Skip to contentSkip to conversation

2.6.9-rc2-mm4

DISCUSSION Borislav Petkov 3 participants 21 messages Sep 30, 2004 – Oct 5, 2004 2 messages not in the archive Block layer

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

#2 lore raw
Quoted 32 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Andrew Morton

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 22 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction but it went
unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Regards,
Boris.

Hi people,

well, I've applied the above patches but no change - same "hdc: lost
interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Regards,
Boris.

#3 lore raw
Quoted 31 lines · Borislav Petkov
Quoted 30 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Andrew Morton

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 20 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction but it went
unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Quoted 7 lines
Quoted 2 lines
Quoted 2 lines

Regards,
Boris.

Hi people,

well, I've applied the above patches but no change - same "hdc: lost
interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Quoted 4 lines

Regards,
Boris.

#4 lore raw
Quoted 35 lines · Bartlomiej Zolnierkiewicz
Quoted 34 lines · Borislav Petkov
Quoted 32 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Andrew Morton

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 22 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction but it went
unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Regards,
Boris.

Quoted 9 lines
Quoted 4 lines

Hi people,

well, I've applied the above patches but no change - same "hdc: lost
interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

Quoted 2 lines
Quoted 2 lines

Regards,
Boris.

Boris.

#5 lore raw
Quoted 43 lines · Borislav Petkov
Quoted 42 lines · Bartlomiej Zolnierkiewicz
Quoted 31 lines · Borislav Petkov
Quoted 30 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Andrew Morton

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 20 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction but it went
unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Quoted 7 lines
Quoted 2 lines
Quoted 2 lines

Regards,
Boris.

Hi people,

well, I've applied the above patches but no change - same "hdc: lost
interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Quoted 8 lines
Quoted 2 lines

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

I'm not aware of any. Jens?

Just a guess but maybe SG_IO filtering is involved? Does cdda2wav work for a root user?

Quoted 6 lines
Quoted 2 lines
Quoted 2 lines

Regards,
Boris.

Boris.

#6 lore raw
Quoted 54 lines · Bartlomiej Zolnierkiewicz
Quoted 45 lines · Borislav Petkov
Quoted 44 lines · Bartlomiej Zolnierkiewicz
Quoted 33 lines · Borislav Petkov
Quoted 32 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Andrew Morton

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 21 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction but it
went unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Quoted 7 lines
Quoted 2 lines
Quoted 2 lines

Regards,
Boris.

Hi people,

well, I've applied the above patches but no change - same "hdc: lost
interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Quoted 7 lines
Quoted 1 line

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

Quoted 5 lines

I'm not aware of any. Jens?

Just a guess but maybe SG_IO filtering is involved? Does cdda2wav work for a root user?

Actually, I ran all the tests only as root... Should I continue testing with the 2.6.7 -mm series?

Regards,
Boris.

#7 lore raw
In reply to Bartlomiej Zolnierkiewicz
Quoted 47 lines · Bartlomiej Zolnierkiewicz
Quoted 46 lines · Borislav Petkov
Quoted 44 lines · Bartlomiej Zolnierkiewicz
Quoted 31 lines · Borislav Petkov
Quoted 30 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Andrew Morton

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 20 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction but it went
unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Quoted 7 lines
Quoted 2 lines
Quoted 2 lines

Regards,
Boris.

Hi people,

well, I've applied the above patches but no change - same "hdc: lost
interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Quoted 7 lines
Quoted 5 lines

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

I'm not aware of any. Jens?

I don't see any changes that could impact this from 2.6.7 to 2.6.8. We tightened the dma alignment (from 4 to 32 bytes), but should not cause problems going in that direction. Unless the other path is buggy, of course.

Does dma make a difference? Please try 2.6.9-rc3 as well.

Signature · 1 line
Jens Axboe
#8 lore raw
Quoted 5 lines · Jens Axboe
Quoted 4 lines · Bartlomiej Zolnierkiewicz
Quoted 3 lines · Borislav Petkov
Quoted 2 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Borislav Petkov

On Thursday 30 September 2004 14:52, Bartlomiej Zolnierkiewicz

Quoted 51 lines · wrote:
Quoted 51 lines
Quoted 50 lines
Quoted 36 lines
Quoted 36 lines
Quoted 35 lines · Andrew Morton
Quoted 1 line

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 23 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction but
it went unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Regards,
Boris.

Quoted 8 lines
Quoted 4 lines

Hi people,

well, I've applied the above patches but no change - same "hdc:
lost interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

Quoted 8 lines
Quoted 1 line

I'm not aware of any. Jens?

I don't see any changes that could impact this from 2.6.7 to 2.6.8. We tightened the dma alignment (from 4 to 32 bytes), but should not cause problems going in that direction. Unless the other path is buggy, of course.

Does dma make a difference? Please try 2.6.9-rc3 as well.

Sorry guys,

still a no go. Tested today 2.6.8.1 and 2.6.9-rc3 both with DMA on/off. same lost interrupt message. How about a hardware problem? Maybe the cd-drive is showing some hidden "features" under certain conditions, although it is highly unlikely since 2.6.7 runs fine. strange...

Boris.

#9 lore raw
Quoted 67 lines · Borislav Petkov
Quoted 5 lines · Jens Axboe
Quoted 4 lines · Bartlomiej Zolnierkiewicz
Quoted 3 lines · Borislav Petkov
Quoted 2 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Borislav Petkov

On Thursday 30 September 2004 14:52, Bartlomiej Zolnierkiewicz

Quoted 51 lines · wrote:
Quoted 51 lines
Quoted 50 lines
Quoted 36 lines
Quoted 36 lines
Quoted 35 lines · Andrew Morton
Quoted 1 line

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 23 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction but
it went unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Regards,
Boris.

Quoted 8 lines
Quoted 4 lines

Hi people,

well, I've applied the above patches but no change - same "hdc:
lost interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

Quoted 8 lines
Quoted 1 line

I'm not aware of any. Jens?

I don't see any changes that could impact this from 2.6.7 to 2.6.8. We tightened the dma alignment (from 4 to 32 bytes), but should not cause problems going in that direction. Unless the other path is buggy, of course.

Does dma make a difference? Please try 2.6.9-rc3 as well.

Quoted 7 lines

Sorry guys,

still a no go. Tested today 2.6.8.1 and 2.6.9-rc3 both with DMA on/off. same lost interrupt message. How about a hardware problem? Maybe the cd-drive is showing some hidden "features" under certain conditions, although it is highly unlikely since 2.6.7 runs fine. strange...

I can't say, probably you need to look outside of ide changes to locate the problem. Have you tried disabling acpi on your box?

Signature · 1 line
Jens Axboe
#10 lore raw
Quoted 5 lines · Jens Axboe
Quoted 4 lines · Borislav Petkov
Quoted 3 lines · Jens Axboe
Quoted 2 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Borislav Petkov

On Thursday 30 September 2004 18:25, Bartlomiej Zolnierkiewicz

Quoted 69 lines · wrote:
Quoted 69 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines · Borislav Petkov
Quoted 1 line

On Thursday 30 September 2004 14:52, Bartlomiej Zolnierkiewicz

Quoted 54 lines · wrote:
Quoted 54 lines
Quoted 53 lines
Quoted 39 lines
Quoted 39 lines
Quoted 36 lines · Andrew Morton
Quoted 1 line

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 24 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction
but it went unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Regards,
Boris.

Hi people,

Quoted 7 lines
Quoted 3 lines

well, I've applied the above patches but no change - same
"hdc: lost interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

Quoted 8 lines
Quoted 1 line

I'm not aware of any. Jens?

I don't see any changes that could impact this from 2.6.7 to 2.6.8. We tightened the dma alignment (from 4 to 32 bytes), but should not cause problems going in that direction. Unless the other path is buggy, of course.

Does dma make a difference? Please try 2.6.9-rc3 as well.

Sorry guys,

Quoted 8 lines
Quoted 5 lines

still a no go. Tested today 2.6.8.1 and 2.6.9-rc3 both with DMA on/off. same lost interrupt message. How about a hardware problem? Maybe the cd-drive is showing some hidden "features" under certain conditions, although it is highly unlikely since 2.6.7 runs fine. strange...

I can't say, probably you need to look outside of ide changes to locate the problem. Have you tried disabling acpi on your box?

I'm not sure whether adding the boot option acpi=off is enough to disable ACPI in 2.6, but if this is the case 2.6.9-rc3 is still a no go with acpi disabled. How about APIC?

Boris.

#11 lore raw
Quoted 15 lines · Borislav Petkov
Quoted 15 lines
Quoted 14 lines
Quoted 6 lines

I don't see any changes that could impact this from 2.6.7 to 2.6.8. We tightened the dma alignment (from 4 to 32 bytes), but should not cause problems going in that direction. Unless the other path is buggy, of course.

Does dma make a difference? Please try 2.6.9-rc3 as well.

Sorry guys,

still a no go. Tested today 2.6.8.1 and 2.6.9-rc3 both with DMA on/off. same lost interrupt message. How about a hardware problem? Maybe the cd-drive is showing some hidden "features" under certain conditions, although it is highly unlikely since 2.6.7 runs fine. strange...

Quoted 6 lines
Quoted 2 lines

I can't say, probably you need to look outside of ide changes to locate the problem. Have you tried disabling acpi on your box?

I'm not sure whether adding the boot option acpi=off is enough to disable ACPI in 2.6, but if this is the case 2.6.9-rc3 is still a no go with acpi disabled. How about APIC?

maybe it would be more useful if you tried the -bk snapshots in between 2.6.7 and 2.6.8 to locate where the problem started. try the 2.6.8-rc first, then try the snapshots in between when you find where the problem was introduced.

Signature · 1 line
Jens Axboe
#12 lore raw
In reply to Borislav Petkov
Quoted 84 lines · Borislav Petkov
Quoted 5 lines · Jens Axboe
Quoted 4 lines · Borislav Petkov
Quoted 3 lines · Jens Axboe
Quoted 2 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Borislav Petkov

On Thursday 30 September 2004 18:25, Bartlomiej Zolnierkiewicz

Quoted 76 lines · wrote:
Quoted 75 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines · Borislav Petkov
Quoted 2 lines

On Thursday 30 September 2004 14:52, Bartlomiej
Zolnierkiewicz

Quoted 54 lines · wrote:
Quoted 54 lines
Quoted 53 lines
Quoted 39 lines
Quoted 39 lines
Quoted 36 lines · Andrew Morton
Quoted 1 line

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 24 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio extraction
but it went unnoticed. Here's a recount: When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Regards,
Boris.

Hi people,

Quoted 7 lines
Quoted 3 lines

well, I've applied the above patches but no change - same
"hdc: lost interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

Quoted 8 lines
Quoted 1 line

I'm not aware of any. Jens?

I don't see any changes that could impact this from 2.6.7 to 2.6.8. We tightened the dma alignment (from 4 to 32 bytes), but should not cause problems going in that direction. Unless the other path is buggy, of course.

Does dma make a difference? Please try 2.6.9-rc3 as well.

Sorry guys,

still a no go. Tested today 2.6.8.1 and 2.6.9-rc3 both with DMA on/off. same lost interrupt message. How about a hardware problem? Maybe the cd-drive is showing some hidden "features" under certain conditions, although it is highly unlikely since 2.6.7 runs fine. strange...

Quoted 6 lines
Quoted 2 lines

I can't say, probably you need to look outside of ide changes to locate the problem. Have you tried disabling acpi on your box?

I'm not sure whether adding the boot option acpi=off is enough to disable ACPI in 2.6, but if this is the case 2.6.9-rc3 is still a no go with acpi disabled. How about APIC?

After booting with "acpi=off noapic", I found something that might bring us further:

<snip>
PCI: Probing PCI hardware
PCI: Probing PCI hardware (bus 00)
PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.1
PCI: Transparent bridge - 0000:00:1e.0 <----
PCI: Using IRQ router PIIX/ICH [8086/24c0] at 0000:00:1f.0
PCI: IRQ 0 for device 0000:00:1f.1 doesn't match PIRQ mask - try pci=usepirqmask
PCI: Found IRQ 5 for device 0000:00:1f.1
PCI: Sharing IRQ 5 with 0000:00:1d.2
</snip>

Well, 0000:00:1f.0 is the IDE controller so later today I'll try the pci=usepirqmask boot option. Comments?

Boris.

#13 lore raw
Quoted 106 lines · Borislav Petkov
Quoted 84 lines · Borislav Petkov
Quoted 5 lines · Jens Axboe
Quoted 4 lines · Borislav Petkov
Quoted 3 lines · Jens Axboe
Quoted 2 lines · Bartlomiej Zolnierkiewicz
Quoted 1 line · Borislav Petkov

On Thursday 30 September 2004 18:25, Bartlomiej Zolnierkiewicz

Quoted 76 lines · wrote:
Quoted 75 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines · Borislav Petkov
Quoted 2 lines

On Thursday 30 September 2004 14:52, Bartlomiej
Zolnierkiewicz

Quoted 54 lines · wrote:
Quoted 54 lines
Quoted 53 lines
Quoted 39 lines
Quoted 39 lines
Quoted 36 lines · Andrew Morton
Quoted 1 line

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 24 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio
extraction but it went unnoticed. Here's a recount:
When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Regards,
Boris.

Hi people,

Quoted 7 lines
Quoted 3 lines

well, I've applied the above patches but no change - same
"hdc: lost interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

Quoted 8 lines
Quoted 1 line

I'm not aware of any. Jens?

I don't see any changes that could impact this from 2.6.7 to 2.6.8. We tightened the dma alignment (from 4 to 32 bytes), but should not cause problems going in that direction. Unless the other path is buggy, of course.

Does dma make a difference? Please try 2.6.9-rc3 as well.

Sorry guys,

still a no go. Tested today 2.6.8.1 and 2.6.9-rc3 both with DMA on/off. same lost interrupt message. How about a hardware problem? Maybe the cd-drive is showing some hidden "features" under certain conditions, although it is highly unlikely since 2.6.7 runs fine. strange...

Quoted 6 lines
Quoted 2 lines

I can't say, probably you need to look outside of ide changes to locate the problem. Have you tried disabling acpi on your box?

I'm not sure whether adding the boot option acpi=off is enough to disable ACPI in 2.6, but if this is the case 2.6.9-rc3 is still a no go with acpi disabled. How about APIC?

After booting with "acpi=off noapic", I found something that might bring us further:

<snip>
PCI: Probing PCI hardware
PCI: Probing PCI hardware (bus 00)
PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.1
PCI: Transparent bridge - 0000:00:1e.0 <----
PCI: Using IRQ router PIIX/ICH [8086/24c0] at 0000:00:1f.0
PCI: IRQ 0 for device 0000:00:1f.1 doesn't match PIRQ mask - try pci=usepirqmask PCI: Found IRQ 5 for device 0000:00:1f.1
PCI: Sharing IRQ 5 with 0000:00:1d.2
</snip>

Quoted 3 lines

Well, 0000:00:1f.0 is the IDE controller so later today I'll try the pci=usepirqmask boot option. Comments?

Still no luck with different boot options. Beginning to test the 2.6.8-rc series...

Boris.

#14 lore raw
Quoted 17 lines · Borislav Petkov
Quoted 16 lines · Borislav Petkov
Quoted 15 lines · Borislav Petkov
Quoted 6 lines · Jens Axboe
Quoted 5 lines · Borislav Petkov
Quoted 4 lines · Jens Axboe
Quoted 3 lines · Bartlomiej Zolnierkiewicz
Quoted 2 lines · Borislav Petkov

On Thursday 30 September 2004 18:25, Bartlomiej
Zolnierkiewicz

Quoted 6 lines · wrote:
Quoted 6 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines · Borislav Petkov
Quoted 2 lines

On Thursday 30 September 2004 14:52, Bartlomiej
Zolnierkiewicz

Quoted 1 line · wrote:
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

On Thursday 30 September 2004 06:46, Andrew Morton

Quoted 92 lines · wrote:
Quoted 92 lines
Quoted 71 lines
Quoted 71 lines
Quoted 70 lines
Quoted 54 lines
Quoted 54 lines
Quoted 53 lines
Quoted 39 lines
Quoted 39 lines
Quoted 36 lines
Quoted 1 line

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from -mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 25 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio
extraction but it went unnoticed. Here's a recount:
When I attempt to read an audio cd into wavs with cdda2wav, the process starts but after a while the completion meter freezes and klogd says "hdc: lost interrupt" and cdda2wav hangs itself. Disabling DMA doesn't help as well as the boot option "pci=routeirq" too. Older kernels like 2.6.7 do not show such behavior and there audio extraction runs fine. Sysinfo attached.

Regards,
Boris.

Hi people,

Quoted 7 lines
Quoted 3 lines

well, I've applied the above patches but no change -
same "hdc: lost interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

Quoted 8 lines
Quoted 1 line

I'm not aware of any. Jens?

I don't see any changes that could impact this from 2.6.7 to 2.6.8. We tightened the dma alignment (from 4 to 32 bytes), but should not cause problems going in that direction. Unless the other path is buggy, of course.

Does dma make a difference? Please try 2.6.9-rc3 as well.

Sorry guys,

still a no go. Tested today 2.6.8.1 and 2.6.9-rc3 both with DMA on/off. same lost interrupt message. How about a hardware problem? Maybe the cd-drive is showing some hidden "features" under certain conditions, although it is highly unlikely since 2.6.7 runs fine. strange...

Quoted 6 lines
Quoted 2 lines

I can't say, probably you need to look outside of ide changes to locate the problem. Have you tried disabling acpi on your box?

I'm not sure whether adding the boot option acpi=off is enough to disable ACPI in 2.6, but if this is the case 2.6.9-rc3 is still a no go with acpi disabled. How about APIC?

After booting with "acpi=off noapic", I found something that might bring us further:

<snip>
PCI: Probing PCI hardware
PCI: Probing PCI hardware (bus 00)
PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.1
PCI: Transparent bridge - 0000:00:1e.0 <----
PCI: Using IRQ router PIIX/ICH [8086/24c0] at 0000:00:1f.0
PCI: IRQ 0 for device 0000:00:1f.1 doesn't match PIRQ mask - try pci=usepirqmask PCI: Found IRQ 5 for device 0000:00:1f.1
PCI: Sharing IRQ 5 with 0000:00:1d.2
</snip>

Quoted 5 lines
Quoted 2 lines

Well, 0000:00:1f.0 is the IDE controller so later today I'll try the pci=usepirqmask boot option. Comments?

Still no luck with different boot options. Beginning to test the 2.6.8-rc series...

Testing log:

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc4: BUG!

OK, we got the two releases inbetween which the problem got introduced. Gonna continue testing with the 4 incremental bk snapshots patch-2.6.8-rc3-bk[1-4] now. To be continued...

Boris.

#15 lore raw
Quoted 13 lines · Borislav Petkov
Quoted 12 lines · Borislav Petkov
Quoted 11 lines · Borislav Petkov
Quoted 10 lines · Borislav Petkov
Quoted 6 lines · Jens Axboe
Quoted 5 lines · Borislav Petkov
Quoted 4 lines · Jens Axboe
Quoted 3 lines · Bartlomiej Zolnierkiewicz
Quoted 2 lines · Borislav Petkov

On Thursday 30 September 2004 18:25, Bartlomiej Zolnierkiewicz

Quoted 1 line · wrote:
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

On Thursday 30 September 2004 17:32, Borislav Petkov

Quoted 107 lines · wrote:
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

On Thursday 30 September 2004 14:52, Bartlomiej
Zolnierkiewicz

Quoted 1 line · wrote:
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

On Thursday 30 September 2004 06:46, Andrew Morton

Quoted 92 lines · wrote:
Quoted 92 lines
Quoted 71 lines
Quoted 71 lines
Quoted 70 lines
Quoted 54 lines
Quoted 54 lines
Quoted 53 lines
Quoted 39 lines
Quoted 39 lines
Quoted 36 lines
Quoted 1 line

ide broke :( Maybe Bart's bk tree?

no, disk works just fine ;) If it is my tree I will happilly fix it.

Borislav, could you apply only these patches from
-mm4 and retest?

linus.patch
bk-ide-dev.patch

Quoted 25 lines

Begin forwarded message:

Date: Wed, 29 Sep 2004 12:43:35 +0200
From: Borislav Petkov <[email protected]>
To: Andrew Morton <[email protected]>
Cc: [email protected]
Subject: Re: 2.6.9-rc2-mm4

<snip>

Hello,
I've already posted about problems with audio
extraction but it went unnoticed. Here's a recount:
When I attempt to read an audio cd into wavs with
cdda2wav, the process starts but after a while the
completion meter freezes and klogd says "hdc: lost
interrupt" and cdda2wav hangs itself. Disabling DMA
doesn't help as well as the boot option
"pci=routeirq" too. Older kernels like 2.6.7 do not
show such behavior and there audio extraction runs
fine. Sysinfo attached.

Regards,
Boris.

Hi people,

Quoted 7 lines
Quoted 3 lines

well, I've applied the above patches but no change -
same "hdc: lost interrupt" message. 2.6.9-rc3 behaves the same, as expected.

Well, if 2.6.9-rc3 fails then it is not my tree...

Please find kernel version which introduces this bug.

Just compiled 2.6.8.1 and tested audio extraction. The bug is there. After that, reran the test with 2.6.7. Everything went fine. So it must have been between 2.6.7 and 2.6.8.1 when the bug got introduced. Any additional debugging options in the ATA/IDE cd driver i could turn on so that I could get more verbose messages while executing cdda2wav?

Quoted 8 lines
Quoted 1 line

I'm not aware of any. Jens?

I don't see any changes that could impact this from 2.6.7 to 2.6.8. We tightened the dma alignment (from 4 to 32 bytes), but should not cause problems going in that direction. Unless the other path is buggy, of course.

Does dma make a difference? Please try 2.6.9-rc3 as well.

Sorry guys,

still a no go. Tested today 2.6.8.1 and 2.6.9-rc3 both with DMA on/off. same lost interrupt message. How about a hardware problem? Maybe the cd-drive is showing some hidden "features" under certain conditions, although it is highly unlikely since 2.6.7 runs fine. strange...

Quoted 6 lines
Quoted 2 lines

I can't say, probably you need to look outside of ide changes to locate the problem. Have you tried disabling acpi on your box?

I'm not sure whether adding the boot option acpi=off is enough to disable ACPI in 2.6, but if this is the case 2.6.9-rc3 is still a no go with acpi disabled. How about APIC?

After booting with "acpi=off noapic", I found something that might bring us further:

<snip>
PCI: Probing PCI hardware
PCI: Probing PCI hardware (bus 00)
PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.1
PCI: Transparent bridge - 0000:00:1e.0 <----
PCI: Using IRQ router PIIX/ICH [8086/24c0] at 0000:00:1f.0
PCI: IRQ 0 for device 0000:00:1f.1 doesn't match PIRQ mask - try pci=usepirqmask PCI: Found IRQ 5 for device 0000:00:1f.1
PCI: Sharing IRQ 5 with 0000:00:1d.2
</snip>

Quoted 5 lines
Quoted 2 lines

Well, 0000:00:1f.0 is the IDE controller so later today I'll try the pci=usepirqmask boot option. Comments?

Still no luck with different boot options. Beginning to test the 2.6.8-rc series...

Testing log:

Quoted 9 lines

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc4: BUG!

OK, we got the two releases inbetween which the problem got introduced. Gonna continue testing with the 4 incremental bk snapshots patch-2.6.8-rc3-bk[1-4] now. To be continued...

Ok here we go,

final results:

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc3-bk1: OK
2.6.8-rc3-bk2: OK
2.6.8-rc3-bk3: OK
2.6.8-rc3-bk4: OK
2.6.8-rc4: BUG!

So, assuming that everything went fine during testing, the bug got introduced in the transition between 2.6.8-rc3-bk4 and 2.6.8-rc4.

Comments?

Boris.

#16 lore raw
Quoted 13 lines · Borislav Petkov

Ok here we go,

final results:

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc3-bk1: OK
2.6.8-rc3-bk2: OK
2.6.8-rc3-bk3: OK
2.6.8-rc3-bk4: OK
2.6.8-rc4: BUG!

Quoted 2 lines

So, assuming that everything went fine during testing, the bug got introduced in the transition between 2.6.8-rc3-bk4 and 2.6.8-rc4.

That's some nice testing, thank you. Try backing out this hunk:

drivers/block/scsi_ioctl.c +1 -1
diff -urp linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c linux-2.6.8-rc4/drivers/block/scsi_ioctl.c
@@ -90,7 +90,7 @@ static int sg_set_reserved_size(request_
 	if (size < 0)
 		return -EINVAL;
 	if (size > (q->max_sectors << 9))
-		return -EINVAL;
+		size = q->max_sectors << 9;
 
 	q->sg_reserved_size = size;
 	return 0;

It's the only thing that sticks out, and it could easily explain it if your cd ripper starts issuing requests that are too big. Maybe even add a printk() here, so it will look like this in the kernel you test:

	if (size > (q->sectors << 9)) {
		printk("%u rejected\n", size);
		return -EINVAL;
	}

to verify.

Signature · 1 line
Jens Axboe
#17 lore raw
Quoted 38 lines · Jens Axboe
Quoted 13 lines · Borislav Petkov

Ok here we go,

final results:

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc3-bk1: OK
2.6.8-rc3-bk2: OK
2.6.8-rc3-bk3: OK
2.6.8-rc3-bk4: OK
2.6.8-rc4: BUG!

Quoted 2 lines

So, assuming that everything went fine during testing, the bug got introduced in the transition between 2.6.8-rc3-bk4 and 2.6.8-rc4.

That's some nice testing, thank you. Try backing out this hunk:

diff -urp linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c linux-2.6.8-rc4/drivers/block/scsi_ioctl.c --- linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c 2004-08-03 23:28:51.000000000 +0200 +++ linux-2.6.8-rc4/drivers/block/scsi_ioctl.c 2004-08-10 04:24:08.000000000 +0200 @@ -90,7 +90,7 @@ static int sg_set_reserved_size(request_
if (size < 0)
return -EINVAL;
if (size > (q->max_sectors << 9))

drivers/block/scsi_ioctl.c +1 -1
-  return -EINVAL;
+  size = q->max_sectors << 9;
 
  q->sg_reserved_size = size;
  return 0;

It's the only thing that sticks out, and it could easily explain it if your cd ripper starts issuing requests that are too big. Maybe even add a printk() here, so it will look like this in the kernel you test:

Quoted 6 lines

if (size > (q->sectors << 9)) {
printk("%u rejected\n", size);
return -EINVAL;
}

to verify.

Yeah, that was it. Two lines in the log:

Oct 4 22:07:04 zmei kernel: 3145728 rejected
Oct 4 22:07:04 zmei kernel: 3145728 rejected

Hmm, so this means that my dvd drive is sending too big requests. What do we
do: firmware upgrade?

Regards,
Boris.

#18 lore raw
Quoted 46 lines · Borislav Petkov
Quoted 38 lines · Jens Axboe
Quoted 13 lines · Borislav Petkov

Ok here we go,

final results:

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc3-bk1: OK
2.6.8-rc3-bk2: OK
2.6.8-rc3-bk3: OK
2.6.8-rc3-bk4: OK
2.6.8-rc4: BUG!

Quoted 2 lines

So, assuming that everything went fine during testing, the bug got introduced in the transition between 2.6.8-rc3-bk4 and 2.6.8-rc4.

That's some nice testing, thank you. Try backing out this hunk:

diff -urp linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c linux-2.6.8-rc4/drivers/block/scsi_ioctl.c --- linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c 2004-08-03 23:28:51.000000000 +0200 +++ linux-2.6.8-rc4/drivers/block/scsi_ioctl.c 2004-08-10 04:24:08.000000000 +0200 @@ -90,7 +90,7 @@ static int sg_set_reserved_size(request_
if (size < 0)
return -EINVAL;
if (size > (q->max_sectors << 9))

drivers/block/scsi_ioctl.c +1 -1
-  return -EINVAL;
+  size = q->max_sectors << 9;
 
  q->sg_reserved_size = size;
  return 0;

It's the only thing that sticks out, and it could easily explain it if your cd ripper starts issuing requests that are too big. Maybe even add a printk() here, so it will look like this in the kernel you test:

Quoted 6 lines

if (size > (q->sectors << 9)) {
printk("%u rejected\n", size);
return -EINVAL;
}

to verify.

Quoted 7 lines

Yeah, that was it. Two lines in the log:

Oct 4 22:07:04 zmei kernel: 3145728 rejected
Oct 4 22:07:04 zmei kernel: 3145728 rejected

Hmm, so this means that my dvd drive is sending too big requests. What do we
do: firmware upgrade?

It actually means we have a little discrepancy between what programs expact of the api. What program are you using? They are supposed to read back what value has been set with SG_GET_RESERVED_SIZE, I guess this one does not.

What's a little extra strange is that this command apparently doesn't have ->dxfer_len set to the correct size, or it should be caught in sg_io() at issue time.

Signature · 1 line
Jens Axboe
#19 lore raw
Quoted 52 lines · Jens Axboe
Quoted 51 lines · Borislav Petkov
Quoted 38 lines · Jens Axboe
Quoted 13 lines · Borislav Petkov

Ok here we go,

final results:

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc3-bk1: OK
2.6.8-rc3-bk2: OK
2.6.8-rc3-bk3: OK
2.6.8-rc3-bk4: OK
2.6.8-rc4: BUG!

Quoted 2 lines

So, assuming that everything went fine during testing, the bug got introduced in the transition between 2.6.8-rc3-bk4 and 2.6.8-rc4.

That's some nice testing, thank you. Try backing out this hunk:

diff -urp linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c linux-2.6.8-rc4/drivers/block/scsi_ioctl.c --- linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c 2004-08-03 23:28:51.000000000 +0200 +++ linux-2.6.8-rc4/drivers/block/scsi_ioctl.c 2004-08-10 04:24:08.000000000 +0200 @@ -90,7 +90,7 @@ static int sg_set_reserved_size(request_ if (size < 0)
return -EINVAL;
if (size > (q->max_sectors << 9))

drivers/block/scsi_ioctl.c +1 -1
-  return -EINVAL;
+  size = q->max_sectors << 9;
 
  q->sg_reserved_size = size;
  return 0;

It's the only thing that sticks out, and it could easily explain it if your cd ripper starts issuing requests that are too big. Maybe even add a printk() here, so it will look like this in the kernel you test:

Quoted 6 lines

if (size > (q->sectors << 9)) {
printk("%u rejected\n", size);
return -EINVAL;
}

to verify.

Yeah, that was it. Two lines in the log:

Oct 4 22:07:04 zmei kernel: 3145728 rejected
Oct 4 22:07:04 zmei kernel: 3145728 rejected

Quoted 8 lines
Quoted 2 lines

Hmm, so this means that my dvd drive is sending too big requests. What do we do: firmware upgrade?

It actually means we have a little discrepancy between what programs expact of the api. What program are you using? They are supposed to read back what value has been set with SG_GET_RESERVED_SIZE, I guess this one does not

It is called cdda2wav and it is part of the cdrtools package by Joerg Schilling.

Quoted 4 lines

What's a little extra strange is that this command apparently doesn't have ->dxfer_len set to the correct size, or it should be caught in sg_io() at issue time.

Boris.

#20 lore raw
Quoted 56 lines · Borislav Petkov
Quoted 55 lines · Jens Axboe
Quoted 53 lines · Borislav Petkov
Quoted 38 lines · Jens Axboe
Quoted 13 lines · Borislav Petkov

Ok here we go,

final results:

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc3-bk1: OK
2.6.8-rc3-bk2: OK
2.6.8-rc3-bk3: OK
2.6.8-rc3-bk4: OK
2.6.8-rc4: BUG!

Quoted 2 lines

So, assuming that everything went fine during testing, the bug got introduced in the transition between 2.6.8-rc3-bk4 and 2.6.8-rc4.

That's some nice testing, thank you. Try backing out this hunk:

diff -urp linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c linux-2.6.8-rc4/drivers/block/scsi_ioctl.c --- linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c 2004-08-03 23:28:51.000000000 +0200 +++ linux-2.6.8-rc4/drivers/block/scsi_ioctl.c 2004-08-10 04:24:08.000000000 +0200 @@ -90,7 +90,7 @@ static int sg_set_reserved_size(request_ if (size < 0)
return -EINVAL;
if (size > (q->max_sectors << 9))

drivers/block/scsi_ioctl.c +1 -1
-  return -EINVAL;
+  size = q->max_sectors << 9;
 
  q->sg_reserved_size = size;
  return 0;

It's the only thing that sticks out, and it could easily explain it if your cd ripper starts issuing requests that are too big. Maybe even add a printk() here, so it will look like this in the kernel you test:

Quoted 6 lines

if (size > (q->sectors << 9)) {
printk("%u rejected\n", size);
return -EINVAL;
}

to verify.

Yeah, that was it. Two lines in the log:

Oct 4 22:07:04 zmei kernel: 3145728 rejected
Oct 4 22:07:04 zmei kernel: 3145728 rejected

Hmm, so this means that my dvd drive is sending too big requests. What do we do: firmware upgrade?

Quoted 7 lines
Quoted 5 lines

It actually means we have a little discrepancy between what programs expact of the api. What program are you using? They are supposed to read back what value has been set with SG_GET_RESERVED_SIZE, I guess this one does not

It is called cdda2wav and it is part of the cdrtools package by Joerg Schilling.

Then it's a bug for sure. Not because it's Joerg, but because the semantics in the newer kernel is what he wanted. And what sg has been doing for a long time. The difference is that if you go through sg/ide-scsi, the scsi mid layer will handle requeueing a request for you. Most other drivers don't support requests larger than what the drive can handle in one operation.

You don't have an old version or anything like that, do you?

Signature · 1 line
Jens Axboe
#21 lore raw
Quoted 67 lines · Jens Axboe
Quoted 58 lines · Borislav Petkov
Quoted 57 lines · Jens Axboe
Quoted 55 lines · Borislav Petkov
Quoted 40 lines · Jens Axboe
Quoted 13 lines · Borislav Petkov

Ok here we go,

final results:

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc3-bk1: OK
2.6.8-rc3-bk2: OK
2.6.8-rc3-bk3: OK
2.6.8-rc3-bk4: OK
2.6.8-rc4: BUG!

Quoted 3 lines

So, assuming that everything went fine during testing, the bug got introduced in the transition between 2.6.8-rc3-bk4 and 2.6.8-rc4.

That's some nice testing, thank you. Try backing out this hunk:

diff -urp linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c linux-2.6.8-rc4/drivers/block/scsi_ioctl.c --- linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c 2004-08-03 23:28:51.000000000 +0200 +++ linux-2.6.8-rc4/drivers/block/scsi_ioctl.c 2004-08-10 04:24:08.000000000 +0200 @@ -90,7 +90,7 @@ static int sg_set_reserved_size(request_ if (size < 0)
return -EINVAL;
if (size > (q->max_sectors << 9))

drivers/block/scsi_ioctl.c +1 -1
-  return -EINVAL;
+  size = q->max_sectors << 9;
 
  q->sg_reserved_size = size;
  return 0;

It's the only thing that sticks out, and it could easily explain it if your cd ripper starts issuing requests that are too big. Maybe even add a printk() here, so it will look like this in the kernel you test:

Quoted 6 lines

if (size > (q->sectors << 9)) {
printk("%u rejected\n", size);
return -EINVAL;
}

to verify.

Yeah, that was it. Two lines in the log:

Oct 4 22:07:04 zmei kernel: 3145728 rejected
Oct 4 22:07:04 zmei kernel: 3145728 rejected

Hmm, so this means that my dvd drive is sending too big requests. What do we do: firmware upgrade?

Quoted 7 lines
Quoted 4 lines

It actually means we have a little discrepancy between what programs expact of the api. What program are you using? They are supposed to read back what value has been set with SG_GET_RESERVED_SIZE, I guess this one does not

It is called cdda2wav and it is part of the cdrtools package by Joerg Schilling.

Quoted 8 lines

Then it's a bug for sure. Not because it's Joerg, but because the semantics in the newer kernel is what he wanted. And what sg has been doing for a long time. The difference is that if you go through sg/ide-scsi, the scsi mid layer will handle requeueing a request for you. Most other drivers don't support requests larger than what the drive can handle in one operation.

You don't have an old version or anything like that, do you?

You mean cdda2wav? No I think it is a bleeding edge version since I run debian unstable:

Package: cdda2wav
Status: install ok installed
Priority: optional
Section: sound
Installed-Size: 324
Maintainer: Joerg Jaspert <[email protected]>
Architecture: i386
Source: cdrtools
Version: 4:2.0+a38-1
Depends: libc6 (>= 2.3.2.ds1-4)
Suggests: vorbis-tools, cdrtools-doc
Conflicts: xcdroast (<< 0.98+0alpha11)
Description: Creates WAV files from audio CDs
cdda2wav lets you digitally copy audio tracks from a CD-ROM, avoiding the distortion that is introduced when recording via a sound card. Data can be dumped into raw (cdr), wav or sun format sound files. Options control the recording format (stereo/mono; 8/16 bits; sampling rate, etc). .
Please install cdrtools-doc if you want most of the documentation and Readme-files.

It is one of the latest alpha versions according to the cdrtools website.

Boris.

#22 lore raw
In reply to Jens Axboe
Quoted 55 lines · Jens Axboe
Quoted 46 lines · Borislav Petkov
Quoted 38 lines · Jens Axboe
Quoted 13 lines · Borislav Petkov

Ok here we go,

final results:

2.6.8-rc1: OK
2.6.8-rc2: OK
2.6.8-rc3: OK
2.6.8-rc3-bk1: OK
2.6.8-rc3-bk2: OK
2.6.8-rc3-bk3: OK
2.6.8-rc3-bk4: OK
2.6.8-rc4: BUG!

Quoted 2 lines

So, assuming that everything went fine during testing, the bug got introduced in the transition between 2.6.8-rc3-bk4 and 2.6.8-rc4.

That's some nice testing, thank you. Try backing out this hunk:

diff -urp linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c linux-2.6.8-rc4/drivers/block/scsi_ioctl.c --- linux-2.6.8-rc3-bk4/drivers/block/scsi_ioctl.c 2004-08-03 23:28:51.000000000 +0200 +++ linux-2.6.8-rc4/drivers/block/scsi_ioctl.c 2004-08-10 04:24:08.000000000 +0200 @@ -90,7 +90,7 @@ static int sg_set_reserved_size(request_ if (size < 0)
return -EINVAL;
if (size > (q->max_sectors << 9))

drivers/block/scsi_ioctl.c +1 -1
-  return -EINVAL;
+  size = q->max_sectors << 9;
 
  q->sg_reserved_size = size;
  return 0;

It's the only thing that sticks out, and it could easily explain it if your cd ripper starts issuing requests that are too big. Maybe even add a printk() here, so it will look like this in the kernel you test:

Quoted 6 lines

if (size > (q->sectors << 9)) {
printk("%u rejected\n", size);
return -EINVAL;
}

to verify.

Quoted 7 lines

Yeah, that was it. Two lines in the log:

Oct 4 22:07:04 zmei kernel: 3145728 rejected
Oct 4 22:07:04 zmei kernel: 3145728 rejected

Hmm, so this means that my dvd drive is sending too big requests. What do we do: firmware upgrade?

Quoted 8 lines

It actually means we have a little discrepancy between what programs expact of the api. What program are you using? They are supposed to read back what value has been set with SG_GET_RESERVED_SIZE, I guess this one does not.

What's a little extra strange is that this command apparently doesn't have ->dxfer_len set to the correct size, or it should be caught in sg_io() at issue time.

Sorry if I've flooded your mails but I've been playing with Postfix ... 8)

Boris.