Hi,
My PC (AMD Bulldozer + Asus SABERTOOTH 990FX) booted fine from UEFI
and it broke between v3.5 and v3.6-rc1.
Other machines with old BIOSes booted fine so I looked into EFI-related
patches trying to revert them, because I didn't know what else to do.
Bingo, bacef661: x86-64/efi: Use EFI to deal with platform wall clock.
At the moment I reverted this commit after v3.6-rc1-133-g42a579a,
and it boots fine.
This really not my domain so tell me if I can help testing.
Regards,
--
cJ
Note, probably irrelevant: I use grub2 with no boot parameters
except root= and the standard .config EFI options.
On 08/05/2012 02:29 PM, Jérôme Carretero wrote:
> Hi,
>
> My PC (AMD Bulldozer + Asus SABERTOOTH 990FX) booted fine from UEFI
> and it broke between v3.5 and v3.6-rc1.
> Other machines with old BIOSes booted fine so I looked into EFI-related
> patches trying to revert them, because I didn't know what else to do.
>
> Bingo, bacef661: x86-64/efi: Use EFI to deal with platform wall clock.
>
> At the moment I reverted this commit after v3.6-rc1-133-g42a579a,
> and it boots fine.
>
> This really not my domain so tell me if I can help testing.
>
Thank you... we were aware of the problem but had not been able to
reproduce it, so we had hoped someone would bisect or otherwise identify
the faulty patch.
-hpa
--
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel. I don't speak on their behalf.
>>> On 06.08.12 at 00:28, "H. Peter Anvin" <[email protected]> wrote:
> On 08/05/2012 02:29 PM, Jérôme Carretero wrote:
>> Hi,
>>
>> My PC (AMD Bulldozer + Asus SABERTOOTH 990FX) booted fine from UEFI
>> and it broke between v3.5 and v3.6-rc1.
>> Other machines with old BIOSes booted fine so I looked into EFI-related
>> patches trying to revert them, because I didn't know what else to do.
>>
>> Bingo, bacef661: x86-64/efi: Use EFI to deal with platform wall clock.
>>
>> At the moment I reverted this commit after v3.6-rc1-133-g42a579a,
>> and it boots fine.
>>
>> This really not my domain so tell me if I can help testing.
>>
>
> Thank you... we were aware of the problem but had not been able to
> reproduce it, so we had hoped someone would bisect or otherwise identify
> the faulty patch.
Faulty? Without technical detail I'd be careful with this, as there's
too many broken EFI implementation around.
The only change that has a (very low) potential for causing
problems by itself is the earlier calling of efi_enter_virtual_mode(),
which was requested/recommended by Matthew.
I am e.g. (meanwhile) aware of (Intel) systems that use floating
point instructions in the UEFI runtime code, which is clearly a
violation of the spec; having the kernel continue to be not spec
compliant is a questionable tradeoff.
In any case, without having seen _how_ things break I don't
think a decision should be taken if/how to address this
(apparent) regression.
Jan
On Mon, Aug 06, 2012 at 07:44:34AM +0100, Jan Beulich wrote:
> In any case, without having seen _how_ things break I don't
> think a decision should be taken if/how to address this
> (apparent) regression.
Machines that previously worked no longer work. That's a pretty strong
argument in favour of reverting until we can identify a workaround.
--
Matthew Garrett | [email protected]
>>> On 06.08.12 at 14:52, Matthew Garrett <[email protected]> wrote:
> On Mon, Aug 06, 2012 at 07:44:34AM +0100, Jan Beulich wrote:
>
>> In any case, without having seen _how_ things break I don't
>> think a decision should be taken if/how to address this
>> (apparent) regression.
>
> Machines that previously worked no longer work. That's a pretty strong
> argument in favour of reverting until we can identify a workaround.
Yes, one can view it that way. But then again things should have
been the way they are now from the beginning - it appears
rather questionable how someone would have come up with the
strange idea to stick some #ifdef CONFIG_X86_32 in there, and
not even bother to comment this hackery. Hence my position is
that this really very likely should have never worked on the box
in question, it was merely one bug hiding another. But of course
all this is guesswork without knowing the technical details of the
observed crash.
Plus I'm pretty convinced that once the change gets reverted,
the code will remain that way for another couple of years (and
quite likely whatever UEFI implementation it is that we have
the problem with here will too), as then there's no incentive for
anyone to get this done properly (until the then very sad day
where some EFI system shows up that, being legacy free,
_really_ doesn't have a CMOS RTC anymore - with the change
at hand I merely tried to be proactive).
Jan
On Mon, 06 Aug 2012 14:08:03 +0100
"Jan Beulich" <[email protected]> wrote:
> with the change at hand I merely tried to be proactive).
Jan,
If it helps:
- I can bisect the patch further down (might be a bit silly because
I don't quite understand it),
- you can suggest some modifications and at least I can test them
HTH,
--
cJ
>>> On 06.08.12 at 15:16, JérômeCarretero <[email protected]> wrote:
> If it helps:
>
> - I can bisect the patch further down (might be a bit silly because
> I don't quite understand it),
> - you can suggest some modifications and at least I can test them
What would help most would be the full kernel log up to the crash,
including the register and stack dumps that are presumably there.
Without that there's nothing I can suggest.
Jan
On Mon, 06 Aug 2012 14:25:33 +0100
"Jan Beulich" <[email protected]> wrote:
> >>> On 06.08.12 at 15:16, JérômeCarretero <[email protected]> wrote:
> > If it helps:
> >
> > - I can bisect the patch further down (might be a bit silly because
> > I don't quite understand it),
> > - you can suggest some modifications and at least I can test them
>
> What would help most would be the full kernel log up to the crash,
> including the register and stack dumps that are presumably there.
> Without that there's nothing I can suggest.
There is absolutely no graphical output at this point (or I'd have provided
more info); I didn't look at the code yet, I'll see if I can do something, though.
--
cJ
On Mon, 6 Aug 2012 09:16:27 -0400
Jérôme Carretero <[email protected]> wrote:
> - I can bisect the patch further down (might be a bit silly because
> I don't quite understand it),
For troubleshooting purposes I edited over your patch.
So far:
- (in arch/x86/platform/efi/efi.c) when making efi_get_time()
return mach_get_cmos_time(), the system boots.
- then I tried to return mach_get_cmos_time() when efi.get_time()
fails, ie. if (status != EFI_SUCCESS).
The system does not boot: efi.get_time(), aka virt_efi_get_time(),
does not return.
Maybe I can get more...
--
cJ
On Mon, 6 Aug 2012 22:32:08 -0400
Jérôme Carretero <[email protected]> wrote:
> For troubleshooting purposes I edited over your patch.
> So far:
> [...]
> Maybe I can get more...
With the following:
diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c
index 2dc29f5..46729f3 100644
--- a/arch/x86/platform/efi/efi.c
+++ b/arch/x86/platform/efi/efi.c
@@ -97,8 +97,9 @@ static efi_status_t virt_efi_get_time(efi_time_t *tm, efi_time_cap_t *tc)
unsigned long flags;
efi_status_t status;
+ printk("%s: get_time=0x%p\n", __func__, efi.systab->runtime->get_time);
spin_lock_irqsave(&rtc_lock, flags);
- status = efi_call_virt2(get_time, tm, tc);
+ status = EFI_SUCCESS + 1;// efi_call_virt2(get_time, tm, tc);
spin_unlock_irqrestore(&rtc_lock, flags);
return status;
}
@@ -270,8 +271,10 @@ static unsigned long efi_get_time(void)
efi_time_cap_t cap;
status = efi.get_time(&eft, &cap);
- if (status != EFI_SUCCESS)
- pr_err("Oops: efitime: can't read time!\n");
+ if (status != EFI_SUCCESS) {
+ /* fall back to RTC time */
+ return mach_get_cmos_time();
+ }
return mktime(eft.year, eft.month, eft.day, eft.hour,
eft.minute, eft.second);
The system boots, at that point... I would say my BIOS is broken,
but it can be expected that others can have the same issue.
--
cJ
PS: Here's the log until the printk
[ 0.000000] Initializing cgroup subsys cpuset
[ 0.000000] Initializing cgroup subsys cpu
[ 0.000000] Linux version 3.6.0-rc1-Bidule-00133-g42a579a-dirty (cJ@Bidule) (gcc version 4.7.1 (Gentoo 4.7.1 p1.0, pie-0.5.3) ) #30 SMP PREEMPT Mon Aug 6 22:53:47 EDT 2012
[ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz root=/dev/sda3 video=efifb printk.time=1
[ 0.000000] KERNEL supported cpus:
[ 0.000000] Intel GenuineIntel
[ 0.000000] AMD AuthenticAMD
[ 0.000000] e820: BIOS-provided physical RAM map:
[ 0.000000] BIOS-e820: [mem 0x0000000000000000-0x000000000009ffff] usable
[ 0.000000] BIOS-e820: [mem 0x0000000000100000-0x00000000bd831fff] usable
[ 0.000000] BIOS-e820: [mem 0x00000000bd832000-0x00000000bd884fff] ACPI NVS
[ 0.000000] BIOS-e820: [mem 0x00000000bd885000-0x00000000bd88ffff] ACPI data
[ 0.000000] BIOS-e820: [mem 0x00000000bd890000-0x00000000bdadbfff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000bdadc000-0x00000000bdaecfff] ACPI NVS
[ 0.000000] BIOS-e820: [mem 0x00000000bdaed000-0x00000000bdafffff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000bdb00000-0x00000000bdb01fff] ACPI NVS
[ 0.000000] BIOS-e820: [mem 0x00000000bdb02000-0x00000000bdb0afff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000bdb0b000-0x00000000bdb10fff] ACPI NVS
[ 0.000000] BIOS-e820: [mem 0x00000000bdb11000-0x00000000bdb72fff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000bdb73000-0x00000000bdd75fff] ACPI NVS
[ 0.000000] BIOS-e820: [mem 0x00000000bdd76000-0x00000000bdefffff] usable
[ 0.000000] BIOS-e820: [mem 0x00000000fec00000-0x00000000fec00fff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000fec10000-0x00000000fec10fff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000fec20000-0x00000000fec20fff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000fed00000-0x00000000fed00fff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000fed61000-0x00000000fed70fff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000fed80000-0x00000000fed8ffff] reserved
[ 0.000000] BIOS-e820: [mem 0x00000000fef00000-0x00000000ffffffff] reserved
[ 0.000000] BIOS-e820: [mem 0x0000000100001000-0x000000043effffff] usable
[ 0.000000] NX (Execute Disable) protection: active
[ 0.000000] efi: EFI v2.00 by American Megatrends
[ 0.000000] efi: SMBIOS=0xbdb71818 ACPI=0xbd885000 ACPI 2.0=0xbd885000
[ 0.000000] efi: mem00: type=3, attr=0xf, range=[0x0000000000000000-0x0000000000008000) (0MB)
[ 0.000000] efi: mem01: type=7, attr=0xf, range=[0x0000000000008000-0x000000000004f000) (0MB)
[ 0.000000] efi: mem02: type=4, attr=0xf, range=[0x000000000004f000-0x0000000000060000) (0MB)
[ 0.000000] efi: mem03: type=3, attr=0xf, range=[0x0000000000060000-0x00000000000a0000) (0MB)
[ 0.000000] efi: mem04: type=7, attr=0xf, range=[0x0000000000100000-0x0000000001000000) (15MB)
[ 0.000000] efi: mem05: type=2, attr=0xf, range=[0x0000000001000000-0x0000000001100000) (1MB)
[ 0.000000] efi: mem06: type=4, attr=0xf, range=[0x0000000001100000-0x000000000170d000) (6MB)
[ 0.000000] efi: mem07: type=3, attr=0xf, range=[0x000000000170d000-0x0000000001713000) (0MB)
[ 0.000000] efi: mem08: type=4, attr=0xf, range=[0x0000000001713000-0x0000000001729000) (0MB)
[ 0.000000] efi: mem09: type=3, attr=0xf, range=[0x0000000001729000-0x000000000172c000) (0MB)
[ 0.000000] efi: mem10: type=4, attr=0xf, range=[0x000000000172c000-0x000000000172d000) (0MB)
[ 0.000000] efi: mem11: type=3, attr=0xf, range=[0x000000000172d000-0x000000000172e000) (0MB)
[ 0.000000] efi: mem12: type=4, attr=0xf, range=[0x000000000172e000-0x0000000001794000) (0MB)
[ 0.000000] efi: mem13: type=3, attr=0xf, range=[0x0000000001794000-0x000000000179e000) (0MB)
[ 0.000000] efi: mem14: type=4, attr=0xf, range=[0x000000000179e000-0x00000000017a1000) (0MB)
[ 0.000000] efi: mem15: type=3, attr=0xf, range=[0x00000000017a1000-0x00000000017a4000) (0MB)
[ 0.000000] efi: mem16: type=4, attr=0xf, range=[0x00000000017a4000-0x00000000017a5000) (0MB)
[ 0.000000] efi: mem17: type=3, attr=0xf, range=[0x00000000017a5000-0x00000000017b1000) (0MB)
[ 0.000000] efi: mem18: type=4, attr=0xf, range=[0x00000000017b1000-0x00000000017b3000) (0MB)
[ 0.000000] efi: mem19: type=3, attr=0xf, range=[0x00000000017b3000-0x00000000017b5000) (0MB)
[ 0.000000] efi: mem20: type=4, attr=0xf, range=[0x00000000017b5000-0x00000000017bb000) (0MB)
[ 0.000000] efi: mem21: type=3, attr=0xf, range=[0x00000000017bb000-0x00000000017be000) (0MB)
[ 0.000000] efi: mem22: type=4, attr=0xf, range=[0x00000000017be000-0x00000000017c2000) (0MB)
[ 0.000000] efi: mem23: type=3, attr=0xf, range=[0x00000000017c2000-0x00000000017c4000) (0MB)
[ 0.000000] efi: mem24: type=4, attr=0xf, range=[0x00000000017c4000-0x00000000017c9000) (0MB)
[ 0.000000] efi: mem25: type=3, attr=0xf, range=[0x00000000017c9000-0x00000000017ca000) (0MB)
[ 0.000000] efi: mem26: type=4, attr=0xf, range=[0x00000000017ca000-0x00000000017d3000) (0MB)
[ 0.000000] efi: mem27: type=3, attr=0xf, range=[0x00000000017d3000-0x00000000017d6000) (0MB)
[ 0.000000] efi: mem28: type=4, attr=0xf, range=[0x00000000017d6000-0x0000000001bec000) (4MB)
[ 0.000000] efi: mem29: type=3, attr=0xf, range=[0x0000000001bec000-0x0000000001bed000) (0MB)
[ 0.000000] efi: mem30: type=4, attr=0xf, range=[0x0000000001bed000-0x0000000001bf1000) (0MB)
[ 0.000000] efi: mem31: type=3, attr=0xf, range=[0x0000000001bf1000-0x0000000001bf3000) (0MB)
[ 0.000000] efi: mem32: type=4, attr=0xf, range=[0x0000000001bf3000-0x0000000001bf5000) (0MB)
[ 0.000000] efi: mem33: type=3, attr=0xf, range=[0x0000000001bf5000-0x0000000001bf8000) (0MB)
[ 0.000000] efi: mem34: type=4, attr=0xf, range=[0x0000000001bf8000-0x0000000001c14000) (0MB)
[ 0.000000] efi: mem35: type=3, attr=0xf, range=[0x0000000001c14000-0x0000000001c1c000) (0MB)
[ 0.000000] efi: mem36: type=4, attr=0xf, range=[0x0000000001c1c000-0x0000000001c1f000) (0MB)
[ 0.000000] efi: mem37: type=3, attr=0xf, range=[0x0000000001c1f000-0x0000000001c22000) (0MB)
[ 0.000000] efi: mem38: type=4, attr=0xf, range=[0x0000000001c22000-0x0000000001c2c000) (0MB)
[ 0.000000] efi: mem39: type=3, attr=0xf, range=[0x0000000001c2c000-0x0000000001c31000) (0MB)
[ 0.000000] efi: mem40: type=4, attr=0xf, range=[0x0000000001c31000-0x0000000001c41000) (0MB)
[ 0.000000] efi: mem41: type=3, attr=0xf, range=[0x0000000001c41000-0x0000000001c49000) (0MB)
[ 0.000000] efi: mem42: type=4, attr=0xf, range=[0x0000000001c49000-0x0000000001c51000) (0MB)
[ 0.000000] efi: mem43: type=3, attr=0xf, range=[0x0000000001c51000-0x0000000001c52000) (0MB)
[ 0.000000] efi: mem44: type=4, attr=0xf, range=[0x0000000001c52000-0x0000000001c8c000) (0MB)
[ 0.000000] efi: mem45: type=3, attr=0xf, range=[0x0000000001c8c000-0x0000000001c90000) (0MB)
[ 0.000000] efi: mem46: type=4, attr=0xf, range=[0x0000000001c90000-0x0000000001c93000) (0MB)
[ 0.000000] efi: mem47: type=3, attr=0xf, range=[0x0000000001c93000-0x0000000001cb2000) (0MB)
[ 0.000000] efi: mem48: type=4, attr=0xf, range=[0x0000000001cb2000-0x0000000001cc2000) (0MB)
[ 0.000000] efi: mem49: type=3, attr=0xf, range=[0x0000000001cc2000-0x0000000001cda000) (0MB)
[ 0.000000] efi: mem50: type=4, attr=0xf, range=[0x0000000001cda000-0x0000000001ce1000) (0MB)
[ 0.000000] efi: mem51: type=3, attr=0xf, range=[0x0000000001ce1000-0x0000000001cea000) (0MB)
[ 0.000000] efi: mem52: type=4, attr=0xf, range=[0x0000000001cea000-0x0000000001cef000) (0MB)
[ 0.000000] efi: mem53: type=3, attr=0xf, range=[0x0000000001cef000-0x0000000001cf7000) (0MB)
[ 0.000000] efi: mem54: type=4, attr=0xf, range=[0x0000000001cf7000-0x0000000001cf8000) (0MB)
[ 0.000000] efi: mem55: type=3, attr=0xf, range=[0x0000000001cf8000-0x0000000001cfa000) (0MB)
[ 0.000000] efi: mem56: type=4, attr=0xf, range=[0x0000000001cfa000-0x0000000001cfe000) (0MB)
[ 0.000000] efi: mem57: type=3, attr=0xf, range=[0x0000000001cfe000-0x0000000001d00000) (0MB)
[ 0.000000] efi: mem58: type=4, attr=0xf, range=[0x0000000001d00000-0x0000000001d02000) (0MB)
[ 0.000000] efi: mem59: type=3, attr=0xf, range=[0x0000000001d02000-0x0000000001d04000) (0MB)
[ 0.000000] efi: mem60: type=4, attr=0xf, range=[0x0000000001d04000-0x0000000001d0f000) (0MB)
[ 0.000000] efi: mem61: type=3, attr=0xf, range=[0x0000000001d0f000-0x0000000001d19000) (0MB)
[ 0.000000] efi: mem62: type=4, attr=0xf, range=[0x0000000001d19000-0x0000000001d1e000) (0MB)
[ 0.000000] efi: mem63: type=3, attr=0xf, range=[0x0000000001d1e000-0x0000000001d26000) (0MB)
[ 0.000000] efi: mem64: type=4, attr=0xf, range=[0x0000000001d26000-0x0000000001d32000) (0MB)
[ 0.000000] efi: mem65: type=3, attr=0xf, range=[0x0000000001d32000-0x0000000001d45000) (0MB)
[ 0.000000] efi: mem66: type=4, attr=0xf, range=[0x0000000001d45000-0x0000000001d7c000) (0MB)
[ 0.000000] efi: mem67: type=3, attr=0xf, range=[0x0000000001d7c000-0x0000000001d86000) (0MB)
[ 0.000000] efi: mem68: type=4, attr=0xf, range=[0x0000000001d86000-0x0000000001d87000) (0MB)
[ 0.000000] efi: mem69: type=3, attr=0xf, range=[0x0000000001d87000-0x0000000001d8b000) (0MB)
[ 0.000000] efi: mem70: type=4, attr=0xf, range=[0x0000000001d8b000-0x0000000001d8c000) (0MB)
[ 0.000000] efi: mem71: type=3, attr=0xf, range=[0x0000000001d8c000-0x0000000001d8f000) (0MB)
[ 0.000000] efi: mem72: type=4, attr=0xf, range=[0x0000000001d8f000-0x0000000001d90000) (0MB)
[ 0.000000] efi: mem73: type=3, attr=0xf, range=[0x0000000001d90000-0x0000000001d93000) (0MB)
[ 0.000000] efi: mem74: type=4, attr=0xf, range=[0x0000000001d93000-0x0000000001d94000) (0MB)
[ 0.000000] efi: mem75: type=3, attr=0xf, range=[0x0000000001d94000-0x0000000001dc4000) (0MB)
[ 0.000000] efi: mem76: type=4, attr=0xf, range=[0x0000000001dc4000-0x0000000001ddc000) (0MB)
[ 0.000000] efi: mem77: type=3, attr=0xf, range=[0x0000000001ddc000-0x0000000001de2000) (0MB)
[ 0.000000] efi: mem78: type=4, attr=0xf, range=[0x0000000001de2000-0x0000000001dee000) (0MB)
[ 0.000000] efi: mem79: type=3, attr=0xf, range=[0x0000000001dee000-0x0000000001df1000) (0MB)
[ 0.000000] efi: mem80: type=4, attr=0xf, range=[0x0000000001df1000-0x0000000001df3000) (0MB)
[ 0.000000] efi: mem81: type=3, attr=0xf, range=[0x0000000001df3000-0x0000000001e06000) (0MB)
[ 0.000000] efi: mem82: type=4, attr=0xf, range=[0x0000000001e06000-0x0000000001eb2000) (0MB)
[ 0.000000] efi: mem83: type=3, attr=0xf, range=[0x0000000001eb2000-0x0000000001eb8000) (0MB)
[ 0.000000] efi: mem84: type=4, attr=0xf, range=[0x0000000001eb8000-0x0000000001ebe000) (0MB)
[ 0.000000] efi: mem85: type=3, attr=0xf, range=[0x0000000001ebe000-0x0000000001ec1000) (0MB)
[ 0.000000] efi: mem86: type=4, attr=0xf, range=[0x0000000001ec1000-0x0000000001ecb000) (0MB)
[ 0.000000] efi: mem87: type=3, attr=0xf, range=[0x0000000001ecb000-0x0000000001ed0000) (0MB)
[ 0.000000] efi: mem88: type=4, attr=0xf, range=[0x0000000001ed0000-0x0000000001ed4000) (0MB)
[ 0.000000] efi: mem89: type=3, attr=0xf, range=[0x0000000001ed4000-0x0000000001eda000) (0MB)
[ 0.000000] efi: mem90: type=4, attr=0xf, range=[0x0000000001eda000-0x0000000001edc000) (0MB)
[ 0.000000] efi: mem91: type=3, attr=0xf, range=[0x0000000001edc000-0x0000000001ede000) (0MB)
[ 0.000000] efi: mem92: type=4, attr=0xf, range=[0x0000000001ede000-0x0000000001ee1000) (0MB)
[ 0.000000] efi: mem93: type=3, attr=0xf, range=[0x0000000001ee1000-0x0000000001ee3000) (0MB)
[ 0.000000] efi: mem94: type=4, attr=0xf, range=[0x0000000001ee3000-0x0000000001f07000) (0MB)
[ 0.000000] efi: mem95: type=3, attr=0xf, range=[0x0000000001f07000-0x0000000001f1b000) (0MB)
[ 0.000000] efi: mem96: type=4, attr=0xf, range=[0x0000000001f1b000-0x0000000001f24000) (0MB)
[ 0.000000] efi: mem97: type=3, attr=0xf, range=[0x0000000001f24000-0x0000000001f33000) (0MB)
[ 0.000000] efi: mem98: type=4, attr=0xf, range=[0x0000000001f33000-0x0000000001f38000) (0MB)
[ 0.000000] efi: mem99: type=3, attr=0xf, range=[0x0000000001f38000-0x0000000001f3f000) (0MB)
[ 0.000000] efi: mem100: type=4, attr=0xf, range=[0x0000000001f3f000-0x0000000001f42000) (0MB)
[ 0.000000] efi: mem101: type=3, attr=0xf, range=[0x0000000001f42000-0x0000000001f44000) (0MB)
[ 0.000000] efi: mem102: type=4, attr=0xf, range=[0x0000000001f44000-0x0000000001f60000) (0MB)
[ 0.000000] efi: mem103: type=3, attr=0xf, range=[0x0000000001f60000-0x0000000001f64000) (0MB)
[ 0.000000] efi: mem104: type=4, attr=0xf, range=[0x0000000001f64000-0x0000000001f69000) (0MB)
[ 0.000000] efi: mem105: type=3, attr=0xf, range=[0x0000000001f69000-0x0000000001f7a000) (0MB)
[ 0.000000] efi: mem106: type=4, attr=0xf, range=[0x0000000001f7a000-0x0000000001fdb000) (0MB)
[ 0.000000] efi: mem107: type=7, attr=0xf, range=[0x0000000001fdb000-0x0000000001fe0000) (0MB)
[ 0.000000] efi: mem108: type=4, attr=0xf, range=[0x0000000001fe0000-0x0000000001ffe000) (0MB)
[ 0.000000] efi: mem109: type=3, attr=0xf, range=[0x0000000001ffe000-0x0000000002412000) (4MB)
[ 0.000000] efi: mem110: type=4, attr=0xf, range=[0x0000000002412000-0x0000000002911000) (4MB)
[ 0.000000] efi: mem111: type=3, attr=0xf, range=[0x0000000002911000-0x0000000002912000) (0MB)
[ 0.000000] efi: mem112: type=4, attr=0xf, range=[0x0000000002912000-0x000000000293d000) (0MB)
[ 0.000000] efi: mem113: type=3, attr=0xf, range=[0x000000000293d000-0x000000000293f000) (0MB)
[ 0.000000] efi: mem114: type=4, attr=0xf, range=[0x000000000293f000-0x0000000002941000) (0MB)
[ 0.000000] efi: mem115: type=7, attr=0xf, range=[0x0000000002941000-0x0000000002942000) (0MB)
[ 0.000000] efi: mem116: type=3, attr=0xf, range=[0x0000000002942000-0x0000000002960000) (0MB)
[ 0.000000] efi: mem117: type=4, attr=0xf, range=[0x0000000002960000-0x0000000002965000) (0MB)
[ 0.000000] efi: mem118: type=7, attr=0xf, range=[0x0000000002965000-0x0000000002982000) (0MB)
[ 0.000000] efi: mem119: type=1, attr=0xf, range=[0x0000000002982000-0x000000000299f000) (0MB)
[ 0.000000] efi: mem120: type=7, attr=0xf, range=[0x000000000299f000-0x00000000029e8000) (0MB)
[ 0.000000] efi: mem121: type=4, attr=0xf, range=[0x00000000029e8000-0x0000000002d62000) (3MB)
[ 0.000000] efi: mem122: type=2, attr=0xf, range=[0x0000000002d62000-0x0000000003ce9000) (15MB)
[ 0.000000] efi: mem123: type=7, attr=0xf, range=[0x0000000003ce9000-0x0000000004000000) (3MB)
[ 0.000000] efi: mem124: type=2, attr=0xf, range=[0x0000000004000000-0x0000000004f87000) (15MB)
[ 0.000000] efi: mem125: type=7, attr=0xf, range=[0x0000000004f87000-0x000000008e9a4000) (2202MB)
[ 0.000000] efi: mem126: type=2, attr=0xf, range=[0x000000008e9a4000-0x00000000bd832000) (750MB)
[ 0.000000] efi: mem127: type=10, attr=0xf, range=[0x00000000bd832000-0x00000000bd885000) (0MB)
[ 0.000000] efi: mem128: type=9, attr=0xf, range=[0x00000000bd885000-0x00000000bd890000) (0MB)
[ 0.000000] efi: mem129: type=0, attr=0xf, range=[0x00000000bd890000-0x00000000bdab6000) (2MB)
[ 0.000000] efi: mem130: type=6, attr=0x800000000000000f, range=[0x00000000bdab6000-0x00000000bdaba000) (0MB)
[ 0.000000] efi: mem131: type=0, attr=0xf, range=[0x00000000bdaba000-0x00000000bdadc000) (0MB)
[ 0.000000] efi: mem132: type=10, attr=0xf, range=[0x00000000bdadc000-0x00000000bdaed000) (0MB)
[ 0.000000] efi: mem133: type=6, attr=0x800000000000000f, range=[0x00000000bdaed000-0x00000000bdafe000) (0MB)
[ 0.000000] efi: mem134: type=5, attr=0x800000000000000f, range=[0x00000000bdafe000-0x00000000bdaff000) (0MB)
[ 0.000000] efi: mem135: type=6, attr=0x800000000000000f, range=[0x00000000bdaff000-0x00000000bdb00000) (0MB)
[ 0.000000] efi: mem136: type=10, attr=0xf, range=[0x00000000bdb00000-0x00000000bdb02000) (0MB)
[ 0.000000] efi: mem137: type=5, attr=0x800000000000000f, range=[0x00000000bdb02000-0x00000000bdb03000) (0MB)
[ 0.000000] efi: mem138: type=6, attr=0x800000000000000f, range=[0x00000000bdb03000-0x00000000bdb04000) (0MB)
[ 0.000000] efi: mem139: type=5, attr=0x800000000000000f, range=[0x00000000bdb04000-0x00000000bdb0b000) (0MB)
[ 0.000000] efi: mem140: type=10, attr=0xf, range=[0x00000000bdb0b000-0x00000000bdb11000) (0MB)
[ 0.000000] efi: mem141: type=5, attr=0x800000000000000f, range=[0x00000000bdb11000-0x00000000bdb1e000) (0MB)
[ 0.000000] efi: mem142: type=6, attr=0x800000000000000f, range=[0x00000000bdb1e000-0x00000000bdb60000) (0MB)
[ 0.000000] efi: mem143: type=5, attr=0x800000000000000f, range=[0x00000000bdb60000-0x00000000bdb70000) (0MB)
[ 0.000000] efi: mem144: type=6, attr=0x800000000000000f, range=[0x00000000bdb70000-0x00000000bdb73000) (0MB)
[ 0.000000] efi: mem145: type=10, attr=0xf, range=[0x00000000bdb73000-0x00000000bdd76000) (2MB)
[ 0.000000] efi: mem146: type=3, attr=0xf, range=[0x00000000bdd76000-0x00000000bdee4000) (1MB)
[ 0.000000] efi: mem147: type=4, attr=0xf, range=[0x00000000bdee4000-0x00000000bdeeb000) (0MB)
[ 0.000000] efi: mem148: type=3, attr=0xf, range=[0x00000000bdeeb000-0x00000000bdf00000) (0MB)
[ 0.000000] efi: mem149: type=7, attr=0xf, range=[0x0000000100001000-0x000000043f000000) (13295MB)
[ 0.000000] efi: mem150: type=11, attr=0x8000000000000001, range=[0x00000000fec00000-0x00000000fec01000) (0MB)
[ 0.000000] efi: mem151: type=11, attr=0x8000000000000001, range=[0x00000000fec10000-0x00000000fec11000) (0MB)
[ 0.000000] efi: mem152: type=11, attr=0x8000000000000001, range=[0x00000000fec20000-0x00000000fec21000) (0MB)
[ 0.000000] efi: mem153: type=11, attr=0x8000000000000001, range=[0x00000000fed00000-0x00000000fed01000) (0MB)
[ 0.000000] efi: mem154: type=11, attr=0x8000000000000001, range=[0x00000000fed61000-0x00000000fed71000) (0MB)
[ 0.000000] efi: mem155: type=11, attr=0x8000000000000001, range=[0x00000000fed80000-0x00000000fed90000) (0MB)
[ 0.000000] efi: mem156: type=11, attr=0x8000000000000001, range=[0x00000000fef00000-0x0000000100000000) (17MB)
[ 0.000000] DMI 2.7 present.
[ 0.000000] DMI: To be filled by O.E.M. To be filled by O.E.M./SABERTOOTH 990FX, BIOS 0813 10/24/2011
[ 0.000000] e820: update [mem 0x00000000-0x0000ffff] usable ==> reserved
[ 0.000000] e820: remove [mem 0x000a0000-0x000fffff] usable
[ 0.000000] No AGP bridge found
[ 0.000000] e820: last_pfn = 0x43f000 max_arch_pfn = 0x400000000
[ 0.000000] MTRR default type: uncachable
[ 0.000000] MTRR fixed ranges enabled:
[ 0.000000] 00000-9FFFF write-back
[ 0.000000] A0000-BFFFF write-through
[ 0.000000] C0000-DAFFF write-protect
[ 0.000000] DB000-EBFFF uncachable
[ 0.000000] EC000-FFFFF write-protect
[ 0.000000] MTRR variable ranges enabled:
[ 0.000000] 0 base 000000000000 mask FFFF80000000 write-back
[ 0.000000] 1 base 000080000000 mask FFFFC0000000 write-back
[ 0.000000] 2 base 0000BDF00000 mask FFFFFFF00000 uncachable
[ 0.000000] 3 base 0000BE000000 mask FFFFFE000000 uncachable
[ 0.000000] 4 disabled
[ 0.000000] 5 disabled
[ 0.000000] 6 disabled
[ 0.000000] 7 disabled
[ 0.000000] TOM2: 000000043f000000 aka 17392M
[ 0.000000] x86 PAT enabled: cpu 0, old 0x7040600070406, new 0x7010600070106
[ 0.000000] e820: update [mem 0xbdf00000-0xffffffff] usable ==> reserved
[ 0.000000] e820: last_pfn = 0xbdf00 max_arch_pfn = 0x400000000
[ 0.000000] initial memory mapped: [mem 0x00000000-0x1fffffff]
[ 0.000000] Base memory trampoline at [ffff880000097000] 97000 size 24576
[ 0.000000] Using GB pages for direct mapping
[ 0.000000] init_memory_mapping: [mem 0x00000000-0xbdefffff]
[ 0.000000] [mem 0x00000000-0x7fffffff] page 1G
[ 0.000000] [mem 0x80000000-0xbddfffff] page 2M
[ 0.000000] [mem 0xbde00000-0xbdefffff] page 4k
[ 0.000000] kernel direct mapping tables up to 0xbdefffff @ [mem 0x1fbfd000-0x1fffffff]
[ 0.000000] early_memtest: # of tests: 1
[ 0.000000] 0000010000 - 000004a000 pattern 0000000000000000
[ 0.000000] 000004bd70 - 000004f000 pattern 0000000000000000
[ 0.000000] 0000060000 - 0000097000 pattern 0000000000000000
[ 0.000000] 000009d000 - 000009d800 pattern 0000000000000000
[ 0.000000] 0000100000 - 0001100000 pattern 0000000000000000
[ 0.000000] 0001fdb000 - 0001fe0000 pattern 0000000000000000
[ 0.000000] 0002941000 - 0002942000 pattern 0000000000000000
[ 0.000000] 0002965000 - 00029e8000 pattern 0000000000000000
[ 0.000000] 0002d62000 - 0004000000 pattern 0000000000000000
[ 0.000000] 0004c4c43e - 001fbfd000 pattern 0000000000000000
[ 0.000000] 001fbff000 - 00bd832000 pattern 0000000000000000
[ 0.000000] init_memory_mapping: [mem 0x100000000-0x43effffff]
[ 0.000000] [mem 0x100000000-0x3ffffffff] page 1G
[ 0.000000] [mem 0x400000000-0x43effffff] page 2M
[ 0.000000] kernel direct mapping tables up to 0x43effffff @ [mem 0xbd830000-0xbd831fff]
[ 0.000000] early_memtest: # of tests: 1
[ 0.000000] 0100001000 - 043f000000 pattern 0000000000000000
[ 0.000000] ACPI: RSDP 00000000bd885000 00024 (v02 ALASKA)
[ 0.000000] ACPI: XSDT 00000000bd885068 0004C (v01 ALASKA A M I 01072009 AMI 00010013)
[ 0.000000] ACPI: FACP 00000000bd88dc30 000F4 (v04 ALASKA A M I 01072009 AMI 00010013)
[ 0.000000] ACPI BIOS Bug: Warning: Optional FADT field Pm2ControlBlock has zero address or length: 0x0000000000000000/0x1 (20120711/tbfadt-598)
[ 0.000000] ACPI: DSDT 00000000bd885140 08AEC (v02 ALASKA A M I 00000000 INTL 20051117)
[ 0.000000] ACPI: FACS 00000000bdb0bf80 00040
[ 0.000000] ACPI: APIC 00000000bd88dd28 0009E (v03 ALASKA A M I 01072009 AMI 00010013)
[ 0.000000] ACPI: MCFG 00000000bd88ddc8 0003C (v01 ALASKA A M I 01072009 MSFT 00010013)
[ 0.000000] ACPI: HPET 00000000bd88de08 00038 (v01 ALASKA A M I 01072009 AMI 00000004)
[ 0.000000] ACPI: SSDT 00000000bd88de40 01714 (v01 AMD POWERNOW 00000001 AMD 00000001)
[ 0.000000] ACPI: Local APIC address 0xfee00000
[ 0.000000] No NUMA configuration found
[ 0.000000] Faking a node at [mem 0x0000000000000000-0x000000043effffff]
[ 0.000000] Initmem setup node 0 [mem 0x00000000-0x43effffff]
[ 0.000000] NODE_DATA [mem 0x43effe000-0x43effffff]
[ 0.000000] [ffffea0000000000-ffffea0010ffffff] PMD -> [ffff88042e600000-ffff88043e5fffff] on node 0
[ 0.000000] Zone ranges:
[ 0.000000] DMA [mem 0x00010000-0x00ffffff]
[ 0.000000] DMA32 [mem 0x01000000-0xffffffff]
[ 0.000000] Normal [mem 0x100000000-0x43effffff]
[ 0.000000] Movable zone start for each node
[ 0.000000] Early memory node ranges
[ 0.000000] node 0: [mem 0x00010000-0x0009ffff]
[ 0.000000] node 0: [mem 0x00100000-0xbd831fff]
[ 0.000000] node 0: [mem 0xbdd76000-0xbdefffff]
[ 0.000000] node 0: [mem 0x100001000-0x43effffff]
[ 0.000000] On node 0 totalpages: 4180299
[ 0.000000] DMA zone: 64 pages used for memmap
[ 0.000000] DMA zone: 28 pages reserved
[ 0.000000] DMA zone: 3892 pages, LIFO batch:0
[ 0.000000] DMA32 zone: 16320 pages used for memmap
[ 0.000000] DMA32 zone: 756220 pages, LIFO batch:31
[ 0.000000] Normal zone: 53184 pages used for memmap
[ 0.000000] Normal zone: 3350591 pages, LIFO batch:31
[ 0.000000] ACPI: PM-Timer IO Port: 0x808
[ 0.000000] ACPI: Local APIC address 0xfee00000
[ 0.000000] ACPI: LAPIC (acpi_id[0x01] lapic_id[0x10] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x02] lapic_id[0x11] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x03] lapic_id[0x12] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x04] lapic_id[0x13] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x05] lapic_id[0x14] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x06] lapic_id[0x15] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x07] lapic_id[0x16] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x08] lapic_id[0x17] enabled)
[ 0.000000] ACPI: LAPIC_NMI (acpi_id[0xff] high edge lint[0x1])
[ 0.000000] ACPI: IOAPIC (id[0x09] address[0xfec00000] gsi_base[0])
[ 0.000000] IOAPIC[0]: apic_id 9, version 33, address 0xfec00000, GSI 0-23
[ 0.000000] ACPI: IOAPIC (id[0x0a] address[0xfec20000] gsi_base[24])
[ 0.000000] IOAPIC[1]: apic_id 10, version 33, address 0xfec20000, GSI 24-55
[ 0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 0 global_irq 2 dfl dfl)
[ 0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 low level)
[ 0.000000] ACPI: IRQ0 used by override.
[ 0.000000] ACPI: IRQ2 used by override.
[ 0.000000] ACPI: IRQ9 used by override.
[ 0.000000] Using ACPI (MADT) for SMP configuration information
[ 0.000000] ACPI: HPET id: 0xffffffff base: 0xfed00000
[ 0.000000] smpboot: Allowing 8 CPUs, 0 hotplug CPUs
[ 0.000000] nr_irqs_gsi: 72
[ 0.000000] PM: Registered nosave memory: 00000000000a0000 - 0000000000100000
[ 0.000000] PM: Registered nosave memory: 00000000bd832000 - 00000000bd885000
[ 0.000000] PM: Registered nosave memory: 00000000bd885000 - 00000000bd890000
[ 0.000000] PM: Registered nosave memory: 00000000bd890000 - 00000000bdadc000
[ 0.000000] PM: Registered nosave memory: 00000000bdadc000 - 00000000bdaed000
[ 0.000000] PM: Registered nosave memory: 00000000bdaed000 - 00000000bdb00000
[ 0.000000] PM: Registered nosave memory: 00000000bdb00000 - 00000000bdb02000
[ 0.000000] PM: Registered nosave memory: 00000000bdb02000 - 00000000bdb0b000
[ 0.000000] PM: Registered nosave memory: 00000000bdb0b000 - 00000000bdb11000
[ 0.000000] PM: Registered nosave memory: 00000000bdb11000 - 00000000bdb73000
[ 0.000000] PM: Registered nosave memory: 00000000bdb73000 - 00000000bdd76000
[ 0.000000] PM: Registered nosave memory: 00000000bdf00000 - 00000000fec00000
[ 0.000000] PM: Registered nosave memory: 00000000fec00000 - 00000000fec01000
[ 0.000000] PM: Registered nosave memory: 00000000fec01000 - 00000000fec10000
[ 0.000000] PM: Registered nosave memory: 00000000fec10000 - 00000000fec11000
[ 0.000000] PM: Registered nosave memory: 00000000fec11000 - 00000000fec20000
[ 0.000000] PM: Registered nosave memory: 00000000fec20000 - 00000000fec21000
[ 0.000000] PM: Registered nosave memory: 00000000fec21000 - 00000000fed00000
[ 0.000000] PM: Registered nosave memory: 00000000fed00000 - 00000000fed01000
[ 0.000000] PM: Registered nosave memory: 00000000fed01000 - 00000000fed61000
[ 0.000000] PM: Registered nosave memory: 00000000fed61000 - 00000000fed71000
[ 0.000000] PM: Registered nosave memory: 00000000fed71000 - 00000000fed80000
[ 0.000000] PM: Registered nosave memory: 00000000fed80000 - 00000000fed90000
[ 0.000000] PM: Registered nosave memory: 00000000fed90000 - 00000000fef00000
[ 0.000000] PM: Registered nosave memory: 00000000fef00000 - 0000000100000000
[ 0.000000] PM: Registered nosave memory: 0000000100000000 - 0000000100001000
[ 0.000000] e820: [mem 0xbdf00000-0xfebfffff] available for PCI devices
[ 0.000000] setup_percpu: NR_CPUS:8 nr_cpumask_bits:8 nr_cpu_ids:8 nr_node_ids:1
[ 0.000000] PERCPU: Embedded 27 pages/cpu @ffff88043ec00000 s78976 r8192 d23424 u262144
[ 0.000000] pcpu-alloc: s78976 r8192 d23424 u262144 alloc=1*2097152
[ 0.000000] pcpu-alloc: [0] 0 1 2 3 4 5 6 7
[ 0.000000] Built 1 zonelists in Zone order, mobility grouping on. Total pages: 4110703
[ 0.000000] Policy zone: Normal
[ 0.000000] Kernel command line: BOOT_IMAGE=/boot/vmlinuz root=/dev/sda3 video=efifb memtest=1 rootflags=ssd,discard,compress printk.time=1 init=/usr/bin/systemd rw
[ 0.000000] PID hash table entries: 4096 (order: 3, 32768 bytes)
[ 0.000000] __ex_table already sorted, skipping sort
[ 0.000000] xsave: enabled xstate_bv 0x7, cntxt size 0x340
[ 0.000000] Checking aperture...
[ 0.000000] No AGP bridge found
[ 0.000000] Node 0: aperture @ b4000000 size 32 MB
[ 0.000000] Aperture pointing to e820 RAM. Ignoring.
[ 0.000000] Your BIOS doesn't leave a aperture memory hole
[ 0.000000] Please enable the IOMMU option in the BIOS setup
[ 0.000000] This costs you 64 MB of RAM
[ 0.000000] Mapping aperture over 65536 KB of RAM @ b4000000
[ 0.000000] PM: Registered nosave memory: 00000000b4000000 - 00000000b8000000
[ 0.000000] Memory: 16283624k/17809408k available (6054k kernel code, 1088212k absent, 437572k reserved, 4805k data, 744k init)
[ 0.000000] SLUB: Genslabs=15, HWalign=64, Order=0-3, MinObjects=0, CPUs=8, Nodes=1
[ 0.000000] Preemptible hierarchical RCU implementation.
[ 0.000000] Dump stacks of tasks blocking RCU-preempt GP.
[ 0.000000] NR_IRQS:4352 nr_irqs:1288 16
[ 0.000000] virt_efi_get_time: get_time=0xffff8800bdb1a594
>>> On 07.08.12 at 01:29, JérômeCarretero <[email protected]> wrote:
> On Mon, 06 Aug 2012 14:25:33 +0100
> "Jan Beulich" <[email protected]> wrote:
>
>> >>> On 06.08.12 at 15:16, JérômeCarretero <[email protected]> wrote:
>> > If it helps:
>> >
>> > - I can bisect the patch further down (might be a bit silly because
>> > I don't quite understand it),
>> > - you can suggest some modifications and at least I can test them
>>
>> What would help most would be the full kernel log up to the crash,
>> including the register and stack dumps that are presumably there.
>> Without that there's nothing I can suggest.
>
> There is absolutely no graphical output at this point (or I'd have provided
> more info); I didn't look at the code yet, I'll see if I can do something,
> though.
Yes, the system isn't far enough for that. You'd need a serial
(and presumably early) console set up to collect the full set of
messages.
Jan
>>> On 07.08.12 at 05:06, JérômeCarretero <[email protected]> wrote:
> On Mon, 6 Aug 2012 22:32:08 -0400
> Jérôme Carretero <[email protected]> wrote:
>
>> For troubleshooting purposes I edited over your patch.
>> So far:
>> [...]
>> Maybe I can get more...
>
> With the following:
>
> diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c
> index 2dc29f5..46729f3 100644
> --- a/arch/x86/platform/efi/efi.c
> +++ b/arch/x86/platform/efi/efi.c
> @@ -97,8 +97,9 @@ static efi_status_t virt_efi_get_time(efi_time_t *tm,
> efi_time_cap_t *tc)
> unsigned long flags;
> efi_status_t status;
>
> + printk("%s: get_time=0x%p\n", __func__,
> efi.systab->runtime->get_time);
> spin_lock_irqsave(&rtc_lock, flags);
> - status = efi_call_virt2(get_time, tm, tc);
> + status = EFI_SUCCESS + 1;// efi_call_virt2(get_time, tm, tc);
> spin_unlock_irqrestore(&rtc_lock, flags);
> return status;
> }
> @@ -270,8 +271,10 @@ static unsigned long efi_get_time(void)
> efi_time_cap_t cap;
>
> status = efi.get_time(&eft, &cap);
> - if (status != EFI_SUCCESS)
> - pr_err("Oops: efitime: can't read time!\n");
> + if (status != EFI_SUCCESS) {
> + /* fall back to RTC time */
> + return mach_get_cmos_time();
> + }
>
> return mktime(eft.year, eft.month, eft.day, eft.hour,
> eft.minute, eft.second);
>
> The system boots, at that point...
That's not surprising. The question really is what goes wrong
when the call is being made - page fault, some other fault, or
silent hang. A page fault would point to an incorrect memory
map as the prime candidate for causing the problem. My
primary suspect would be #NM, i.e. the implementation using
floating point (SSE to be precise) instructions when they're
unavailable.
> I would say my BIOS is broken,
> but it can be expected that others can have the same issue.
Likely. The question is whether we could make Linux be spec
compliant on sane systems _and_ tolerate broken ones like
this. But whether e.g. adding a command line option (or DMI-
based quirk) is appropriate depends on whether this really is
a firmware issue or a flaw in the patch.
Jan
On Tue, 2012-08-07 at 08:14 +0100, Jan Beulich wrote:
> That's not surprising. The question really is what goes wrong
> when the call is being made - page fault, some other fault, or
> silent hang. A page fault would point to an incorrect memory
> map as the prime candidate for causing the problem. My
> primary suspect would be #NM, i.e. the implementation using
> floating point (SSE to be precise) instructions when they're
> unavailable.
I managed to find a machine to reproduce this on and it looks like the
ASUS firmware engineers are upto their old tricks of referencing
physical addresses after we've taken control of the memory map,
[ 20.740414] BUG: unable to handle kernel paging request at 00000000fed1f410
[ 20.740426] IP: [<ffff8800badf2134>] 0xffff8800badf2133
[ 20.740436] PGD 138a81067 PUD 0
[ 20.740443] Oops: 0000 [#1] PREEMPT SMP
[ 20.740452] Modules linked in:
[ 20.740457] CPU 1
[ 20.740464] Pid: 1051, comm: bash Not tainted 3.5.0-1.2-desktop+ #82 System manufacturer System Product Name/P8H67-M LE
[ 20.740475] RIP: 0010:[<ffff8800badf2134>] [<ffff8800badf2134>] 0xffff8800badf2133
[ 20.740485] RSP: 0018:ffff880138bc5c48 EFLAGS: 00010002
[ 20.740491] RAX: ffff880138bc0000 RBX: ffff880138bc5e0b RCX: 00000000fed1f410
[ 20.740499] RDX: 0000000000000010 RSI: ffff880138bc5e00 RDI: ffff880138bc5e18
[ 20.740506] RBP: ffff880138bc5df8 R08: 0000000000000000 R09: 0000000000000000
[ 20.740514] R10: 0000000000000001 R11: 0000000000000001 R12: 00000000fed1f410
[ 20.740521] R13: 0000000000000000 R14: ffff880138bc5e18 R15: 0000000000000002
[ 20.740529] FS: 00007fe8f1f5a700(0000) GS:ffff88013fa80000(0000) knlGS:0000000000000000
[ 20.740538] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 20.740544] CR2: 00000000fed1f410 CR3: 0000000138b54000 CR4: 00000000000407e0
[ 20.740552] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 20.740560] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[ 20.740568] Process bash (pid: 1051, threadinfo ffff880138bc4000, task ffff880137d60040)
[ 20.740576] Stack:
[ 20.740579] ffff8800badf139d ffffffff00000000 ffffffff81c1f360 0000000000000086
[ 20.740591] ffffffff81c1f2d8 0000000000000007 ffff880138bc5e08 ffff880138bc5e18
[ 20.740603] ffff880138bc5e08 ffff880138bc5e00 ffff8800badf16f3 00000000000003ad
[ 20.740615] Call Trace:
[ 20.740624] [<ffffffff8109f35d>] ? trace_hardirqs_off+0xd/0x10
[ 20.740636] [<ffffffff81037740>] ? efi_call2+0x40/0x70
[ 20.740645] [<ffffffff810374e7>] ? virt_efi_get_time+0x57/0x90
[ 20.740654] [<ffffffff8103716a>] efi_get_time+0x4a/0x50
[ 20.740664] [<ffffffff8100ab22>] read_persistent_clock+0x12/0x30
[ 20.740674] [<ffffffff81048a0f>] mjf_proc_dointvec+0x3f/0x50
[ 20.740683] [<ffffffff810478e0>] ? iomem_is_exclusive+0xb0/0xb0
[ 20.740694] [<ffffffff811d1972>] ? sysctl_perm+0x32/0x90
[ 20.740704] [<ffffffff811d1a8c>] proc_sys_call_handler+0xbc/0xd0
[ 20.740715] [<ffffffff811d1aaf>] proc_sys_write+0xf/0x20
[ 20.740724] [<ffffffff81162a76>] vfs_write+0xc6/0x180
[ 20.740732] [<ffffffff81162d8c>] sys_write+0x4c/0x90
[ 20.740743] [<ffffffff8173e8a2>] system_call_fastpath+0x16/0x1b
Relevant bits of the dmesg for the offending address,
[ 0.000000] BIOS-e820: [mem 0x00000000fed1c000-0x00000000fed1ffff] reserved
[ 0.000000] efi: mem140: type=11, attr=0x8000000000000001, range=[0x00000000fed1c000-0x00000000fed20000) (0MB)
[ 0.000000] ACPI: HPET id: 0x8086a701 base: 0xfed00000
[ 0.000000] PM: Registered nosave memory: 00000000fed1c000 - 00000000fed20000
[ 0.251154] hpet0: at MMIO 0xfed00000, IRQs 2, 8, 0, 0, 0, 0, 0, 0
[ 0.251163] hpet0: 8 comparators, 64-bit 14.318180 MHz counter
[ 0.265436] pnp 00:08: [mem 0xfed1c000-0xfed1ffff]
[ 0.265523] system 00:08: [mem 0xfed1c000-0xfed1ffff] has been reserved
>>> On 07.08.12 at 11:30, Matt Fleming <[email protected]> wrote:
> On Tue, 2012-08-07 at 08:14 +0100, Jan Beulich wrote:
>> That's not surprising. The question really is what goes wrong
>> when the call is being made - page fault, some other fault, or
>> silent hang. A page fault would point to an incorrect memory
>> map as the prime candidate for causing the problem. My
>> primary suspect would be #NM, i.e. the implementation using
>> floating point (SSE to be precise) instructions when they're
>> unavailable.
>
> I managed to find a machine to reproduce this on and it looks like the
> ASUS firmware engineers are upto their old tricks of referencing
> physical addresses after we've taken control of the memory map,
Yippie. On such systems we simply can't do any runtime calls.
Should we add a command line option forcing efi_native to false,
thus suppressing all runtime calls? Or would the "noefi" one be
enough already?
Jan
On Tue, 2012-08-07 at 11:50 +0100, Jan Beulich wrote:
> >
> > I managed to find a machine to reproduce this on and it looks like the
> > ASUS firmware engineers are upto their old tricks of referencing
> > physical addresses after we've taken control of the memory map,
>
> Yippie. On such systems we simply can't do any runtime calls.
> Should we add a command line option forcing efi_native to false,
> thus suppressing all runtime calls? Or would the "noefi" one be
> enough already?
I think a better solution for this, seeing as there appear to be *so*
many ASUS machines in the wild with this inability to do virtual EFI
calls, is to provide a 1:1 mapping as well as our regular virt->phys
mapping for the benefit of the firmware. We can load our special page
table in efi_call_*, etc.
One thing to note is that because of breakage seen on Apple machines
last time Matthew tried this approach, we (the kernel) can't actually
access the 1:1 mapping, it would exist purely for the benefit of
firmware that was broken enough to reference physical addresses after
SetVirtualAddressMap().
FYI I tried to contact support, here's their answer:
Dear Valued Customer,
Thank you for contacting ASUS Customer Service.
My name is Carter and it's my pleasure to help you with your problem.
I am afraid to say this board doesn't support linux,so we can't asure the compability.
It is depengding on the actual situation.
If you continue to experience issues in the future, please do not hesitate
to contact us.
Best Regard
Carter
---------- Original Message ----------
>From : [email protected]
Sent : 2012/8/7 ?? 01:18:04
To : "[email protected]"
Subject : <TSD> Motherboard SABERTOOTH 990FX
[CASEID=WTM201208072118475482]
Apply date : 8/7/2012 1:18:47 PM(UTC Time)
[Contact Information]
*Name : Jérôme Carretero
*Email Address : [email protected]
[Product Information]
*Product Type : Motherboard
*Product Model : SABERTOOTH 990FX
*Product S/N : MB-1234567890
[Motherboard Specification]
*Motherboard Revision : 1.xx
*Motherboard BIOS Revision : 813
*Operating System : Linux
[Problem Description]
Hi,
A proposed patch to Linux uses EFI GetTime service to get the wall clock time.
It seems this function does not work properly.
See http://www.gossamer-threads.com/lists/linux/kernel/1576468
I know I don't have the latest bios but the change logs don't mention any EFI change,
and upgrades make the settings lost, so I didn't try them.
Thanks,
--
cJ
On Thu, Aug 9, 2012 at 1:51 AM, Matt Fleming <[email protected]> wrote:
> On Tue, 2012-08-07 at 11:50 +0100, Jan Beulich wrote:
>> >
>> > I managed to find a machine to reproduce this on and it looks like the
>> > ASUS firmware engineers are upto their old tricks of referencing
>> > physical addresses after we've taken control of the memory map,
>>
>> Yippie. On such systems we simply can't do any runtime calls.
>> Should we add a command line option forcing efi_native to false,
>> thus suppressing all runtime calls? Or would the "noefi" one be
>> enough already?
>
> I think a better solution for this, seeing as there appear to be *so*
> many ASUS machines in the wild with this inability to do virtual EFI
> calls, is to provide a 1:1 mapping as well as our regular virt->phys
> mapping for the benefit of the firmware. We can load our special page
> table in efi_call_*, etc.
>
> One thing to note is that because of breakage seen on Apple machines
> last time Matthew tried this approach, we (the kernel) can't actually
> access the 1:1 mapping, it would exist purely for the benefit of
> firmware that was broken enough to reference physical addresses after
> SetVirtualAddressMap().
>
What is solution for this regression?
It seems Jan's commit broke our setup with UEFI too.
Assume other systems with AMI code base would have same problem.
Thanks
Yinghai
On Fri, Aug 10, 2012 at 12:22:12PM -0700, Yinghai Lu wrote:
> What is solution for this regression?
Revert the patch for now, we'll add it back once we've got the UEFI
pagetable set up.
--
Matthew Garrett | [email protected]
Commit-ID: f026cfa82f628db24b8cea41b9d6202af104cecb
Gitweb: http://git.kernel.org/tip/f026cfa82f628db24b8cea41b9d6202af104cecb
Author: H. Peter Anvin <[email protected]>
AuthorDate: Tue, 14 Aug 2012 09:53:38 -0700
Committer: H. Peter Anvin <[email protected]>
CommitDate: Tue, 14 Aug 2012 09:58:25 -0700
Revert "x86-64/efi: Use EFI to deal with platform wall clock"
This reverts commit bacef661acdb634170a8faddbc1cf28e8f8b9eee.
This commit has been found to cause serious regressions on a number of
ASUS machines at the least. We probably need to provide a 1:1 map in
addition to the EFI virtual memory map in order for this to work.
Signed-off-by: H. Peter Anvin <[email protected]>
Reported-and-bisected-by: Jérôme Carretero <[email protected]>
Cc: Jan Beulich <[email protected]>
Cc: Matt Fleming <[email protected]>
Cc: Matthew Garrett <[email protected]>
Cc: Linus Torvalds <[email protected]>
Cc: Andrew Morton <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
---
arch/x86/mm/pageattr.c | 10 ++++------
arch/x86/platform/efi/efi.c | 30 ++++++++++++++++++++++++++----
include/linux/efi.h | 2 ++
init/main.c | 8 ++++----
4 files changed, 36 insertions(+), 14 deletions(-)
diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c
index 931930a..a718e0d 100644
--- a/arch/x86/mm/pageattr.c
+++ b/arch/x86/mm/pageattr.c
@@ -919,13 +919,11 @@ static int change_page_attr_set_clr(unsigned long *addr, int numpages,
/*
* On success we use clflush, when the CPU supports it to
- * avoid the wbindv. If the CPU does not support it, in the
- * error case, and during early boot (for EFI) we fall back
- * to cpa_flush_all (which uses wbinvd):
+ * avoid the wbindv. If the CPU does not support it and in the
+ * error case we fall back to cpa_flush_all (which uses
+ * wbindv):
*/
- if (early_boot_irqs_disabled)
- __cpa_flush_all((void *)(long)cache);
- else if (!ret && cpu_has_clflush) {
+ if (!ret && cpu_has_clflush) {
if (cpa.flags & (CPA_PAGES_ARRAY | CPA_ARRAY)) {
cpa_flush_array(addr, numpages, cache,
cpa.flags, pages);
diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c
index 2dc29f5..92660eda 100644
--- a/arch/x86/platform/efi/efi.c
+++ b/arch/x86/platform/efi/efi.c
@@ -234,7 +234,22 @@ static efi_status_t __init phys_efi_set_virtual_address_map(
return status;
}
-static int efi_set_rtc_mmss(unsigned long nowtime)
+static efi_status_t __init phys_efi_get_time(efi_time_t *tm,
+ efi_time_cap_t *tc)
+{
+ unsigned long flags;
+ efi_status_t status;
+
+ spin_lock_irqsave(&rtc_lock, flags);
+ efi_call_phys_prelog();
+ status = efi_call_phys2(efi_phys.get_time, virt_to_phys(tm),
+ virt_to_phys(tc));
+ efi_call_phys_epilog();
+ spin_unlock_irqrestore(&rtc_lock, flags);
+ return status;
+}
+
+int efi_set_rtc_mmss(unsigned long nowtime)
{
int real_seconds, real_minutes;
efi_status_t status;
@@ -263,7 +278,7 @@ static int efi_set_rtc_mmss(unsigned long nowtime)
return 0;
}
-static unsigned long efi_get_time(void)
+unsigned long efi_get_time(void)
{
efi_status_t status;
efi_time_t eft;
@@ -606,13 +621,18 @@ static int __init efi_runtime_init(void)
}
/*
* We will only need *early* access to the following
- * EFI runtime service before set_virtual_address_map
+ * two EFI runtime services before set_virtual_address_map
* is invoked.
*/
+ efi_phys.get_time = (efi_get_time_t *)runtime->get_time;
efi_phys.set_virtual_address_map =
(efi_set_virtual_address_map_t *)
runtime->set_virtual_address_map;
-
+ /*
+ * Make efi_get_time can be called before entering
+ * virtual mode.
+ */
+ efi.get_time = phys_efi_get_time;
early_iounmap(runtime, sizeof(efi_runtime_services_t));
return 0;
@@ -700,10 +720,12 @@ void __init efi_init(void)
efi_enabled = 0;
return;
}
+#ifdef CONFIG_X86_32
if (efi_native) {
x86_platform.get_wallclock = efi_get_time;
x86_platform.set_wallclock = efi_set_rtc_mmss;
}
+#endif
#if EFI_DEBUG
print_efi_memmap();
diff --git a/include/linux/efi.h b/include/linux/efi.h
index 103adc6..ec45ccd 100644
--- a/include/linux/efi.h
+++ b/include/linux/efi.h
@@ -503,6 +503,8 @@ extern u64 efi_mem_attribute (unsigned long phys_addr, unsigned long size);
extern int __init efi_uart_console_only (void);
extern void efi_initialize_iomem_resources(struct resource *code_resource,
struct resource *data_resource, struct resource *bss_resource);
+extern unsigned long efi_get_time(void);
+extern int efi_set_rtc_mmss(unsigned long nowtime);
extern void efi_reserve_boot_services(void);
extern struct efi_memory_map memmap;
diff --git a/init/main.c b/init/main.c
index e60679d..b286730 100644
--- a/init/main.c
+++ b/init/main.c
@@ -461,10 +461,6 @@ static void __init mm_init(void)
percpu_init_late();
pgtable_cache_init();
vmalloc_init();
-#ifdef CONFIG_X86
- if (efi_enabled)
- efi_enter_virtual_mode();
-#endif
}
asmlinkage void __init start_kernel(void)
@@ -606,6 +602,10 @@ asmlinkage void __init start_kernel(void)
calibrate_delay();
pidmap_init();
anon_vma_init();
+#ifdef CONFIG_X86
+ if (efi_enabled)
+ efi_enter_virtual_mode();
+#endif
thread_info_cache_init();
cred_init();
fork_init(totalram_pages);
On Fri, 2012-08-10 at 20:26 +0100, Matthew Garrett wrote:
> On Fri, Aug 10, 2012 at 12:22:12PM -0700, Yinghai Lu wrote:
>
> > What is solution for this regression?
>
> Revert the patch for now, we'll add it back once we've got the UEFI
> pagetable set up.
FYI I've got a proof of concept patch that seems to work on my ASUS
machine.
I'll try and get it sent out before the end of the week, but failing
that it'll probably be after Plumbers.