apply_alternatives uses memcpy() to apply alternatives. Which has the
unfortunate effect that while applying memcpy alternative to memcpy
itself it tries to overwrite itself with nops - which causes #UD fault
as it overwrites half of an instruction in copy loop, and from this
point on only possible outcome is triplefault and reboot.
So let's overwrite only first two instructions of memcpy - as long as
the main memcpy loop is not in first two bytes it will work fine.
Signed-off-by: Petr Vandrovec <petr@vandrovec.name>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
There was OpenVZ specific bug rendering some cpufreq drivers unusable on SMP.
In short, when cpufreq code thinks it confined itself to needed cpu by means
of set_cpus_allowed() to execute rdmsr, some "virtual cpu" feature can migrate
process to anywhere. This triggers bugons and does wrong things in general.
This got fixed by introducing rdmsr_on_cpu and wrmsr_on_cpu executing rdmsr
and wrmsr on given physical cpu by means of smp_call_function_single().
Dave Jones mentioned cpufreq might be not only user of rdmsr_on_cpu() and
wrmsr_on_cpu(), so I'm putting them into arch/{i386,x86_64}/lib/ .
Signed-off-by: Alexey Dobriyan <adobriyan@openvz.org>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Dave Jones <davej@redhat.com>
This does user copies in fs write() into the page cache with write combining.
This pushes the destination out of the CPU's cache, but allows higher bandwidth
in some case.
The theory is that the page cache data is usually not touched by the
CPU again and it's better to not pollute the cache with it. Also it is a little
faster.
Signed-off-by: Andi Kleen <ak@suse.de>
Port two patches from i386 to x86_64 delay.c to make sure all rounding is done
upward instead of downward.
There is no sign in commit messages that the mismatch was done on purpose, and
"delay() guarantees sleeping at least for the specified time" is still a valid
rule IMHO.
The original x86 patches are both from pre-GIT era, i.e.:
"[PATCH] round up in __udelay()" in commit
54c7e1f5cc6771ff644d7bc21a2b829308bd126f
"[PATCH] add 1 in __const_udelay()" in commit
42c77a9801b8877d8b90f65f75db758822a0bccc
(both commits are from converted BK repository to x86_64).
AK: fixed gcc warning
linux/arch/x86_64/lib/delay.c:43: warning: suggest parentheses around + or - inside shift
(did this actually work?)
Signed-off-by: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Signed-off-by: Andi Kleen <ak@suse.de>
* sanitize prototypes, annotate
* usual ntohs->shift
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
- Don't zero for __copy_from_user_inatomic following i386.
This will prevent spurious zeros for parallel file system writers when
one does a exception
- The string instruction version didn't zero the output on
exception. Oops.
Also I cleaned up the code a bit while I was at it and added a minor
optimization to the string instruction path.
Signed-off-by: Andi Kleen <ak@suse.de>
Add unwind annotations to arch/x86_64/lib/*.S, and also use the macros
provided by linux/linkage.h where-ever possible.
Some of the alternative instructions handling needed to be adjusted so
that the replacement code would also have valid unwind information.
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Andi Kleen <ak@suse.de>
- Move the slow path fallbacks to their own assembly files
This makes them much easier to read and is needed for the next change.
- Add CFI annotations for unwinding (XXX need review)
- Remove constant case which can never happen with out of line spinlocks
- Use patchable LOCK prefixes
- Don't use lock sections anymore for inline code because they can't
be expressed by the unwinder (this adds one taken jump to the lock
fast path)
Cc: jbeulich@novell.com
Signed-off-by: Andi Kleen <ak@suse.de>
- Use tail call from clear_user to __clear_user to save some code size
- Use standard memcpy for forward memmove
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Only exports for assembler files are left in x8664_ksyms.c
Originally inspired by a patch from Al Viro
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
They cause quite bad performance regressions on Netburst
This is temporary until we can get new optimized functions
for these CPUs.
This undoes changes that were done in 2.6.15 and in 2.6.16-rc1,
essentially bringing the code back to 2.6.14 level. Only change
is I renamed the X86_FEATURE_K8_C flag to X86_FEATURE_REP_GOOD
and fixed the check for the flag and also fixed some comments.
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This assembly version is measurably faster than the generic version in
lib/iomap_copy.c.
Signed-off-by: Bryan O'Sullivan <bos@pathscale.com>
Cc: Andi Kleen <ak@muc.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
When building in a separate objtree, file names produced by BUG() & Co. can
get fairly long; printing only the first 50 characters may thus result in
(almost) no useful information. The following change makes it so that rather
the last 50 characters of the filename get printed.
Signed-Off-By: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
- Remove optimization for old B stepping Opteron
- Make the fast path for copies with a multiple of eight length faster.
- Minor instruction rearrangement to hopefully avoid a pipeline
stall or two.
- Add comment about errata to consider.
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The attempt to avoid overflow in __delay caused varying precision
on different CPUs depending on differences in the CPU speed.
We should be able to do this multiplication with out overflowing
provided the
cpu is running at less than about 128 GHz. xloops < 20000 * 0x10c6.
loops_per_jiffy * HZ <= cpu_clock_speed. So if the cpu clock speed
< 2^64/(20000 * 0x10c6) = 2^64/ 51E6CC0 < 2^64/2^27 = 2^37 = 128G we
will not overflow the calculation.
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
B stepping were the first shipping Opterons. memcpy/memset/copy_page/
clear_page had special optimized version for them. These are really
old and in the minority now and the difference to the generic versions
(using rep microcode) is not that big anyways. So just remove them.
TODO: figure out optimized versions for Intel Netburst based EM64T
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This fixes the x86-64 find_[first|next]_zero_bit() function for the
end-of-range case. It didn't test for a zero size, and the "rep scas"
would do entirely the wrong thing.
Signed-off-by: Alexandre Oliva <oliva@lsd.ic.unicamp.br>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Several implementations were essentialy a common piece of C code using
the cmpxchg() macro. Put the implementation in one spot that everyone
can share, and convert sparc64 over to using this.
Alpha is the lone arch-specific implementation, which codes up a
special fast path for the common case in order to avoid GP reloading
which a pure C version would require.
Signed-off-by: David S. Miller <davem@davemloft.net>
I was observing reproducible crashes on the "movw %bx,(%rsi)" instruction
below while a process in a recvfrom() system call was copying packet data
to user space. The patch below fixes the exception table and causes the
crash to no longer reproduce. Please apply.
Acked-by: Andi Kleen <ak@muc.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
