Impact: make nr_irqs depend more on cards used in a system
depend on nr_irq_gsi more, and have a ratio for MSI.
v2: make nr_irqs less than NR_VECTORS * nr_cpu_ids
aka if only one cpu, we only can support nr_irqs = NR_VECTORS
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: clean up
Now that a generic in_nmi is available, this patch removes the
special code in the ring_buffer and implements the in_nmi generic
version instead.
With this change, I was also able to rename the "arch_ftrace_nmi_enter"
back to "ftrace_nmi_enter" and remove the code from the ring buffer.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
The function graph tracer piggy backed onto the dynamic ftracer
to use the in_nmi custom code for dynamic tracing. The problem
was (as Andrew Morton pointed out) it really only wanted to bail
out if the context of the current CPU was in NMI context. But the
dynamic ftrace in_nmi custom code was true if _any_ CPU happened
to be in NMI context.
Now that we have a generic in_nmi interface, this patch changes
the function graph code to use it instead of the dynamic ftarce
custom code.
Reported-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Impact: clean up
The in_nmi variable in x86 arch ftrace.c is a misnomer.
Andrew Morton pointed out that the in_nmi variable is incremented
by all CPUS. It can be set when another CPU is running an NMI.
Since this is actually intentional, the fix is to rename it to
what it really is: "nmi_running"
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Impact: prevent deadlock in NMI
The ring buffers are not yet totally lockless with writing to
the buffer. When a writer crosses a page, it grabs a per cpu spinlock
to protect against a reader. The spinlocks taken by a writer are not
to protect against other writers, since a writer can only write to
its own per cpu buffer. The spinlocks protect against readers that
can touch any cpu buffer. The writers are made to be reentrant
with the spinlocks disabling interrupts.
The problem arises when an NMI writes to the buffer, and that write
crosses a page boundary. If it grabs a spinlock, it can be racing
with another writer (since disabling interrupts does not protect
against NMIs) or with a reader on the same CPU. Luckily, most of the
users are not reentrant and protects against this issue. But if a
user of the ring buffer becomes reentrant (which is what the ring
buffers do allow), if the NMI also writes to the ring buffer then
we risk the chance of a deadlock.
This patch moves the ftrace_nmi_enter called by nmi_enter() to the
ring buffer code. It replaces the current ftrace_nmi_enter that is
used by arch specific code to arch_ftrace_nmi_enter and updates
the Kconfig to handle it.
When an NMI is called, it will set a per cpu variable in the ring buffer
code and will clear it when the NMI exits. If a write to the ring buffer
crosses page boundaries inside an NMI, a trylock is used on the spin
lock instead. If the spinlock fails to be acquired, then the entry
is discarded.
This bug appeared in the ftrace work in the RT tree, where event tracing
is reentrant. This workaround solved the deadlocks that appeared there.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Fix compile problem:
CC arch/x86/kernel/early_printk.o
In file included from /home/jeremy/hg/xen/paravirt/linux/arch/x86/kernel/early_printk.c:17:
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h: In function 'pmd_page':
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h:516: error: implicit declaration of function '__pfn_to_section'
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h:516: warning: initialization makes pointer from integer without a cast
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h:516: error: implicit declaration of function '__section_mem_map_addr'
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h:516: warning: return makes pointer from integer without a cast
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h: In function 'pud_page':
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h:586: warning: initialization makes pointer from integer without a cast
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h:586: warning: return makes pointer from integer without a cast
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h: In function 'pgd_page':
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h:625: warning: initialization makes pointer from integer without a cast
/home/jeremy/hg/xen/paravirt/linux/arch/x86/include/asm/pgtable.h:625: warning: return makes pointer from integer without a cast
This is a cycling dependency between asm/pgtable.h and linux/mmzone.h
when using CONFIG_SPARSEMEM. Rather than hacking up the headers some
more, remove asm/pgtable.h, since early_printk.c doesn't actually need
it.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Implement Linus's suggestion: introduce the hpet_cnt_ahead()
helper function to compare hpet time values - like other
wrapping counter comparisons are abstracted away elsewhere.
(jiffies, ktime_t, etc.)
Reported-by: Kirill Korotaev <dev@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup
only leave _default_ipi_xx etc in .h
Beyond the cleanup factor, this saves a bit of code size as well:
text data bss dec hex filename
7281931 1630144 1463304 10375379 9e50d3 vmlinux.before
7281753 1630144 1463304 10375201 9e5021 vmlinux.after
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup
disable_ioapic_setup() in init/main.c is ugly as the function is
x86-specific. The #ifdef inline prototype there is ugly too.
Replace it with a generic arch_disable_smp_support() function - which
has a weak alias for non-x86 architectures and for non-ioapic x86 builds.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
At this time, the PowerNow! driver for K8 uses an experimentally
derived formula to calculate transition latency. The value it
provides is orders of magnitude too large on modern systems.
This patch replaces the formula with ACPI _PSS latency values
for more accuracy and better performance.
I've tested it on two 2nd generation Opteron systems, a 3rd
generation Operton system, and a Turion X2 without seeing any
stability problems.
Signed-off-by: Mark Langsdorf <mark.langsdorf@amd.com>
Signed-off-by: Thomas Renninger <trenn@suse.de>
Signed-off-by: Dave Jones <davej@redhat.com>
There's a small problem with hpet_rtc_reinit function - it checks
for the:
hpet_readl(HPET_COUNTER) - hpet_t1_cmp > 0
to continue increasing both the HPET_T1_CMP (register) and the
hpet_t1_cmp (variable).
But since the HPET_COUNTER is always 32-bit, if the hpet_t1_cmp
is 64-bit this condition will always be FALSE once the latter hits
the 32-bit boundary, and we can have a situation, when we don't
increase the HPET_T1_CMP register high enough.
The result - timer stops ticking, since HPET_T1_CMP becomes less,
than the COUNTER and never increased again.
The solution is (based on Linus's suggestion) to not compare 64-bits
(on 64-bit x86), but to do the comparison on 32-bit signed
integers.
Reported-by: Kirill Korotaev <dev@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch echoes what we already do on 32-bit since
90f7d25c6b, and prints the DMI
product name in show_regs, so that system specific problems can be
easily identified.
Signed-off-by: Kyle McMartin <kyle@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
With oprofile as a module, and unloaded by profiling script,
both oprofile and kerneltop work fine.. unless you leave kerneltop
running when you start profiling, then you may see badness.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
They were long enough set deprecated...
Update Documentation/cpu-freq/users-guide.txt:
The deprecated files listed there seen not to exist for some time anymore
already.
Signed-off-by: Thomas Renninger <trenn@suse.de>
Signed-off-by: Len Brown <len.brown@intel.com>
Impact: reduce kernel BSS size by 7 pages, improve code readability
Two page tables are used in current x86_64 kexec implementation. One
is used to jump from kernel virtual address to identity map address,
the other is used to map all physical memory. In fact, on x86_64,
there is no conflict between kernel virtual address space and physical
memory space, so just one page table is sufficient. The page table
pages used to map control page are dynamically allocated to save
memory if kexec image is not loaded. ASM code used to map control page
is replaced by C code too.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: fix to enable APIC for AMD Fam10h on chipsets with a missing/b0rked
ACPI MP table (MADT)
Booting a 32bit kernel on an AMD Fam10h CPU running on chipsets with
missing/b0rked MP table leads to a hang pretty early in the boot process
due to the APIC not being initialized. Fix that by falling back to the
default APIC base address in 32bit code, as it is done in the 64bit
codepath.
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: Fixes dumpstack and KDB on 64 bits
This re-adds the old stack pointer to the top of the irqstack to help
with unwinding. It was removed in commit d99015b1ab
as part of the save_args out-of-line work.
Both dumpstack and KDB require this information.
Signed-off-by: Martin Hicks <mort@sgi.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Zach says:
> Enable/Disable have no clobbers at all.
> Save clobbers only return value, %eax
> Restore also clobbers nothing.
This is precisely compatible with the calling convention, so we can
just call them directly without wrapping.
(Compile tested only.)
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Fix:
arch/x86/kernel/irqinit_32.c:124: warning: 'smp_intr_init' defined but not used
Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Eric Paris reported:
> I have an hp dl785g5 which is unable to successfully run
> 2.6.29-0.66.rc3.fc11.x86_64 or 2.6.29-rc2-next-20090126. During bootup
> (early in userspace daemons starting) I get the below BUG, which quickly
> renders the machine dead. I assume it is because sparse_irq_lock never
> gets released when the BUG kills that task.
Adjust lock sequence when migrating a descriptor with
CONFIG_NUMA_MIGRATE_IRQ_DESC enabled.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, ds, bts: cleanup/fix DS configuration
ring-buffer: reset timestamps when ring buffer is reset
trace: set max latency variable to zero on default
trace: stop all recording to ring buffer on ftrace_dump
trace: print ftrace_dump at KERN_EMERG log level
ring_buffer: reset write when reserve buffer fail
tracing/function-graph-tracer: fix a regression while suspend to disk
ring-buffer: fix alignment problem
* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86 setup: fix asm constraints in vesa_store_edid
xen: make sysfs files behave as their names suggest
x86: tone down mtrr_trim_uncached_memory() warning
x86: correct the CPUID pattern for MSR_IA32_MISC_ENABLE availability
[
mingo@elte.hu: these fixes are a subset of changes cherry-picked from:
git://git.kernel.org:/pub/scm/linux/kernel/git/jejb/voyager-2.6.git
They fix various problems that recent x86 changes caused in the Voyager
subarchitecture: both APIC changes and cpumask changes and certain
cleanups caused subarch assumptions to break.
Most of these changes are obsolete as the subarch code has been removed
from the x86 development tree - but we merge them upstream to make Voyager
build and boot.
]
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: split out a function, no functional change
Xen needs to be able to access percpu data from very early on. For
various reasons, it cannot also load the gdt at that time. It does,
however, have a pefectly functional gdt at that point, so there's no
pressing need to reload the gdt.
Split the function to load the segment registers off, so Xen can call
it directly.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Impact: cleanup, prepare for xen boot fix.
Xen needs to call this function very early to setup the GDT and
per-cpu segments. Remove the call to smp_processor_id() and just
pass in the cpu number.
Signed-off-by: Brian Gerst <brgerst@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Impact: fix possible tlb mis-flushing on UV
uv_flush_send_and_wait() should return a pointer if the broadcast
remote tlb shootdown requests fail. That causes the conventional IPI
method of shootdown to be used.
Signed-off-by: Cliff Wickman <cpw@sgi.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Impact: Optimization
In the native case, pte_val, make_pte, etc are all just identity
functions, so there's no need to clobber a lot of registers over them.
(This changes the 32-bit callee-save calling convention to return both
EAX and EDX so functions can return 64-bit values.)
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: Optimization
One of the problems with inserting a pile of C calls where previously
there were none is that the register pressure is greatly increased.
The C calling convention says that the caller must expect a certain
set of registers may be trashed by the callee, and that the callee can
use those registers without restriction. This includes the function
argument registers, and several others.
This patch seeks to alleviate this pressure by introducing wrapper
thunks that will do the register saving/restoring, so that the
callsite doesn't need to worry about it, but the callee function can
be conventional compiler-generated code. In many cases (particularly
performance-sensitive cases) the callee will be in assembler anyway,
and need not use the compiler's calling convention.
Standard calling convention is:
arguments return scratch
x86-32 eax edx ecx eax ?
x86-64 rdi rsi rdx rcx rax r8 r9 r10 r11
The thunk preserves all argument and scratch registers. The return
register is not preserved, and is available as a scratch register for
unwrapped callee code (and of course the return value).
Wrapped function pointers are themselves wrapped in a struct
paravirt_callee_save structure, in order to get some warning from the
compiler when functions with mismatched calling conventions are used.
The most common paravirt ops, both statically and dynamically, are
interrupt enable/disable/save/restore, so handle them first. This is
particularly easy since their calls are handled specially anyway.
XXX Deal with VMI. What's their calling convention?
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: Fix latent bug
The clobber is trying to say that anything except RDI is available for
clobbering, but actually clobbers everything. This hasn't mattered
because the clobbers were basically ignored, but subsequent patches
will rely on them.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: Optimization
Several paravirt ops implementations simply return their arguments,
the most obvious being the make_pte/pte_val class of operations on
native.
On 32-bit, the identity function is literally a no-op, as the calling
convention uses the same registers for the first argument and return.
On 64-bit, it can be implemented with a single "mov".
This patch adds special identity functions for 32 and 64 bit argument,
and machinery to recognize them and replace them with either nops or a
mov as appropriate.
At the moment, the only users for the identity functions are the
pagetable entry conversion functions.
The result is a measureable improvement on pagetable-heavy benchmarks
(2-3%, reducing the pvops overhead from 5 to 2%).
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: fix rare crash on 32-bit
The 32-bit APIC drivers had their send_IPI_self vectors set to NULL,
but ioapic_retrigger_irq() depends on it being always set. Fix it.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
just like 64 bit switch from flat logical APIC messages to
flat physical mode automatically.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: 32-bit should use logical version
there are two version: for default_send_IPI_mask_sequence/allbutself
one in ipi.h and one in ipi.c for 32bit
it seems .h version overwrote ipi.c for a while.
restore it so 32 bit could use its old logical version.
also remove dupicated functions in .c
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
