Speeding Up The Linux Kernel With Transparent Hugepage Support

[Smorg]

Personlly, since the latest test updates of BFS, I'm using 2.6.36.1 but with the most recent CK "test" patches, because I found the "automatic group scheduler" (AGS) patch a slower than BFS with very heavy I/O loads (eg. when I was moving large files between different partitions and at the same time browsing the internet, the page loading was very slow (but not that much, in comparision with BFS); even in more slow I/O loads, these "AGS" patches were moving the files "by segments", while on BFS the files were being moved more "smoothly").

These new patches you put there, currently only work with 2.6.37-rc kernels (I tried to create a BFS "testing" + Huge Transparent Page patched kernel, but it was giving me failed non-reversible hunks...)

Cheers!

BFS is the exact opposite for me. Interactivity on BFS under heavy IO load is abysmal in comparison to auto group scheduler when doing ccache compilation on many threads while browsing under conditions that I wouldn't call "unrealistic workloads" as CK puts it. BFS seems like what you want if you care only about throughput and not about not about interactivity. (tweaking priorities with schedtool doesn't help this one iota btw).

I'm on the zen kernel which unfortunately no longer maintains their unstable branch so I'll probably have to wait til this is either backported or 2.6.37 goes final.

[evolution]

BFS is the exact opposite for me. Interactivity on BFS under heavy IO load is abysmal in comparison to auto group scheduler when doing ccache compilation on many threads while browsing under conditions that I wouldn't call "unrealistic workloads" as CK puts it. BFS seems like what you want if you care only about throughput and not about not about interactivity. (tweaking priorities with schedtool doesn't help this one iota btw).

I'm on the zen kernel which unfortunately no longer maintains their unstable branch so I'll probably have to wait til this is either backported or 2.6.37 goes final.

In the last weeks, I've been compiling my own kernel builds. I used Zen-Kernel (a very well optimized kernel for desktops ) until last month because their BFS patches are old (click here and here and see the comments).

Futhermore, those BFS patches added to Zen Kernel above were removed in recent CK patchsets...

Cheers

[tettamanti]

According to the kernel docs, in most cases memory isn't wasted. If you were to alloc 4MB of ram and only use 1 byte it would waste ram. If you alloc 4K of ram and and use just one byte it would be no worse than it is now.

Well it's worse because unless you're using the latest a greatest CPU the number of TLB entries for hugepages is pretty small:

TLB info
 Instruction TLB: 4x 4MB page entries, or 8x 2MB pages entries, 4-way associative
 Instruction TLB: 4K pages, 4-way associative, 128 entries.
 Data TLB: 4MB pages, 4-way associative, 32 entries
 L1 Data TLB: 4KB pages, 4-way set associative, 16 entries
 L1 Data TLB: 4MB pages, 4-way set associative, 16 entries
 Data TLB: 4K pages, 4-way associative, 256 entries.

This is a C2D T7500; so wasting a 4MB page creates more TLB pressure than wasting a 4KB page.

[Obscene_CNN]

Well it's worse because unless you're using the latest a greatest CPU the number of TLB entries for hugepages is pretty small:

Actually until the latest x86 processors the TLB cache entries for huge and small pages were the same silcon (why they would change it and split them up is a mystery to me).

TLB info
 Instruction TLB: 4x 4MB page entries, or 8x 2MB pages entries, 4-way associative
 Instruction TLB: 4K pages, 4-way associative, 128 entries.
 Data TLB: 4MB pages, 4-way associative, 32 entries
 L1 Data TLB: 4KB pages, 4-way set associative, 16 entries
 L1 Data TLB: 4MB pages, 4-way set associative, 16 entries
 Data TLB: 4K pages, 4-way associative, 256 entries.

This is a C2D T7500; so wasting a 4MB page creates more TLB pressure than wasting a 4KB page.

Well sloppy software that allocates 2M of ram just to use 4k or less of it should probably be rewritten anyway. Fortunately from what I have seen so far this is not the norm.

Look at it this way, it would take 512 4K TLB entries to cover a single 2MB page. So it relieves pressure on the small page TLB cache. Further more the frequency of a huge page miss is far lower than that of a 4K page miss. At least youlbe using the huge TLB cache for something instead of it just sitting around doing nothing.

[squirrl]

Try copying a 9gig file across HDD's in windows, then try todo something CPU/IO heavy, watch as explorer crashes or everything slows down to 386 speeds.

This of cause doesnt always happen, but I have seen it often enough with XP/Vista/7.

I think window's problem is that it spends to much time trying to guess how long its going to take, and even then it never gets it right

Works good for me. I think Microsoft uses the FreeBSD scheduler.

[toplinuxsir]

I dowloaded the linux-2.6.37-rc4 and transparent_hugepage-31.bz2 file from www.kernel.org, and extract all the source file to apply the patch
But I got the below output(errors) :

root@lenovo:/usr/src/linux-2.6.37-rc4# patch -p1 < transparent_hugepage-31
patching file Documentation/vm/transhuge.txt
patching file arch/alpha/include/asm/mman.h
patching file arch/mips/include/asm/mman.h
patching file arch/parisc/include/asm/mman.h
patching file arch/powerpc/mm/gup.c
patching file arch/x86/include/asm/kvm_host.h
Hunk #1 succeeded at 788 (offset 39 lines).
patching file arch/x86/include/asm/paravirt.h
Hunk #1 succeeded at 435 (offset -5 lines).
Hunk #2 succeeded at 447 (offset -5 lines).
Hunk #3 succeeded at 554 (offset -5 lines).
patching file arch/x86/include/asm/paravirt_types.h
Hunk #1 succeeded at 265 (offset -1 lines).
patching file arch/x86/include/asm/pgtable-2level.h
patching file arch/x86/include/asm/pgtable-3level.h
patching file arch/x86/include/asm/pgtable.h
Hunk #1 succeeded at 35 (offset 2 lines).
Hunk #2 succeeded at 60 (offset 2 lines).
Hunk #3 succeeded at 97 (offset 2 lines).
Hunk #4 succeeded at 150 (offset 2 lines).
Hunk #5 succeeded at 241 (offset 2 lines).
Hunk #6 succeeded at 330 (offset 2 lines).
Hunk #7 succeeded at 383 (offset 2 lines).
Hunk #8 succeeded at 434 (offset 2 lines).
Hunk #9 succeeded at 608 (offset 2 lines).
Hunk #10 succeeded at 705 with fuzz 2 (offset 4 lines).
patching file arch/x86/include/asm/pgtable_64.h
patching file arch/x86/include/asm/pgtable_types.h
patching file arch/x86/kernel/paravirt.c
Hunk #1 succeeded at 421 (offset -1 lines).
patching file arch/x86/kernel/tboot.c
patching file arch/x86/kernel/vm86_32.c
patching file arch/x86/kvm/mmu.c
Hunk #1 succeeded at 543 (offset 45 lines).
Hunk #2 succeeded at 960 (offset 39 lines).
Hunk #3 succeeded at 1009 (offset 39 lines).
Hunk #4 succeeded at 2328 (offset 150 lines).
Hunk #5 succeeded at 2341 (offset 150 lines).
Hunk #6 succeeded at 2646 (offset 239 lines).
patching file arch/x86/kvm/paging_tmpl.h
Hunk #1 succeeded at 561 (offset 109 lines).
patching file arch/x86/mm/gup.c
patching file arch/x86/mm/pgtable.c
Hunk #1 succeeded at 320 (offset 10 lines).
Hunk #2 succeeded at 354 (offset 10 lines).
Hunk #3 succeeded at 383 (offset 10 lines).
patching file arch/xtensa/include/asm/mman.h
patching file drivers/base/node.c
patching file fs/Kconfig
Hunk #1 succeeded at 147 (offset 2 lines).
patching file fs/exec.c
Hunk #1 FAILED at 54.
Hunk #2 succeeded at 248 (offset 7 lines).
Hunk #3 succeeded at 519 (offset 7 lines).
Hunk #4 succeeded at 533 (offset 7 lines).
Hunk #5 succeeded at 595 (offset 7 lines).
Hunk #6 succeeded at 665 (offset 7 lines).
Hunk #7 succeeded at 679 (offset 7 lines).
1 out of 7 hunks FAILED -- saving rejects to file fs/exec.c.rej
patching file fs/proc/meminfo.c
patching file fs/proc/page.c
patching file include/asm-generic/mman-common.h
patching file include/asm-generic/pgtable.h
Hunk #7 succeeded at 470 (offset 4 lines).
patching file include/linux/compaction.h
patching file include/linux/gfp.h
Hunk #1 succeeded at 81 with fuzz 2 (offset 21 lines).
Hunk #2 succeeded at 108 (offset 21 lines).
Hunk #3 succeeded at 330 (offset 21 lines).
patching file include/linux/huge_mm.h
patching file include/linux/khugepaged.h
patching file include/linux/kvm_host.h
Hunk #1 succeeded at 295 (offset 4 lines).
Hunk #2 succeeded at 326 (offset 6 lines).
patching file include/linux/memory_hotplug.h
patching file include/linux/mm.h
Hunk #4 succeeded at 244 (offset 1 line).
Hunk #5 succeeded at 308 (offset 1 line).
Hunk #6 succeeded at 356 (offset 1 line).
Hunk #7 succeeded at 384 (offset 1 line).
Hunk #8 succeeded at 442 (offset 1 line).
Hunk #9 succeeded at 1143 (offset 19 lines).
Hunk #10 succeeded at 1213 (offset 19 lines).
Hunk #11 succeeded at 1497 (offset 21 lines).
Hunk #12 succeeded at 1601 (offset 21 lines).
patching file include/linux/mm_inline.h
patching file include/linux/mm_types.h
Hunk #1 succeeded at 183 with fuzz 2 (offset -127 lines).
patching file include/linux/mmu_notifier.h
patching file include/linux/mmzone.h
Hunk #1 succeeded at 114 (offset 2 lines).
patching file include/linux/page-flags.h
patching file include/linux/sched.h
Hunk #1 succeeded at 431 (offset 3 lines).
patching file include/linux/swap.h
patching file kernel/fork.c
Hunk #1 FAILED at 65.
Hunk #2 succeeded at 328 (offset 1 line).
Hunk #3 succeeded at 530 (offset 2 lines).
Hunk #4 succeeded at 547 (offset 2 lines).
Hunk #5 succeeded at 674 (offset 2 lines).
1 out of 5 hunks FAILED -- saving rejects to file kernel/fork.c.rej
patching file kernel/futex.c
Hunk #1 succeeded at 219 (offset 1 line).
Hunk #2 succeeded at 251 (offset 1 line).
Hunk #3 succeeded at 301 (offset 1 line).
patching file mm/Kconfig
Hunk #1 FAILED at 301.
1 out of 1 hunk FAILED -- saving rejects to file mm/Kconfig.rej
patching file mm/Makefile
Hunk #1 succeeded at 42 (offset -5 lines).
patching file mm/compaction.c
patching file mm/huge_memory.c
patching file mm/hugetlb.c
Reversed (or previously applied) patch detected! Assume -R? [n]
Apply anyway? [n] y
Hunk #1 FAILED at 394.
Hunk #2 FAILED at 2383.
Hunk #3 succeeded at 2558 (offset 71 lines).
2 out of 3 hunks FAILED -- saving rejects to file mm/hugetlb.c.rej
patching file mm/ksm.c
patching file mm/madvise.c
patching file mm/memcontrol.c
Hunk #1 succeeded at 1027 (offset 86 lines).
Hunk #2 succeeded at 1091 (offset 86 lines).
Hunk #3 succeeded at 1666 (offset 156 lines).
Hunk #4 succeeded at 1891 (offset 196 lines).
Hunk #5 succeeded at 1923 (offset 196 lines).
Hunk #6 succeeded at 1947 (offset 196 lines).
Hunk #7 succeeded at 1988 (offset 196 lines).
Hunk #8 succeeded at 2009 (offset 196 lines).
Hunk #9 succeeded at 2038 (offset 196 lines).
Hunk #10 succeeded at 2095 (offset 196 lines).
Hunk #11 succeeded at 2123 (offset 196 lines).
Hunk #12 succeeded at 2196 (offset 196 lines).
Hunk #13 succeeded at 2257 (offset 196 lines).
Hunk #14 succeeded at 2285 (offset 196 lines).
Hunk #15 succeeded at 2296 (offset 196 lines).
Hunk #16 succeeded at 2309 (offset 196 lines).
Hunk #17 succeeded at 2414 (offset 196 lines).
Hunk #18 succeeded at 2436 (offset 196 lines).
Hunk #19 succeeded at 2485 (offset 196 lines).
Hunk #20 succeeded at 2525 (offset 196 lines).
Hunk #21 succeeded at 2544 (offset 196 lines).
Hunk #22 succeeded at 2556 (offset 196 lines).
Hunk #23 succeeded at 2592 (offset 196 lines).
Hunk #24 succeeded at 2614 (offset 196 lines).
Hunk #25 succeeded at 2809 (offset 196 lines).
Hunk #26 succeeded at 2859 (offset 196 lines).
Hunk #27 succeeded at 2886 (offset 196 lines).
Hunk #28 succeeded at 4497 (offset 194 lines).
Hunk #29 succeeded at 4660 (offset 194 lines).
Hunk #30 succeeded at 4827 (offset 211 lines).
patching file mm/memory-failure.c
Hunk #1 succeeded at 385 (offset 4 lines).
patching file mm/memory.c
Hunk #9 succeeded at 1648 (offset 1 line).
Hunk #10 succeeded at 1870 (offset 1 line).
Hunk #11 succeeded at 2101 (offset 1 line).
Hunk #12 succeeded at 3187 (offset 8 lines).
Hunk #13 succeeded at 3268 (offset 8 lines).
Hunk #14 succeeded at 3439 (offset 8 lines).
Hunk #15 succeeded at 3680 (offset 19 lines).
patching file mm/memory_hotplug.c
patching file mm/mempolicy.c
Hunk #2 succeeded at 1794 (offset 9 lines).
Hunk #3 succeeded at 1803 (offset 9 lines).
Hunk #4 succeeded at 1817 (offset 9 lines).
Hunk #5 succeeded at 1830 (offset 9 lines).
Hunk #6 succeeded at 1839 (offset 9 lines).
Hunk #7 succeeded at 1848 (offset 9 lines).
patching file mm/migrate.c
Hunk #1 FAILED at 104.
Hunk #2 succeeded at 630 (offset 65 lines).
Hunk #3 succeeded at 1043 with fuzz 1 (offset 196 lines).
1 out of 3 hunks FAILED -- saving rejects to file mm/migrate.c.rej
patching file mm/mincore.c
patching file mm/mmap.c
Hunk #1 FAILED at 28.
Hunk #2 succeeded at 588 (offset 1 line).
Hunk #3 succeeded at 817 (offset 1 line).
Hunk #4 succeeded at 836 (offset 1 line).
Hunk #5 succeeded at 1765 (offset 2 lines).
Hunk #6 succeeded at 1813 (offset 2 lines).
1 out of 6 hunks FAILED -- saving rejects to file mm/mmap.c.rej
patching file mm/mmu_notifier.c
patching file mm/mprotect.c
patching file mm/mremap.c
patching file mm/page_alloc.c
Hunk #1 succeeded at 352 (offset 1 line).
Hunk #2 succeeded at 444 (offset 1 line).
Hunk #3 succeeded at 478 (offset 1 line).
Hunk #4 succeeded at 647 (offset 1 line).
Hunk #5 succeeded at 1941 with fuzz 1 (offset 1 line).
Hunk #6 succeeded at 2001 (offset 1 line).
Hunk #7 succeeded at 5516 (offset 57 lines).
Hunk #8 succeeded at 5563 (offset 57 lines).
patching file mm/pagewalk.c
patching file mm/rmap.c
Hunk #2 succeeded at 364 (offset 2 lines).
Hunk #3 succeeded at 439 (offset 2 lines).
Hunk #4 succeeded at 495 (offset 2 lines).
Hunk #5 succeeded at 514 (offset 2 lines).
Hunk #6 succeeded at 882 (offset -7 lines).
Hunk #7 succeeded at 916 (offset -7 lines).
Hunk #8 succeeded at 972 (offset -7 lines).
Hunk #9 succeeded at 1232 (offset -7 lines).
Hunk #10 succeeded at 1260 (offset -7 lines).
Hunk #11 succeeded at 1402 (offset -7 lines).
patching file mm/sparse.c
patching file mm/swap.c
Hunk #3 succeeded at 465 (offset 1 line).
patching file mm/swap_state.c
patching file mm/swapfile.c
Hunk #1 succeeded at 964 (offset 5 lines).
patching file mm/vmscan.c
Hunk #2 FAILED at 79.
Hunk #3 FAILED at 580.
Hunk #4 succeeded at 1022 (offset 59 lines).
Hunk #5 succeeded at 1080 (offset 59 lines).
Hunk #6 succeeded at 1135 (offset 64 lines).
Hunk #7 succeeded at 1253 (offset 64 lines).
Hunk #8 FAILED at 1249.
Hunk #9 FAILED at 1308.
Hunk #10 FAILED at 1321.
Hunk #11 succeeded at 1462 (offset 61 lines).
Hunk #12 succeeded at 1530 (offset 61 lines).
Hunk #13 FAILED at 1724.
Hunk #14 FAILED at 1753.
Hunk #15 succeeded at 2251 (offset 54 lines).
Hunk #16 succeeded at 2281 (offset 54 lines).
7 out of 16 hunks FAILED -- saving rejects to file mm/vmscan.c.rej
patching file mm/vmstat.c
Hunk #1 succeeded at 761 (offset 8 lines).
patching file virt/kvm/iommu.c
patching file virt/kvm/kvm_main.c
Hunk #4 succeeded at 929 (offset -2 lines).
Hunk #5 succeeded at 938 (offset -2 lines).
Hunk #6 FAILED at 991.
1 out of 6 hunks FAILED -- saving rejects to file virt/kvm/kvm_main.c.rej

I don not know the phoronix how to complete the test in the article ?
Can anyone give me some ideas to resolve the problem ?
Thanks !

[toplinuxsir]

I download the linux-2.6.37-rc4 and transparent_hugepage-31

but I got the errors , why ? How the phoronix finish his testing about the article (Speeding up linux ......)
root@lenovo:/usr/src/linux-2.6.37-rc4# patch -p1 < transparent_hugepage-31
patching file Documentation/vm/transhuge.txt
patching file arch/alpha/include/asm/mman.h
patching file arch/mips/include/asm/mman.h
patching file arch/parisc/include/asm/mman.h
patching file arch/powerpc/mm/gup.c
patching file arch/x86/include/asm/kvm_host.h
Hunk #1 succeeded at 788 (offset 39 lines).
patching file arch/x86/include/asm/paravirt.h
Hunk #1 succeeded at 435 (offset -5 lines).
Hunk #2 succeeded at 447 (offset -5 lines).
Hunk #3 succeeded at 554 (offset -5 lines).
patching file arch/x86/include/asm/paravirt_types.h
Hunk #1 succeeded at 265 (offset -1 lines).
patching file arch/x86/include/asm/pgtable-2level.h
patching file arch/x86/include/asm/pgtable-3level.h
patching file arch/x86/include/asm/pgtable.h
Hunk #1 succeeded at 35 (offset 2 lines).
Hunk #2 succeeded at 60 (offset 2 lines).
Hunk #3 succeeded at 97 (offset 2 lines).
Hunk #4 succeeded at 150 (offset 2 lines).
Hunk #5 succeeded at 241 (offset 2 lines).
Hunk #6 succeeded at 330 (offset 2 lines).
Hunk #7 succeeded at 383 (offset 2 lines).
Hunk #8 succeeded at 434 (offset 2 lines).
Hunk #9 succeeded at 608 (offset 2 lines).
Hunk #10 succeeded at 705 with fuzz 2 (offset 4 lines).
patching file arch/x86/include/asm/pgtable_64.h
patching file arch/x86/include/asm/pgtable_types.h
patching file arch/x86/kernel/paravirt.c
Hunk #1 succeeded at 421 (offset -1 lines).
patching file arch/x86/kernel/tboot.c
patching file arch/x86/kernel/vm86_32.c
patching file arch/x86/kvm/mmu.c
Hunk #1 succeeded at 543 (offset 45 lines).
Hunk #2 succeeded at 960 (offset 39 lines).
Hunk #3 succeeded at 1009 (offset 39 lines).
Hunk #4 succeeded at 2328 (offset 150 lines).
Hunk #5 succeeded at 2341 (offset 150 lines).
Hunk #6 succeeded at 2646 (offset 239 lines).
patching file arch/x86/kvm/paging_tmpl.h
Hunk #1 succeeded at 561 (offset 109 lines).
patching file arch/x86/mm/gup.c
patching file arch/x86/mm/pgtable.c
Hunk #1 succeeded at 320 (offset 10 lines).
Hunk #2 succeeded at 354 (offset 10 lines).
Hunk #3 succeeded at 383 (offset 10 lines).
patching file arch/xtensa/include/asm/mman.h
patching file drivers/base/node.c
patching file fs/Kconfig
Hunk #1 succeeded at 147 (offset 2 lines).
patching file fs/exec.c
Hunk #1 FAILED at 54.
Hunk #2 succeeded at 248 (offset 7 lines).
Hunk #3 succeeded at 519 (offset 7 lines).
Hunk #4 succeeded at 533 (offset 7 lines).
Hunk #5 succeeded at 595 (offset 7 lines).
Hunk #6 succeeded at 665 (offset 7 lines).
Hunk #7 succeeded at 679 (offset 7 lines).
1 out of 7 hunks FAILED -- saving rejects to file fs/exec.c.rej
patching file fs/proc/meminfo.c
patching file fs/proc/page.c
patching file include/asm-generic/mman-common.h
patching file include/asm-generic/pgtable.h
Hunk #7 succeeded at 470 (offset 4 lines).
patching file include/linux/compaction.h
patching file include/linux/gfp.h
Hunk #1 succeeded at 81 with fuzz 2 (offset 21 lines).
Hunk #2 succeeded at 108 (offset 21 lines).
Hunk #3 succeeded at 330 (offset 21 lines).
patching file include/linux/huge_mm.h
patching file include/linux/khugepaged.h
patching file include/linux/kvm_host.h
Hunk #1 succeeded at 295 (offset 4 lines).
Hunk #2 succeeded at 326 (offset 6 lines).
patching file include/linux/memory_hotplug.h
patching file include/linux/mm.h
Hunk #4 succeeded at 244 (offset 1 line).
Hunk #5 succeeded at 308 (offset 1 line).
Hunk #6 succeeded at 356 (offset 1 line).
Hunk #7 succeeded at 384 (offset 1 line).
Hunk #8 succeeded at 442 (offset 1 line).
Hunk #9 succeeded at 1143 (offset 19 lines).
Hunk #10 succeeded at 1213 (offset 19 lines).
Hunk #11 succeeded at 1497 (offset 21 lines).
Hunk #12 succeeded at 1601 (offset 21 lines).
patching file include/linux/mm_inline.h
patching file include/linux/mm_types.h
Hunk #1 succeeded at 183 with fuzz 2 (offset -127 lines).
patching file include/linux/mmu_notifier.h
patching file include/linux/mmzone.h
Hunk #1 succeeded at 114 (offset 2 lines).
patching file include/linux/page-flags.h
patching file include/linux/sched.h
Hunk #1 succeeded at 431 (offset 3 lines).
patching file include/linux/swap.h
patching file kernel/fork.c
Hunk #1 FAILED at 65.
Hunk #2 succeeded at 328 (offset 1 line).
Hunk #3 succeeded at 530 (offset 2 lines).
Hunk #4 succeeded at 547 (offset 2 lines).
Hunk #5 succeeded at 674 (offset 2 lines).
1 out of 5 hunks FAILED -- saving rejects to file kernel/fork.c.rej
patching file kernel/futex.c
Hunk #1 succeeded at 219 (offset 1 line).
Hunk #2 succeeded at 251 (offset 1 line).
Hunk #3 succeeded at 301 (offset 1 line).
patching file mm/Kconfig
Hunk #1 FAILED at 301.
1 out of 1 hunk FAILED -- saving rejects to file mm/Kconfig.rej
patching file mm/Makefile
Hunk #1 succeeded at 42 (offset -5 lines).
patching file mm/compaction.c
patching file mm/huge_memory.c
patching file mm/hugetlb.c
Reversed (or previously applied) patch detected! Assume -R? [n]
Apply anyway? [n] y
Hunk #1 FAILED at 394.
Hunk #2 FAILED at 2383.
Hunk #3 succeeded at 2558 (offset 71 lines).
2 out of 3 hunks FAILED -- saving rejects to file mm/hugetlb.c.rej
patching file mm/ksm.c
patching file mm/madvise.c
patching file mm/memcontrol.c
Hunk #1 succeeded at 1027 (offset 86 lines).
Hunk #2 succeeded at 1091 (offset 86 lines).
Hunk #3 succeeded at 1666 (offset 156 lines).
Hunk #4 succeeded at 1891 (offset 196 lines).
Hunk #5 succeeded at 1923 (offset 196 lines).
Hunk #6 succeeded at 1947 (offset 196 lines).
Hunk #7 succeeded at 1988 (offset 196 lines).
Hunk #8 succeeded at 2009 (offset 196 lines).
Hunk #9 succeeded at 2038 (offset 196 lines).
Hunk #10 succeeded at 2095 (offset 196 lines).
Hunk #11 succeeded at 2123 (offset 196 lines).
Hunk #12 succeeded at 2196 (offset 196 lines).
Hunk #13 succeeded at 2257 (offset 196 lines).
Hunk #14 succeeded at 2285 (offset 196 lines).
Hunk #15 succeeded at 2296 (offset 196 lines).
Hunk #16 succeeded at 2309 (offset 196 lines).
Hunk #17 succeeded at 2414 (offset 196 lines).
Hunk #18 succeeded at 2436 (offset 196 lines).
Hunk #19 succeeded at 2485 (offset 196 lines).
Hunk #20 succeeded at 2525 (offset 196 lines).
Hunk #21 succeeded at 2544 (offset 196 lines).
Hunk #22 succeeded at 2556 (offset 196 lines).
Hunk #23 succeeded at 2592 (offset 196 lines).
Hunk #24 succeeded at 2614 (offset 196 lines).
Hunk #25 succeeded at 2809 (offset 196 lines).
Hunk #26 succeeded at 2859 (offset 196 lines).
Hunk #27 succeeded at 2886 (offset 196 lines).
Hunk #28 succeeded at 4497 (offset 194 lines).
Hunk #29 succeeded at 4660 (offset 194 lines).
Hunk #30 succeeded at 4827 (offset 211 lines).
patching file mm/memory-failure.c
Hunk #1 succeeded at 385 (offset 4 lines).
patching file mm/memory.c
Hunk #9 succeeded at 1648 (offset 1 line).
Hunk #10 succeeded at 1870 (offset 1 line).
Hunk #11 succeeded at 2101 (offset 1 line).
Hunk #12 succeeded at 3187 (offset 8 lines).
Hunk #13 succeeded at 3268 (offset 8 lines).
Hunk #14 succeeded at 3439 (offset 8 lines).
Hunk #15 succeeded at 3680 (offset 19 lines).
patching file mm/memory_hotplug.c
patching file mm/mempolicy.c
Hunk #2 succeeded at 1794 (offset 9 lines).
Hunk #3 succeeded at 1803 (offset 9 lines).
Hunk #4 succeeded at 1817 (offset 9 lines).
Hunk #5 succeeded at 1830 (offset 9 lines).
Hunk #6 succeeded at 1839 (offset 9 lines).
Hunk #7 succeeded at 1848 (offset 9 lines).
patching file mm/migrate.c
Hunk #1 FAILED at 104.
Hunk #2 succeeded at 630 (offset 65 lines).
Hunk #3 succeeded at 1043 with fuzz 1 (offset 196 lines).
1 out of 3 hunks FAILED -- saving rejects to file mm/migrate.c.rej
patching file mm/mincore.c
patching file mm/mmap.c
Hunk #1 FAILED at 28.
Hunk #2 succeeded at 588 (offset 1 line).
Hunk #3 succeeded at 817 (offset 1 line).
Hunk #4 succeeded at 836 (offset 1 line).
Hunk #5 succeeded at 1765 (offset 2 lines).
Hunk #6 succeeded at 1813 (offset 2 lines).
1 out of 6 hunks FAILED -- saving rejects to file mm/mmap.c.rej
patching file mm/mmu_notifier.c
patching file mm/mprotect.c
patching file mm/mremap.c
patching file mm/page_alloc.c
Hunk #1 succeeded at 352 (offset 1 line).
Hunk #2 succeeded at 444 (offset 1 line).
Hunk #3 succeeded at 478 (offset 1 line).
Hunk #4 succeeded at 647 (offset 1 line).
Hunk #5 succeeded at 1941 with fuzz 1 (offset 1 line).
Hunk #6 succeeded at 2001 (offset 1 line).
Hunk #7 succeeded at 5516 (offset 57 lines).
Hunk #8 succeeded at 5563 (offset 57 lines).
patching file mm/pagewalk.c
patching file mm/rmap.c
Hunk #2 succeeded at 364 (offset 2 lines).
Hunk #3 succeeded at 439 (offset 2 lines).
Hunk #4 succeeded at 495 (offset 2 lines).
Hunk #5 succeeded at 514 (offset 2 lines).
Hunk #6 succeeded at 882 (offset -7 lines).
Hunk #7 succeeded at 916 (offset -7 lines).
Hunk #8 succeeded at 972 (offset -7 lines).
Hunk #9 succeeded at 1232 (offset -7 lines).
Hunk #10 succeeded at 1260 (offset -7 lines).
Hunk #11 succeeded at 1402 (offset -7 lines).
patching file mm/sparse.c
patching file mm/swap.c
Hunk #3 succeeded at 465 (offset 1 line).
patching file mm/swap_state.c
patching file mm/swapfile.c
Hunk #1 succeeded at 964 (offset 5 lines).
patching file mm/vmscan.c
Hunk #2 FAILED at 79.
Hunk #3 FAILED at 580.
Hunk #4 succeeded at 1022 (offset 59 lines).
Hunk #5 succeeded at 1080 (offset 59 lines).
Hunk #6 succeeded at 1135 (offset 64 lines).
Hunk #7 succeeded at 1253 (offset 64 lines).
Hunk #8 FAILED at 1249.
Hunk #9 FAILED at 1308.
Hunk #10 FAILED at 1321.
Hunk #11 succeeded at 1462 (offset 61 lines).
Hunk #12 succeeded at 1530 (offset 61 lines).
Hunk #13 FAILED at 1724.
Hunk #14 FAILED at 1753.
Hunk #15 succeeded at 2251 (offset 54 lines).
Hunk #16 succeeded at 2281 (offset 54 lines).
7 out of 16 hunks FAILED -- saving rejects to file mm/vmscan.c.rej
patching file mm/vmstat.c
Hunk #1 succeeded at 761 (offset 8 lines).
patching file virt/kvm/iommu.c
patching file virt/kvm/kvm_main.c
Hunk #4 succeeded at 929 (offset -2 lines).
Hunk #5 succeeded at 938 (offset -2 lines).
Hunk #6 FAILED at 991.
1 out of 6 hunks FAILED -- saving rejects to file virt/kvm/kvm_main.c.rej

[Obscene_CNN]

toplinuxsir,

Try the #32 patch found here

http://www.kernel.org/pub/linux/kern...ugepage-32.bz2

[towo2099]

For 2.6.37-rc5, i have a fixed patch uploaded there

http://myfreefilehosting.com/f/43040dc043_0.21MB

[Obscene_CNN]

For 2.6.37-rc5, i have a fixed patch uploaded there

http://myfreefilehosting.com/f/43040dc043_0.21MB

towo2099:

you rock!! Unfortunately for me OpenDNS is blocking that site for phishing