[Bug 623393] New: hard lock at boot with acpi on acer aspire 1310
http://bugzilla.novell.com/show_bug.cgi?id=623393 http://bugzilla.novell.com/show_bug.cgi?id=623393#c0 Summary: hard lock at boot with acpi on acer aspire 1310 Classification: openSUSE Product: openSUSE 11.3 Version: Final Platform: i686 OS/Version: openSUSE 11.3 Status: NEW Severity: Critical Priority: P5 - None Component: Kernel AssignedTo: kernel-maintainers@forge.provo.novell.com ReportedBy: jabber@gmx.org QAContact: qa@suse.de Found By: --- Blocker: --- Created an attachment (id=376715) --> (http://bugzilla.novell.com/attachment.cgi?id=376715) acpi dump User-Agent: Mozilla/5.0 (X11; U; Linux i686; de; rv:1.9.2.6) Gecko/20100626 SUSE/3.6.6-1.2 Firefox/3.6.6 I've a acer aspire 1310. After upgrade from openSUSE 11.2 to openSUSE 11.3 boot hangs unless I'm using acpi=off or acpi=ht at a very early state. The last messages I've got are: [0.535729] ACPI: Core revision 20100121 [0.532496] ACPI: setting ELCR to 0200 (from 0e20) ... [0.564360] ACPI: but type pci registered [0.716290] PCI: Using configuration type 1 for base access [0.716566] mtrr: your CPUs had inconsistent fixed MTRR settings [0.716665] mtrr: propably your BIOS does not setup all CPUs. [0.716762] mtrr: corrected configuration. [0.720655] bio: create slab <bio-0> at 0 [0.722712] ACPI: EC: Loop up EC in DSDT After that it hangs with fan at maximum speed, no keyboard reaction e.g. on Caps-Lock or CTRL-ALT-DEL or Sys-Rq. Only power off after 4s works. With acpi=ht or acpi=off, it without power management, battery control, even power off after halt, I cannot use it. I've also tried acpi.debug_level=0xf in hope to get some more information, but nothing changed. I've also tried out kernel desktop or rt instead of default but it still hangs. Currently I've reinstalled kernel 2.6.31.12-0.2-default from openSUSE 11.2, which works fine with acpi. BTW: Producing the attached acpi dump, acpidump reported: "Wrong checksum for DSDT!" The bios is the newest version from acer. Markus Reproducible: Always Steps to Reproduce: 1. upgrade from openSUSE 11.2 to openSUSE 11.3 2. remove "quiet" from kernel parameters at grub/menu.lst 3. reboot Actual Results: hang Expected Results: boot -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=623393
http://bugzilla.novell.com/show_bug.cgi?id=623393#c1
--- Comment #1 from Markus Kohm
http://bugzilla.novell.com/show_bug.cgi?id=623393
http://bugzilla.novell.com/show_bug.cgi?id=623393#c
Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c2
Tony Jones
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c3
--- Comment #3 from Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c
Jiri Slaby
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c4
Jiri Slaby
BTW: Producing the attached acpi dump, acpidump reported: "Wrong checksum for DSDT!"
This is normal.
1. upgrade from openSUSE 11.2 to openSUSE 11.3 2. remove "quiet" from kernel parameters at grub/menu.lst
Could you add initcall_debug to see where exactly it hangs? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c5
Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c6
Jiri Slaby
Yes, the problem is still the same. With kernel from openSuSE 11.3 or from Kernel:HEAD repository I must deactivate acpi with the result, that apm (e.g. simply power off on halt) cannot be activated, wlan-usbstick are not recognized etc.
Could you specify kernel versions you have been trying? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c7
Jiri Slaby
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c8
Markus Kohm
(In reply to comment #3)
Yes, the problem is still the same. With kernel from openSuSE 11.3 or from Kernel:HEAD repository I must deactivate acpi with the result, that apm (e.g. simply power off on halt) cannot be activated, wlan-usbstick are not recognized etc.
Could you specify kernel versions you have been trying?
I've tried out several kernel versions since openSuSE 11.3 release. Today I've tested again with following i586 versions: a) 2.6.34.7-0.3-debug and 2.6.34.7-0.3-default (old versions from openSuSE 11.3 Update) b) 2.6.34.7-0.4-default and 2.6.34.7-0.4-debug (current version from openSuSE 11.3 Update) c) 2.6.36-rc8.32-debug (current version from openSuSE Kernel:HEAD) With 2.6.34.7-0.3-debug, 2.6.34.7-0.3-default, 2.6.34.7-0.4-debug and 2.6.34.7-0.4-default last message is always "calling acpi_init + 0x0/0xdc @ 1". With 2.6.34.7-0.4-default I get an additional screen corruption after this message with missing pixel rows. Some times I can read the text at least, sometimes not. With 2.6.36-rc8.32-debug last message is "[0.197985] bio: create slab <bio-0> at 0". There's no message "calling acpi_init ...". I write this reply using openSuSE 11.2 kernel 2.6.31.12-0.2-default installed to openSuSE 11.3, because it's the only way to have acpi. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c
Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c9
Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c10
Markus Kohm
Can you also try the SLE11 SP1 kernel matching your system configuration from:
http://ftp.suse.com/pub/projects/kernel/kotd/SLE11-SP1/
I'd like to verify if the regression appeared in 2.6.32 already or later.
Same problem with kernel-debug-2.6.32.23-0.3.99.7.22ef0dd.i586.rpm and kernel-default-2.6.32.23-0.3.99.7.22ef0dd.i586.rpm: hard lock after "calling acpi_init" message. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c
Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c11
Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c12
--- Comment #12 from Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c13
Markus Kohm
Can you try boot an affected kernel with "acpi_osi=" (no value) in the command line and see if that helps?
Tried with debug kernel 2.6.34 and 2.6.32.23. No change, still hard lock after "calling acpi_init". After this I've also added acpi.debug_layer=0xffffffff acpi.debug_level=0xffffffff again using 2.6.32.23 and got a lot of additional messages. I made a screen photo and attached it. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c14
--- Comment #14 from Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c15
--- Comment #15 from Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c16
Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c17
Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c18
Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c19
Markus Kohm
1. Have you tried booting with acpi=noirq instead acpi=off ?
Yes. Same hard lock problem. I've tried out every setting of acpi option and only acpi=off or acpi=ht (not ith 3.6.37, see below) boot.
2. Please attach dmesg from 2.6.34.y with acpi=ht and _without_ initcall_debug or ACPI debug options.
I've done so with kernel 2.6.34. With kernel 2.6.37 I've made the dmesg output with acpi=off.
3. Can you try the latest kernel from: http://ftp.suse.com/pub/projects/kernel/kotd/master/ matching your system configuration and see if the problem is still there?
The problem is even worse. With kernel 2.6.37 from there, I cannot longer boot with acpi=ht. With this option I get a hard lock after "bio: create slab <bio-0> at 0". But acpi=off still boots. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c20
--- Comment #20 from Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c21
Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c22
--- Comment #22 from Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c23
--- Comment #23 from Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c
Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c24
Markus Kohm
Created an attachment (id=408109) --> (http://bugzilla.novell.com/attachment.cgi?id=408109) [details] EC: Skip DSDT scan
@Markus: Does this patch on top of 2.6.37 make any difference (without acpi=off)?
Sorry, needed some time to compile and install the patched new kernel (first compilation failed after about 4 hours because of missing space on partition). But it no difference. Still hard lock after "bio: create slab <bio-0> at 0" without acpi=off. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c25
--- Comment #25 from Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c26
--- Comment #26 from Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c
Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c27
--- Comment #27 from Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c28
--- Comment #28 from Markus Kohm
I wonder what happens if you comment out the acpi_bus_osc_support() entirely?
Nothing. Still hard lock. Only my crude patch to prevent 35th call of acpi_init_one_device from initializing the device (or device subtree) helps booting with acpi. So maybe finding out, what device is handled by this call, may help to fix the problem? Or what else can I do to get a valid fix? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c29
--- Comment #29 from Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c
Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c30
--- Comment #30 from Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c31
--- Comment #31 from Markus Kohm
Function acpi_init_one_device does not exist, there only is acpi_ns_init_one_device in drivers/acpi/acpica/nsinit.c
If you would have attached a real diff
Sorry. Done (see comment #30).
Anyway, if it's really this, something might go wrong when the _STA or _INI function of the affected device is called...
If it's really this function you could also place a debug print at the beginning like at the end: ACPI_EXCEPTION((AE_INFO, status, "during %s._INI" "execution",scope_name)); then we would know which device exactly is no. 35.
scope_name is not defined at the begin of acpi_ns_init_one_device. It will be defined at the function some lines after status = acpi_ns_evaluate(info); but it never returns from acpi_ns_evaluate of 35th call of acpi_ns_init_one_device (see comment #27). The calls of _STA and _INI function are before this line. Because of this, I don't know how to provide the needed info. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c32
--- Comment #32 from Thomas Renninger
From the picture you provided it's hanging when trying to write 0xE3 to APMC. This is the \SB.GSMI function which is called at various places (before entering suspend in _PTS, when exiting suspend in _WAK and in notification handlers(_Lxx).
But it's only called in one INI func: VGA. This "looks" unrelated to EC, but it seems the HW is in a state a possible SMI call (GSMI?!?) does not like. But the new kernel mode setting should be far after this happens, so this cannot be it. What you can try to verify above is to compile the modified DSDT I'll attach into your kernel. You have to place the file into your kernel sources in drivers/acpi/DSDT.hex and then point to it via compile options in .config CONFIG_ACPI_CUSTOM_DSDT_FILE="DSDT.hex" CONFIG_ACPI_CUSTOM_DSDT=y It should not hang (at least not at the same place) anymore... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c33
--- Comment #33 from Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c
Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c34
Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c35
--- Comment #35 from Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c36
--- Comment #36 from Thomas Renninger
No idea how to find out why Only idea I have is bisecting...
-- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c37
--- Comment #37 from Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c38
--- Comment #38 from Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c39
--- Comment #39 from Markus Kohm
Can you blacklist the preloadtrace driver, please.
The preloadtrace driver was never installed for kernel 2.6.37. And lsmod | grep preload showed nothing. Nevertheless I've blacklisted it now, created new initrd and nothing changed. Hard while booting without your patch, or while unplugging/replugging AC adapter with your patch. BTW: I think the hard lock while booting is before usage of initrd. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c40
Thomas Renninger
BTW: I think the hard lock while booting is before usage of initrd. Yes, sorry about that. It came to my mind because this one caused some early boot hangs as well, but yep, it's totally unrelated.
For the records, these are the last two executed instructions before the hang. The first is a PCI config space, the last an IO port access: Device (PCI0) { OperationRegion (NB00, PCI_Config, 0x00, 0x0100) Field (NB00, ByteAcc, NoLock, Preserve) { .. Offset (0xF3), APMD, 8, .. } OperationRegion (PMIO, SystemIO, 0x1000, 0x50) Field (PMIO, ByteAcc, NoLock, Preserve) { .. Offset (0x2F), APMC, 8, .. } Store(Arg0, APMD) Store (0xE3, \APMC) Interesting, there is a pci quirk trying to fix up IO access for PM registers for your HW: 00:11.0 ISA bridge [0601]: VIA Technologies, Inc. VT8235 ISA Bridge [1106:3177] from drivers/pci/quirks.c: DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8235, quirk_vt8235_acpi) Yep the quirk is active (in 2.6.31 already), compare with dmesg: [ 0.089654] pci 0000:00:11.0: quirk: region 1000-107f claimed by vt8235 PM [ 0.089660] pci 0000:00:11.0: quirk: region 1400-140f claimed by vt8235 SMB You can double check with "git annotate drivers/pci/quirks.c" whether the code would fit the time since your machine does not boot anymore or just try to remove it or read up the git changelog why it is there, etc... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c41
--- Comment #41 from Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c42
--- Comment #42 from Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c43
--- Comment #43 from Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c44
--- Comment #44 from Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c45
--- Comment #45 from Markus Kohm
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c46
Suresh Siddha
I've tried git bisect once more. Due to the early hard lock, I've decided to give
make i386_defconfig make isoimage
a chance. This does not result in a bootable system, so I cannot do something like dmsg. But it reduces the code to compile. And I've started with bad v2.6.32 and good v2.6.31 after tried both kernels. So I needed much less time to compile and reboot while reducing the probability of compiling problems.
I needed three days and about 15 build-reboot cycles but the result is:
d0af9eed5aa91b6b7b5049cae69e5ea956fd85c3 is the first bad commit
This sounded like an issue that is fixed with the upstream commit:
commit 68f202e4e87cfab4439568bf397fcc5c7cf8d729
Author: Suresh Siddha
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c47
--- Comment #47 from Suresh Siddha
If "a" and "b" are both true, then I am missing something subtle.
Ok. Does this patch (For mainline) help in anyway? diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 0cbe8c0..92f7b50 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -1156,8 +1156,8 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) if (is_uv_system()) uv_system_init(); - set_mtrr_aps_delayed_init(); out: + set_mtrr_aps_delayed_init(); preempt_enable(); } If this helps, then I have a theory (that bios is changing MTRR's at a point during OS boot, but because of above bug we might be going back to the original state seen during OS boot start. And later SMI or something is not happy of this fact). If this helps, then I can think of a cleaner fix for this issue. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c48
--- Comment #48 from Thomas Renninger
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c49
Markus Kohm
a) The above patch is in 2.6.36 onwards and the previous comments seem to say kernels above 2.6.36 also has the same problem. is that correct?
Even kernel 2.6.37 has the problem.
b) this MTRR rendezvous patch should affect only the systems which have multiple logical cpus (and the fix is addressed by 68f202e4e87cfab4439568bf397fcc5c7cf8d729) and your system seem to have only one logical cpu. Correct?
Correct, but I often see messages like "weird, boot CPU (#0) not listed by the BIOS." Don't know, if this is part of the problem. (In reply to comment #48)
I've built and copied i386 kernel flavors with the bisected patch reverted here: ftp.suse.com/pub/people/trenn/bug_623393_mtrr_rendezvous_patch_reverted (if not already it should get synced and show up there in some hours).
I've tried your desktop kernel (to avoid removing my working default kernel 2.6.31) and it was able to boot. WLAN with WPA2-PSK does not work, but I hope this is only a configuration problem. Was this the same patch Suresh proposed or should I try this one too (compiling my own kernel)? Would it be a solution to remove COMFIG_SMP? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c50
Thomas Renninger
Was this the same patch Suresh proposed No, I just reverted the patch you bisected from our latest 11.4 kernel, to be sure this really is the issue. Thanks a lot for the bisecting, we (at least I) would never have found it otherwise.
or should I try this one too (compiling my own kernel)? That would be great. Reverting the whole thing is not an option, hope Suresh can help to find a proper mainlaine solution. If it's as tiny as proposed in comment #47, the patch would probably get spread to stable@ kernels and find it's way into 11.4 kernel pretty soon.
-- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c51
Markus Kohm
and find it's way into 11.4 kernel pretty soon.
That would be great! -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c52
Thomas Renninger
I've tried Suresh's patch with kernel 2.6.36. Currently I'm working with this patched kernel. So I can say: It works! Is that enough for you Suresh? Do you plan to submit this patch upstream and eventually to stable@? I can already push it into our kernel and it should still show up in 11.4. I still like to wait for statement from Suresh, though.
-- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c53
Suresh Siddha
Hm, with Suresh patch from comment #47 you also get: set_mtrr_aps_delayed_init for UP machines.
that call is just an init code, which actually makes sure that we don't touch the boot cpu's mtrr's later down in the boot path. I will think of a clean and simple patch to fix this issue.
I've tried Suresh's patch with kernel 2.6.36. Currently I'm working with this patched kernel. So I can say: It works!
Thanks for testing it out Markus.
Is that enough for you Suresh? Do you plan to submit this patch upstream and eventually to stable@?
Yes. Will do that mostly by end of day today and will update this bug with my lkml post once I did that.
I can already push it into our kernel and it should still show up in 11.4. I still like to wait for statement from Suresh, though.
Can you wait for couple of more days till the patch gets accepted into -tip tree? At that moment, you can take the patch into your distro. thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c54
--- Comment #54 from Suresh Siddha
Will do that mostly by end of day today and will update this bug with my lkml post once I did that.
http://marc.info/?l=linux-kernel&m=129669498508824&w=2 I posted a different patch to address this. Markus can you please double check one more time to see if it fixes your issue and Ack it on lkml if it works for you? Also I tried to explain the issue in the changelog. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c55
--- Comment #55 from Markus Kohm
I posted a different patch to address this. Markus can you please double check one more time to see if it fixes your issue and Ack it on lkml if it works for you?
Yes, I'll do this today (or tonight, depending on the duration; hope it will compile fast because only few files should be recompiled). Is it OK, if I'll post the Ack without being subscribed to lkml? (BTW: My last name ends up with "m" not "n". But doesn't matter, fixing the kernel problem is much more important. You and the other guys do a real great job!) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c56
--- Comment #56 from Markus Kohm
(In reply to comment #54)
I posted a different patch to address this. Markus can you please double check one more time to see if it fixes your issue
Yes, I'll do this today
Done. The new patch seems to fix the lock, too. I can boot and unplug and replug the AC adapter. BTW:
So suspend/resume might be already broken on this platform for all linux kernel versions.
That's correct. Suspend to RAM (S3) never worked for this notebook. But, because it never worked and was a usual problem, I never used it. Wake from suspend to disc was almost as slow as cold boot, so I also never used this. Hey, I'm very happy! Hope, that the patch becomes part of the kernel soon. Thank you very much to all of you! -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c57
--- Comment #57 from Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c58
Rafael Wysocki
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c59
Marcus Meissner
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c60
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c61
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c62
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c63
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c64
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c65
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c66
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c67
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c68
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=623393
https://bugzilla.novell.com/show_bug.cgi?id=623393#c69
Swamp Workflow Management
participants (1)
-
bugzilla_noreply@novell.com