The patch titled splice: use get_user_pages_fast has been added to the -mm tree. Its filename is splice-use-get_user_pages_fast.patch Before you just go and hit "reply", please: a) Consider who else should be cc'ed b) Prefer to cc a suitable mailing list as well c) Ideally: find the original patch on the mailing list and do a reply-to-all to that, adding suitable additional cc's *** Remember to use Documentation/SubmitChecklist when testing your code *** See http://www.zip.com.au/~akpm/linux/patches/stuff/added-to-mm.txt to find out what to do about this The current -mm tree may be found at http://userweb.kernel.org/~akpm/mmotm/ ------------------------------------------------------ Subject: splice: use get_user_pages_fast From: Nick Piggin <npiggin@xxxxxxx> Use get_user_pages_fast in splice. This reverts some mmap_sem batching there, however the biggest problem with mmap_sem tends to be hold times blocking out other threads rather than cacheline bouncing. Further: on architectures that implement get_user_pages_fast without locks, mmap_sem can be avoided completely anyway. Signed-off-by: Nick Piggin <npiggin@xxxxxxx> Cc: Dave Kleikamp <shaggy@xxxxxxxxxxxxxx> Cc: Andy Whitcroft <apw@xxxxxxxxxxxx> Cc: Ingo Molnar <mingo@xxxxxxx> Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx> Cc: Andi Kleen <andi@xxxxxxxxxxxxxx> Cc: Dave Kleikamp <shaggy@xxxxxxxxxxxxxx> Cc: Badari Pulavarty <pbadari@xxxxxxxxxx> Cc: Zach Brown <zach.brown@xxxxxxxxxx> Cc: Jens Axboe <jens.axboe@xxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- fs/splice.c | 41 +++-------------------------------------- 1 file changed, 3 insertions(+), 38 deletions(-) diff -puN fs/splice.c~splice-use-get_user_pages_fast fs/splice.c --- a/fs/splice.c~splice-use-get_user_pages_fast +++ a/fs/splice.c @@ -1154,36 +1154,6 @@ static long do_splice(struct file *in, l } /* - * Do a copy-from-user while holding the mmap_semaphore for reading, in a - * manner safe from deadlocking with simultaneous mmap() (grabbing mmap_sem - * for writing) and page faulting on the user memory pointed to by src. - * This assumes that we will very rarely hit the partial != 0 path, or this - * will not be a win. - */ -static int copy_from_user_mmap_sem(void *dst, const void __user *src, size_t n) -{ - int partial; - - if (!access_ok(VERIFY_READ, src, n)) - return -EFAULT; - - pagefault_disable(); - partial = __copy_from_user_inatomic(dst, src, n); - pagefault_enable(); - - /* - * Didn't copy everything, drop the mmap_sem and do a faulting copy - */ - if (unlikely(partial)) { - up_read(¤t->mm->mmap_sem); - partial = copy_from_user(dst, src, n); - down_read(¤t->mm->mmap_sem); - } - - return partial; -} - -/* * Map an iov into an array of pages and offset/length tupples. With the * partial_page structure, we can map several non-contiguous ranges into * our ones pages[] map instead of splitting that operation into pieces. @@ -1196,8 +1166,6 @@ static int get_iovec_page_array(const st { int buffers = 0, error = 0; - down_read(¤t->mm->mmap_sem); - while (nr_vecs) { unsigned long off, npages; struct iovec entry; @@ -1206,7 +1174,7 @@ static int get_iovec_page_array(const st int i; error = -EFAULT; - if (copy_from_user_mmap_sem(&entry, iov, sizeof(entry))) + if (copy_from_user(&entry, iov, sizeof(entry))) break; base = entry.iov_base; @@ -1240,9 +1208,8 @@ static int get_iovec_page_array(const st if (npages > PIPE_BUFFERS - buffers) npages = PIPE_BUFFERS - buffers; - error = get_user_pages(current, current->mm, - (unsigned long) base, npages, 0, 0, - &pages[buffers], NULL); + error = get_user_pages_fast((unsigned long)base, npages, + 0, &pages[buffers]); if (unlikely(error <= 0)) break; @@ -1281,8 +1248,6 @@ static int get_iovec_page_array(const st iov++; } - up_read(¤t->mm->mmap_sem); - if (buffers) return buffers; _ Patches currently in -mm which might be from npiggin@xxxxxxx are hugetlb-fix-lockdep-error.patch linux-next.patch spufs-convert-nopfn-to-fault.patch vt-fix-vc_resize-locking.patch mspec-convert-nopfn-to-fault.patch mspec-convert-nopfn-to-fault-fix.patch mm-remove-nopfn.patch mm-remove-double-indirection-on-tlb-parameter-to-free_pgd_range-co.patch x86-implement-pte_special.patch mm-introduce-get_user_pages_fast.patch mm-introduce-get_user_pages_fast-checkpatch-fixes.patch x86-lockless-get_user_pages_fast.patch x86-lockless-get_user_pages_fast-checkpatch-fixes.patch x86-lockless-get_user_pages_fast-fix.patch x86-lockless-get_user_pages_fast-fix-warning.patch dio-use-get_user_pages_fast.patch splice-use-get_user_pages_fast.patch reiser4.patch likeliness-accounting-change-and-cleanup.patch -- To unsubscribe from this list: send the line "unsubscribe mm-commits" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html