
> >CDB: 2A 00 00 00 08 B8 00 00 1F 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)
> >resid: 63488
> >cmd finished after 0.007s timeout 100s
>
>
> In one of my mails, I decribed why there are 2 bugs in the kernel.
> Only one of them so far has been fixed. The sense data is still missing.

That is correct, I'll fix the 2nd bug tomorrow. The new SG_IO transport
is still so new that there are places where I know ide-cd will not do
the right thing wrt sense. In fact I don't think it will every copy
sense back.

--
Jens Axboe

2003-01-23 17:52:14

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23, 2003 at 06:52:24PM +0100, Joerg Schilling wrote:

> In one of my mails, I decribed why there are 2 bugs in the kernel.
> Only one of them so far has been fixed. The sense data is still missing.

Oups, sorry I didn't read enough carefully!!!

Does someone know how to fix the sense data bug?

Thank you very much,

Gr?goire
________________________________________________________________
http://ulima.unil.ch/greg ICQ:16624071 mailto:[email protected]

2003-01-23 18:01:11

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23 2003, Jens Axboe wrote:
> On Thu, Jan 23 2003, Gregoire Favre wrote:
> > On Thu, Jan 23, 2003 at 06:52:24PM +0100, Joerg Schilling wrote:
> >
> > > In one of my mails, I decribed why there are 2 bugs in the kernel.
> > > Only one of them so far has been fixed. The sense data is still missing.
> >
> > Oups, sorry I didn't read enough carefully!!!
> >
> > Does someone know how to fix the sense data bug?
>
> In drivers/ide/ide-cd.c:cdrom_end_request(), try to insert something
> ala:
>
> if ((rq->flags & REQ_SENSE) && uptodate) {
> struct request *failed = (struct request *) rq->buffer;
> struct cdrom_info *info = drive->driver_data;
> void *sense = &info->sense_data;
>
> + if (failed && block_pc_request(failed))
> + printk("%s: failed %p\n", __FUNCTION__, failed->sense);
>
> if (failed && failed->sense)
> sense = failed->sense;
>
> cdrom_analyze_sense_data(drive, failed, sense);
> }
>
> in pseudo-patch form.

oh, and dump failed->sense_len as well!

--
Jens Axboe

2003-01-23 17:58:02

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23 2003, Gregoire Favre wrote:
> On Thu, Jan 23, 2003 at 06:52:24PM +0100, Joerg Schilling wrote:
>
> > In one of my mails, I decribed why there are 2 bugs in the kernel.
> > Only one of them so far has been fixed. The sense data is still missing.
>
> Oups, sorry I didn't read enough carefully!!!
>
> Does someone know how to fix the sense data bug?

In drivers/ide/ide-cd.c:cdrom_end_request(), try to insert something
ala:

if ((rq->flags & REQ_SENSE) && uptodate) {
struct request *failed = (struct request *) rq->buffer;
struct cdrom_info *info = drive->driver_data;
void *sense = &info->sense_data;

+ if (failed && block_pc_request(failed))
+ printk("%s: failed %p\n", __FUNCTION__, failed->sense);

if (failed && failed->sense)
sense = failed->sense;

cdrom_analyze_sense_data(drive, failed, sense);
}

in pseudo-patch form.

--
Jens Axboe

2003-01-23 18:46:45

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23, 2003 at 07:10:02PM +0100, Jens Axboe wrote:

> oh, and dump failed->sense_len as well!

??? I am sorry, I undestood the first one, but not this one...
Could you explain it a little more?

Thank you very much,

Gr?goire
________________________________________________________________
http://ulima.unil.ch/greg ICQ:16624071 mailto:[email protected]

2003-01-23 18:58:22

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23 2003, Gregoire Favre wrote:
> On Thu, Jan 23, 2003 at 07:10:02PM +0100, Jens Axboe wrote:
>
> > oh, and dump failed->sense_len as well!
>
> ??? I am sorry, I undestood the first one, but not this one...
> Could you explain it a little more?

just add a len=%d to the printk line, and failed->sense_len as the
argument.

--
Jens Axboe

2003-01-23 19:12:32

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23, 2003 at 08:07:11PM +0100, Jens Axboe wrote:

> just add a len=%d to the printk line, and failed->sense_len as the
> argument.

Does it show like this:

if ((rq->flags & REQ_SENSE) && uptodate) {
/*
* For REQ_SENSE, "rq->buffer" points to the original failed
* request
*/
struct request *failed = (struct request *) rq->buffer;
struct cdrom_info *info = drive->driver_data;
void *sense = &info->sense_data;

if (failed && block_pc_request(failed))
printk("%s: failed %p len=%d \n", __FUNCTION__, failed->sense,failed->sense_len);

if (failed && failed->sense)
sense = failed->sense;

In that case, I got:

ld -m elf_i386 -e stext -T arch/i386/vmlinux.lds.s arch/i386/kernel/head.o arch/i386/kernel/init_task.o init/built-in.o --start-group usr/built-in.o arch/i386/kernel/built-in.o arch/i386/mm/built-in.o arch/i386/mach-default/built-in.o kernel/built-in.o mm/built-in.o fs/built-in.o ipc/built-in.o security/built-in.o crypto/built-in.o lib/lib.a arch/i386/lib/lib.a drivers/built-in.o sound/built-in.o arch/i386/pci/built-in.o net/built-in.o --end-group -o vmlinux
drivers/built-in.o(.text+0x5c563): In function `cdrom_end_request':
: undefined reference to `block_pc_request'
make: *** [vmlinux] Error 1

Sorry I certainly didn't understand you right...

Gr?goire
________________________________________________________________
http://ulima.unil.ch/greg ICQ:16624071 mailto:[email protected]

2003-01-23 21:09:17

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23 2003, Gregoire Favre wrote:
> ld -m elf_i386 -e stext -T arch/i386/vmlinux.lds.s arch/i386/kernel/head.o arch/i386/kernel/init_task.o init/built-in.o --start-group usr/built-in.o arch/i386/kernel/built-in.o arch/i386/mm/built-in.o arch/i386/mach-default/built-in.o kernel/built-in.o mm/built-in.o fs/built-in.o ipc/built-in.o security/built-in.o crypto/built-in.o lib/lib.a arch/i386/lib/lib.a drivers/built-in.o sound/built-in.o arch/i386/pci/built-in.o net/built-in.o --end-group -o vmlinux
> drivers/built-in.o(.text+0x5c563): In function `cdrom_end_request':
> : undefined reference to `block_pc_request'
> make: *** [vmlinux] Error 1
>
> Sorry I certainly didn't understand you right...

No it's my mistake, should be blk_pc_request(). Sorry about that.

--
Jens Axboe

2003-01-23 21:45:36

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23, 2003 at 10:18:05PM +0100, Jens Axboe wrote:

> No it's my mistake, should be blk_pc_request(). Sorry about that.

Hello again ;-)

I don't have more sucess now:

Cdrecord-ProDVD-Clone 2.0 (i586-pc-linux-gnu) Copyright (C) 1995-2002 J?rg Schilling
Unlocked features: ProDVD Clone
Limited features: speed
This copy of cdrecord is licensed for: private/research/educational_non-commercial_use
TOC Type: 1 = CD-ROM
scsidev: '/dev/hdc'
devname: '/dev/hdc'
scsibus: -2 target: -2 lun: -2
Warning: Open by 'devname' is unintentional and not supported.
Linux sg driver version: 3.5.27
Using libscg version 'schily-0.7'
Driveropts: 'burnfree'
atapi: 1
Device type : Removable CD-ROM
Version : 2
Response Format: 2
Capabilities :
Vendor_info : 'SONY '
Identifikation : 'DVD RW DRU-500A '
Revision : '1.0f'
Device seems to be: Generic mmc2 DVD-R/DVD-RW.
Using generic SCSI-3/mmc-2 DVD-R/DVD-RW driver (mmc_dvd).
Driver flags : DVD SWABAUDIO BURNFREE
Supported modes: TAO PACKET SAO SAO/R96R RAW/R96R
Drive buf size : 8126464 = 7936 KB
FIFO size : 67108864 = 65536 KB
Track 01: data 4001 MB
Total size: 4001 MB = 2048512 sectors
Current Secsize: 2048
Blocks total: 2298496 Blocks current: 2298496 Blocks remaining: 249984
Starting to write CD/DVD at speed 1 in dummy TAO mode for single session.
Last chance to quit, starting dummy write in 9 seconds. 0.24% done, estimate finish Thu Jan 23 22:58:58 2003
8 seconds. 0.49% done, estimate finish Thu Jan 23 22:58:59 2003
7 seconds. 0.73% done, estimate finish Thu Jan 23 22:58:59 2003
6 seconds. 0.98% done, estimate finish Thu Jan 23 22:58:59 2003
5 seconds. 1.22% done, estimate finish Thu Jan 23 22:58:59 2003
4 seconds. 1.46% done, estimate finish Thu Jan 23 22:58:59 2003
0 seconds. Operation starts.
Waiting for reader process to fill input buffer ... input buffer ready.
BURN-Free is ON.
Starting new track at sector: 0
Track 01: 4 of 4001 MB written (fifo 97%) 16.1x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
CDB: 2A 00 00 00 08 B8 00 00 1F 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)
resid: 63488
cmd finished after 0.010s timeout 100s

write track data: error after 4571136 bytes
Sense Bytes: 70 00 00 00 00 00 00 12 00 00 00 00 00 00 00 00 00 00
Writing time: 5.304s
Average write speed 574.4x.
Fixating...
Fixating time: 77.059s
cdrecord-prodvd: fifo had 1095 puts and 73 gets.
cdrecord-prodvd: fifo was 0 times empty and 2 times full, min fill was 96%.
Exit 254

Thank you very much,

Gr?goire
________________________________________________________________
http://ulima.unil.ch/greg ICQ:16624071 mailto:[email protected]

2003-01-24 07:47:04

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Thu, Jan 23 2003, Gregoire Favre wrote:
> On Thu, Jan 23, 2003 at 10:18:05PM +0100, Jens Axboe wrote:
>
> > No it's my mistake, should be blk_pc_request(). Sorry about that.
>
> Hello again ;-)
>
> I don't have more sucess now:

You're not supposed too, but you are supposed to send me the dmesg
output after trying a burn! The patch changes nothing, it just dumps
some info that I need to the logs.

--
Jens Axboe

2003-01-24 09:00:53

by Henning P. Schmiedehausen

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

Jens Axboe <[email protected]> writes:

>In drivers/ide/ide-cd.c:cdrom_end_request(), try to insert something
>ala:

> if ((rq->flags & REQ_SENSE) && uptodate) {
> struct request *failed = (struct request *) rq->buffer;
> struct cdrom_info *info = drive->driver_data;
> void *sense = &info->sense_data;

>+ if (failed && block_pc_request(failed))
>+ printk("%s: failed %p\n", __FUNCTION__, failed->sense);

> if (failed && failed->sense)
> sense = failed->sense;

Shouldn't this be below the 2nd if() and then just test "sense" ?

Like

> if (failed && failed->sense)
> sense = failed->sense;

>+ if (failed && block_pc_request(failed))
>+ printk("%s: failed %p\n", __FUNCTION__, sense);

That makes sure, that you report what is analyzed later here:

> cdrom_analyze_sense_data(drive, failed, sense);

Regards
Henning

--
Dipl.-Inf. (Univ.) Henning P. Schmiedehausen -- Geschaeftsfuehrer
INTERMETA - Gesellschaft fuer Mehrwertdienste mbH [email protected]

Am Schwabachgrund 22 Fon.: 09131 / 50654-0 [email protected]
D-91054 Buckenhof Fax.: 09131 / 50654-20

2003-01-24 09:17:23

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Fri, Jan 24 2003, Henning P. Schmiedehausen wrote:
> Jens Axboe <[email protected]> writes:
>
> >In drivers/ide/ide-cd.c:cdrom_end_request(), try to insert something
> >ala:
>
> > if ((rq->flags & REQ_SENSE) && uptodate) {
> > struct request *failed = (struct request *) rq->buffer;
> > struct cdrom_info *info = drive->driver_data;
> > void *sense = &info->sense_data;
>
> >+ if (failed && block_pc_request(failed))
> >+ printk("%s: failed %p\n", __FUNCTION__, failed->sense);
>
> > if (failed && failed->sense)
> > sense = failed->sense;
>
> Shouldn't this be below the 2nd if() and then just test "sense" ?
>
> Like
>
> > if (failed && failed->sense)
> > sense = failed->sense;
>
> >+ if (failed && block_pc_request(failed))
> >+ printk("%s: failed %p\n", __FUNCTION__, sense);
>
> That makes sure, that you report what is analyzed later here:

The interesting part is whether failed has a sense attached, and if so
what length.
>
> > cdrom_analyze_sense_data(drive, failed, sense);

To avoid confusion, I made the patch. Would have been easier to do in
the first place it seems :)

===== drivers/ide/ide-cd.c 1.35 vs edited =====
--- 1.35/drivers/ide/ide-cd.c Thu Nov 21 22:56:59 2002
+++ edited/drivers/ide/ide-cd.c Fri Jan 24 10:25:53 2003
@@ -649,6 +649,8 @@
struct cdrom_info *info = drive->driver_data;
void *sense = &info->sense_data;

+ if (failed && blk_pc_request(failed))
+ printk("%s: failed, sense %p, len=%d\n", __FUNCTION__, failed->sense, failed->sense_len);
if (failed && failed->sense)
sense = failed->sense;

--
Jens Axboe

2003-01-25 14:16:35

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Fri, Jan 24, 2003 at 10:26:16AM +0100, Jens Axboe wrote:

> The interesting part is whether failed has a sense attached, and if so
> what length.
> >
> > > cdrom_analyze_sense_data(drive, failed, sense);
>
> To avoid confusion, I made the patch. Would have been easier to do in
> the first place it seems :)
>
> ===== drivers/ide/ide-cd.c 1.35 vs edited =====
> --- 1.35/drivers/ide/ide-cd.c Thu Nov 21 22:56:59 2002
> +++ edited/drivers/ide/ide-cd.c Fri Jan 24 10:25:53 2003
> @@ -649,6 +649,8 @@
> struct cdrom_info *info = drive->driver_data;
> void *sense = &info->sense_data;
>
> + if (failed && blk_pc_request(failed))
> + printk("%s: failed, sense %p, len=%d\n", __FUNCTION__, failed->sense, failed->sense_len);
> if (failed && failed->sense)
> sense = failed->sense;

Hello,

sorry I didn't understood that I was supposed to do this...
I have recompiled with your patch, and here the result of dmesg:

cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0
cdrom_end_request: failed, sense ce68de20, len=0

Maybe I should add that, also under 2.4, it goes till 4 Mb and then stay
at 4 Mb for a long time (something like one minute) and then it
continues perfectly (with ide-scsi under 2.4).

And in case still needed:

Cdrecord-ProDVD-Clone 2.0 (i586-pc-linux-gnu) Copyright (C) 1995-2002 J?rg Schilling
Unlocked features: ProDVD Clone
Limited features: speed
This copy of cdrecord is licensed for: private/research/educational_non-commercial_use
TOC Type: 1 = CD-ROM
scsidev: '/dev/hdc'
devname: '/dev/hdc'
scsibus: -2 target: -2 lun: -2
Warning: Open by 'devname' is unintentional and not supported.
Linux sg driver version: 3.5.27
Using libscg version 'schily-0.7'
Driveropts: 'burnfree'
atapi: 1
Device type : Removable CD-ROM
Version : 2
Response Format: 2
Capabilities :
Vendor_info : 'SONY '
Identifikation : 'DVD RW DRU-500A '
Revision : '1.0f'
Device seems to be: Generic mmc2 DVD-R/DVD-RW.
Using generic SCSI-3/mmc-2 DVD-R/DVD-RW driver (mmc_dvd).
Driver flags : DVD SWABAUDIO BURNFREE
Supported modes: TAO PACKET SAO SAO/R96R RAW/R96R
Drive buf size : 8126464 = 7936 KB
FIFO size : 67108864 = 65536 KB
Track 01: data 4001 MB
Total size: 4001 MB = 2048512 sectors
Current Secsize: 2048
Blocks total: 2298496 Blocks current: 2298496 Blocks remaining: 249984
Starting to write CD/DVD at speed 1 in dummy TAO mode for single session.
Last chance to quit, starting dummy write in 9 seconds. 0.24% done, estimate finish Sat Jan 25 15:26:58 2003
8 seconds. 0.49% done, estimate finish Sat Jan 25 15:26:59 2003
7 seconds. 0.73% done, estimate finish Sat Jan 25 15:26:59 2003
6 seconds. 0.98% done, estimate finish Sat Jan 25 15:26:59 2003
5 seconds. 1.22% done, estimate finish Sat Jan 25 15:26:59 2003
4 seconds. 1.46% done, estimate finish Sat Jan 25 15:26:59 2003
0 seconds. Operation starts.
Waiting for reader process to fill input buffer ... input buffer ready.
BURN-Free is ON.
Starting new track at sector: 0
Track 01: 4 of 4001 MB written (fifo 97%) 16.2x.cdrecord-prodvd: Success. write_g1: scsi sendcmd: no error
CDB: 2A 00 00 00 08 B8 00 00 1F 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)
resid: 63488
cmd finished after 0.001s timeout 100s

write track data: error after 4571136 bytes
Sense Bytes: 70 00 00 00 00 00 00 12 00 00 00 00 00 00 00 00 00 00
Writing time: 5.294s
Average write speed 575.6x.
Fixating...
Fixating time: 77.019s
cdrecord-prodvd: fifo had 1095 puts and 73 gets.
cdrecord-prodvd: fifo was 0 times empty and 2 times full, min fill was 96%.
Exit 254

Thank you very much,

Gr?goire
________________________________________________________________
http://ulima.unil.ch/greg ICQ:16624071 mailto:[email protected]

2003-01-26 16:15:01

[permalink] [raw]

Subject: Re: Can't burn DVD under 2.5.59 with ide-cd

On Sat, Jan 25 2003, Gregoire Favre wrote:
> On Fri, Jan 24, 2003 at 10:26:16AM +0100, Jens Axboe wrote:
>
> > The interesting part is whether failed has a sense attached, and if so
> > what length.
> > >
> > > > cdrom_analyze_sense_data(drive, failed, sense);
> >
> > To avoid confusion, I made the patch. Would have been easier to do in
> > the first place it seems :)
> >
> > ===== drivers/ide/ide-cd.c 1.35 vs edited =====
> > --- 1.35/drivers/ide/ide-cd.c Thu Nov 21 22:56:59 2002
> > +++ edited/drivers/ide/ide-cd.c Fri Jan 24 10:25:53 2003
> > @@ -649,6 +649,8 @@
> > struct cdrom_info *info = drive->driver_data;
> > void *sense = &info->sense_data;
> >
> > + if (failed && blk_pc_request(failed))
> > + printk("%s: failed, sense %p, len=%d\n", __FUNCTION__, failed->sense, failed->sense_len);
> > if (failed && failed->sense)
> > sense = failed->sense;
>
> Hello,
>
> sorry I didn't understood that I was supposed to do this...
> I have recompiled with your patch, and here the result of dmesg:
>
> cdrom_end_request: failed, sense ce68de20, len=0
> cdrom_end_request: failed, sense ce68de20, len=0

Thanks, this is the info I wanted. The sense info has not been read from
the drive, it seems.

--
Jens Axboe

2003-01-26 16:28:59