This more or less reverts commits 08be979 (x86: Force HPET
readback_cmp for all ATI chipsets) and 30a564be (x86, hpet: Restrict
read back to affected ATI chipsets) to the status of commit 8da854c
(x86, hpet: Erratum workaround for read after write of HPET
comparator).
The delta to commit 8da854c is mostly comments and the change from
WARN_ONCE to printk_once as we know the call path of this function
already.
This needs really in depth explanation:
First of all the HPET design is a complete failure. Having a counter
compare register which generates an interrupt on matching values
forces the software to do at least one superfluous readback of the
counter register.
While it is nice in theory to program "absolute" time events it is
practically useless because the timer runs at some absurd frequency
which can never be matched to real world units. So we are forced to
calculate a relative delta and this forces a readout of the actual
counter value, adding the delta and programming the compare
register. When the delta is small enough we run into the danger that
we program a compare value which is already in the past. Due to the
compare for equal nature of HPET we need to read back the counter
value after writing the compare rehgister (btw. this is necessary for
absolute timeouts as well) to make sure that we did not miss the timer
event. We try to work around that by setting the minimum delta to a
value which is larger than the theoretical time which elapses between
the counter readout and the compare register write, but that's only
true in theory. A NMI or SMI which hits between the readout and the
write can easily push us beyond that limit. This would result in
waiting for the next HPET timer interrupt until the 32bit wraparound
of the counter happens which takes about 306 seconds.
So we designed the next event function to look like:
match = read_cnt() + delta;
write_compare_ref(match);
return read_cnt() < match ? 0 : -ETIME;
At some point we got into trouble with certain ATI chipsets. Even the
above "safe" procedure failed. The reason was that the write to the
compare register was delayed probably for performance reasons. The
theory was that they wanted to avoid the synchronization of the write
with the HPET clock, which is understandable. So the write does not
hit the compare register directly instead it goes to some intermediate
register which is copied to the real compare register in sync with the
HPET clock. That opens another window for hitting the dreaded "wait
for a wraparound" problem.
To work around that "optimization" we added a read back of the compare
register which either enforced the update of the just written value or
just delayed the readout of the counter enough to avoid the issue. We
unfortunately never got any affirmative info from ATI/AMD about this.
One thing is sure, that we nuked the performance "optimization" that
way completely and I'm pretty sure that the result is worse than
before some HW folks came up with those.
Just for paranoia reasons I added a check whether the read back
compare register value was the same as the value we wrote right
before. That paranoia check triggered a couple of years after it was
added on an Intel ICH9 chipset. Venki added a workaround (commit
8da854c) which was reading the compare register twice when the first
check failed. We considered this to be a penalty in general and
restricted the readback (thus the wasted CPU cycles) to the known to
be affected ATI chipsets.
This turned out to be a utterly wrong decision. 2.6.35 testers
experienced massive problems and finally one of them bisected it down
to commit 30a564be which spured some further investigation.
Finally we got confirmation that the write to the compare register can
be delayed by up to two HPET clock cycles which explains the problems
nicely. All we can do about this is to go back to Venki's initial
workaround in a slightly modified version.
Just for the record I need to say, that all of this could have been
avoided if hardware designers and of course the HPET committee would
have thought about the consequences for a split second. It's out of my
comprehension why designing a working timer is so hard. There are two
ways to achieve it:
1) Use a counter wrap around aware compare_reg <= counter_reg
implementation instead of the easy compare_reg == counter_reg
Downsides:
- It needs more silicon.
- It needs a readout of the counter to apply a relative
timeout. This is necessary as the counter does not run in
any useful (and adjustable) frequency and there is no
guarantee that the counter which is used for timer events is
the same which is used for reading the actual time (and
therefor for calculating the delta)
Upsides:
- None
2) Use a simple down counter for relative timer events
Downsides:
- Absolute timeouts are not possible, which is not a problem
at all in the context of an OS and the expected
max. latencies/jitter (also see Downsides of #1)
Upsides:
- It needs less or equal silicon.
- It works ALWAYS
- It is way faster than a compare register based solution (One
write versus one write plus at least one and up to four
reads)
I would not be so grumpy about all of this, if I would not have been
ignored for many years when pointing out these flaws to various
hardware folks. I really hate timers (at least those which seem to be
designed by janitors).
Though finally we got a reasonable explanation plus a solution and I
want to thank all the folks involved in chasing it down and providing
valuable input to this.
Bisected-by: Nix <nix@esperi.org.uk>
Reported-by: Artur Skawina <art.08.09@gmail.com>
Reported-by: Damien Wyart <damien.wyart@free.fr>
Reported-by: John Drescher <drescherjm@gmail.com>
Cc: Venkatesh Pallipadi <venki@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Andreas Herrmann <andreas.herrmann3@amd.com>
Cc: Borislav Petkov <borislav.petkov@amd.com>
Cc: stable@kernel.org
Acked-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The ACPI/X86_IO_ACPI ifdef isn't necessary at this point,
because it is checked in an outer ifdef level already and has no
effect here.
Cleanup only, no functional effect.
Signed-off-by: Christian Dietrich <qy03fugy@stud.informatik.uni-erlangen.de>
Acked-by: Borislav Petkov <borislav.petkov@amd.com>
Cc: vamos-dev@i4.informatik.uni-erlangen.de
LKML-Reference: <d4376e6d79b8dc0f89a4b3ce4a880904a7b93ead.1283782701.git.qy03fugy@stud.informatik.uni-erlangen.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
commit 30a564be (x86, hpet: Restrict read back to affected ATI
chipset) restricted the workaround for the HPET bug to SMX00
chipsets. This was reasonable as those were the only ones against
which we ever got a bug report.
Stephan Wolf reported now that this patch breaks his IXP400 based
machine. Though it's confirmed to work on other IXP400 based systems.
To error out on the safe side, we force the HPET readback workaround
for all ATI SMbus class chipsets.
Reported-by: Stephan Wolf <stephan@letzte-bankreihe.de>
LKML-Reference: <alpine.LFD.2.00.1007142134140.3321@localhost.localdomain>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Stephan Wolf <stephan@letzte-bankreihe.de>
Acked-by: Borislav Petkov <borislav.petkov@amd.com>
Impact: cleanup
this warning:
arch/x86/kernel/early-quirks.c:99: warning: ‘ati_ixp4x0_rev’ defined but not used
triggers because ati_ixp4x0_rev() is only used in the
ACPI && X86_IO_APIC case.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
We found a situation on Linus' machine that the Nvidia timer quirk hit on
a Intel chipset system. The problem is that the system has a fancy Nvidia
card with an own PCI bridge, and the early-quirks code looking for any
NVidia bridge triggered on it incorrectly. This didn't lead a boot
failure by luck, but the timer routing code selecting the wrong timer
first and some ugly messages. It might lead to real problems on other
systems.
I checked all the devices which are currently checked for by early_quirks
and it turns out they are all located in the root bus zero.
So change the early-quirks loop to only scan bus 0. This incidently also
saves quite some unnecessary scanning work, because early_quirks doesn't
go through all the non root busses.
The graphics card is not on bus 0, so it is not matched anymore.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This reverts commit e51af66308, which was
wrongly hoovered up and submitted about a month after a better fix had
already been merged.
The better fix is commit cbda1ba898
("PCI/iommu: blacklist DMAR on Intel G31/G33 chipsets"), where we do
this blacklisting based on the DMI identification for the offending
motherboard, since sometimes this chipset (or at least a chipset with
the same PCI ID) apparently _does_ actually have an IOMMU.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
On some more HP laptops BIOS reports an IRQ0 override
but the SB600 chipset is configured such that timer
interrupts go to INT0 of IOAPIC.
Check IRQ0 routing and if it is routed to INT0 of IOAPIC skip the
timer override.
http://bugzilla.kernel.org/show_bug.cgi?id=11715http://bugzilla.kernel.org/show_bug.cgi?id=11516
Signed-off-by: Andreas Herrmann <andreas.herrmann3@amd.com>
Signed-off-by: Len Brown <len.brown@intel.com>
On some HP nx6... laptops (e.g. nx6325) BIOS reports an IRQ0 override
but the SB450 chipset is configured such that timer interrupts goe to
INT0 of IOAPIC.
Check IRQ0 routing and if it is routed to INT0 of IOAPIC skip the
timer override.
[ This more generic PCI ID based quirk should alleviate the need for
dmi_ignore_irq0_timer_override DMI quirks. ]
Signed-off-by: Andreas Herrmann <andreas.herrmann3@amd.com>
Acked-by: "Maciej W. Rozycki" <macro@linux-mips.org>
Tested-by: Dmitry Torokhov <dtor@mail.ru>
Cc: <stable@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Some BIOSes (the Intel DG33BU, for example) wrongly claim to have DMAR
when they don't. Avoid the resulting crashes when it doesn't work as
expected.
I'd still be grateful if someone could test it on a DG33BU with the old
BIOS though, since I've killed mine. I tested the DMI version, but not
this one.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'linux-next' of git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6: (72 commits)
Revert "x86/PCI: ACPI based PCI gap calculation"
PCI: remove unnecessary volatile in PCIe hotplug struct controller
x86/PCI: ACPI based PCI gap calculation
PCI: include linux/pm_wakeup.h for device_set_wakeup_capable
PCI PM: Fix pci_prepare_to_sleep
x86/PCI: Fix PCI config space for domains > 0
Fix acpi_pm_device_sleep_wake() by providing a stub for CONFIG_PM_SLEEP=n
PCI: Simplify PCI device PM code
PCI PM: Introduce pci_prepare_to_sleep and pci_back_from_sleep
PCI ACPI: Rework PCI handling of wake-up
ACPI: Introduce new device wakeup flag 'prepared'
ACPI: Introduce acpi_device_sleep_wake function
PCI: rework pci_set_power_state function to call platform first
PCI: Introduce platform_pci_power_manageable function
ACPI: Introduce acpi_bus_power_manageable function
PCI: make pci_name use dev_name
PCI: handle pci_name() being const
PCI: add stub for pci_set_consistent_dma_mask()
PCI: remove unused arch pcibios_update_resource() functions
PCI: fix pci_setup_device()'s sprinting into a const buffer
...
Fixed up conflicts in various files (arch/x86/kernel/setup_64.c,
arch/x86/pci/irq.c, arch/x86/pci/pci.h, drivers/acpi/sleep/main.c,
drivers/pci/pci.c, drivers/pci/pci.h, include/acpi/acpi_bus.h) from x86
and ACPI updates manually.
gart.h has only GART-specific stuff. Only GART code needs it. Other
IOMMU stuff should include iommu.h instead of gart.h.
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Acked-by: Muli Ben-Yehuda <muli@il.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Remove the "disable_8254_timer" and "enable_8254_timer" kernel
parameters. Now that AEOI acknowledgements are no longer needed for
correct timer operation, the 8259A can be kept disabled unconditionally
unless interrupts, either timer or watchdog ones, are actually passed
through it.
Signed-off-by: Maciej W. Rozycki <macro@linux-mips.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
On x86, we do early PCI probing to apply some quirks for chipset bugs.
However, in a recent cleanup (7bcbc78dea) a
thinko was introduced that causes us to probe all subfunctions of even single
function devices (a function was factored out of an inner loop and a "break"
became a "return"). Fix that up by making check_dev_quirk() return a value so
we can keep the factored code intact.
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Fix following warnings:
WARNING: arch/x86/kernel/built-in.o(.text+0x139e1): Section mismatch: reference to .init.data:early_qrk in 'check_dev_quirk'
WARNING: arch/x86/kernel/built-in.o(.text+0x139f5): Section mismatch: reference to .init.data:early_qrk in 'check_dev_quirk'
WARNING: arch/x86/kernel/built-in.o(.text+0x13a0c): Section mismatch: reference to .init.data:early_qrk in 'check_dev_quirk'
WARNING: arch/x86/kernel/built-in.o(.text+0x13a12): Section mismatch: reference to .init.data:early_qrk in 'check_dev_quirk'
WARNING: arch/x86/kernel/built-in.o(.text+0x13a1a): Section mismatch: reference to .init.data:early_qrk in 'check_dev_quirk'
WARNING: arch/x86/kernel/built-in.o(.text+0x13a36): Section mismatch: reference to .init.data:early_qrk in 'check_dev_quirk'
WARNING: arch/x86/kernel/built-in.o(.text+0x13a42): Section mismatch: reference to .init.data:
Warning was caused by access to the __initdata annotated variable
from the non-annotated static function check_dev_quirk().
check_dev_quirk() were only used from a function annotated
__init so add __init annotation to check_dev_quirk() to fix it.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Recently a kdump bug was discovered in which a system would hang inside
calibrate_delay during the booting of the kdump kernel. This was caused
by the fact that the jiffies counter was not being incremented during
timer calibration. The root cause of this problem was found to be a
bios misconfiguration of the hypertransport bus. On system affected by
this hang, the bios had assigned APIC ids which used extended apic bits
(more than the nominal 4 bit ids's), but failed to configure bit 17 of
the hypertransport transaction config register, which indicated that the
mask for the destination field of interrupt packets accross the ht bus
(see section 3.3.9 of
http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/26094.PDF).
If a crash occurs on a cpu with an APIC id that extends beyond 4 bits,
it will not recieve interrupts during the kdump kernel boot, and this
hang will be the result. The fix is to add this patch, whcih add an
early pci quirk check, to forcibly enable this bit in the httcfg
register. This enables all cpus on a system to receive interrupts, and
allows kdump kernel bootup to procede normally.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch renames the 4 symbols iommu_hole_init(), iommu_aperture,
iommu_aperture_allowed, iommu_aperture_disabled. All these symbols are only
used for the GART implementation of IOMMUs.
It adds and additional gart_ prefix to them.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch renames the IOMMU config option to GART_IOMMU because in fact it
means the GART and not general support for an IOMMU on x86.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Acked-by: Muli Ben-Yehuda <muli@il.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch renames the include file asm-x86/iommu.h to asm-x86/gart.h to make
clear to which IOMMU implementation it belongs. The patch also adds "GART" to
the Kconfig line.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Acked-by: Muli Ben-Yehuda <muli@il.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
fix this warning:
arch/x86/kernel/early-quirks.c:40: warning: nvidia_hpet_check defined but not used
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
They were already very similar; just use the same file now.
[ tglx: arch/x86 adaptation ]
Cc: lenb@kernel.org
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Linus Torvalds