Hi,
Reposting for -rc1, with some fixes, and an additional path at the end, too. I've swept through and looked for problems, as well.
Changes since v2 [1]:
* Rebased onto 5.10-rc1
* Fixed an improper ".." include path, indentified by Linus [2].
* Added an "if (cmd == DUMP_USER_PAGES_TEST)" guard to invoking dump_pages_test(). Before, it worked, but it's too subtle to depend merely on struct gup_test.which_pages[] being zeroed out, in order to avoid dumping pages that are not requested to be dumped.
* Added a patch to the end: 2x speedup for run_vmtests.sh
* Tweaked some commit logs and comments slightly
Original cover letter, edited slightly:
Summary: This series provides two main things, and a number of smaller supporting goodies. The two main points are:
1) Add a new sub-test to gup_test, which in turn is a renamed version of gup_benchmark. This sub-test allows nicer testing of dump_pages(), at least on user-space pages.
For quite a while, I was doing a quick hack to gup_test.c whenever I wanted to try out changes to dump_page(). Then Matthew Wilcox asked me what I meant when I said "I used my dump_page() unit test", and I realized that it might be nice to check in a polished up version of that.
Details about how it works and how to use it are in the commit description for patch #6 ("selftests/vm: gup_test: introduce the dump_pages() sub-test").
2) Fixes a limitation of hmm-tests: these tests are incredibly useful, but only if people actually build and run them. And it turns out that libhugetlbfs is a little too effective at throwing a wrench in the works, there. So I've added a little configuration check that removes just two of the 21 hmm-tests, if libhugetlbfs is not available.
Further details in the commit description of patch #8 ("selftests/vm: hmm-tests: remove the libhugetlbfs dependency").
Other smaller things that this series does:
a) Remove code duplication by creating gup_test.h.
b) Clear up the sub-test organization, and their invocation within run_vmtests.sh.
c) Other minor assorted improvements.
[1] v2 is here: https://lore.kernel.org/linux-doc/20200929212747.251804-1-jhubbard@nvidia.co...
[2] https://lore.kernel.org/r/CAHk-=wgh-TMPHLY3jueHX7Y2fWh3D+nMBqVS__AZm6-oorquW...
John Hubbard (9): mm/gup_benchmark: rename to mm/gup_test selftests/vm: use a common gup_test.h selftests/vm: rename run_vmtests --> run_vmtests.sh selftests/vm: minor cleanup: Makefile and gup_test.c selftests/vm: only some gup_test items are really benchmarks selftests/vm: gup_test: introduce the dump_pages() sub-test selftests/vm: run_vmtests.sh: update and clean up gup_test invocation selftests/vm: hmm-tests: remove the libhugetlbfs dependency selftests/vm: 2x speedup for run_vmtests.sh
Documentation/core-api/pin_user_pages.rst | 6 +- arch/s390/configs/debug_defconfig | 2 +- arch/s390/configs/defconfig | 2 +- mm/Kconfig | 21 +- mm/Makefile | 2 +- mm/{gup_benchmark.c => gup_test.c} | 111 ++++++---- mm/gup_test.h | 32 +++ tools/testing/selftests/vm/.gitignore | 3 +- tools/testing/selftests/vm/Makefile | 38 +++- tools/testing/selftests/vm/check_config.sh | 31 +++ tools/testing/selftests/vm/config | 2 +- tools/testing/selftests/vm/gup_benchmark.c | 143 ------------- tools/testing/selftests/vm/gup_test.c | 194 ++++++++++++++++++ tools/testing/selftests/vm/hmm-tests.c | 10 +- .../vm/{run_vmtests => run_vmtests.sh} | 32 ++- 15 files changed, 416 insertions(+), 213 deletions(-) rename mm/{gup_benchmark.c => gup_test.c} (60%) create mode 100644 mm/gup_test.h create mode 100644 tools/testing/selftests/vm/check_config.sh delete mode 100644 tools/testing/selftests/vm/gup_benchmark.c create mode 100644 tools/testing/selftests/vm/gup_test.c rename tools/testing/selftests/vm/{run_vmtests => run_vmtests.sh} (88%)
base-commit: 3650b228f83adda7e5ee532e2b90429c03f7b9ec
Rename nearly every "gup_benchmark" reference and file name to "gup_test". The one exception is for the actual gup benchmark test itself.
The current code already does a *little* bit more than benchmarking, and definitely covers more than get_user_pages_fast(). More importantly, however, subsequent patches are about to add some functionality that is non-benchmark related.
Closely related changes:
* Kconfig: in addition to renaming the options from GUP_BENCHMARK to GUP_TEST, update the help text to reflect that it's no longer a benchmark-only test.
Signed-off-by: John Hubbard jhubbard@nvidia.com Cc: Jonathan Corbet corbet@lwn.net Cc: Jérôme Glisse jglisse@redhat.com Cc: Ralph Campbell rcampbell@nvidia.com Cc: Shuah Khan shuah@kernel.org --- Documentation/core-api/pin_user_pages.rst | 6 ++-- arch/s390/configs/debug_defconfig | 2 +- arch/s390/configs/defconfig | 2 +- mm/Kconfig | 15 +++++--- mm/Makefile | 2 +- mm/{gup_benchmark.c => gup_test.c} | 36 +++++++++---------- tools/testing/selftests/vm/.gitignore | 2 +- tools/testing/selftests/vm/Makefile | 2 +- tools/testing/selftests/vm/config | 2 +- .../vm/{gup_benchmark.c => gup_test.c} | 16 ++++----- tools/testing/selftests/vm/run_vmtests | 8 ++--- 11 files changed, 49 insertions(+), 44 deletions(-) rename mm/{gup_benchmark.c => gup_test.c} (81%) rename tools/testing/selftests/vm/{gup_benchmark.c => gup_test.c} (85%)
diff --git a/Documentation/core-api/pin_user_pages.rst b/Documentation/core-api/pin_user_pages.rst index 7ca8c7bac650..eae972b23224 100644 --- a/Documentation/core-api/pin_user_pages.rst +++ b/Documentation/core-api/pin_user_pages.rst @@ -221,12 +221,12 @@ Unit testing ============ This file::
- tools/testing/selftests/vm/gup_benchmark.c + tools/testing/selftests/vm/gup_test.c
has the following new calls to exercise the new pin*() wrapper functions:
-* PIN_FAST_BENCHMARK (./gup_benchmark -a) -* PIN_BENCHMARK (./gup_benchmark -b) +* PIN_FAST_BENCHMARK (./gup_test -a) +* PIN_BENCHMARK (./gup_test -b)
You can monitor how many total dma-pinned pages have been acquired and released since the system was booted, via two new /proc/vmstat entries: :: diff --git a/arch/s390/configs/debug_defconfig b/arch/s390/configs/debug_defconfig index 0784bf3caf43..c624f4b1ad33 100644 --- a/arch/s390/configs/debug_defconfig +++ b/arch/s390/configs/debug_defconfig @@ -100,7 +100,7 @@ CONFIG_ZSMALLOC_STAT=y CONFIG_DEFERRED_STRUCT_PAGE_INIT=y CONFIG_IDLE_PAGE_TRACKING=y CONFIG_PERCPU_STATS=y -CONFIG_GUP_BENCHMARK=y +CONFIG_GUP_TEST=y CONFIG_NET=y CONFIG_PACKET=y CONFIG_PACKET_DIAG=m diff --git a/arch/s390/configs/defconfig b/arch/s390/configs/defconfig index 905bc8c4cfaf..878b89706998 100644 --- a/arch/s390/configs/defconfig +++ b/arch/s390/configs/defconfig @@ -94,7 +94,7 @@ CONFIG_ZSMALLOC_STAT=y CONFIG_DEFERRED_STRUCT_PAGE_INIT=y CONFIG_IDLE_PAGE_TRACKING=y CONFIG_PERCPU_STATS=y -CONFIG_GUP_BENCHMARK=y +CONFIG_GUP_TEST=y CONFIG_NET=y CONFIG_PACKET=y CONFIG_PACKET_DIAG=m diff --git a/mm/Kconfig b/mm/Kconfig index d42423f884a7..373834274ff5 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -834,13 +834,18 @@ config PERCPU_STATS 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() and related calls benchmarking" +config GUP_TEST + bool "Enable infrastructure for get_user_pages()-related unit tests" help - Provides /sys/kernel/debug/gup_benchmark that helps with testing - performance of get_user_pages() and related calls. + Provides /sys/kernel/debug/gup_test, which in turn provides a way + to make ioctl calls that can launch kernel-based unit tests for + the get_user_pages*() and pin_user_pages*() family of API calls.
- See tools/testing/selftests/vm/gup_benchmark.c + These tests include benchmark testing of the _fast variants of + get_user_pages*() and pin_user_pages*(), as well as smoke tests of + the non-_fast variants. + + See tools/testing/selftests/vm/gup_test.c
config GUP_GET_PTE_LOW_HIGH bool diff --git a/mm/Makefile b/mm/Makefile index d73aed0fc99c..069f216e109e 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -90,7 +90,7 @@ obj-$(CONFIG_PAGE_COUNTER) += page_counter.o obj-$(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_GUP_TEST) += gup_test.o obj-$(CONFIG_MEMORY_FAILURE) += memory-failure.o obj-$(CONFIG_HWPOISON_INJECT) += hwpoison-inject.o obj-$(CONFIG_DEBUG_KMEMLEAK) += kmemleak.o diff --git a/mm/gup_benchmark.c b/mm/gup_test.c similarity index 81% rename from mm/gup_benchmark.c rename to mm/gup_test.c index 8b3e5b5cd8fa..59472ea6aa39 100644 --- a/mm/gup_benchmark.c +++ b/mm/gup_test.c @@ -5,13 +5,13 @@ #include <linux/ktime.h> #include <linux/debugfs.h>
-#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_benchmark) -#define GUP_BENCHMARK _IOWR('g', 2, struct gup_benchmark) -#define PIN_FAST_BENCHMARK _IOWR('g', 3, struct gup_benchmark) -#define PIN_BENCHMARK _IOWR('g', 4, struct gup_benchmark) -#define PIN_LONGTERM_BENCHMARK _IOWR('g', 5, struct gup_benchmark) +#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_test) +#define GUP_BENCHMARK _IOWR('g', 2, struct gup_test) +#define PIN_FAST_BENCHMARK _IOWR('g', 3, struct gup_test) +#define PIN_BENCHMARK _IOWR('g', 4, struct gup_test) +#define PIN_LONGTERM_BENCHMARK _IOWR('g', 5, struct gup_test)
-struct gup_benchmark { +struct gup_test { __u64 get_delta_usec; __u64 put_delta_usec; __u64 addr; @@ -56,7 +56,7 @@ static void verify_dma_pinned(unsigned int cmd, struct page **pages, if (WARN(!page_maybe_dma_pinned(page), "pages[%lu] is NOT dma-pinned\n", i)) {
- dump_page(page, "gup_benchmark failure"); + dump_page(page, "gup_test failure"); break; } } @@ -64,8 +64,8 @@ static void verify_dma_pinned(unsigned int cmd, struct page **pages, } }
-static int __gup_benchmark_ioctl(unsigned int cmd, - struct gup_benchmark *gup) +static int __gup_test_ioctl(unsigned int cmd, + struct gup_test *gup) { ktime_t start_time, end_time; unsigned long i, nr_pages, addr, next; @@ -164,10 +164,10 @@ static int __gup_benchmark_ioctl(unsigned int cmd, return ret; }
-static long gup_benchmark_ioctl(struct file *filep, unsigned int cmd, +static long gup_test_ioctl(struct file *filep, unsigned int cmd, unsigned long arg) { - struct gup_benchmark gup; + struct gup_test gup; int ret;
switch (cmd) { @@ -184,7 +184,7 @@ static long gup_benchmark_ioctl(struct file *filep, unsigned int cmd, if (copy_from_user(&gup, (void __user *)arg, sizeof(gup))) return -EFAULT;
- ret = __gup_benchmark_ioctl(cmd, &gup); + ret = __gup_test_ioctl(cmd, &gup); if (ret) return ret;
@@ -194,17 +194,17 @@ static long gup_benchmark_ioctl(struct file *filep, unsigned int cmd, return 0; }
-static const struct file_operations gup_benchmark_fops = { +static const struct file_operations gup_test_fops = { .open = nonseekable_open, - .unlocked_ioctl = gup_benchmark_ioctl, + .unlocked_ioctl = gup_test_ioctl, };
-static int gup_benchmark_init(void) +static int gup_test_init(void) { - debugfs_create_file_unsafe("gup_benchmark", 0600, NULL, NULL, - &gup_benchmark_fops); + debugfs_create_file_unsafe("gup_test", 0600, NULL, NULL, + &gup_test_fops);
return 0; }
-late_initcall(gup_benchmark_init); +late_initcall(gup_test_init); diff --git a/tools/testing/selftests/vm/.gitignore b/tools/testing/selftests/vm/.gitignore index 849e8226395a..2c8ddcf41c0e 100644 --- a/tools/testing/selftests/vm/.gitignore +++ b/tools/testing/selftests/vm/.gitignore @@ -15,7 +15,7 @@ userfaultfd mlock-intersect-test mlock-random-test virtual_address_range -gup_benchmark +gup_test va_128TBswitch map_fixed_noreplace write_to_hugetlbfs diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile index 30873b19d04b..d1ae706d9927 100644 --- a/tools/testing/selftests/vm/Makefile +++ b/tools/testing/selftests/vm/Makefile @@ -23,7 +23,7 @@ MAKEFLAGS += --no-builtin-rules CFLAGS = -Wall -I ../../../../usr/include $(EXTRA_CFLAGS) LDLIBS = -lrt TEST_GEN_FILES = compaction_test -TEST_GEN_FILES += gup_benchmark +TEST_GEN_FILES += gup_test TEST_GEN_FILES += hmm-tests TEST_GEN_FILES += hugepage-mmap TEST_GEN_FILES += hugepage-shm diff --git a/tools/testing/selftests/vm/config b/tools/testing/selftests/vm/config index 69dd0d1aa30b..60e82da0de85 100644 --- a/tools/testing/selftests/vm/config +++ b/tools/testing/selftests/vm/config @@ -3,4 +3,4 @@ CONFIG_USERFAULTFD=y CONFIG_TEST_VMALLOC=m CONFIG_DEVICE_PRIVATE=y CONFIG_TEST_HMM=m -CONFIG_GUP_BENCHMARK=y +CONFIG_GUP_TEST=y diff --git a/tools/testing/selftests/vm/gup_benchmark.c b/tools/testing/selftests/vm/gup_test.c similarity index 85% rename from tools/testing/selftests/vm/gup_benchmark.c rename to tools/testing/selftests/vm/gup_test.c index 1d4359341e44..00b4731f535e 100644 --- a/tools/testing/selftests/vm/gup_benchmark.c +++ b/tools/testing/selftests/vm/gup_test.c @@ -14,18 +14,18 @@ #define MB (1UL << 20) #define PAGE_SIZE sysconf(_SC_PAGESIZE)
-#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_benchmark) -#define GUP_BENCHMARK _IOWR('g', 2, struct gup_benchmark) +#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_test) +#define GUP_BENCHMARK _IOWR('g', 2, struct gup_test)
/* Similar to above, but use FOLL_PIN instead of FOLL_GET. */ -#define PIN_FAST_BENCHMARK _IOWR('g', 3, struct gup_benchmark) -#define PIN_BENCHMARK _IOWR('g', 4, struct gup_benchmark) -#define PIN_LONGTERM_BENCHMARK _IOWR('g', 5, struct gup_benchmark) +#define PIN_FAST_BENCHMARK _IOWR('g', 3, struct gup_test) +#define PIN_BENCHMARK _IOWR('g', 4, struct gup_test) +#define PIN_LONGTERM_BENCHMARK _IOWR('g', 5, struct gup_test)
/* Just the flags we need, copied from mm.h: */ #define FOLL_WRITE 0x01 /* check pte is writable */
-struct gup_benchmark { +struct gup_test { __u64 get_delta_usec; __u64 put_delta_usec; __u64 addr; @@ -37,7 +37,7 @@ struct gup_benchmark {
int main(int argc, char **argv) { - struct gup_benchmark gup; + struct gup_test gup; unsigned long size = 128 * MB; int i, fd, filed, opt, nr_pages = 1, thp = -1, repeats = 1, write = 0; int cmd = GUP_FAST_BENCHMARK, flags = MAP_PRIVATE; @@ -104,7 +104,7 @@ int main(int argc, char **argv) if (write) gup.flags |= FOLL_WRITE;
- fd = open("/sys/kernel/debug/gup_benchmark", O_RDWR); + fd = open("/sys/kernel/debug/gup_test", O_RDWR); if (fd == -1) { perror("open"); exit(1); diff --git a/tools/testing/selftests/vm/run_vmtests b/tools/testing/selftests/vm/run_vmtests index a3f4f30f0a2e..d1843d5f3c30 100755 --- a/tools/testing/selftests/vm/run_vmtests +++ b/tools/testing/selftests/vm/run_vmtests @@ -124,9 +124,9 @@ else fi
echo "--------------------------------------------" -echo "running 'gup_benchmark -U' (normal/slow gup)" +echo "running 'gup_test -U' (normal/slow gup)" echo "--------------------------------------------" -./gup_benchmark -U +./gup_test -U if [ $? -ne 0 ]; then echo "[FAIL]" exitcode=1 @@ -135,9 +135,9 @@ else fi
echo "------------------------------------------" -echo "running gup_benchmark -b (pin_user_pages)" +echo "running gup_test -b (pin_user_pages)" echo "------------------------------------------" -./gup_benchmark -b +./gup_test -b if [ $? -ne 0 ]; then echo "[FAIL]" exitcode=1
Avoid the need to copy-paste the gup_test ioctl commands and the struct gup_test definition, between the kernel and the user space application, by providing a new header file for these. This allows easier and safer adding of new ioctl calls, as well as reducing the overall line count.
Details: The header file has to be able to compile independently, because of the arguably unfortunate way that the Makefile is written: the Makefile tries to build all of its prerequisites, when really it should be only building the .c files, and leaving the other prerequisites (LOCAL_HDRS) as pure dependencies.
That Makefile limitation is probably not worth fixing, but it explains why one of the includes had to be moved into the new header file.
Also: simplify the ioctl struct (struct gup_test), by deleting the unused __expansion[10] field. This sort of thing is what you might see in a stable ABI, but this low-level, kernel-developer-oriented selftests/vm system is very much not subject to ABI stability. So "expansion" and "reserved" fields are unnecessary here.
Signed-off-by: John Hubbard jhubbard@nvidia.com Cc: Jérôme Glisse jglisse@redhat.com Cc: Jonathan Corbet corbet@lwn.net Cc: Ralph Campbell rcampbell@nvidia.com Cc: Shuah Khan shuah@kernel.org --- mm/gup_test.c | 17 +---------------- mm/gup_test.h | 22 ++++++++++++++++++++++ tools/testing/selftests/vm/Makefile | 2 ++ tools/testing/selftests/vm/gup_test.c | 22 +--------------------- 4 files changed, 26 insertions(+), 37 deletions(-) create mode 100644 mm/gup_test.h
diff --git a/mm/gup_test.c b/mm/gup_test.c index 59472ea6aa39..4c2d70d88f24 100644 --- a/mm/gup_test.c +++ b/mm/gup_test.c @@ -4,22 +4,7 @@ #include <linux/uaccess.h> #include <linux/ktime.h> #include <linux/debugfs.h> - -#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_test) -#define GUP_BENCHMARK _IOWR('g', 2, struct gup_test) -#define PIN_FAST_BENCHMARK _IOWR('g', 3, struct gup_test) -#define PIN_BENCHMARK _IOWR('g', 4, struct gup_test) -#define PIN_LONGTERM_BENCHMARK _IOWR('g', 5, struct gup_test) - -struct gup_test { - __u64 get_delta_usec; - __u64 put_delta_usec; - __u64 addr; - __u64 size; - __u32 nr_pages_per_call; - __u32 flags; - __u64 expansion[10]; /* For future use */ -}; +#include "gup_test.h"
static void put_back_pages(unsigned int cmd, struct page **pages, unsigned long nr_pages) diff --git a/mm/gup_test.h b/mm/gup_test.h new file mode 100644 index 000000000000..931c2f3f477a --- /dev/null +++ b/mm/gup_test.h @@ -0,0 +1,22 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +#ifndef __GUP_TEST_H +#define __GUP_TEST_H + +#include <linux/types.h> + +#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_test) +#define GUP_BENCHMARK _IOWR('g', 2, struct gup_test) +#define PIN_FAST_BENCHMARK _IOWR('g', 3, struct gup_test) +#define PIN_BENCHMARK _IOWR('g', 4, struct gup_test) +#define PIN_LONGTERM_BENCHMARK _IOWR('g', 5, struct gup_test) + +struct gup_test { + __u64 get_delta_usec; + __u64 put_delta_usec; + __u64 addr; + __u64 size; + __u32 nr_pages_per_call; + __u32 flags; +}; + +#endif /* __GUP_TEST_H */ diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile index d1ae706d9927..9cc6bc087461 100644 --- a/tools/testing/selftests/vm/Makefile +++ b/tools/testing/selftests/vm/Makefile @@ -130,3 +130,5 @@ endif $(OUTPUT)/userfaultfd: LDLIBS += -lpthread
$(OUTPUT)/mlock-random-test: LDLIBS += -lcap + +$(OUTPUT)/gup_test: ../../../../mm/gup_test.h diff --git a/tools/testing/selftests/vm/gup_test.c b/tools/testing/selftests/vm/gup_test.c index 00b4731f535e..03f7c4f1beaf 100644 --- a/tools/testing/selftests/vm/gup_test.c +++ b/tools/testing/selftests/vm/gup_test.c @@ -2,39 +2,19 @@ #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> +#include "../../../../mm/gup_test.h"
#define MB (1UL << 20) #define PAGE_SIZE sysconf(_SC_PAGESIZE)
-#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_test) -#define GUP_BENCHMARK _IOWR('g', 2, struct gup_test) - -/* Similar to above, but use FOLL_PIN instead of FOLL_GET. */ -#define PIN_FAST_BENCHMARK _IOWR('g', 3, struct gup_test) -#define PIN_BENCHMARK _IOWR('g', 4, struct gup_test) -#define PIN_LONGTERM_BENCHMARK _IOWR('g', 5, struct gup_test) - /* Just the flags we need, copied from mm.h: */ #define FOLL_WRITE 0x01 /* check pte is writable */
-struct gup_test { - __u64 get_delta_usec; - __u64 put_delta_usec; - __u64 addr; - __u64 size; - __u32 nr_pages_per_call; - __u32 flags; - __u64 expansion[10]; /* For future use */ -}; - int main(int argc, char **argv) { struct gup_test gup;
Rename to *.sh, in order to match the conventions of all of the other items in selftest/vm.
The only reason not to use a .sh suffix a shell script like this, might be to make it look more like a normal program, but that's not an issue here.
Signed-off-by: John Hubbard jhubbard@nvidia.com Cc: Jérôme Glisse jglisse@redhat.com Cc: Jonathan Corbet corbet@lwn.net Cc: Ralph Campbell rcampbell@nvidia.com Cc: Shuah Khan shuah@kernel.org --- tools/testing/selftests/vm/Makefile | 2 +- tools/testing/selftests/vm/{run_vmtests => run_vmtests.sh} | 0 2 files changed, 1 insertion(+), 1 deletion(-) rename tools/testing/selftests/vm/{run_vmtests => run_vmtests.sh} (100%)
diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile index 9cc6bc087461..5a3bd0c497b6 100644 --- a/tools/testing/selftests/vm/Makefile +++ b/tools/testing/selftests/vm/Makefile @@ -69,7 +69,7 @@ TEST_GEN_FILES += virtual_address_range TEST_GEN_FILES += write_to_hugetlbfs endif
-TEST_PROGS := run_vmtests +TEST_PROGS := run_vmtests.sh
TEST_FILES := test_vmalloc.sh
diff --git a/tools/testing/selftests/vm/run_vmtests b/tools/testing/selftests/vm/run_vmtests.sh similarity index 100% rename from tools/testing/selftests/vm/run_vmtests rename to tools/testing/selftests/vm/run_vmtests.sh
A few cleanups that don't deserve separate patches, but that also should not clutter up other functional changes:
1. Remove an unnecessary #include <prctl.h>
2. Restore the sorted order of TEST_GEN_FILES.
3. Add -lpthread to the common LDLIBS, as it is harmless and several tests use it. This gets rid of one special rule already.
Signed-off-by: John Hubbard jhubbard@nvidia.com Cc: Jérôme Glisse jglisse@redhat.com Cc: Jonathan Corbet corbet@lwn.net Cc: Ralph Campbell rcampbell@nvidia.com Cc: Shuah Khan shuah@kernel.org --- tools/testing/selftests/vm/Makefile | 10 ++++------ tools/testing/selftests/vm/gup_test.c | 1 - 2 files changed, 4 insertions(+), 7 deletions(-)
diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile index 5a3bd0c497b6..2579242386ac 100644 --- a/tools/testing/selftests/vm/Makefile +++ b/tools/testing/selftests/vm/Makefile @@ -21,14 +21,15 @@ MACHINE ?= $(shell echo $(uname_M) | sed -e 's/aarch64.*/arm64/') MAKEFLAGS += --no-builtin-rules
CFLAGS = -Wall -I ../../../../usr/include $(EXTRA_CFLAGS) -LDLIBS = -lrt +LDLIBS = -lrt -lpthread TEST_GEN_FILES = compaction_test TEST_GEN_FILES += gup_test TEST_GEN_FILES += hmm-tests TEST_GEN_FILES += hugepage-mmap TEST_GEN_FILES += hugepage-shm -TEST_GEN_FILES += map_hugetlb +TEST_GEN_FILES += khugepaged TEST_GEN_FILES += map_fixed_noreplace +TEST_GEN_FILES += map_hugetlb TEST_GEN_FILES += map_populate TEST_GEN_FILES += mlock-random-test TEST_GEN_FILES += mlock2-tests @@ -37,7 +38,6 @@ TEST_GEN_FILES += on-fault-limit TEST_GEN_FILES += thuge-gen TEST_GEN_FILES += transhuge-stress TEST_GEN_FILES += userfaultfd -TEST_GEN_FILES += khugepaged
ifeq ($(ARCH),x86_64) CAN_BUILD_I386 := $(shell ./../x86/check_cc.sh $(CC) ../x86/trivial_32bit_program.c -m32) @@ -76,7 +76,7 @@ TEST_FILES := test_vmalloc.sh KSFT_KHDR_INSTALL := 1 include ../lib.mk
-$(OUTPUT)/hmm-tests: LDLIBS += -lhugetlbfs -lpthread +$(OUTPUT)/hmm-tests: LDLIBS += -lhugetlbfs
ifeq ($(ARCH),x86_64) BINARIES_32 := $(patsubst %,$(OUTPUT)/%,$(BINARIES_32)) @@ -127,8 +127,6 @@ warn_32bit_failure: endif endif
-$(OUTPUT)/userfaultfd: LDLIBS += -lpthread - $(OUTPUT)/mlock-random-test: LDLIBS += -lcap
$(OUTPUT)/gup_test: ../../../../mm/gup_test.h diff --git a/tools/testing/selftests/vm/gup_test.c b/tools/testing/selftests/vm/gup_test.c index 03f7c4f1beaf..1a54771ad97e 100644 --- a/tools/testing/selftests/vm/gup_test.c +++ b/tools/testing/selftests/vm/gup_test.c @@ -4,7 +4,6 @@ #include <unistd.h> #include <sys/ioctl.h> #include <sys/mman.h> -#include <sys/prctl.h> #include <sys/stat.h> #include <sys/types.h> #include "../../../../mm/gup_test.h"
Therefore, some minor cleanup and improvements are in order:
1. Rename the other items appropriately.
2. Stop reporting timing information on the non-benchmark items. It's still being recorded and is available, but there's no point in cluttering up the report with data that no one reasonably needs to check.
3. Don't do iterations, for non-benchmark items.
4. Print out a shorter, more appropriate report for the non-benchmark tests.
5. Add the command that was run, to the report. This really helps, as there are quite a lot of options now.
6. Use a larger integer type for cmd, now that it's being compared Otherwise it doesn't work, because in this case cmd is about 3 billion, which is the perfect size for problems with signed vs unsigned int.
Signed-off-by: John Hubbard jhubbard@nvidia.com Cc: Jérôme Glisse jglisse@redhat.com Cc: Jonathan Corbet corbet@lwn.net Cc: Ralph Campbell rcampbell@nvidia.com Cc: Shuah Khan shuah@kernel.org --- Documentation/core-api/pin_user_pages.rst | 2 +- mm/gup_test.c | 14 +++---- mm/gup_test.h | 8 ++-- tools/testing/selftests/vm/gup_test.c | 47 +++++++++++++++++++---- 4 files changed, 51 insertions(+), 20 deletions(-)
diff --git a/Documentation/core-api/pin_user_pages.rst b/Documentation/core-api/pin_user_pages.rst index eae972b23224..fcf605be43d0 100644 --- a/Documentation/core-api/pin_user_pages.rst +++ b/Documentation/core-api/pin_user_pages.rst @@ -226,7 +226,7 @@ This file:: has the following new calls to exercise the new pin*() wrapper functions:
* PIN_FAST_BENCHMARK (./gup_test -a) -* PIN_BENCHMARK (./gup_test -b) +* PIN_BASIC_TEST (./gup_test -b)
You can monitor how many total dma-pinned pages have been acquired and released since the system was booted, via two new /proc/vmstat entries: :: diff --git a/mm/gup_test.c b/mm/gup_test.c index 4c2d70d88f24..173bb38f3688 100644 --- a/mm/gup_test.c +++ b/mm/gup_test.c @@ -13,13 +13,13 @@ static void put_back_pages(unsigned int cmd, struct page **pages,
switch (cmd) { case GUP_FAST_BENCHMARK: - case GUP_BENCHMARK: + case GUP_BASIC_TEST: for (i = 0; i < nr_pages; i++) put_page(pages[i]); break;
case PIN_FAST_BENCHMARK: - case PIN_BENCHMARK: + case PIN_BASIC_TEST: case PIN_LONGTERM_BENCHMARK: unpin_user_pages(pages, nr_pages); break; @@ -34,7 +34,7 @@ static void verify_dma_pinned(unsigned int cmd, struct page **pages,
switch (cmd) { case PIN_FAST_BENCHMARK: - case PIN_BENCHMARK: + case PIN_BASIC_TEST: case PIN_LONGTERM_BENCHMARK: for (i = 0; i < nr_pages; i++) { page = pages[i]; @@ -94,7 +94,7 @@ static int __gup_test_ioctl(unsigned int cmd, nr = get_user_pages_fast(addr, nr, gup->flags, pages + i); break; - case GUP_BENCHMARK: + case GUP_BASIC_TEST: nr = get_user_pages(addr, nr, gup->flags, pages + i, NULL); break; @@ -102,7 +102,7 @@ static int __gup_test_ioctl(unsigned int cmd, nr = pin_user_pages_fast(addr, nr, gup->flags, pages + i); break; - case PIN_BENCHMARK: + case PIN_BASIC_TEST: nr = pin_user_pages(addr, nr, gup->flags, pages + i, NULL); break; @@ -157,10 +157,10 @@ static long gup_test_ioctl(struct file *filep, unsigned int cmd,
switch (cmd) { case GUP_FAST_BENCHMARK: - case GUP_BENCHMARK: case PIN_FAST_BENCHMARK: - case PIN_BENCHMARK: case PIN_LONGTERM_BENCHMARK: + case GUP_BASIC_TEST: + case PIN_BASIC_TEST: break; default: return -EINVAL; diff --git a/mm/gup_test.h b/mm/gup_test.h index 931c2f3f477a..921b4caad8ef 100644 --- a/mm/gup_test.h +++ b/mm/gup_test.h @@ -5,10 +5,10 @@ #include <linux/types.h>
#define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_test) -#define GUP_BENCHMARK _IOWR('g', 2, struct gup_test) -#define PIN_FAST_BENCHMARK _IOWR('g', 3, struct gup_test) -#define PIN_BENCHMARK _IOWR('g', 4, struct gup_test) -#define PIN_LONGTERM_BENCHMARK _IOWR('g', 5, struct gup_test) +#define PIN_FAST_BENCHMARK _IOWR('g', 2, struct gup_test) +#define PIN_LONGTERM_BENCHMARK _IOWR('g', 3, struct gup_test) +#define GUP_BASIC_TEST _IOWR('g', 4, struct gup_test) +#define PIN_BASIC_TEST _IOWR('g', 5, struct gup_test)
struct gup_test { __u64 get_delta_usec; diff --git a/tools/testing/selftests/vm/gup_test.c b/tools/testing/selftests/vm/gup_test.c index 1a54771ad97e..f9163e1bb57a 100644 --- a/tools/testing/selftests/vm/gup_test.c +++ b/tools/testing/selftests/vm/gup_test.c @@ -14,12 +14,30 @@ /* Just the flags we need, copied from mm.h: */ #define FOLL_WRITE 0x01 /* check pte is writable */
+static char *cmd_to_str(unsigned long cmd) +{ + switch (cmd) { + case GUP_FAST_BENCHMARK: + return "GUP_FAST_BENCHMARK"; + case PIN_FAST_BENCHMARK: + return "PIN_FAST_BENCHMARK"; + case PIN_LONGTERM_BENCHMARK: + return "PIN_LONGTERM_BENCHMARK"; + case GUP_BASIC_TEST: + return "GUP_BASIC_TEST"; + case PIN_BASIC_TEST: + return "PIN_BASIC_TEST"; + } + return "Unknown command"; +} + int main(int argc, char **argv) { struct gup_test gup; unsigned long size = 128 * MB; int i, fd, filed, opt, nr_pages = 1, thp = -1, repeats = 1, write = 0; - int cmd = GUP_FAST_BENCHMARK, flags = MAP_PRIVATE; + unsigned long cmd = GUP_FAST_BENCHMARK; + int flags = MAP_PRIVATE; char *file = "/dev/zero"; char *p;
@@ -29,7 +47,7 @@ int main(int argc, char **argv) cmd = PIN_FAST_BENCHMARK; break; case 'b': - cmd = PIN_BENCHMARK; + cmd = PIN_BASIC_TEST; break; case 'L': cmd = PIN_LONGTERM_BENCHMARK; @@ -50,7 +68,7 @@ int main(int argc, char **argv) thp = 0; break; case 'U': - cmd = GUP_BENCHMARK; + cmd = GUP_BASIC_TEST; break; case 'u': cmd = GUP_FAST_BENCHMARK; @@ -104,18 +122,31 @@ int main(int argc, char **argv) for (; (unsigned long)p < gup.addr + size; p += PAGE_SIZE) p[0] = 0;
- for (i = 0; i < repeats; i++) { + /* Only report timing information on the *_BENCHMARK commands: */ + if ((cmd == PIN_FAST_BENCHMARK) || (cmd == GUP_FAST_BENCHMARK) || + (cmd == PIN_LONGTERM_BENCHMARK)) { + for (i = 0; i < repeats; i++) { + gup.size = size; + if (ioctl(fd, cmd, &gup)) + perror("ioctl"), exit(1); + + printf("%s: Time: get:%lld put:%lld us", + cmd_to_str(cmd), gup.get_delta_usec, + gup.put_delta_usec); + if (gup.size != size) + printf(", truncated (size: %lld)", gup.size); + printf("\n"); + } + } else { gup.size = size; if (ioctl(fd, cmd, &gup)) { perror("ioctl"); exit(1); }
- printf("Time: get:%lld put:%lld us", gup.get_delta_usec, - gup.put_delta_usec); + printf("%s: done\n", cmd_to_str(cmd)); if (gup.size != size) - printf(", truncated (size: %lld)", gup.size); - printf("\n"); + printf("Truncated (size: %lld)\n", gup.size); }
return 0;
For quite a while, I was doing a quick hack to gup_test.c (previously, gup_benchmark.c) whenever I wanted to try out my changes to dump_page(). This makes that hack unnecessary, and instead allows anyone to easily get the same coverage from a user space program. That saves a lot of time because you don't have to change the kernel, in order to test different pages and options.
The new sub-test takes advantage of the existing gup_test infrastructure, which already provides a simple user space program, some allocated user space pages, an ioctl call, pinning of those pages (via either get_user_pages or pin_user_pages) and a corresponding kernel-side test invocation. There's not much more required, mainly just a couple of inputs from the user.
In fact, the new test re-uses the existing command line options in order to get various helpful combinations (THP or normal, _fast or slow gup, gup vs. pup, and more).
New command line options are: which pages to dump, and what type of "get/pin" to use.
In order to figure out which pages to dump, the logic is:
* If the user doesn't specify anything, the page 0 (the first page in the address range that the program sets up for testing) is dumped.
* Or, the user can type up to 8 page indices anywhere on the command line. If you type more than 8, then it uses the first 8 and ignores the remaining items.
For example:
./gup_test -ct -F 1 0 19 0x1000
Meaning: -c: dump pages sub-test -t: use THP pages -F 1: use pin_user_pages() instead of get_user_pages() 0 19 0x1000: dump pages 0, 19, and 4096
Signed-off-by: John Hubbard jhubbard@nvidia.com Cc: Jérôme Glisse jglisse@redhat.com Cc: Jonathan Corbet corbet@lwn.net Cc: Ralph Campbell rcampbell@nvidia.com Cc: Shuah Khan shuah@kernel.org --- mm/Kconfig | 6 +++ mm/gup_test.c | 56 ++++++++++++++++++++++++++- mm/gup_test.h | 10 +++++ tools/testing/selftests/vm/gup_test.c | 45 ++++++++++++++++++++- 4 files changed, 113 insertions(+), 4 deletions(-)
diff --git a/mm/Kconfig b/mm/Kconfig index 373834274ff5..01b0ae0cd9d3 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -845,6 +845,12 @@ config GUP_TEST get_user_pages*() and pin_user_pages*(), as well as smoke tests of the non-_fast variants.
+ There is also a sub-test that allows running dump_page() on any + of up to eight pages (selected by command line args) within the + range of user-space addresses. These pages are either pinned via + pin_user_pages*(), or pinned via get_user_pages*(), as specified + by other command line arguments. + See tools/testing/selftests/vm/gup_test.c
config GUP_GET_PTE_LOW_HIGH diff --git a/mm/gup_test.c b/mm/gup_test.c index 173bb38f3688..087ddaf30eb4 100644 --- a/mm/gup_test.c +++ b/mm/gup_test.c @@ -7,7 +7,7 @@ #include "gup_test.h"
static void put_back_pages(unsigned int cmd, struct page **pages, - unsigned long nr_pages) + unsigned long nr_pages, unsigned int gup_test_flags) { unsigned long i;
@@ -23,6 +23,15 @@ static void put_back_pages(unsigned int cmd, struct page **pages, case PIN_LONGTERM_BENCHMARK: unpin_user_pages(pages, nr_pages); break; + case DUMP_USER_PAGES_TEST: + if (gup_test_flags & GUP_TEST_FLAG_DUMP_PAGES_USE_PIN) { + unpin_user_pages(pages, nr_pages); + } else { + for (i = 0; i < nr_pages; i++) + put_page(pages[i]); + + } + break; } }
@@ -49,6 +58,37 @@ static void verify_dma_pinned(unsigned int cmd, struct page **pages, } }
+static void dump_pages_test(struct gup_test *gup, struct page **pages, + unsigned long nr_pages) +{ + unsigned int index_to_dump; + unsigned int i; + + /* + * Zero out any user-supplied page index that is out of range. Remember: + * .which_pages[] contains a 1-based set of page indices. + */ + for (i = 0; i < GUP_TEST_MAX_PAGES_TO_DUMP; i++) { + if (gup->which_pages[i] > nr_pages) { + pr_warn("ZEROING due to out of range: .which_pages[%u]: %u\n", + i, gup->which_pages[i]); + gup->which_pages[i] = 0; + } + } + + for (i = 0; i < GUP_TEST_MAX_PAGES_TO_DUMP; i++) { + index_to_dump = gup->which_pages[i]; + + if (index_to_dump) { + index_to_dump--; // Decode from 1-based, to 0-based + pr_info("---- page #%u, starting from user virt addr: 0x%llx\n", + index_to_dump, gup->addr); + dump_page(pages[index_to_dump], + "gup_test: dump_pages() test"); + } + } +} + static int __gup_test_ioctl(unsigned int cmd, struct gup_test *gup) { @@ -111,6 +151,14 @@ static int __gup_test_ioctl(unsigned int cmd, gup->flags | FOLL_LONGTERM, pages + i, NULL); break; + case DUMP_USER_PAGES_TEST: + if (gup->flags & GUP_TEST_FLAG_DUMP_PAGES_USE_PIN) + nr = pin_user_pages(addr, nr, gup->flags, + pages + i, NULL); + else + nr = get_user_pages(addr, nr, gup->flags, + pages + i, NULL); + break; default: ret = -EINVAL; goto unlock; @@ -134,9 +182,12 @@ static int __gup_test_ioctl(unsigned int cmd, */ verify_dma_pinned(cmd, pages, nr_pages);
+ if (cmd == DUMP_USER_PAGES_TEST) + dump_pages_test(gup, pages, nr_pages); + start_time = ktime_get();
- put_back_pages(cmd, pages, nr_pages); + put_back_pages(cmd, pages, nr_pages, gup->flags);
end_time = ktime_get(); gup->put_delta_usec = ktime_us_delta(end_time, start_time); @@ -161,6 +212,7 @@ static long gup_test_ioctl(struct file *filep, unsigned int cmd, case PIN_LONGTERM_BENCHMARK: case GUP_BASIC_TEST: case PIN_BASIC_TEST: + case DUMP_USER_PAGES_TEST: break; default: return -EINVAL; diff --git a/mm/gup_test.h b/mm/gup_test.h index 921b4caad8ef..90a6713d50eb 100644 --- a/mm/gup_test.h +++ b/mm/gup_test.h @@ -9,6 +9,11 @@ #define PIN_LONGTERM_BENCHMARK _IOWR('g', 3, struct gup_test) #define GUP_BASIC_TEST _IOWR('g', 4, struct gup_test) #define PIN_BASIC_TEST _IOWR('g', 5, struct gup_test) +#define DUMP_USER_PAGES_TEST _IOWR('g', 6, struct gup_test) + +#define GUP_TEST_MAX_PAGES_TO_DUMP 8 + +#define GUP_TEST_FLAG_DUMP_PAGES_USE_PIN 0x1
struct gup_test { __u64 get_delta_usec; @@ -17,6 +22,11 @@ struct gup_test { __u64 size; __u32 nr_pages_per_call; __u32 flags; + /* + * Each non-zero entry is the number of the page (1-based: first page is + * page 1, so that zero entries mean "do nothing") from the .addr base. + */ + __u32 which_pages[GUP_TEST_MAX_PAGES_TO_DUMP]; };
#endif /* __GUP_TEST_H */ diff --git a/tools/testing/selftests/vm/gup_test.c b/tools/testing/selftests/vm/gup_test.c index f9163e1bb57a..6c6336dd3b7f 100644 --- a/tools/testing/selftests/vm/gup_test.c +++ b/tools/testing/selftests/vm/gup_test.c @@ -27,13 +27,15 @@ static char *cmd_to_str(unsigned long cmd) return "GUP_BASIC_TEST"; case PIN_BASIC_TEST: return "PIN_BASIC_TEST"; + case DUMP_USER_PAGES_TEST: + return "DUMP_USER_PAGES_TEST"; } return "Unknown command"; }
int main(int argc, char **argv) { - struct gup_test gup; + struct gup_test gup = { 0 }; unsigned long size = 128 * MB; int i, fd, filed, opt, nr_pages = 1, thp = -1, repeats = 1, write = 0; unsigned long cmd = GUP_FAST_BENCHMARK; @@ -41,7 +43,7 @@ int main(int argc, char **argv) char *file = "/dev/zero"; char *p;
- while ((opt = getopt(argc, argv, "m:r:n:f:abtTLUuwSH")) != -1) { + while ((opt = getopt(argc, argv, "m:r:n:F:f:abctTLUuwSH")) != -1) { switch (opt) { case 'a': cmd = PIN_FAST_BENCHMARK; @@ -52,6 +54,21 @@ int main(int argc, char **argv) case 'L': cmd = PIN_LONGTERM_BENCHMARK; break; + case 'c': + cmd = DUMP_USER_PAGES_TEST; + /* + * Dump page 0 (index 1). May be overridden later, by + * user's non-option arguments. + * + * .which_pages is zero-based, so that zero can mean "do + * nothing". + */ + gup.which_pages[0] = 1; + break; + case 'F': + /* strtol, so you can pass flags in hex form */ + gup.flags = strtol(optarg, 0, 0); + break; case 'm': size = atoi(optarg) * MB; break; @@ -91,6 +108,30 @@ int main(int argc, char **argv) } }
+ if (optind < argc) { + int extra_arg_count = 0; + /* + * For example: + * + * ./gup_test -c 0 1 0x1001 + * + * ...to dump pages 0, 1, and 4097 + */ + + while ((optind < argc) && + (extra_arg_count < GUP_TEST_MAX_PAGES_TO_DUMP)) { + /* + * Do the 1-based indexing here, so that the user can + * use normal 0-based indexing on the command line. + */ + long page_index = strtol(argv[optind], 0, 0) + 1; + + gup.which_pages[extra_arg_count] = page_index; + extra_arg_count++; + optind++; + } + } + filed = open(file, O_RDWR|O_CREAT); if (filed < 0) { perror("open");
Run benchmarks on the _fast variants of gup and pup, as originally intended.
Run the new gup_test sub-test: dump pages. In addition to exercising the dump_page() call, it also demonstrates the various options you can use to specify which pages to dump, and how.
Signed-off-by: John Hubbard jhubbard@nvidia.com Cc: Jérôme Glisse jglisse@redhat.com Cc: Jonathan Corbet corbet@lwn.net Cc: Ralph Campbell rcampbell@nvidia.com Cc: Shuah Khan shuah@kernel.org --- tools/testing/selftests/vm/run_vmtests.sh | 28 ++++++++++++++++------- 1 file changed, 20 insertions(+), 8 deletions(-)
diff --git a/tools/testing/selftests/vm/run_vmtests.sh b/tools/testing/selftests/vm/run_vmtests.sh index d1843d5f3c30..4ac84b350d9f 100755 --- a/tools/testing/selftests/vm/run_vmtests.sh +++ b/tools/testing/selftests/vm/run_vmtests.sh @@ -123,10 +123,10 @@ else echo "[PASS]" fi
-echo "--------------------------------------------" -echo "running 'gup_test -U' (normal/slow gup)" -echo "--------------------------------------------" -./gup_test -U +echo "------------------------------------------------------" +echo "running: gup_test -u # get_user_pages_fast() benchmark" +echo "------------------------------------------------------" +./gup_test -u if [ $? -ne 0 ]; then echo "[FAIL]" exitcode=1 @@ -134,10 +134,22 @@ else echo "[PASS]" fi
-echo "------------------------------------------" -echo "running gup_test -b (pin_user_pages)" -echo "------------------------------------------" -./gup_test -b +echo "------------------------------------------------------" +echo "running: gup_test -a # pin_user_pages_fast() benchmark" +echo "------------------------------------------------------" +./gup_test -a +if [ $? -ne 0 ]; then + echo "[FAIL]" + exitcode=1 +else + echo "[PASS]" +fi + +echo "------------------------------------------------------------" +echo "# Dump pages 0, 19, and 4096, using pin_user_pages:" +echo "running: gup_test -ct -F 0x1 0 19 0x1000 # dump_page() test" +echo "------------------------------------------------------------" +./gup_test -ct -F 0x1 0 19 0x1000 if [ $? -ne 0 ]; then echo "[FAIL]" exitcode=1
HMM selftests are incredibly useful, but they are only effective if people actually build and run them. All the other tests in selftests/vm can be built with very standard, always-available libraries: libpthread, librt. The hmm-tests.c program, on the other hand, requires something that is (much) less readily available: libhugetlbfs. And so the build will typically fail for many developers.
A simple attempt to install libhugetlbfs will also run into complications on some common distros these days: Fedora and Arch Linux (yes, Arch AUR has it, but that's fragile, as always with AUR). The library is not maintained actively enough at the moment, for distros to deal with it. I had to build it from source, for Fedora, and that didn't go too smoothly either.
It turns out that, out of 21 tests in hmm-tests.c, only 2 actually require functionality from libhugetlbfs. Therefore, if libhugetlbfs is missing, simply ifdef those two tests out and allow the developer to at least have the other 19 tests, if they don't want to pause to work through the above issues. Also issue a warning, so that it's clear that there is an imperfection in the build.
In order to do that, a tiny shell script (check_config.sh) runs a quick compile (not link, that's too prone to false failures with library paths), and basically, if the compiler doesn't find hugetlbfs.h in its standard locations, then the script concludes that libhugetlbfs is not available. The output is in two files, one for inclusion in hmm-test.c (local_config.h), and one for inclusion in the Makefile (local_config.mk).
Signed-off-by: John Hubbard jhubbard@nvidia.com Cc: Ralph Campbell rcampbell@nvidia.com Cc: Jérôme Glisse jglisse@redhat.com Cc: Jonathan Corbet corbet@lwn.net Cc: Shuah Khan shuah@kernel.org --- tools/testing/selftests/vm/.gitignore | 1 + tools/testing/selftests/vm/Makefile | 24 +++++++++++++++-- tools/testing/selftests/vm/check_config.sh | 31 ++++++++++++++++++++++ tools/testing/selftests/vm/hmm-tests.c | 10 ++++++- 4 files changed, 63 insertions(+), 3 deletions(-) create mode 100644 tools/testing/selftests/vm/check_config.sh
diff --git a/tools/testing/selftests/vm/.gitignore b/tools/testing/selftests/vm/.gitignore index 2c8ddcf41c0e..e90d28bcd518 100644 --- a/tools/testing/selftests/vm/.gitignore +++ b/tools/testing/selftests/vm/.gitignore @@ -20,3 +20,4 @@ va_128TBswitch map_fixed_noreplace write_to_hugetlbfs hmm-tests +local_config.* diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile index 2579242386ac..a9332a7cf33f 100644 --- a/tools/testing/selftests/vm/Makefile +++ b/tools/testing/selftests/vm/Makefile @@ -1,5 +1,8 @@ # SPDX-License-Identifier: GPL-2.0 # Makefile for vm selftests + +include local_config.mk + uname_M := $(shell uname -m 2>/dev/null || echo not) MACHINE ?= $(shell echo $(uname_M) | sed -e 's/aarch64.*/arm64/')
@@ -76,8 +79,6 @@ TEST_FILES := test_vmalloc.sh KSFT_KHDR_INSTALL := 1 include ../lib.mk
-$(OUTPUT)/hmm-tests: LDLIBS += -lhugetlbfs - ifeq ($(ARCH),x86_64) BINARIES_32 := $(patsubst %,$(OUTPUT)/%,$(BINARIES_32)) BINARIES_64 := $(patsubst %,$(OUTPUT)/%,$(BINARIES_64)) @@ -130,3 +131,22 @@ endif $(OUTPUT)/mlock-random-test: LDLIBS += -lcap
$(OUTPUT)/gup_test: ../../../../mm/gup_test.h + +$(OUTPUT)/hmm-tests: local_config.h + +# HMM_EXTRA_LIBS may get set in local_config.mk, or it may be left empty. +$(OUTPUT)/hmm-tests: LDLIBS += $(HMM_EXTRA_LIBS) + +local_config.mk local_config.h: check_config.sh + /bin/sh ./check_config.sh $(CC) + +EXTRA_CLEAN += local_config.mk local_config.h + +ifeq ($(HMM_EXTRA_LIBS),) +all: warn_missing_hugelibs + +warn_missing_hugelibs: + @echo ; \ + echo "Warning: missing libhugetlbfs support. Some HMM tests will be skipped." ; \ + echo +endif diff --git a/tools/testing/selftests/vm/check_config.sh b/tools/testing/selftests/vm/check_config.sh new file mode 100644 index 000000000000..079c8a40b85d --- /dev/null +++ b/tools/testing/selftests/vm/check_config.sh @@ -0,0 +1,31 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 +# +# Probe for libraries and create header files to record the results. Both C +# header files and Makefile include fragments are created. + +OUTPUT_H_FILE=local_config.h +OUTPUT_MKFILE=local_config.mk + +# libhugetlbfs +tmpname=$(mktemp) +tmpfile_c=${tmpname}.c +tmpfile_o=${tmpname}.o + +echo "#include <sys/types.h>" > $tmpfile_c +echo "#include <hugetlbfs.h>" >> $tmpfile_c +echo "int func(void) { return 0; }" >> $tmpfile_c + +CC=${1:?"Usage: $0 <compiler> # example compiler: gcc"} +$CC -c $tmpfile_c -o $tmpfile_o >/dev/null 2>&1 + +if [ -f $tmpfile_o ]; then + echo "#define LOCAL_CONFIG_HAVE_LIBHUGETLBFS 1" > $OUTPUT_H_FILE + echo "HMM_EXTRA_LIBS = -lhugetlbfs" > $OUTPUT_MKFILE +else + echo "// No libhugetlbfs support found" > $OUTPUT_H_FILE + echo "# No libhugetlbfs support found, so:" > $OUTPUT_MKFILE + echo "HMM_EXTRA_LIBS = " >> $OUTPUT_MKFILE +fi + +rm ${tmpname}.* diff --git a/tools/testing/selftests/vm/hmm-tests.c b/tools/testing/selftests/vm/hmm-tests.c index c9404ef9698e..5d1ac691b9f4 100644 --- a/tools/testing/selftests/vm/hmm-tests.c +++ b/tools/testing/selftests/vm/hmm-tests.c @@ -21,12 +21,16 @@ #include <strings.h> #include <time.h> #include <pthread.h> -#include <hugetlbfs.h> #include <sys/types.h> #include <sys/stat.h> #include <sys/mman.h> #include <sys/ioctl.h>
+#include "./local_config.h" +#ifdef LOCAL_CONFIG_HAVE_LIBHUGETLBFS +#include <hugetlbfs.h> +#endif + /* * This is a private UAPI to the kernel test module so it isn't exported * in the usual include/uapi/... directory. @@ -662,6 +666,7 @@ TEST_F(hmm, anon_write_huge) hmm_buffer_free(buffer); }
+#ifdef LOCAL_CONFIG_HAVE_LIBHUGETLBFS /* * Write huge TLBFS page. */ @@ -720,6 +725,7 @@ TEST_F(hmm, anon_write_hugetlbfs) buffer->ptr = NULL; hmm_buffer_free(buffer); } +#endif /* LOCAL_CONFIG_HAVE_LIBHUGETLBFS */
/* * Read mmap'ed file memory. @@ -1336,6 +1342,7 @@ TEST_F(hmm2, snapshot) hmm_buffer_free(buffer); }
+#ifdef LOCAL_CONFIG_HAVE_LIBHUGETLBFS /* * Test the hmm_range_fault() HMM_PFN_PMD flag for large pages that * should be mapped by a large page table entry. @@ -1411,6 +1418,7 @@ TEST_F(hmm, compound) buffer->ptr = NULL; hmm_buffer_free(buffer); } +#endif /* LOCAL_CONFIG_HAVE_LIBHUGETLBFS */
/* * Test two devices reading the same memory (double mapped).
Each invocation of userfaultfd for "anon" and "shmem" was taking about 6.5 sec to run, contributing to an overall run time of about 22 sec for run_vmtests.sh.
Reduce the size and bounce input values to the userfaultfd invocation within run_vmtests.sh, enough to get each invocation down to about 1.0 sec. This should still provide a reasonable smoke test, while staying within a nominal time budget of around 1 second or so per test. And this brings the overall running time of run_vmtests.sh down to 11 second.
Cc: Mike Rapoport rppt@linux.vnet.ibm.com Cc: Shuah Khan shuah@kernel.org Signed-off-by: John Hubbard jhubbard@nvidia.com --- tools/testing/selftests/vm/run_vmtests.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/testing/selftests/vm/run_vmtests.sh b/tools/testing/selftests/vm/run_vmtests.sh index 4ac84b350d9f..9e8837768ee2 100755 --- a/tools/testing/selftests/vm/run_vmtests.sh +++ b/tools/testing/selftests/vm/run_vmtests.sh @@ -160,7 +160,7 @@ fi echo "-------------------" echo "running userfaultfd" echo "-------------------" -./userfaultfd anon 128 32 +./userfaultfd anon 20 16 if [ $? -ne 0 ]; then echo "[FAIL]" exitcode=1 @@ -185,7 +185,7 @@ rm -f $mnt/ufd_test_file echo "-------------------------" echo "running userfaultfd_shmem" echo "-------------------------" -./userfaultfd shmem 128 32 +./userfaultfd shmem 20 16 if [ $? -ne 0 ]; then echo "[FAIL]" exitcode=1
linux-kselftest-mirror@lists.linaro.org
-
John Hubbard