x86: Change x86 to use the generic find_next_bit implementation The versions ... micro-benchmark, that is). Patch is against -x86#testing. It compiles ... size) return __find_first_bit(addr,size); } -/** - * find_next_bit - find the first set bit ...

linux-kernel - Alexander van Heukelum - Mar 9 2008 - 16:01

... .com> wrote: > x86: Change x86 to use the generic find_next_bit implementation > ... too much. Have you tried to build with: CONFIG_CC_OPTIMIZE_FOR_SIZE=y CONFIG_OPTIMIZE_INLINING=y (the latter only available in x86.git) > Patch is against - ...

linux-kernel - Ingo Molnar - Mar 9 2008 - 16:10

... Heukelum wrote: > > x86: Change x86 to use the generic find_next_bit implementation ... set > > (cleared) bit is found, find_next_bit (find_next_zero_bit) returns a > > value outside ... the range [0,size]. The generic version always returns > > exactly size ...

linux-kernel - Alexander van Heukelum - Mar 13 2008 - 10:27

... 's rather convincing. > > the generic version in lib/find_next_bit.c is open-coded ... is probably that the old versions use find_first_bit and find_first_zero_bit, which are ... results) > (the latter only available in x86.git) > > > Patch is against - ...

linux-kernel - Alexander van Heukelum - Mar 9 2008 - 17:13

... Heukelum wrote: > x86: Change x86 to use the generic find_next_bit implementation ... set > (cleared) bit is found, find_next_bit (find_next_zero_bit) returns a > value outside ... .c have some examples. -aneesh -- To unsubscribe from this list: send ...

linux-kernel - Aneesh Kumar K.V - Mar 13 2008 - 08:44

... #define BITOP_WORD(nr) ((nr) / BITS_PER_LONG) > +#undef find_next_bit > +#undef find_next_zero_bit this bit looks weird ... need it for testing? Ingo -- To unsubscribe from this list: send ... the body of a message to majordomo@vger.kernel.org More ...

linux-kernel - Ingo Molnar - Mar 9 2008 - 16:11

... #define BITOP_WORD(nr) ((nr) / BITS_PER_LONG) > > > +#undef find_next_bit > > > +#undef find_next_zero_bit > > > > this bit looks weird ... needs to be pushed into generic code so that 32-bit x86 and ... can make use of it as well. The lib/find_next_bit.c ...

linux-kernel - Ingo Molnar - Mar 9 2008 - 16:51

... BITOP_WORD(nr) ((nr) / BITS_PER_LONG) > > +#undef find_next_bit > > +#undef find_next_zero_bit > > this bit looks ... /asm-x86/bitops_64.h (which gets included). They are used to ... an email service should be -- To unsubscribe from this list: send ...

linux-kernel - Alexander van Heukelum - Mar 9 2008 - 16:31

... writes: > > the generic version in lib/find_next_bit.c is open-coded C ... arch/x86/lib/bitops_32.c mostly use > the special x86 'bit search forward' ( ... cased and optimized for each. -Andi -- To unsubscribe from this list: send the ...

linux-kernel - Andi Kleen - Mar 9 2008 - 17:03

... 's rather convincing. > > > > the generic version in lib/find_next_bit.c is open-coded C ... /x86/lib/bitops_32.c mostly > > use the special x86 'bit search forward' ( ... that it has nontrivial overhead. Ingo -- To unsubscribe from this list: send the ...

linux-kernel - Ingo Molnar - Mar 10 2008 - 02:29

... and no set > (cleared) bit is found, find_next_bit (find_next_zero_bit) returns a > value outside of the ... ]. The generic version always returns > exactly size. With that change it is ... was just a workaround for the x86 quirk. I suspect with such a ...

linux-kernel - Andi Kleen - Mar 9 2008 - 16:28

... > writes: > > ok - but this needs to be solved in a cleaner way. ... to be pushed into generic code so that 32-bit x86 ... too large open coded. -Andi -- To unsubscribe from this list: send ... the body of a message to majordomo@vger.kernel.org More ...

linux-kernel - Andi Kleen - Mar 9 2008 - 17:29

... Heukelum writes: > > Benchmark using: gcc -fomit-frame-pointer -Os. ... what bitmap sizes a real kernel uses (should be easy enough with some ... searches are common for example ... -Andi -- To unsubscribe from this list: send the ...

linux-kernel - Andi Kleen - Mar 9 2008 - 17:31

... > I had a patchkit some time ago to special case the max_bit <= 63 case ... never ... at some point, but fixed later. -Andi -- To unsubscribe from this list: send the line " ... " in the body of a message to majordomo@vger.kernel.org More majordomo info ...

linux-kernel - Andi Kleen - Mar 9 2008 - 17:32