This adds a const to the definitions vvar.c makes, so that the vdso_*
variables go into .rodata instead of .data. This is essentially a
cosmetic change, just giving the section headers in the vDSO file more
pleasing flags. These variables are read-only from the perspective of
the vDSO itself and user mode, even though the contents of the DSO image
were adjusted at boot.
Signed-off-by: Roland McGrath <roland@redhat.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Now that we check for translation enabled/disabled based on the presence
of the IOMMU translation table, we can get rid of translate_phb.
Signed-off-by: Muli Ben-Yehuda <muli@il.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
- Fix this:
include/asm/io.h: In function `memcpy_fromio':
include/asm/io.h:208: warning: passing argument 2 of `__memcpy' discards qualifiers from pointer target type
- Clean up code a bit
Reported-by: Uwe Bugla <uwe.bugla@gmx.de>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
"extern inline" will have different semantics with gcc 4.3.
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Acked-by: Andrey Panin <pazke@donpac.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
In x86_64 and i386 architectures most arrays that are sized using
NR_CPUS lay in local memory on node 0. Not only will most (99%?) of the
systems not use all the slots in these arrays, particularly when NR_CPUS
is increased to accommodate future very high cpu count systems, but a
number of cache lines are passed unnecessarily on the system bus when
these arrays are referenced by cpus on other nodes.
Typically, the values in these arrays are referenced by the cpu
accessing it's own values, though when passing IPI interrupts, the cpu
does access the data relevant to the targeted cpu/node. Of course, if
the referencing cpu is not on node 0, then the reference will still
require cross node exchanges of cache lines. A common use of this is
for an interrupt service routine to pass the interrupt to other cpus
local to that node.
Ideally, all the elements in these arrays should be moved to the per_cpu
data area. In some cases (such as x86_cpu_to_apicid) the array is
referenced before the per_cpu data areas are setup. In this case, a
static array is declared in the __initdata area and initialized by the
booting cpu (BSP). The values are then moved to the per_cpu area after
it is initialized and the original static array is freed with the rest
of the __initdata.
This patch:
Fix four instances where cpu_to_node is referenced by array instead of
via the cpu_to_node macro. This is preparation to moving it to the
per_cpu data area.
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Andi Kleen <ak@suse.de>
Cc: Christoph Lameter <clameter@sgi.com>
Cc: "Siddha, Suresh B" <suresh.b.siddha@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
sys_iopl is long gone and there is no reason to declare
sys_rt_sigaction here.
Remove it all together and fix the whitespace mess as well.
It's worth the trouble: 25897 -> 21337 bytes, the win is
larger than the memory of my first computer :)
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
It needs to be automatic. The HPET legacy interrupt mode disconnects
the RTC interrupt and connects the interrupt of the second HPET channel.
[ tglx: arch/x86 adaptation and comment fixup]
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
.. as they're, with a single exception, never written to.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Move the = into the __setup line.
Document the option in kernel-parameters.txt by adding a pointer
to the x86-64 specific documentation.
[ tglx: arch/x86 adaptation ]
Pointed out by Robert Day
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
These build warnings:
In file included from include/asm/thread_info.h:16,
from include/linux/thread_info.h:21,
from include/linux/preempt.h:9,
from include/linux/spinlock.h:49,
from include/linux/vmalloc.h:4,
from arch/i386/boot/compressed/misc.c:14:
include/asm/processor.h: In function cpuid_count
include/asm/processor.h:615: warning: pointer targets in passing argument 1 of native_cpuid differ in signedness
include/asm/processor.h:615: warning: pointer targets in passing argument 2 of native_cpuid differ in signedness
include/asm/processor.h:615: warning: pointer targets in passing argument 3 of native_cpuid differ in signedness
include/asm/processor.h:615: warning: pointer targets in passing argument 4 of native_cpuid differ in signedness
come because the arguments have been specified as pointers to (signed) int
types, not unsigned. So let's specify those as unsigned. Do some codingstyle
here and there while at it.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Satyam Sharma <satyam@infradead.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>
.i is an ending used for preprocessed stuff.
This patch therefore renames assembler include files to .h and guards
the contents with an #ifdef __ASSEMBLY__.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Adrian Bunk <bunk@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>
It is not good taste to have macros with additions that do not have
parenthesises around them. This patch parethesizes the IRQ vector
macros for x86_64 arch.
Note, this caused me a bit of heart-ache debugging lguest64.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.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>
The return type of __scanbit() doesn't match the return type of
find_{first,next}_bit(). Thus when you construct something like
this:
boolean ? __scanbit() : find_first_bit()
you get an unsigned long result if "boolean" is true, and a signed
long result if "boolean" is false.
In file included from /home/cel/src/linux/include/linux/mmzone.h:15,
from /home/cel/src/linux/include/linux/gfp.h:4,
from /home/cel/src/linux/include/linux/slab.h:14,
from /home/cel/src/linux/include/linux/percpu.h:5,
from
/home/cel/src/linux/include/linux/rcupdate.h:41,
from /home/cel/src/linux/include/linux/dcache.h:10,
from /home/cel/src/linux/include/linux/fs.h:275,
from /home/cel/src/linux/fs/nfs/sysctl.c:9:
/home/cel/src/linux/include/linux/nodemask.h: In function
â__first_nodeâ:
/home/cel/src/linux/include/linux/nodemask.h:229: warning: signed and
unsigned type in conditional expression
/home/cel/src/linux/include/linux/nodemask.h: In function
â__next_nodeâ:
/home/cel/src/linux/include/linux/nodemask.h:235: warning: signed and
unsigned type in conditional expression
/home/cel/src/linux/include/linux/nodemask.h: In function
â__first_unset_nodeâ:
/home/cel/src/linux/include/linux/nodemask.h:253: warning: signed and
unsigned type in conditional expression
[ tglx: arch/x86 adaptation ]
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch removes the __STR() and STR() macros from x86_64 header files.
They seem to be legacy, and has no more users. Even if there were users,
they should use __stringify() instead.
In fact, there were one third place in which this macro was defined
(ia32_binfmt.c), and used just below. In this file, usage was properly
converted to __stringify()
[ tglx: arch/x86 adaptation ]
Signed-off-by: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Remove the x86_cpu_to_log_apicid array. It is set in
arch/x86_64/kernel/genapic_flat.c:flat_init_apic_ldr() and
arch/x86_64/kernel/smpboot.c:do_boot_cpu() but it is never
referenced.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.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>
The constraints in the inline assembler implementation of i386
strrchr() were incorrect and break the build with recent gcc 4.3.
Since there are only very few callers of strrchr() and none of them
are performance relevant just remove the assembler implementation
and use the C fallback instead.
[ tglx: arch/x86 adaptation ]
Cc: rguenther@suse.de
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The volatile keyword has already been removed from the declaration of atomic_t
on x86_64. For consistency, remove it from atomic64_t as well.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Chris Snook <csnook@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
CC: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
smp_call_function_single() now knows how to call the function on the
current cpu.
[ tglx: arch/x86 adaptation ]
Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
smp_call_function_single() now knows how to call the function on the
current cpu.
[ tglx: arch/x86 adaptation ]
Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
As long as there's no write access to this variable there's no reason to
let gcc check it at runtime.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch contains the following cleanups:
- make some needlessly global functions static
- #if 0 the unused es7000_stop_cpu()
AK: actually removed es7000_stop_cpu
AK: fixed a non ISO prototype too
[ tglx: arch/x86 adaptation ]
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.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>
smp_call_function_single handles the call to local CPU case correctly
now, no need to handle this in the caller.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Andrew Morton <akpm@linux-foundation.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>
Create an inline function for clflush(), with the proper arguments,
and use it instead of hard-coding the instruction.
This also removes one instance of hard-coded wbinvd, based on a patch
by Bauder de Oliveira Costa.
[ tglx: arch/x86 adaptation ]
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Glauber de Oliveira Costa <gcosta@redhat.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
.. as they're never written to.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Miscellaneous x86 stuff that can live in .rodata.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The i386 irqstat per cpu conversion left an bogus export of the old
irqstat array in the header file. Remove it.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
- It was redundant with sync_core()
- It was unused
- It was broken: no input arguments to cpuid; could fault randomly
depending on eax contents.
Now it's gone.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
cpuid_class_cpu_callback() is callback function of a CPU hotplug
notifier_block (that is already marked as __cpuinitdata). Therefore
it can safely be marked as __cpuinit.
cpuid_device_create() is only referenced from other functions that
are __cpuinit or __init. So it can also be safely marked __cpuinit.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Satyam Sharma <satyam@infradead.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>
msr_class_cpu_callback() can be marked __cpuinit, being the notifier
callback for a __cpuinitdata notifier_block. So can be marked
msr_device_create() too, called only from the newly-__cpuinit
msr_class_cpu_callback() or from __init-marked msr_init().
[ tglx: arch/x86 adaptation ]
Signed-off-by: Satyam Sharma <satyam@infradead.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>
add cpu core name for arch/i386/Kconfig.cpu:Pentium 4 sections help
add Pentium D for arch/i386/Kconfig.cpu
add Pentium D for arch/x86_64/Kconfig
AK: Clarified some of the descriptions
[ tglx: arch/x86 adaptation ]
Signed-off-by: Oliver Pinter <oliver.pntr@gmail.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch makes some needlessly global variables static.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This brings x86_64 into line with all other architectures by only defining
cond_syscall() when __KERNEL__ is defined.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.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>
This patch makes the needlessly global struct apic_probe static.
[ tglx: arch/x86 adaptation ]
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.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>
Remove rogue default m in drivers/video/Kconfig
default m is near always wrong, like here. For some reason ACPI
likes to reintroduce these and I like to immediately squash them again
before they pollute too many .configs.
Cc: len.brown@intel.com
Cc: luming.yu@gmail.com
Acked-by: Len Brown <len.brown@intel.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
cache_shared_cpu_map_setup() and cache_remove_shared_cpu_map()
are functions called from another function that is __cpuinit. But the
!CONFIG_SMP empty-body stubs of these functions are unconditionally
marked __init, which is actively wrong, and will lead to oops. But we
never saw this oops, because they always managed to get inlined in their
callsites, by virtue of being empty-body stubs! They should still be
__cpuinit, of course.
assocs[], levels[] and types[] are only referenced from function that is
__cpuinit. So these are candidates for being marked __cpuinitdata.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Satyam Sharma <satyam@infradead.org>
Cc: Andi Kleen <ak@suse.de>
Cc: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
use dev_to_node() to get node for device in dma_alloc_pages().
Signed-off-by: Yinghai Lu <yinghai.lu@sun.com>
Acked-by: Christoph Lameter <clameter@sgi.com>
Cc: Christoph Lameter <clameter@sgi.com>
Cc: Andy Whitcroft <apw@shadowen.org>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
HP ProLiant systems DL385 G2 and DL585 G2 need pci=bfsort to enumerate PCI
devices in the expected order.
Matt sayeth:
biosdevname is a userspace app I wrote to help solve this so we don't need
to patch the kernel for future systems. It's not integrated into any
distributions properly yet, but is included in openSUSE 10.3 and Fedora 8
for people who want to download and install it there. It acts as a udev
helper.
For the time being, patching the kernel is necessary. I really hope
biosdevname eliminates that need in future distributions.
http://linux.dell.com/biosdevname/
Signed-off-by: Michal Schmidt <mschmidt@redhat.com>
Cc: <john.cagle@hp.com>
Cc: Matt Domsch <Matt_Domsch@dell.com>
Cc: Andy Gospodarek <andy@greyhouse.net>
Cc: Greg KH <greg@kroah.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The old check we used based on dev->bus->number is wrong for devices on
CalIOC2. Instead look whether we have an IOMMU table for that bus - if
not, translation is disabled.
Thanks to Murillo Fernandes Bernardes <bernarde@br.ibm.com> for
spotting, suggesting a fix and testing.
Signed-off-by: Muli Ben-Yehuda <muli@il.ibm.com>
Acked-by: Murillo Fernandes Bernardes <bernarde@br.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch fixes a bug of change_page_attr/change_page_attr_addr on
Intel x86_64 CPUs. After changing page attribute to be executable with
these functions, the page remains un-executable on Intel x86_64 CPU.
Because on Intel x86_64 CPU, only if the "NX" bits of all four level
page tables are cleared, the corresponding page is executable (refer to
section 4.13.2 of Intel 64 and IA-32 Architectures Software Developer's
Manual). So, the bug is fixed through clearing the "NX" bit of PMD when
splitting the huge PMD.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
A while ago Randy Dunlap and Adrian Bunk suggested we simply prevent UP
voyager building. I resisted this on the grounds that the nagging was the
only thing that was going to cause me to look at this. However, now I
think we should probably take this course.
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
When dumping memory via sysrq-m it is possible to take a bogus NMI
watchdog or softlockup watchdog because the dump can take a long time on
big memory systems.
Occasionally tickle the watchdog when doing the dump.
Signed-off-by: Prarit Bhargava <prarit@redhat.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
if CONFIG_PAGEALLOC is enabled then X86_FEATURE_PSE is disabled and all
the kernel physical RAM pagetables are set up as 4K pages. This is
needed so that CONFIG_PAGEALLOC can do finegrained mapping and unmapping
of pages.
as a side-effect though, the total size of memory allocated as kernel
pagetables increases significantly. All these pagetables are allocated
via alloc_bootmem_low_pages(), straight out of the lowmem DMA pool. If
the system has enough RAM and a large kernel image then almost all of
the 16 MB lowmem DMA pool is allocated to the image and to pagetables -
leaving no space for __GFP_DMA allocations.
this results in drivers failing and the bootup hanging:
swapper invoked oom-killer: gfp_mask=0x80d1, order=0, oomkilladj=0
[<4015059f>] out_of_memory+0x17f/0x1c0
[<40151f3c>] __alloc_pages+0x37c/0x3a0
[<40168cd7>] slob_new_page+0x37/0x50
[<40168dff>] slob_alloc+0x10f/0x190
[<40169010>] __kmalloc_node+0x80/0x90
[<405a17e3>] scsi_host_alloc+0x33/0x2c0
[<405a1a82>] scsi_register+0x12/0x60
[<40d5889e>] aha1542_detect+0x9e/0x940
[<405c5ba5>] ultrastor_detect+0x265/0x5f0
[<401352f5>] getnstimeofday+0x35/0xf0
[<40d58751>] init_this_scsi_driver+0x41/0xf0
[<40d0b856>] kernel_init+0x136/0x310
[<40d58710>] init_this_scsi_driver+0x0/0xf0
[<40d0b720>] kernel_init+0x0/0x310
[<40105547>] kernel_thread_helper+0x7/0x10
=======================
the fix is to first allocate from above the DMA pool, and if that fails
(for example due to it being a machine with less than 16 MB of RAM),
allocate from the DMA pool as a fallback.
With this fix applied i was able to boot a PAGEALLOC=y kernel that would
hang before.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
To preserve the DMA pool in CONFIG_DEBUG_PAGEALLOC=y kernels, we'll
allocate pagetables from above the 16MB DMA limit, so we'll have to set
up boot pagetables to cover 16MB more RAM (worst-case).
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
