[2/6] 2.6.21-rc2: known regressions

[2/6] 2.6.21-rc2: known regressions

[Adrian Bunk]

This email lists some known regressions in 2.6.21-rc2 compared to 2.6.20
that are not yet fixed in Linus' tree.

If you find your name in the Cc header, you are either submitter of one
of the bugs, maintainer of an affectected subsystem or driver, a patch
of you caused a breakage or I'm considering you in any other way
possibly involved with one or more of these issues.

Due to the huge amount of recipients, please trim the Cc when answering.


Subject    : ThinkPad doesn't resume from suspend to RAM
References : http://lkml.org/lkml/2007/2/27/80
             http://lkml.org/lkml/2007/2/28/348
Submitter  : Jens Axboe <jens.axboe@oracle.com>
             Jeff Chua <jeff.chua.linux@gmail.com>
Status     : unknown


Subject    : beeps get longer after suspend
References : http://lkml.org/lkml/2007/2/26/276
Submitter  : Pavel Machek <pavel@ucw.cz>
Status     : unknown


Subject    : ThinkPad T60: no screen after suspend to RAM
References : http://lkml.org/lkml/2007/2/22/391
Submitter  : Michael S. Tsirkin <mst@mellanox.co.il>
Status     : unknown


Subject    : ThinkPad Z60m: usb mouse stops working after suspend to ram
References : http://lkml.org/lkml/2007/2/21/413
             http://lkml.org/lkml/2007/2/28/172
Submitter  : Arkadiusz Miskiewicz <arekm@maven.pl>
Caused-By  : Konstantin Karasyov <konstantin.a.karasyov@intel.com>
             commit 0a6139027f3986162233adc17285151e78b39cac
Handled-By : Konstantin Karasyov <konstantin.a.karasyov@intel.com>
Status     : problem is being debugged


Subject    : AE_NOT_FOUND ACPI messages
References : http://bugzilla.kernel.org/show_bug.cgi?id=8066
             http://lkml.org/lkml/2007/2/27/263
Submitter  : Thomas Meyer <thomas.mey@web.de>
             Meelis Roos <mroos@linux.ee>
Handled-By : Alexey Starikovskiy <alexey.y.starikovskiy@linux.intel.com>
Patch      : http://bugzilla.kernel.org/show_bug.cgi?id=8066
Status     : patch available


Subject    : Asus M6Ne notebook: ACPI: First battery is not detected
References : http://bugzilla.kernel.org/show_bug.cgi?id=8080
Submitter  : Janosch Machowinski <jmachowinski@gmx.de>
Status     : unknown


Subject    : AT keyboard only works with pci=noacpi
References : http://lkml.org/lkml/2007/3/3/68
Submitter  : Ash Milsted <thatistosayiseenem@gawab.com>
Status     : unknown

Re: [2/6] 2.6.21-rc2: known regressions

[Jeff Garzik]

Adrian Bunk wrote:
> Subject    : AT keyboard only works with pci=noacpi
> References : http://lkml.org/lkml/2007/3/3/68
> Submitter  : Ash Milsted <thatistosayiseenem@gawab.com>
> Status     : unknown


sounds like a BIOS bug, even though it appears to be a regression?

Re: [2/6] 2.6.21-rc2: known regressions

[Linus Torvalds]

On Wed, 7 Mar 2007, Jeff Garzik wrote:
>
> Adrian Bunk wrote:
> > Subject    : AT keyboard only works with pci=noacpi
> > References : http://lkml.org/lkml/2007/3/3/68
> > Submitter  : Ash Milsted <thatistosayiseenem@gawab.com>
> > Status     : unknown
> 
> sounds like a BIOS bug, even though it appears to be a regression?

So?

BIOS bugs are not bugs we can fix, they are things we have to work around. 
They are like hardware bugs in a network chip: a "driver" that doesn't 
work around a BIOS bug is simply a *buggy* driver, exactly the same way a 
network driver has to work around errata in the hardware.

So it doesn't really matter if something is a BIOS bug or not. It's not 
reasonable to expect people to upgrade their BIOS'es - even if such an 
upgrade were to exist (which is fairly rare in itself). 

If it used to work, that just makes it (a) doubly important to try to fix 
it, since regressions are BAD BAD BAD and (b) a fair amount *easier* to 
fix, since we can hopefully get an idea of what broke it.

Ash, can you try to use "git bisect" to figure where it started? But 
perhaps just try -rc3 first to see if it's been fixed?

The working setups (whether with irqfixup or with pci=noacpi seem to both 
have a nice

	input: AT Translated Set 2 keyboard as /class/input/inputX

but the nonworking one does not. But even the nonworking one actually 
*found* the controller:

	PNP: PS/2 Controller [PNP0303:PS2K] at 0x60,0x64 irq 1
	PNP: PS/2 controller doesn't have AUX irq; using default 12
	serio: i8042 KBD port at 0x60,0x64 irq 1
	serio: i8042 AUX port at 0x60,0x64 irq 12
	mice: PS/2 mouse device common for all mice

which probably means that the "atkbd_probe()" function fails (probably 
because it never gets a reply, which in turn is probably because the 
interrupt routing is broken).

If you use "atkbd.reset=1" on the kernel command line, you would probably 
get a message like "keyboard reset failed"..

Now, the most likely cause is obviously just the ACPI changes that mess up 
irq routing somehow, but it would be important to figure out *what* makes 
it happen, which is why "git bisect" would be wonderful to try.

So Ash, if you get the git tree, just start with

	git bisect start
	git bisect good v2.6.20
	git bisect bad v2.6.21-rc1

and off you go.. git isn't really that hard to use any more.

		Linus

Re: [2/6] 2.6.21-rc2: known regressions

[Ash Milsted]

On Wed, 07 Mar 2007 06:09:06 -0500
Jeff Garzik <jeff@garzik.org> wrote:

> Adrian Bunk wrote:
> > Subject    : AT keyboard only works with pci=noacpi
> > References : http://lkml.org/lkml/2007/3/3/68
> > Submitter  : Ash Milsted <thatistosayiseenem@gawab.com>
> > Status     : unknown
> 
> 
> sounds like a BIOS bug, even though it appears to be a regression?
> 

Resolved here:
http://marc.theaimsgroup.com/?l=linux-kernel&m=117332735404395&w=2

Was a bug in the i8042 driver it seems.

Ash

Re: [2/6] 2.6.21-rc2: known regressions

[Michael S. Tsirkin]

[-- Attachment #1: Type: text/plain, Size: 996 bytes --]

> Quoting Adrian Bunk <bunk@stusta.de>:
> Subject: [2/6] 2.6.21-rc2: known regressions
> 
> Subject    : ThinkPad T60: no screen after suspend to RAM
> References : http://lkml.org/lkml/2007/2/22/391
> Submitter  : Michael S. Tsirkin <mst@mellanox.co.il>
> Status     : unknown

Here's the status with -rc3: better, but still does not work as well as 2.6.20.

1. Switching to console VT, and running s2ram, I am able to resume to console
   (I have not tested with echo mem > /sys/power/state yet, both work OK
    with 2.6.20)
   This seems to take about as long as with 2.6.20.
   During this time, I see some messages on console (see attached log)

2. First disk access after resume takes a couple of minutes
   (seemed instant with 2.6.20) during this time no new messages show on console

3. When I switch to X (CTRL-ALT-F7), X hangs after drawing a couple of windows
   after waiting for some 10 min, I rebooted.
   no new messages showed up in /var/log/messages

log attached
   
-- 
MST


[-- Attachment #2: newlog --]
[-- Type: text/plain, Size: 20275 bytes --]

Mar  8 08:47:20 localhost kernel: Inspecting /boot/System.map-2.6.21-rc3-work
Mar  8 08:47:20 localhost kernel: Loaded 31265 symbols from /boot/System.map-2.6.21-rc3-work.
Mar  8 08:47:20 localhost kernel: Symbols match kernel version 2.6.21.
Mar  8 08:47:20 localhost kernel: No module symbols loaded - kernel modules not enabled. 
Mar  8 08:47:20 localhost kernel: 1
Mar  8 08:47:20 localhost kernel: [   15.317823] ata2: SATA link down (SStatus 0 SControl 0)
Mar  8 08:47:20 localhost kernel: [   15.317897] scsi2 : ahci
Mar  8 08:47:20 localhost kernel: [   15.379836] ata3: SATA link down (SStatus 0 SControl 0)
Mar  8 08:47:20 localhost kernel: [   15.379909] scsi3 : ahci
Mar  8 08:47:20 localhost kernel: [   15.441854] ata4: SATA link down (SStatus 0 SControl 0)
Mar  8 08:47:20 localhost kernel: [   15.441999] scsi 0:0:0:0: Direct-Access     ATA      HTS541080G9SA00  MB4I PQ: 0 ANSI: 5
Mar  8 08:47:20 localhost kernel: [   15.442216] SCSI device sda: 156301488 512-byte hdwr sectors (80026 MB)
Mar  8 08:47:20 localhost kernel: [   15.442292] sda: Write Protect is off
Mar  8 08:47:20 localhost kernel: [   15.443138] SCSI device sda: write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Mar  8 08:47:20 localhost kernel: [   15.443256] SCSI device sda: 156301488 512-byte hdwr sectors (80026 MB)
Mar  8 08:47:20 localhost kernel: [   15.443330] sda: Write Protect is off
Mar  8 08:47:20 localhost kernel: [   15.443406] SCSI device sda: write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Mar  8 08:47:20 localhost kernel: [   15.443490]  sda: sda1 sda2 sda3 sda4 < sda5 sda6 >
Mar  8 08:47:20 localhost kernel: [   15.845145] sd 0:0:0:0: Attached scsi disk sda
Mar  8 08:47:20 localhost kernel: [   15.845290] sd 0:0:0:0: Attached scsi generic sg0 type 0
Mar  8 08:47:20 localhost kernel: [   15.845797] ACPI: PCI Interrupt 0000:00:1d.7[D] -> GSI 19 (level, low) -> IRQ 21
Mar  8 08:47:20 localhost kernel: [   15.845946] ehci_hcd 0000:00:1d.7: EHCI Host Controller
Mar  8 08:47:20 localhost kernel: [   15.846106] ehci_hcd 0000:00:1d.7: new USB bus registered, assigned bus number 1
Mar  8 08:47:20 localhost kernel: [   15.846229] ehci_hcd 0000:00:1d.7: debug port 1
Mar  8 08:47:20 localhost kernel: [   15.846315] ehci_hcd 0000:00:1d.7: irq 21, io mem 0xee444000
Mar  8 08:47:20 localhost kernel: [   15.850269] ehci_hcd 0000:00:1d.7: USB 2.0 started, EHCI 1.00, driver 10 Dec 2004
Mar  8 08:47:20 localhost kernel: [   15.850508] usb usb1: configuration #1 chosen from 1 choice
Mar  8 08:47:20 localhost kernel: [   15.850630] hub 1-0:1.0: USB hub found
Mar  8 08:47:20 localhost kernel: [   15.850697] hub 1-0:1.0: 8 ports detected
Mar  8 08:47:20 localhost kernel: [   15.862005] USB Universal Host Controller Interface driver v3.0
Mar  8 08:47:20 localhost kernel: [   15.862135] ACPI: PCI Interrupt 0000:00:1d.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:47:20 localhost kernel: [   15.862268] uhci_hcd 0000:00:1d.0: UHCI Host Controller
Mar  8 08:47:20 localhost kernel: [   15.862381] uhci_hcd 0000:00:1d.0: new USB bus registered, assigned bus number 2
Mar  8 08:47:20 localhost kernel: [   15.862484] uhci_hcd 0000:00:1d.0: irq 20, io base 0x00001820
Mar  8 08:47:20 localhost kernel: [   15.862702] usb usb2: configuration #1 chosen from 1 choice
Mar  8 08:47:20 localhost kernel: [   15.862821] hub 2-0:1.0: USB hub found
Mar  8 08:47:20 localhost kernel: [   15.862887] hub 2-0:1.0: 2 ports detected
Mar  8 08:47:20 localhost kernel: [   15.892543] ACPI: PCI Interrupt 0000:00:1d.1[B] -> GSI 17 (level, low) -> IRQ 22
Mar  8 08:47:20 localhost kernel: [   15.892676] uhci_hcd 0000:00:1d.1: UHCI Host Controller
Mar  8 08:47:20 localhost kernel: [   15.892789] uhci_hcd 0000:00:1d.1: new USB bus registered, assigned bus number 3
Mar  8 08:47:20 localhost kernel: [   15.892894] uhci_hcd 0000:00:1d.1: irq 22, io base 0x00001840
Mar  8 08:47:20 localhost kernel: [   15.893113] usb usb3: configuration #1 chosen from 1 choice
Mar  8 08:47:20 localhost kernel: [   15.893229] hub 3-0:1.0: USB hub found
Mar  8 08:47:20 localhost kernel: [   15.893296] hub 3-0:1.0: 2 ports detected
Mar  8 08:47:20 localhost kernel: [   15.933182] ACPI: PCI Interrupt 0000:00:1d.2[C] -> GSI 18 (level, low) -> IRQ 23
Mar  8 08:47:20 localhost kernel: [   15.933314] uhci_hcd 0000:00:1d.2: UHCI Host Controller
Mar  8 08:47:20 localhost kernel: [   15.933427] uhci_hcd 0000:00:1d.2: new USB bus registered, assigned bus number 4
Mar  8 08:47:20 localhost kernel: [   15.933577] uhci_hcd 0000:00:1d.2: irq 23, io base 0x00001860
Mar  8 08:47:20 localhost kernel: [   15.933797] usb usb4: configuration #1 chosen from 1 choice
Mar  8 08:47:20 localhost kernel: [   15.933915] hub 4-0:1.0: USB hub found
Mar  8 08:47:20 localhost kernel: [   15.933981] hub 4-0:1.0: 2 ports detected
Mar  8 08:47:20 localhost kernel: [   15.976494] ACPI: PCI Interrupt 0000:00:1d.3[D] -> GSI 19 (level, low) -> IRQ 21
Mar  8 08:47:20 localhost kernel: [   15.976627] uhci_hcd 0000:00:1d.3: UHCI Host Controller
Mar  8 08:47:20 localhost kernel: [   15.976740] uhci_hcd 0000:00:1d.3: new USB bus registered, assigned bus number 5
Mar  8 08:47:20 localhost kernel: [   15.976843] uhci_hcd 0000:00:1d.3: irq 21, io base 0x00001880
Mar  8 08:47:20 localhost kernel: [   15.977060] usb usb5: configuration #1 chosen from 1 choice
Mar  8 08:47:20 localhost kernel: [   15.977196] hub 5-0:1.0: USB hub found
Mar  8 08:47:20 localhost kernel: [   15.977263] hub 5-0:1.0: 2 ports detected
Mar  8 08:47:20 localhost kernel: [   16.040824] Initializing USB Mass Storage driver...
Mar  8 08:47:20 localhost kernel: [   16.157808] usb 5-2: new full speed USB device using uhci_hcd and address 2
Mar  8 08:47:20 localhost kernel: [   16.221329] usb 5-2: configuration #1 chosen from 1 choice
Mar  8 08:47:20 localhost kernel: [   16.224625] usbcore: registered new interface driver usb-storage
Mar  8 08:47:20 localhost kernel: [   16.224692] USB Mass Storage support registered.
Mar  8 08:47:20 localhost kernel: [   16.224886] PNP: PS/2 Controller [PNP0303:KBD,PNP0f13:MOU] at 0x60,0x64 irq 1,12
Mar  8 08:47:20 localhost kernel: [   16.233142] serio: i8042 KBD port at 0x60,0x64 irq 1
Mar  8 08:47:20 localhost kernel: [   16.233211] serio: i8042 AUX port at 0x60,0x64 irq 12
Mar  8 08:47:20 localhost kernel: [   16.233366] mice: PS/2 mouse device common for all mice
Mar  8 08:47:20 localhost kernel: [   16.233716] Advanced Linux Sound Architecture Driver Version 1.0.14rc3 (Tue Mar 06 13:10:00 2007 UTC).
Mar  8 08:47:20 localhost kernel: [   16.234221] ACPI: PCI Interrupt 0000:00:1b.0[B] -> GSI 17 (level, low) -> IRQ 22
Mar  8 08:47:20 localhost kernel: [   16.236937] input: AT Translated Set 2 keyboard as /class/input/input3
Mar  8 08:47:20 localhost kernel: [   16.444361] ALSA device list:
Mar  8 08:47:20 localhost kernel: [   16.444428]   #0: HDA Intel at 0xee240000 irq 22
Mar  8 08:47:20 localhost kernel: [   16.444525] nf_conntrack version 0.5.0 (7168 buckets, 57344 max)
Mar  8 08:47:20 localhost kernel: [   16.444736] ip_tables: (C) 2000-2006 Netfilter Core Team
Mar  8 08:47:20 localhost kernel: [   16.444809] TCP cubic registered
Mar  8 08:47:20 localhost kernel: [   16.444887] NET: Registered protocol family 1
Mar  8 08:47:20 localhost kernel: [   16.444953] NET: Registered protocol family 17
Mar  8 08:47:20 localhost kernel: [   16.445019] ieee80211: 802.11 data/management/control stack, git-1.1.13
Mar  8 08:47:20 localhost kernel: [   16.445087] ieee80211: Copyright (C) 2004-2005 Intel Corporation <jketreno@linux.intel.com>
Mar  8 08:47:20 localhost kernel: [   16.445185] Starting balanced_irq
Mar  8 08:47:20 localhost kernel: [   16.445254] Using IPI No-Shortcut mode
Mar  8 08:47:20 localhost kernel: [   16.472663] kjournald starting.  Commit interval 5 seconds
Mar  8 08:47:20 localhost kernel: [   16.472735] EXT3-fs: mounted filesystem with ordered data mode.
Mar  8 08:47:20 localhost kernel: [   16.472864] VFS: Mounted root (ext3 filesystem) readonly.
Mar  8 08:47:20 localhost kernel: [   16.473029] Freeing unused kernel memory: 184k freed
Mar  8 08:47:20 localhost kernel: [   16.473122] Write protecting the kernel read-only data: 902k
Mar  8 08:47:20 localhost kernel: [   24.658491] Intel(R) PRO/1000 Network Driver - version 7.3.20-k2
Mar  8 08:47:20 localhost kernel: [   24.658576] Copyright (c) 1999-2006 Intel Corporation.
Mar  8 08:47:20 localhost kernel: [   24.658761] ACPI: PCI Interrupt 0000:02:00.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:47:20 localhost kernel: [   24.733801] e1000: 0000:02:00.0: e1000_probe: (PCI Express:2.5Gb/s:Width x1) 00:16:41:54:6c:47
Mar  8 08:47:20 localhost kernel: [   24.815441] e1000: eth0: e1000_probe: Intel(R) PRO/1000 Network Connection
Mar  8 08:47:20 localhost kernel: [   25.282292] Adding 1951824k swap on /dev/sda6.  Priority:-1 extents:1 across:1951824k
Mar  8 08:47:20 localhost kernel: [   25.343370] EXT3 FS on sda3, internal journal
Mar  8 08:47:20 localhost kernel: [   25.697608] Synaptics Touchpad, model: 1, fw: 6.2, id: 0x81a0b1, caps: 0xa04793/0x300000
Mar  8 08:47:20 localhost kernel: [   25.697710] serio: Synaptics pass-through port at isa0060/serio1/input0
Mar  8 08:47:20 localhost kernel: [   25.742946] input: SynPS/2 Synaptics TouchPad as /class/input/input4
Mar  8 08:47:20 localhost kernel: [   30.419349] NTFS volume version 3.1.
Mar  8 08:47:20 localhost kernel: [   31.156204] IBM TrackPoint firmware: 0x0e, buttons: 3/3
Mar  8 08:47:20 localhost kernel: [   31.426442] input: TPPS/2 IBM TrackPoint as /class/input/input5
Mar  8 08:47:22 localhost hpiod: 0.9.7 accepting connections at 50899... 
Mar  8 08:47:26 localhost kernel: [   39.389386] [drm] Initialized i915 1.6.0 20060119 on minor 0
Mar  8 08:48:07 localhost kernel: <1] pcieport-driver 0000:00:1c.3: LATE suspend
Mar  8 08:48:07 localhost kernel: [    2.183443] Intel machine check architecture supported.
Mar  8 08:48:07 localhost kernel: [    2.183451] Intel machine check reporting enabled on CPU#0.
Mar  8 08:48:07 localhost kernel: [    2.183982] Enabling non-boot CPUs ...
Mar  8 08:48:07 localhost kernel: [    2.321035] SMP alternatives: switching to SMP code
Mar  8 08:48:07 localhost kernel: [    2.321241] Booting processor 1/1 eip 3000
Mar  8 08:48:07 localhost kernel: [    2.332199] Initializing CPU#1
Mar  8 08:48:07 localhost kernel: [    3.143727] Calibrating delay using timer specific routine.. 20089.15 BogoMIPS (lpj=100445793)
Mar  8 08:48:07 localhost kernel: [    3.143746] monitor/mwait feature present.
Mar  8 08:48:07 localhost kernel: [    3.143751] CPU: L1 I cache: 32K, L1 D cache: 32K
Mar  8 08:48:07 localhost kernel: [    3.143755] CPU: L2 cache: 2048K
Mar  8 08:48:07 localhost kernel: [    3.143759] CPU: Physical Processor ID: 0
Mar  8 08:48:07 localhost kernel: [    3.143761] CPU: Processor Core ID: 1
Mar  8 08:48:07 localhost kernel: [    3.143774] Intel machine check architecture supported.
Mar  8 08:48:07 localhost kernel: [    3.143783] Intel machine check reporting enabled on CPU#1.
Mar  8 08:48:07 localhost kernel: [    3.144104] CPU1: Intel Genuine Intel(R) CPU           T2400  @ 1.83GHz stepping 08
Mar  8 08:48:07 localhost kernel: [    3.144892] speedstep-centrino with X86_SPEEDSTEP_CENTRINO_ACPI config is deprecated.
Mar  8 08:48:07 localhost kernel: [    3.144896]  Use X86_ACPI_CPUFREQ (acpi-cpufreq) instead.
Mar  8 08:48:07 localhost kernel: [    3.144936] CPU1 is up
Mar  8 08:48:07 localhost kernel: [    3.440273] ACPI: PCI Interrupt 0000:00:02.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:07 localhost kernel: [    3.527272] ACPI: PCI Interrupt 0000:00:1b.0[B] -> GSI 17 (level, low) -> IRQ 22
Mar  8 08:48:07 localhost kernel: [    3.912917] ACPI: PCI Interrupt 0000:00:1d.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:07 localhost kernel: [    3.912966] usb usb2: root hub lost power or was reset
Mar  8 08:48:07 localhost kernel: [    3.912988] ACPI: PCI Interrupt 0000:00:1d.1[B] -> GSI 17 (level, low) -> IRQ 22
Mar  8 08:48:07 localhost kernel: [    3.913036] usb usb3: root hub lost power or was reset
Mar  8 08:48:07 localhost kernel: [    3.913055] ACPI: PCI Interrupt 0000:00:1d.2[C] -> GSI 18 (level, low) -> IRQ 23
Mar  8 08:48:07 localhost kernel: [    3.913102] usb usb4: root hub lost power or was reset
Mar  8 08:48:07 localhost kernel: [    3.913120] ACPI: PCI Interrupt 0000:00:1d.3[D] -> GSI 19 (level, low) -> IRQ 21
Mar  8 08:48:07 localhost kernel: [    3.913168] usb usb5: root hub lost power or was reset
Mar  8 08:48:07 localhost kernel: [    3.914015] ACPI: PCI Interrupt 0000:00:1d.7[D] -> GSI 19 (level, low) -> IRQ 21
Mar  8 08:48:07 localhost kernel: [    3.914205] ACPI: PCI Interrupt 0000:00:1f.1[C] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:07 localhost kernel: [    3.914407] ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:07 localhost kernel: [    3.922880] ACPI: PCI Interrupt 0000:02:00.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:07 localhost kernel: [    3.928614] ACPI: PCI Interrupt 0000:15:00.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:07 localhost kernel: [    3.928638] pnp: Device 00:08 does not support activation.
Mar  8 08:48:07 localhost kernel: [    3.928641] pnp: Device 00:09 does not support activation.
Mar  8 08:48:07 localhost kernel: [    4.209165] ata2: SATA link down (SStatus 0 SControl 0)
Mar  8 08:48:07 localhost kernel: [    4.209179] ata3: SATA link down (SStatus 0 SControl 0)
Mar  8 08:48:07 localhost kernel: [    4.209191] ata4: SATA link down (SStatus 0 SControl 0)
Mar  8 08:48:07 localhost kernel: [    4.849894] atkbd.c: Unknown key released (translated set 2, code 0xe0 on isa0060/serio0).
Mar  8 08:48:07 localhost kernel: [    4.849896] atkbd.c: Use 'setkeycodes e060 <keycode>' to make it known.
Mar  8 08:48:07 localhost kernel: [    4.850297] atkbd.c: Unknown key pressed (translated set 2, code 0x63 on isa0060/serio0).
Mar  8 08:48:07 localhost kernel: [    4.850300] atkbd.c: Use 'setkeycodes 63 <keycode>' to make it known.
Mar  8 08:48:07 localhost kernel: [    5.826771] Restarting tasks ... <6>usb 5-2: USB disconnect, address 2
Mar  8 08:48:07 localhost kernel: [    5.839206] done.
Mar  8 08:48:07 localhost kernel: [    6.702776] ata1: waiting for device to spin up (7 secs)
Mar  8 08:48:07 localhost kernel: [    6.775788] usb 5-2: new full speed USB device using uhci_hcd and address 3
Mar  8 08:48:07 localhost kernel: [    6.866806] usb 5-2: configuration #1 chosen from 1 choice
Mar  8 08:48:07 localhost kernel: [   11.334773] ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
Mar  8 08:48:07 localhost kernel: [   11.673734] ata1.00: configured for UDMA/100
Mar  8 08:48:07 localhost kernel: [   11.696311] SCSI device sda: 156301488 512-byte hdwr sectors (80026 MB)
Mar  8 08:48:07 localhost kernel: [   11.697691] sda: Write Protect is off
Mar  8 08:48:07 localhost kernel: [   11.708467] SCSI device sda: write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Mar  8 08:48:16 localhost kernel: <>[   19.841599] button button_power:00: suspend
Mar  8 08:48:16 localhost kernel: [   19.841607] Disabling non-boot CPUs ...
Mar  8 08:48:16 localhost kernel: [   19.847811] Breaking affinity for irq 12
Mar  8 08:48:16 localhost kernel: [   20.444974] CPU 1 is now offline
Mar  8 08:48:16 localhost kernel: [   20.444980] SMP alternatives: switching to UP code
Mar  8 08:48:16 localhost kernel: [   20.445823] CPU1 is down
Mar  8 08:48:16 localhost kernel: [    2.423903] Intel machine check architecture supported.
Mar  8 08:48:16 localhost kernel: [    2.423911] Intel machine check reporting enabled on CPU#0.
Mar  8 08:48:16 localhost kernel: [    2.424444] Enabling non-boot CPUs ...
Mar  8 08:48:16 localhost kernel: [    2.534642] SMP alternatives: switching to SMP code
Mar  8 08:48:16 localhost kernel: [    2.534847] Booting processor 1/1 eip 3000
Mar  8 08:48:16 localhost kernel: [    2.545805] Initializing CPU#1
Mar  8 08:48:16 localhost kernel: [    3.357337] Calibrating delay using timer specific routine.. 20089.13 BogoMIPS (lpj=100445659)
Mar  8 08:48:16 localhost kernel: [    3.357357] monitor/mwait feature present.
Mar  8 08:48:16 localhost kernel: [    3.357362] CPU: L1 I cache: 32K, L1 D cache: 32K
Mar  8 08:48:16 localhost kernel: [    3.357365] CPU: L2 cache: 2048K
Mar  8 08:48:16 localhost kernel: [    3.357369] CPU: Physical Processor ID: 0
Mar  8 08:48:16 localhost kernel: [    3.357372] CPU: Processor Core ID: 1
Mar  8 08:48:16 localhost kernel: [    3.357384] Intel machine check architecture supported.
Mar  8 08:48:16 localhost kernel: [    3.357394] Intel machine check reporting enabled on CPU#1.
Mar  8 08:48:16 localhost kernel: [    3.357711] CPU1: Intel Genuine Intel(R) CPU           T2400  @ 1.83GHz stepping 08
Mar  8 08:48:16 localhost kernel: [    3.358503] speedstep-centrino with X86_SPEEDSTEP_CENTRINO_ACPI config is deprecated.
Mar  8 08:48:16 localhost kernel: [    3.358507]  Use X86_ACPI_CPUFREQ (acpi-cpufreq) instead.
Mar  8 08:48:16 localhost kernel: [    3.358548] CPU1 is up
Mar  8 08:48:16 localhost kernel: [    3.653795] ACPI: PCI Interrupt 0000:00:02.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:16 localhost kernel: [    3.740900] ACPI: PCI Interrupt 0000:00:1b.0[B] -> GSI 17 (level, low) -> IRQ 22
Mar  8 08:48:16 localhost kernel: [    4.126569] ACPI: PCI Interrupt 0000:00:1d.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:16 localhost kernel: [    4.126619] usb usb2: root hub lost power or was reset
Mar  8 08:48:16 localhost kernel: [    4.126639] ACPI: PCI Interrupt 0000:00:1d.1[B] -> GSI 17 (level, low) -> IRQ 22
Mar  8 08:48:16 localhost kernel: [    4.126686] usb usb3: root hub lost power or was reset
Mar  8 08:48:16 localhost kernel: [    4.126705] ACPI: PCI Interrupt 0000:00:1d.2[C] -> GSI 18 (level, low) -> IRQ 23
Mar  8 08:48:16 localhost kernel: [    4.126752] usb usb4: root hub lost power or was reset
Mar  8 08:48:16 localhost kernel: [    4.126772] ACPI: PCI Interrupt 0000:00:1d.3[D] -> GSI 19 (level, low) -> IRQ 21
Mar  8 08:48:16 localhost kernel: [    4.126819] usb usb5: root hub lost power or was reset
Mar  8 08:48:16 localhost kernel: [    4.127659] ACPI: PCI Interrupt 0000:00:1d.7[D] -> GSI 19 (level, low) -> IRQ 21
Mar  8 08:48:16 localhost kernel: [    4.127848] ACPI: PCI Interrupt 0000:00:1f.1[C] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:16 localhost kernel: [    4.128049] ACPI: PCI Interrupt 0000:00:1f.2[B] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:16 localhost kernel: [    4.136510] ACPI: PCI Interrupt 0000:02:00.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:16 localhost kernel: [    4.142247] ACPI: PCI Interrupt 0000:15:00.0[A] -> GSI 16 (level, low) -> IRQ 20
Mar  8 08:48:16 localhost kernel: [    4.142271] pnp: Device 00:08 does not support activation.
Mar  8 08:48:16 localhost kernel: [    4.142274] pnp: Device 00:09 does not support activation.
Mar  8 08:48:16 localhost kernel: [    4.421975] ata2: SATA link down (SStatus 0 SControl 0)
Mar  8 08:48:16 localhost kernel: [    4.421989] ata3: SATA link down (SStatus 0 SControl 0)
Mar  8 08:48:16 localhost kernel: [    4.422001] ata4: SATA link down (SStatus 0 SControl 0)
Mar  8 08:48:16 localhost kernel: [    5.971047] Restarting tasks ... done.
Mar  8 08:48:16 localhost kernel: [    6.012513] usb 5-2: USB disconnect, address 3
Mar  8 08:48:16 localhost kernel: [    6.891677] ata1: waiting for device to spin up (7 secs)
Mar  8 08:48:16 localhost kernel: [    6.958587] usb 5-2: new full speed USB device using uhci_hcd and address 4
Mar  8 08:48:16 localhost kernel: [    7.049506] usb 5-2: configuration #1 chosen from 1 choice
Mar  8 08:48:16 localhost kernel: [   11.547844] ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
Mar  8 08:48:16 localhost kernel: [   11.885459] ata1.00: configured for UDMA/100
Mar  8 08:48:16 localhost kernel: [   11.913910] SCSI device sda: 156301488 512-byte hdwr sectors (80026 MB)
Mar  8 08:48:16 localhost kernel: [   11.914575] sda: Write Protect is off
Mar  8 08:48:16 localhost kernel: [   11.914868] SCSI device sda: write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Mar  8 08:48:16 localhost shutdown[4092]: shutting down for system halt

Re: [2/6] 2.6.21-rc2: known regressions

[Jeff Chua]

On 3/8/07, Michael S. Tsirkin <mst@mellanox.co.il> wrote:
> > Quoting Adrian Bunk <bunk@stusta.de>:
> > Subject: [2/6] 2.6.21-rc2: known regressions
> >
> > Subject    : ThinkPad T60: no screen after suspend to RAM
> > References : http://lkml.org/lkml/2007/2/22/391
> > Submitter  : Michael S. Tsirkin <mst@mellanox.co.il>
> > Status     : unknown
>
> Here's the status with -rc3: better, but still does not work as well as 2.6.20.

> 3. When I switch to X (CTRL-ALT-F7), X hangs after drawing a couple of windows
>    after waiting for some 10 min, I rebooted.
>    no new messages showed up in /var/log/messages

In my case, I can "suspend" and "resume", but seems the clock died
after resume. "date" always returns the same time! I had to disable
CONFIG_NO_HZ and CONFIG_SYSFS_DEPRECATED in order to suspend.

Try to execute "date" and see if it changes after resume.

Thanks,
Jeff.

Re: [2/6] 2.6.21-rc2: known regressions

[Linus Torvalds]

[ Eric, Ingo, can you double-check the timer initialization after resume? 
  We appear to have several reports of date not advancing, and while this 
  could be some SATA issue, it could easily be a timer tick issue too ]

On Thu, 8 Mar 2007, Michael S. Tsirkin wrote:
> 
> Here's the status with -rc3: better, but still does not work as well as 2.6.20.

Ok. I think we mostly solved the irq-related stuff, but you might want to 
check whether you have CONFIG_NOHZ on or off and whether that makes a 
difference.

> 2. First disk access after resume takes a couple of minutes
>    (seemed instant with 2.6.20) during this time no new messages show on console

Yeah, there is some problem with SATA resume. It would be beautiful if the 
people who actually see this could narrow it down with bisection. "It 
works for me" is clearly the case for many people, but not all.

But before blaming SATA, check if you have NO_HZ enabled and whether 
disabling that makes it work ok. If timeouts don't work right (or are 
*extremely* slow) things that should be instant won't be.

> 3. When I switch to X (CTRL-ALT-F7), X hangs after drawing a couple of windows
>    after waiting for some 10 min, I rebooted.
>    no new messages showed up in /var/log/messages

I think this is likely just more of the disk being buggered, but it could 
again be related to NO_HZ (people report time not advancing, and that 
would make any X timeout taking forever, and you'd see exactly your 
behaviour).

		Linus

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Linus Torvalds <torvalds@linux-foundation.org> wrote:

> > 3. When I switch to X (CTRL-ALT-F7), X hangs after drawing a couple of windows
> >    after waiting for some 10 min, I rebooted. no new messages showed 
> >    up in /var/log/messages
> 
> I think this is likely just more of the disk being buggered, but it 
> could again be related to NO_HZ (people report time not advancing, and 
> that would make any X timeout taking forever, and you'd see exactly 
> your behaviour).

Michael - does your 'date' output advance after resume? If not then i'd 
say it's a NO_HZ related problem. If yes then i'd guess it's the SATA 
problem.

	Ingo

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Ingo Molnar <mingo@elte.hu> wrote:

> Michael - does your 'date' output advance after resume? If not then 
> i'd say it's a NO_HZ related problem. [...]

in that case please do this on such a 'frozen date' system:

	echo q > /proc/sysrq-trigger

and then send us the hw-timers info.

	Ingo

Re: [2/6] 2.6.21-rc2: known regressions

[Michael S. Tsirkin]

> Quoting Ingo Molnar <mingo@elte.hu>:
> Subject: Re: [2/6] 2.6.21-rc2: known regressions
> 
> 
> * Linus Torvalds <torvalds@linux-foundation.org> wrote:
> 
> > > 3. When I switch to X (CTRL-ALT-F7), X hangs after drawing a couple of windows
> > >    after waiting for some 10 min, I rebooted. no new messages showed 
> > >    up in /var/log/messages
> > 
> > I think this is likely just more of the disk being buggered, but it 
> > could again be related to NO_HZ (people report time not advancing, and 
> > that would make any X timeout taking forever, and you'd see exactly 
> > your behaviour).
> 
> Michael - does your 'date' output advance after resume? If not then i'd 
> say it's a NO_HZ related problem. If yes then i'd guess it's the SATA 
> problem.

I'll test, but I have NO_HZ off for now.

-- 
MST

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Michael S. Tsirkin <mst@mellanox.co.il> wrote:

> > Michael - does your 'date' output advance after resume? If not then 
> > i'd say it's a NO_HZ related problem. If yes then i'd guess it's the 
> > SATA problem.
> 
> I'll test, but I have NO_HZ off for now.

there can still be effects of it (the regression we fixed in -rc3 was 
such an effect too), so please test it.

	Ingo

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Linus Torvalds <torvalds@linux-foundation.org> wrote:

> > 2. First disk access after resume takes a couple of minutes
> >    (seemed instant with 2.6.20) during this time no new messages 
> >    show on console
> 
> Yeah, there is some problem with SATA resume. It would be beautiful if 
> the people who actually see this could narrow it down with bisection. 
> "It works for me" is clearly the case for many people, but not all.

Thomas found a new twist to this today: applying the patch below (which 
turns on ATA_DEBUG) made the SATA problem go away on his laptop. 
Michael, could you try this patch, does it change the behavior of your 
laptop in any way?

	Ingo

---
 include/linux/libata.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: linux/include/linux/libata.h
===================================================================
--- linux.orig/include/linux/libata.h
+++ linux/include/linux/libata.h
@@ -51,7 +51,7 @@
  * compile-time options: to be removed as soon as all the drivers are
  * converted to the new debugging mechanism
  */
-#undef ATA_DEBUG		/* debugging output */
+#define ATA_DEBUG 1		/* debugging output */
 #undef ATA_VERBOSE_DEBUG	/* yet more debugging output */
 #undef ATA_IRQ_TRAP		/* define to ack screaming irqs */
 #undef ATA_NDEBUG		/* define to disable quick runtime checks */

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Ingo Molnar <mingo@elte.hu> wrote:

> Thomas found a new twist to this today: applying the patch below 
> (which turns on ATA_DEBUG) made the SATA problem go away on his 
> laptop. Michael, could you try this patch, does it change the behavior 
> of your laptop in any way?

Here's another suspend/resume artifact: one of my boxes wouldnt resume, 
it hangs at:

[    1.456633] pci 0000:00:18.2: resuming
[    1.456641] pci 0000:00:18.3: resuming
[    1.456648] 8139too 0000:05:07.0: resuming
[    1.456667] radeonfb 0000:01:00.0: resuming

the box is pingable but otherwise no console and it's hung as well. [I 
got the log above via netconsole]

disabling the following radeonfb options in the .config made resume work 
again:

 CONFIG_FB_RADEON=y
 CONFIG_FB_RADEON_I2C=y
 CONFIG_FB_RADEON_BACKLIGHT=y

	Ingo

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Ingo Molnar <mingo@elte.hu> wrote:

> Here's another suspend/resume artifact: one of my boxes wouldnt 
> resume, it hangs at:
> 
> [    1.456633] pci 0000:00:18.2: resuming
> [    1.456641] pci 0000:00:18.3: resuming
> [    1.456648] 8139too 0000:05:07.0: resuming
> [    1.456667] radeonfb 0000:01:00.0: resuming
> 
> the box is pingable but otherwise no console and it's hung as well. [I 
> got the log above via netconsole]

unfortunately there's no way to get any backtrace out of the system at 
this point - the timer interrupts are not active yet here, so i cannot 
use the softlockup_tick patch to sample the lockup. Since it's pingable 
i'll try to hack a dump_stack() into icmp_reply() ;-)

	Ingo

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Ingo Molnar <mingo@elte.hu> wrote:

> > Here's another suspend/resume artifact: one of my boxes wouldnt 
> > resume, it hangs at:
> > 
> > [    1.456633] pci 0000:00:18.2: resuming
> > [    1.456641] pci 0000:00:18.3: resuming
> > [    1.456648] 8139too 0000:05:07.0: resuming
> > [    1.456667] radeonfb 0000:01:00.0: resuming
> > 
> > the box is pingable but otherwise no console and it's hung as well. [I 
> > got the log above via netconsole]
> 
> unfortunately there's no way to get any backtrace out of the system at 
> this point - the timer interrupts are not active yet here, so i cannot 
> use the softlockup_tick patch to sample the lockup. Since it's 
> pingable i'll try to hack a dump_stack() into icmp_reply() ;-)

didnt succeed at that - no netconsole output during the hang. But the 
RADEON_FB .config options definitely make the difference. The video card 
in question is:

01:00.0 VGA compatible controller: ATI Technologies Inc RV370 5B60 [Radeon X300 (PCIE)]
01:00.1 Display controller: ATI Technologies Inc RV370 [Radeon X300SE]

	Ingo

Re: [2/6] 2.6.21-rc2: known regressions

[Linus Torvalds]

On Fri, 9 Mar 2007, Ingo Molnar wrote:
> 
> disabling the following radeonfb options in the .config made resume work 
> again:

In general, don't even *try* to use radeonfb for suspend/resume.

I don't think it has ever worked, except on some very rare laptops 
(largely PPC Macs) where people had enough information to set up the 
PLL's.

I don't think the other framebuffer drivers are much better.

You're better off using the VGA console, and lettign X re-initialize the 
graphics device. That generally at least has a reasonably good chance of 
working.

Re-initializing graphics modes really is very hard. You can try with the 
BIOS video hack (I forget the kernel command line to turn it on), but we 
really do end up depending on X doing it better.

Some day we may have modesetting support in the kernel for some graphics 
hw, right now it's pretty damn spotty.

		Linus

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Linus Torvalds <torvalds@linux-foundation.org> wrote:

> > disabling the following radeonfb options in the .config made resume 
> > work again:
> 
> In general, don't even *try* to use radeonfb for suspend/resume.
> 
> I don't think it has ever worked, except on some very rare laptops 
> (largely PPC Macs) where people had enough information to set up the 
> PLL's.
> 
> I don't think the other framebuffer drivers are much better.
> 
> You're better off using the VGA console, and lettign X re-initialize 
> the graphics device. That generally at least has a reasonably good 
> chance of working.
> 
> Re-initializing graphics modes really is very hard. You can try with 
> the BIOS video hack (I forget the kernel command line to turn it on), 
> but we really do end up depending on X doing it better.

it's the s3_sleep boot option and /proc/sys/kernel/acpi_video_mode, but 
that didnt make a difference.

> Some day we may have modesetting support in the kernel for some 
> graphics hw, right now it's pretty damn spotty.

having no video is what i'd have expected - but getting a /hang/ is not 
what i'd have expected.

	Ingo

Re: [2/6] 2.6.21-rc2: known regressions

[Linus Torvalds]

On Fri, 9 Mar 2007, Ingo Molnar wrote:
> > Some day we may have modesetting support in the kernel for some 
> > graphics hw, right now it's pretty damn spotty.
> 
> having no video is what i'd have expected - but getting a /hang/ is not 
> what i'd have expected.

I debugged a case exactly like this (with the TRACE_RESUME() thing) back 
last November on my laptop.

The problem is that radeonfb actually *tries* to re-initialize the 
graphics chips, but because it doesn't know all the details, it does it 
wrong, and hangs the whole chip.

I forget exactly where it happened, and it might well be different for 
you. But this is exactly what TRACE_RESUME can be used to pinpoint if you 
want to try to debug it. 

			Linus

Re: [2/6] 2.6.21-rc2: known regressions

[Pavel Machek]

Hi!

> > disabling the following radeonfb options in the .config made resume work 
> > again:
> 
> In general, don't even *try* to use radeonfb for suspend/resume.
> 
> I don't think it has ever worked, except on some very rare laptops 
> (largely PPC Macs) where people had enough information to set up the 
> PLL's.

It worked ok on thinkpad x32. BIOS did the setup in resume case (with
acpi_sleep=..., anyway), and radeonfb could pick the card up from there.

> I don't think the other framebuffer drivers are much better.
> 
> You're better off using the VGA console, and lettign X re-initialize the 
> graphics device. That generally at least has a reasonably good chance of 
> working.

suspend.sf.net, s2ram there has a long list of tricks. If you invent
new one, please add it there.

> Re-initializing graphics modes really is very hard. You can try with the 
> BIOS video hack (I forget the kernel command line to turn it on), but we 
> really do end up depending on X doing it better.

...or you can try vbetool; it is similar hack to acpi_sleep=... , but
it works for more people.

> Some day we may have modesetting support in the kernel for some graphics 
> hw, right now it's pretty damn spotty.

Yep, that's the way to go.
									Pavel

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Pavel Machek <pavel@ucw.cz> wrote:

> > Some day we may have modesetting support in the kernel for some 
> > graphics hw, right now it's pretty damn spotty.
> 
> Yep, that's the way to go.

hey, i wildly supported this approach ever since 1996, when GGI came up
:-/

	Ingo

Re: [linux-pm] [2/6] 2.6.21-rc2: known regressions

[Jim Gettys]

On Sun, 2007-03-18 at 17:07 +0100, Ingo Molnar wrote:
> * Pavel Machek <pavel@ucw.cz> wrote:
> 
> > > Some day we may have modesetting support in the kernel for some 
> > > graphics hw, right now it's pretty damn spotty.
> > 
> > Yep, that's the way to go.
> 
> hey, i wildly supported this approach ever since 1996, when GGI came up
> :-/
> 

So wildly you wrote tons of code.... ;-).

More seriously, at the time, XFree86 would have spat in your face for
any such thing.  Thankfully, times are changing.

Also more seriously, a somewhat hybrid approach is in order for mode
setting: simple mode setting isn't much code and is required for sane
behavior on crash (it is nice to get oopses onto a screen); but the full
blown mode setting/configuration problem is so large that on some
hardware, it is likely left best left to a helper process (not the X
server).

Also key to get sane behavior out of the scheduler is to get the X
server to yield (sleep in the kernel) rather than busy waiting when the
GPU is busy; a standardized interface for this for both fbdev and dri is
in order.  Right now, X is a misbehaving compute bound process rather
than the properly interactive process it can/should/will be, releasing
the CPU whenever the hardware is busy. Needless to say, this wastes
cycles and hurts interactivity with just about any scheduler you can
devise. It isn't as if this is hard; on UNIX systems we did it in 1984
or thereabouts.

Of course, in 1996, XFree86 would have ignored any such interfaces, in
its insane quest for operating system independent user space drivers
requiring no standard kernel interfaces.... (it is the second part of
this where the true insanity lay).
                                  - Jim

-- 
Jim Gettys
One Laptop Per Child

Re: [linux-pm] [2/6] 2.6.21-rc2: known regressions

[Bill Davidsen]

Jim Gettys wrote:
> On Sun, 2007-03-18 at 17:07 +0100, Ingo Molnar wrote:
>> * Pavel Machek <pavel@ucw.cz> wrote:
>>
>>>> Some day we may have modesetting support in the kernel for some 
>>>> graphics hw, right now it's pretty damn spotty.
>>> Yep, that's the way to go.
>> hey, i wildly supported this approach ever since 1996, when GGI came up
>> :-/
>>
> 
> So wildly you wrote tons of code.... ;-).
> 
> More seriously, at the time, XFree86 would have spat in your face for
> any such thing.  Thankfully, times are changing.
> 
> Also more seriously, a somewhat hybrid approach is in order for mode
> setting: simple mode setting isn't much code and is required for sane
> behavior on crash (it is nice to get oopses onto a screen); but the full
> blown mode setting/configuration problem is so large that on some
> hardware, it is likely left best left to a helper process (not the X
> server).
> 
> Also key to get sane behavior out of the scheduler is to get the X
> server to yield (sleep in the kernel) rather than busy waiting when the
> GPU is busy; a standardized interface for this for both fbdev and dri is
> in order.  Right now, X is a misbehaving compute bound process rather
> than the properly interactive process it can/should/will be, releasing
> the CPU whenever the hardware is busy. Needless to say, this wastes
> cycles and hurts interactivity with just about any scheduler you can
> devise. It isn't as if this is hard; on UNIX systems we did it in 1984
> or thereabouts.

What you say sounds good, assuming that the cost of a sleep is less than 
the cost of the busy wait. But this may be hardware, the waits may be 
very small and frequent, and if it's hitting a small hardware window 
like retrace, delays in response will cause the time period to be missed 
completely. This probably less critical with very smart cards, many of 
us don't run them.
> 
> Of course, in 1996, XFree86 would have ignored any such interfaces, in
> its insane quest for operating system independent user space drivers
> requiring no standard kernel interfaces.... (it is the second part of
> this where the true insanity lay).
>                                   - Jim
> 


-- 
Bill Davidsen <davidsen@tmr.com>
   "We have more to fear from the bungling of the incompetent than from
the machinations of the wicked."  - from Slashdot

Re: [linux-pm] [2/6] 2.6.21-rc2: known regressions

[Jim Gettys]

On Mon, 2007-03-19 at 16:33 -0400, Bill Davidsen wrote:

> 
> What you say sounds good, assuming that the cost of a sleep is less than 
> the cost of the busy wait. But this may be hardware, the waits may be 
> very small and frequent, and if it's hitting a small hardware window 
> like retrace, delays in response will cause the time period to be missed 
> completely. This probably less critical with very smart cards, many of 
> us don't run them.
> > 

Actually, various strategies involving short busy waiting, or looking at
DMA address registers before sleeping were commonplace.  But a
syscall/sleep/wakeup is/was pretty fast.  If you have an operation
blitting the screen (e.g. scrolling), it takes a bit of time for the GPU
to execute the command.  I see this right now on OLPC, where a wonderful
music application needs to scroll (most of) the screen left),
periodically, and we're losing samples sometimes at those operation.

Remember also, that being nice to everyone else by sleeping, there are
more cycles to go around, and the scheduler can nicely boost the X
server's priority as it will for "interactive" processes that are being
cooperative.
                                - Jim

-- 
Jim Gettys
One Laptop Per Child

Re: [linux-pm] [2/6] 2.6.21-rc2: known regressions

[Bill Davidsen]

Jim Gettys wrote:
> On Mon, 2007-03-19 at 16:33 -0400, Bill Davidsen wrote:
>
>   
>> What you say sounds good, assuming that the cost of a sleep is less than 
>> the cost of the busy wait. But this may be hardware, the waits may be 
>> very small and frequent, and if it's hitting a small hardware window 
>> like retrace, delays in response will cause the time period to be missed 
>> completely. This probably less critical with very smart cards, many of 
>> us don't run them.
>>     
>
> Actually, various strategies involving short busy waiting, or looking at
> DMA address registers before sleeping were commonplace.  But a
> syscall/sleep/wakeup is/was pretty fast.  If you have an operation
> blitting the screen (e.g. scrolling), it takes a bit of time for the GPU
> to execute the command.  I see this right now on OLPC, where a wonderful
> music application needs to scroll (most of) the screen left),
> periodically, and we're losing samples sometimes at those operation.
>   
None of that conflicts with what I said, but what works on an LCD may 
not be appropriate for a CRT. With even moderate 1024x768@70 timing the 
horizontal retrace happens ~50k/sec, and that's not an appropriate 
syscall rate. I'm just pointing out that some things a video interface 
does with simple hardware involve lots of very small windows. Don't read 
that as "don't do it," just "be careful HOW you do it."
> Remember also, that being nice to everyone else by sleeping, there are
> more cycles to go around, and the scheduler can nicely boost the X
> server's priority as it will for "interactive" processes that are being
> cooperative.
I'm going to cautiously guess that the problem might be not "how much" 
but "how soon." That is, latency might be more important than giving the 
server a lot of CPU.

-- 
bill davidsen <davidsen@tmr.com>
  CTO TMR Associates, Inc
  Doing interesting things with small computers since 1979

Re: [2/6] 2.6.21-rc2: known regressions

[Johannes Stezenbach]

On Thu, Mar 08, 2007, Linus Torvalds wrote:
> On Fri, 9 Mar 2007, Ingo Molnar wrote:
> > 
> > disabling the following radeonfb options in the .config made resume work 
> > again:
> 
> In general, don't even *try* to use radeonfb for suspend/resume.
> 
> I don't think it has ever worked, except on some very rare laptops 
> (largely PPC Macs) where people had enough information to set up the 
> PLL's.
> 
> I don't think the other framebuffer drivers are much better.
> 
> You're better off using the VGA console, and lettign X re-initialize the 
> graphics device. That generally at least has a reasonably good chance of 
> working.
> 
> Re-initializing graphics modes really is very hard. You can try with the 
> BIOS video hack (I forget the kernel command line to turn it on), but we 
> really do end up depending on X doing it better.

acpi_sleep=s3_bios has always worked for me on my ThinkPad T42p.

Even if one doesn't use the fb console at all, radeonfb apparently
is still required on some ThinkPad models to work around BIOS bugs:

http://www.thinkwiki.org/wiki/Problem_with_high_power_drain_in_ACPI_sleep#Radeon_GPU_not_powered_off


Johannes

Re: [2/6] 2.6.21-rc2: known regressions

[Pavel Machek]

Hi!

> > You're better off using the VGA console, and lettign X re-initialize the 
> > graphics device. That generally at least has a reasonably good chance of 
> > working.
> > 
> > Re-initializing graphics modes really is very hard. You can try with the 
> > BIOS video hack (I forget the kernel command line to turn it on), but we 
> > really do end up depending on X doing it better.
> 
> acpi_sleep=s3_bios has always worked for me on my ThinkPad T42p.
> 
> Even if one doesn't use the fb console at all, radeonfb apparently
> is still required on some ThinkPad models to work around BIOS bugs:
> 
> http://www.thinkwiki.org/wiki/Problem_with_high_power_drain_in_ACPI_sleep#Radeon_GPU_not_powered_off


s2ram should be able to work around this, it has parts from
radeontool. (suspend.sf.net).


-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: [2/6] 2.6.21-rc2: known regressions

[Ingo Molnar]

* Pavel Machek <pavel@suse.cz> wrote:

> > Even if one doesn't use the fb console at all, radeonfb apparently 
> > is still required on some ThinkPad models to work around BIOS bugs:
> > 
> > http://www.thinkwiki.org/wiki/Problem_with_high_power_drain_in_ACPI_sleep#Radeon_GPU_not_powered_off
> 
> 
> s2ram should be able to work around this, it has parts from 
> radeontool. (suspend.sf.net).

i'm wondering, do you have any idea how Windows handles the 
suspend/resume quirks problem area? Do they "curse BIOS vendors and 
maintain a large DB of DMI-driven exceptions", or do they perhaps have 
some fundamentally better approach than us? If it's the former, then we 
might as well try to bring more automatism (and more of your database) 
into the kernel itself.

btw., the s2ram database seems quite a bit spotty:

 $ ./s2ram -n
 Machine is unknown.
 This machine can be identified by:
     sys_vendor   = "System manufacturer"
     sys_product  = "System Product Name"
     sys_version  = "System Version"
     bios_version = "ASUS A8N-E ACPI BIOS Revision 1008"

 $ ./s2ram -n
 Machine is unknown.
 This machine can be identified by:
    sys_vendor   = "Hewlett-Packard "
    sys_product  = "compaq nx9030 (PG630ET#ABD)       "
    sys_version  = "Rev 1           "
    bios_version = "F.15    "
 See http://en.opensuse.org/S2ram for details.

even at the link above i didnt find any clear algorithm about how to 
extend the quirks-list and the white-list - while i expect that most 
people experience what i did: that s2ram doesnt know their boxes. 
(otherwise they would not visit that URL at all i suspect)

	Ingo

s2ram (was Re: [2/6] 2.6.21-rc2: known regressions)

[Pavel Machek]

Hi!

> > > Even if one doesn't use the fb console at all, radeonfb apparently 
> > > is still required on some ThinkPad models to work around BIOS bugs:
> > > 
> > > http://www.thinkwiki.org/wiki/Problem_with_high_power_drain_in_ACPI_sleep#Radeon_GPU_not_powered_off
> > 
> > 
> > s2ram should be able to work around this, it has parts from 
> > radeontool. (suspend.sf.net).
> 
> i'm wondering, do you have any idea how Windows handles the 
> suspend/resume quirks problem area? Do they "curse BIOS vendors and 

Windows actually have (kernel) graphics drivers that know how to
resume the video. If you boot save mode, they go w/o graphics drivers,
and have similar problems to us.

> btw., the s2ram database seems quite a bit spotty:
> 
>  $ ./s2ram -n
>  Machine is unknown.
>  This machine can be identified by:
>      sys_vendor   = "System manufacturer"
>      sys_product  = "System Product Name"
>      sys_version  = "System Version"
>      bios_version = "ASUS A8N-E ACPI BIOS Revision 1008"
> 
>  $ ./s2ram -n
>  Machine is unknown.
>  This machine can be identified by:
>     sys_vendor   = "Hewlett-Packard "
>     sys_product  = "compaq nx9030 (PG630ET#ABD)       "
>     sys_version  = "Rev 1           "
>     bios_version = "F.15    "
>  See http://en.opensuse.org/S2ram for details.

Desktops are the problem; but that nx9030 should be reasonably easy to
add.

> even at the link above i didnt find any clear algorithm about how to 
> extend the quirks-list and the white-list - while i expect that most 
> people experience what i did: that s2ram doesnt know their boxes. 
> (otherwise they would not visit that URL at all i suspect)

Did you get options for s2ram that work on your systems? Mail them to
seife@suse.de, and he'll extend the whitelist ;-). In the meantime,
just use -f <whatever>
								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: [2/6] 2.6.21-rc2: known regressions

[Michael S. Tsirkin]

> Quoting Linus Torvalds <torvalds@linux-foundation.org>:
> > 
> > Here's the status with -rc3: better, but still does not work as well as 2.6.20.
> 
> Ok. I think we mostly solved the irq-related stuff, but you might want to 
> check whether you have CONFIG_NOHZ on or off and whether that makes a 
> difference.

This testing was done without NO_HZ.

-- 
MST

Re: [2/6] 2.6.21-rc2: known regressions

[Michael S. Tsirkin]

> > 2. First disk access after resume takes a couple of minutes
> >    (seemed instant with 2.6.20) during this time no new messages show on console
> 
> Yeah, there is some problem with SATA resume. It would be beautiful if the 
> people who actually see this could narrow it down with bisection. "It 
> works for me" is clearly the case for many people, but not all.

Problem is, there seem to be multiple problems some of which got
fixed between rc2 and rc3.
With rc2 I didn't get as far as getting to console.

-- 
MST