The patch titled
Subject: mm: add infrastructure for get_user_pages_fast() benchmarking
has been added to the -mm tree. Its filename is
mm-add-infrastructure-for-get_user_pages_fast-benchmarking.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/mm-add-infrastructure-for-get_user_pages_fast-benchmarking.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/mm-add-infrastructure-for-get_user_pages_fast-benchmarking.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 ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: "Kirill A. Shutemov" <kirill.shutemov@xxxxxxxxxxxxxxx>
Subject: mm: add infrastructure for get_user_pages_fast() benchmarking
Performance of get_user_pages_fast() is critical for some workloads, but
it's tricky to test it directly.
This patch provides /sys/kernel/debug/gup_benchmark that helps with
testing performance of it.
See tools/testing/selftests/vm/gup_benchmark.c for userspace
counterpart.
Link: http://lkml.kernel.org/r/20170908215603.9189-2-kirill.shutemov@xxxxxxxxxxxxxxx
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
Cc: Shuah Khan <shuah@xxxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Thorsten Leemhuis <regressions@xxxxxxxxxxxxx>
Cc: Jonathan Corbet <corbet@xxxxxxx>
Cc: Huang Ying <ying.huang@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/Kconfig | 9 +
mm/Makefile | 1
mm/gup_benchmark.c | 100 +++++++++++++++++++
tools/testing/selftests/vm/Makefile | 1
tools/testing/selftests/vm/gup_benchmark.c | 91 +++++++++++++++++
5 files changed, 202 insertions(+)
diff -puN /dev/null mm/gup_benchmark.c
--- /dev/null
+++ a/mm/gup_benchmark.c
@@ -0,0 +1,100 @@
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/slab.h>
+#include <linux/uaccess.h>
+#include <linux/ktime.h>
+#include <linux/debugfs.h>
+
+#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_benchmark)
+
+struct gup_benchmark {
+ __u64 delta_usec;
+ __u64 addr;
+ __u64 size;
+ __u32 nr_pages_per_call;
+ __u32 flags;
+};
+
+static int __gup_benchmark_ioctl(unsigned int cmd,
+ struct gup_benchmark *gup)
+{
+ ktime_t start_time, end_time;
+ unsigned long i, nr, nr_pages, addr, next;
+ struct page **pages;
+
+ nr_pages = gup->size / PAGE_SIZE;
+ pages = kvmalloc(sizeof(void *) * nr_pages, GFP_KERNEL);
+ if (!pages)
+ return -ENOMEM;
+
+ i = 0;
+ nr = gup->nr_pages_per_call;
+ start_time = ktime_get();
+ for (addr = gup->addr; addr < gup->addr + gup->size; addr = next) {
+ if (nr != gup->nr_pages_per_call)
+ break;
+
+ next = addr + nr * PAGE_SIZE;
+ if (next > gup->addr + gup->size) {
+ next = gup->addr + gup->size;
+ nr = (next - addr) / PAGE_SIZE;
+ }
+
+ nr = get_user_pages_fast(addr, nr, gup->flags & 1, pages + i);
+ i += nr;
+ }
+ end_time = ktime_get();
+
+ gup->delta_usec = ktime_us_delta(end_time, start_time);
+ gup->size = addr - gup->addr;
+
+ for (i = 0; i < nr_pages; i++) {
+ if (!pages[i])
+ break;
+ put_page(pages[i]);
+ }
+
+ kvfree(pages);
+ return 0;
+}
+
+static long gup_benchmark_ioctl(struct file *filep, unsigned int cmd,
+ unsigned long arg)
+{
+ struct gup_benchmark gup;
+ int ret;
+
+ if (cmd != GUP_FAST_BENCHMARK)
+ return -EINVAL;
+
+ if (copy_from_user(&gup, (void __user *)arg, sizeof(gup)))
+ return -EFAULT;
+
+ ret = __gup_benchmark_ioctl(cmd, &gup);
+ if (ret)
+ return ret;
+
+ if (copy_to_user((void __user *)arg, &gup, sizeof(gup)))
+ return -EFAULT;
+
+ return 0;
+}
+
+static const struct file_operations gup_benchmark_fops = {
+ .open = nonseekable_open,
+ .unlocked_ioctl = gup_benchmark_ioctl,
+};
+
+static int gup_benchmark_init(void)
+{
+ void *ret;
+
+ ret = debugfs_create_file_unsafe("gup_benchmark", 0600, NULL, NULL,
+ &gup_benchmark_fops);
+ if (!ret)
+ pr_warn("Failed to create gup_benchmark in debugfs");
+
+ return 0;
+}
+
+late_initcall(gup_benchmark_init);
diff -puN mm/Kconfig~mm-add-infrastructure-for-get_user_pages_fast-benchmarking mm/Kconfig
--- a/mm/Kconfig~mm-add-infrastructure-for-get_user_pages_fast-benchmarking
+++ a/mm/Kconfig
@@ -756,3 +756,12 @@ config PERCPU_STATS
This feature collects and exposes statistics via debugfs. The
information includes global and per chunk statistics, which can
be used to help understand percpu memory usage.
+
+config GUP_BENCHMARK
+ bool "Enable infrastructure for get_user_pages_fast() benchmarking"
+ default n
+ help
+ Provides /sys/kernel/debug/gup_benchmark that helps with testing
+ performance of get_user_pages_fast().
+
+ See tools/testing/selftests/vm/gup_benchmark.c
diff -puN mm/Makefile~mm-add-infrastructure-for-get_user_pages_fast-benchmarking mm/Makefile
--- a/mm/Makefile~mm-add-infrastructure-for-get_user_pages_fast-benchmarking
+++ a/mm/Makefile
@@ -81,6 +81,7 @@ obj-$(CONFIG_PAGE_COUNTER) += page_count
obj-$(CONFIG_MEMCG) += memcontrol.o vmpressure.o
obj-$(CONFIG_MEMCG_SWAP) += swap_cgroup.o
obj-$(CONFIG_CGROUP_HUGETLB) += hugetlb_cgroup.o
+obj-$(CONFIG_GUP_BENCHMARK) += gup_benchmark.o
obj-$(CONFIG_MEMORY_FAILURE) += memory-failure.o
obj-$(CONFIG_HWPOISON_INJECT) += hwpoison-inject.o
obj-$(CONFIG_DEBUG_KMEMLEAK) += kmemleak.o
diff -puN /dev/null tools/testing/selftests/vm/gup_benchmark.c
--- /dev/null
+++ a/tools/testing/selftests/vm/gup_benchmark.c
@@ -0,0 +1,91 @@
+#include <fcntl.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <unistd.h>
+
+#include <sys/ioctl.h>
+#include <sys/mman.h>
+#include <sys/prctl.h>
+#include <sys/stat.h>
+#include <sys/types.h>
+
+#include <linux/types.h>
+
+#define MB (1UL << 20)
+#define PAGE_SIZE sysconf(_SC_PAGESIZE)
+
+#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_benchmark)
+
+struct gup_benchmark {
+ __u64 delta_usec;
+ __u64 addr;
+ __u64 size;
+ __u32 nr_pages_per_call;
+ __u32 flags;
+};
+
+int main(int argc, char **argv)
+{
+ struct gup_benchmark gup;
+ unsigned long size = 128 * MB;
+ int i, fd, opt, nr_pages = 1, thp = -1, repeats = 1, write = 0;
+ char *p;
+
+ while ((opt = getopt(argc, argv, "m:r:n:tT")) != -1) {
+ switch (opt) {
+ case 'm':
+ size = atoi(optarg) * MB;
+ break;
+ case 'r':
+ repeats = atoi(optarg);
+ break;
+ case 'n':
+ nr_pages = atoi(optarg);
+ break;
+ case 't':
+ thp = 1;
+ break;
+ case 'T':
+ thp = 0;
+ break;
+ case 'w':
+ write = 1;
+ default:
+ return -1;
+ }
+ }
+
+ gup.nr_pages_per_call = nr_pages;
+ gup.flags = write;
+
+ fd = open("/sys/kernel/debug/gup_benchmark", O_RDWR);
+ if (fd == -1)
+ perror("open"), exit(1);
+
+ p = mmap(NULL, size, PROT_READ | PROT_WRITE,
+ MAP_ANONYMOUS | MAP_PRIVATE, -1, 0);
+ if (p == MAP_FAILED)
+ perror("mmap"), exit(1);
+ gup.addr = (unsigned long)p;
+
+ if (thp == 1)
+ madvise(p, size, MADV_HUGEPAGE);
+ else if (thp == 0)
+ madvise(p, size, MADV_NOHUGEPAGE);
+
+ for (; (unsigned long)p < gup.addr + size; p += PAGE_SIZE)
+ p[0] = 0;
+
+ for (i = 0; i < repeats; i++) {
+ gup.size = size;
+ if (ioctl(fd, GUP_FAST_BENCHMARK, &gup))
+ perror("ioctl"), exit(1);
+
+ printf("Time: %lld us", gup.delta_usec);
+ if (gup.size != size)
+ printf(", truncated (size: %lld)", gup.size);
+ printf("\n");
+ }
+
+ return 0;
+}
diff -puN tools/testing/selftests/vm/Makefile~mm-add-infrastructure-for-get_user_pages_fast-benchmarking tools/testing/selftests/vm/Makefile
--- a/tools/testing/selftests/vm/Makefile~mm-add-infrastructure-for-get_user_pages_fast-benchmarking
+++ a/tools/testing/selftests/vm/Makefile
@@ -17,6 +17,7 @@ TEST_GEN_FILES += transhuge-stress
TEST_GEN_FILES += userfaultfd
TEST_GEN_FILES += mlock-random-test
TEST_GEN_FILES += virtual_address_range
+TEST_GEN_FILES += gup_benchmark
TEST_PROGS := run_vmtests
_
Patches currently in -mm which might be from kirill.shutemov@xxxxxxxxxxxxxxx are
mm-fix-typo-in-vm_mpx-definition.patch
mm-account-pud-page-tables.patch
mm-add-infrastructure-for-get_user_pages_fast-benchmarking.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
