Andre Berger <[EMAIL PROTECTED]> writes:
> > I get the timeouts with both 2.2.17pre13-ben2 und
> > 2.2.17pre15-ben1 when using pmud.
>
> (on PB 3400)
>
> > > hda: Enabling MultiWord DMA 2
> > > hda: dma_intr: status=0x00 { }
> > > media bay 0 contains a floppy disk drive
> > > hda: Enabling Mult
Andre Berger <[EMAIL PROTECTED]> writes:
> I get the timeouts with both 2.2.17pre13-ben2 und
> 2.2.17pre15-ben1 when using pmud.
(on PB 3400)
> > hda: Enabling MultiWord DMA 2
> > hda: dma_intr: status=0x00 { }
> > media bay 0 contains a floppy disk drive
> > hda: Enabling MultiWord DMA 2
With
Andre Berger <[EMAIL PROTECTED]> writes:
I get the timeouts with both 2.2.17pre13-ben2 und
2.2.17pre15-ben1 when using pmud.
> Claus <[EMAIL PROTECTED]> writes:
>
> > now it's 2.2.17pre13-ben2
>
> hda: Enabling MultiWord DMA 2
> hda: dma_intr: status=0x00 { }
> media bay 0 contains a floppy dis
Claus <[EMAIL PROTECTED]> writes:
> now it's 2.2.17pre13-ben2
hda: Enabling MultiWord DMA 2
hda: dma_intr: status=0x00 { }
media bay 0 contains a floppy disk drive
hda: Enabling MultiWord DMA 2
But no crashes with pmud so far.
Andre
> >: No problem here on Lombard with the 'same' kernel. Forced crash, you say -
> >: how so, and how would be rebooting from a crash differ from just booting
> >: using yaboot?
>
>
> :)
>
> in my case there was a problem with 2.2.16, sometimes lost interrupts,
> especially during boot when p
>: >
>: > there's no timeout in normal cases without extended fsck'ing,
>: > so I think benh's modifications do prevent lost interrupts etc.
>:
>: No problem here on Lombard with the 'same' kernel. Forced crash, you say -
>: how so, and how would be rebooting from a crash differ from just b
> now it's 2.2.17pre13-ben2
>
> after a forced crash:
> Jul 31 11:56:45 apple pmud: PMU version 9: 2400/3400/3500
> Jul 31 11:56:46 apple pmud: pmud startup succeeded
> Jul 31 11:56:46 apple pmud[306]: running /etc/power/pwrctl maximum ac
> Jul 31 11:56:46 apple pcmcia: Starting PCMCIA services:
>
>what's with your problems, -raf , Andre?
I've not had time to do any kernel compiles, so I haven't been trying out
any of this stuff. But if somebody has a pointer to a precompiled kernel,
I'd be happy to give it a whirl...
__
>: hda: Enabling MultiWord DMA 2
>: ide_pmace: wait_for_ready, error status: 51
>: pmac_ide_do_setfeature disk not ready after SET_FEATURE !
>: hda: dma_intr: status=0x00 { }
>: hda: Failed !
>:
>:
>: [9][22:56]uname -r
>: 2.2.16pmac_stable
now it's 2.2.17pre13-ben2
after a forced cra
>: Hast Du eigentlich auch die IDE-Probleme? Und was mich am meisten
saw it this time:
Registered ide 1 for media bay 0
hda: IBM-DARA-212000, ATA DISK drive
hdc: CD-ROMFFREDDIE+, ATAPI CDROM drive
ide0 at 0x102-0x1020007,0x1020160 on irq 13
ide1 at 0x1021000-0x1021007,0x1021160 on irq 14
hda
> > Also, to get rid of the interrupt, can you try changing
> > pmac_ide_do_setfeature() this way: (hand-made diff)
>
> I'll give it a try, but as it takes almost an hour to compile
> a kernel here, it may take some time.
Just recompiling the kernel after changing a few lines in ide-pmac won't
ta
>>
>Let me first say that
>>> In function dma_bits_to_command(), comment out the MDMA 2 case
>gave me Multiword DMA 1 instead of 2, but didn't have any other effect.
Well, I wanted to be sure you driver didn't report a bogus capability
about MDMA 2.
Ben.
Benjamin Herrenschmidt <[EMAIL PROTECTED]> writes:
> Thanks. Now, I'd like if you could try to get more precise values because
> those I sent you are really too high.
>
Let me first say that
>> In function dma_bits_to_command(), comment out the MDMA 2 case
gave me Multiword DMA 1 instead of 2, bu
Benjamin Herrenschmidt wrote:
> Ok, I've pushed a slightly modified version, can you test again ?
No luck, unplugging the CD-ROM during boot (while it waits) helps. The book
boots without CD-ROM.
I commented out the line 641: OUT_BYTE(drive->ctl|2,IDE_CONTROL_REG);
Then it boots again.
Andrea
Benjamin Herrenschmidt wrote:
> Interesting... obviously that should not happen. Let me double check the
> code. Note that the timeout is set to 10 secs, I suppose you have waited
> more than that ?
About two minutes.
>
> Ok, I've pushed a slightly modified version, can you test again ?
I resy
>
>A synced tree from about 14:00 results here in a hanger on a wallstreet I.
>
>-> hdc: Enabling MultiWord DMA 2
>nothing more..
>
>I resync again.
Interesting... obviously that should not happen. Let me double check the
code. Note that the timeout is set to 10 secs, I suppose you have waited
Benjamin Herrenschmidt wrote:
> >media bay 0 contains a CD-ROM drive
> >media-bay 0 is ide 1
> >hda: Enabling MultiWord DMA 2
> >hdc: Enabling MultiWord DMA 2
> >ide1: unexpected interrupt, status=0x00, count=1
>
> Hum... looking at the code again, it appears that the IDE_WAKEUP_DELAY_MS
> is n
>> - First, try increasing the wait_for_ready to a larger value
>> line 597:
>> int timeout = 5000 (was 2000)
>> - Try increasing the IDE_WAKEUP_DELAY_MS definition to a much larger
>> > value (1 ?)
>> line 102:
>> #define IDE_WAKEUP_DELAY_MS 1 (was 1500)
>
>Those two measures seem to hav
>Those two measures seem to have solved the problem for now. I just get
>the following message when waking the machine up from pmud:
Thanks. Now, I'd like if you could try to get more precise values because
those I sent you are really too high.
Can you try changing only one of them, and then try
Andre Berger <[EMAIL PROTECTED]> writes:
> - First, try increasing the wait_for_ready to a larger value
> line 597:
> int timeout = 5000 (was 2000)
> - Try increasing the IDE_WAKEUP_DELAY_MS definition to a much larger
> > value (1 ?)
> line 102:
> #define IDE_WAKEUP_DELAY_MS 1 (was 1500
Benjamin Herrenschmidt <[EMAIL PROTECTED]> writes:
> >I have recompiled my 2.2.17pre11 kernel with int timeout=5000; nothing
> >has changed (PB 3400):
> Sounds like a bogus hard disk, it's returning an error status from the
> SET_FEATURE command while advertising some DMA support. I'll double chec
> >Anything in the interrupt code, or the arch independent IDE code since
> >then? From the descriptions posted, there was a stray IDE interrupt after
> >the set multiword DMA failure in most cases. Should the SET_FEATURE
> >command that performs the set multiword DMA 2 operation generate an
> >int
>
>Anything in the interrupt code, or the arch independent IDE code since
>then? From the descriptions posted, there was a stray IDE interrupt after
>the set multiword DMA failure in most cases. Should the SET_FEATURE
>command that performs the set multiword DMA 2 operation generate an
>interrupt?
> >the second try get me running, don't know why.
> >kernel 2.2.15pre14 didn't do this for sure, pmud is the same plus
> >garcia's pmu patch (enabling wmapm)
>
> I don't think I changed anything to either IDE or serial sleep code since
> 2.2.15pre14. Weird...
Anything in the interrupt code, or th
>
>the second try get me running, don't know why.
>kernel 2.2.15pre14 didn't do this for sure, pmud is the same plus
>garcia's pmu patch (enabling wmapm)
I don't think I changed anything to either IDE or serial sleep code since
2.2.15pre14. Weird...
Ben.
>:
>: I have recompiled my 2.2.17pre11 kernel with int timeout=5000; nothing
>: has changed (PB 3400):
I just ran into the described hangup, timeout=2000 :)
the machine was in 'snooze' for about two hours; then I wake it up and tried to
connect to my ISP (isdn modem, ttyS0). The connector was
>I have recompiled my 2.2.17pre11 kernel with int timeout=5000; nothing
>has changed (PB 3400):
>
>ide0 at 0x102-0x1020007,0x1020160 on irq 13
>ide1 at 0x1021000-0x1021007,0x1021160 on irq 14
>hda: Enabling MultiWord DMA 2
>hda: IBM-DMCA-21440, 1378MB w/96kB Cache, CHS=2800/16/63, DMA
>hdc: Ena
Claus <[EMAIL PROTECTED]> writes:
> drivers/block/ide-pmac.c line 597
>
> wait_for_ready(ide_drive_t *drive)
> {
> /* Timeout bumped for some powerbooks */
> int timeout = 2000;
> byte stat;
I have recompiled my 2.2.17pre11 kernel with int timeout=5000; nothing
has changed (P
> >: could you name the file to fix?
>
> sorry,
> I guess it's (2.2.16 here)
> drivers/block/ide-pmac.c line 597
>
> wait_for_ready(ide_drive_t *drive)
> {
> /* Timeout bumped for some powerbooks */
> int timeout = 2000;
> byte stat;
That's the one, yes.
Michael
> > New hypothesis: generic too short timeout - please increase timeout to
> > 5000 in wait_for_ready.
> >
> > Michael
>
> Where could I increase that timeout?
In drivers/block/ide-pmac.d:wait_for_ready(), it's set to 2000 right now.
Recompile and try the resulting kernel.
Michael
>:
>: >:
>: >: New hypothesis: generic too short timeout - please increase timeout to
>: >: 5000 in wait_for_ready.
>:
>: could you name the file to fix?
sorry,
I guess it's (2.2.16 here)
drivers/block/ide-pmac.c line 597
wait_for_ready(ide_drive_t *drive)
{
/* Timeout bum
On Tue, Jul 25, 2000 at 08:18:51PM +0200, Michael Schmitz:
>: > information in the pot and say I often see this sort of error during boot,
>: > and sometimes the system recovers and continues, and othertimes it locks
in my case it never locks at boot
>: > up before the login prompt appears
Michael Schmitz <[EMAIL PROTECTED]> writes:
> New hypothesis: generic too short timeout - please increase timeout to
> 5000 in wait_for_ready.
>
> Michael
Where could I increase that timeout?
--
Andre
> New hypothesis: generic too short timeout - please increase timeout to
> 5000 in wait_for_ready.
There may be something here. I just updated to 2.2.17 as packaged in
frozen (compile date of Jul 18th). I got some form of the usual IDE
errors on boot, from which it finally recovered, which incl
> > There's been a dma_intr status=0x00 message for the hda case as well,
> > apparently enabling the multiword DMA will generate a DMA interrupt.
> > Please note that the set multiword DMA 2 command succeeded earlier for hdc
> > during driver init. Seems the problem is only with this function bein
> There's been a dma_intr status=0x00 message for the hda case as well,
> apparently enabling the multiword DMA will generate a DMA interrupt.
> Please note that the set multiword DMA 2 command succeeded earlier for hdc
> during driver init. Seems the problem is only with this function being
> used
> Well there seems to be a multiword DMA 2 problem. All -d1... commands
> on /dev/hda result in the orig. error; -d0 solves the problem. Note the
Ok, that sorts out the timeout hypothesis. If the -d1 always fails I don't
understand why the problem doesn't happen during kernel init though.
You can
Michael Schmitz <[EMAIL PROTECTED]> writes:
> You could try the following to narrow down the problem further:
>
> - omit the -p in the wakeup hdparm command, to make sure we're not running
> into a problem with PIO mode autotune here (the kernel option is kind of
> defeated by that option for all
> > You could try the following to narrow down the problem further:
> >
> > - omit the -p in the wakeup hdparm command, to make sure we're not running
> > into a problem with PIO mode autotune here (the kernel option is kind of
> > defeated by that option for all I can see).
>
> Sorry, which one
Michael Schmitz <[EMAIL PROTECTED]> writes:
> > > Your error message looks like the disk doesn't like one of the feature set
> > > commands the kernel sends it. I'll have to look at the IDE driver to see
> > > if that's the DMA setting or some other drive command. Did you try
> > > hda=noautotune,
> > Your error message looks like the disk doesn't like one of the feature set
> > commands the kernel sends it. I'll have to look at the IDE driver to see
> > if that's the DMA setting or some other drive command. Did you try
> > hda=noautotune, and does this happen with all kernel versions?
>
>
> >hda: Enabling MultiWord DMA 2
> >ide_pmace: wait_for_ready, error status: 51
> >pmac_ide_do_setfeature disk not ready after SET_FEATURE !
> >hda: dma_intr: status=0x00 { }
> >hda: Failed !
>
> I get similar ide related errors or lockups on boot on my PB3400, with both
> the packaged kernels ava
Michael Schmitz <[EMAIL PROTECTED]> writes:
> > hda: Enabling MultiWord DMA 2
> > ide_pmace: wait_for_ready, error status: 51
> > pmac_ide_do_setfeature disk not ready after SET_FEATURE !
> > hda: dma_intr: status=0x00 { }
> > hda: Failed !
> >
> > I have an 'alien'ized pmud 0.6.1 installed and i
> hda: Enabling MultiWord DMA 2
> ide_pmace: wait_for_ready, error status: 51
> pmac_ide_do_setfeature disk not ready after SET_FEATURE !
> hda: dma_intr: status=0x00 { }
> hda: Failed !
>
> I have an 'alien'ized pmud 0.6.1 installed and its script
You can install the pmu 0.6 deb from unstable. I
PB3400, with both
the packaged kernels available in frozen (.15pre? and .17pre?) but the
error message I see most often is "lost interrupt" I haven't been able to
figure out why. I can say that I ran several different older kernel
versions under LinuxPPC without ever having any IDE
I get the follpwing message every now and then on my potato box, a
PowerBook 3400 (custom 2.2.17pre11, IDE settings not changed from pmac
standard), or even a complete lockup at boot time which forces me to
press the reset btn, and has already resulted in data loss(t+found):
hda: Enabling MultiW
46 matches
Mail list logo
