2
0
mirror of https://github.com/edk2-porting/linux-next.git synced 2024-12-17 01:34:00 +08:00
linux-next/lib/interval_tree_test_main.c

107 lines
2.3 KiB
C
Raw Normal View History

rbtree: add prio tree and interval tree tests Patch 1 implements support for interval trees, on top of the augmented rbtree API. It also adds synthetic tests to compare the performance of interval trees vs prio trees. Short answers is that interval trees are slightly faster (~25%) on insert/erase, and much faster (~2.4 - 3x) on search. It is debatable how realistic the synthetic test is, and I have not made such measurements yet, but my impression is that interval trees would still come out faster. Patch 2 uses a preprocessor template to make the interval tree generic, and uses it as a replacement for the vma prio_tree. Patch 3 takes the other prio_tree user, kmemleak, and converts it to use a basic rbtree. We don't actually need the augmented rbtree support here because the intervals are always non-overlapping. Patch 4 removes the now-unused prio tree library. Patch 5 proposes an additional optimization to rb_erase_augmented, now providing it as an inline function so that the augmented callbacks can be inlined in. This provides an additional 5-10% performance improvement for the interval tree insert/erase benchmark. There is a maintainance cost as it exposes augmented rbtree users to some of the rbtree library internals; however I think this cost shouldn't be too high as I expect the augmented rbtree will always have much less users than the base rbtree. I should probably add a quick summary of why I think it makes sense to replace prio trees with augmented rbtree based interval trees now. One of the drivers is that we need augmented rbtrees for Rik's vma gap finding code, and once you have them, it just makes sense to use them for interval trees as well, as this is the simpler and more well known algorithm. prio trees, in comparison, seem *too* clever: they impose an additional 'heap' constraint on the tree, which they use to guarantee a faster worst-case complexity of O(k+log N) for stabbing queries in a well-balanced prio tree, vs O(k*log N) for interval trees (where k=number of matches, N=number of intervals). Now this sounds great, but in practice prio trees don't realize this theorical benefit. First, the additional constraint makes them harder to update, so that the kernel implementation has to simplify things by balancing them like a radix tree, which is not always ideal. Second, the fact that there are both index and heap properties makes both tree manipulation and search more complex, which results in a higher multiplicative time constant. As it turns out, the simple interval tree algorithm ends up running faster than the more clever prio tree. This patch: Add two test modules: - prio_tree_test measures the performance of lib/prio_tree.c, both for insertion/removal and for stabbing searches - interval_tree_test measures the performance of a library of equivalent functionality, built using the augmented rbtree support. In order to support the second test module, lib/interval_tree.c is introduced. It is kept separate from the interval_tree_test main file for two reasons: first we don't want to provide an unfair advantage over prio_tree_test by having everything in a single compilation unit, and second there is the possibility that the interval tree functionality could get some non-test users in kernel over time. Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Hillf Danton <dhillf@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-10-09 07:31:23 +08:00
#include <linux/module.h>
#include <linux/interval_tree.h>
#include <linux/random.h>
#include <asm/timex.h>
#define NODES 100
#define PERF_LOOPS 100000
#define SEARCHES 100
#define SEARCH_LOOPS 10000
static struct rb_root root = RB_ROOT;
static struct interval_tree_node nodes[NODES];
static u32 queries[SEARCHES];
static struct rnd_state rnd;
static inline unsigned long
search(unsigned long query, struct rb_root *root)
{
struct interval_tree_node *node;
unsigned long results = 0;
for (node = interval_tree_iter_first(root, query, query); node;
node = interval_tree_iter_next(node, query, query))
results++;
return results;
}
static void init(void)
{
int i;
for (i = 0; i < NODES; i++) {
random32: rename random32 to prandom This renames all random32 functions to have 'prandom_' prefix as follows: void prandom_seed(u32 seed); /* rename from srandom32() */ u32 prandom_u32(void); /* rename from random32() */ void prandom_seed_state(struct rnd_state *state, u64 seed); /* rename from prandom32_seed() */ u32 prandom_u32_state(struct rnd_state *state); /* rename from prandom32() */ The purpose of this renaming is to prevent some kernel developers from assuming that prandom32() and random32() might imply that only prandom32() was the one using a pseudo-random number generator by prandom32's "p", and the result may be a very embarassing security exposure. This concern was expressed by Theodore Ts'o. And furthermore, I'm going to introduce new functions for getting the requested number of pseudo-random bytes. If I continue to use both prandom32 and random32 prefixes for these functions, the confusion is getting worse. As a result of this renaming, "prandom_" is the common prefix for pseudo-random number library. Currently, srandom32() and random32() are preserved because it is difficult to rename too many users at once. Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Robert Love <robert.w.love@intel.com> Cc: Michel Lespinasse <walken@google.com> Cc: Valdis Kletnieks <valdis.kletnieks@vt.edu> Cc: David Laight <david.laight@aculab.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Artem Bityutskiy <dedekind1@gmail.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-12-18 08:04:23 +08:00
u32 a = prandom_u32_state(&rnd);
u32 b = prandom_u32_state(&rnd);
rbtree: add prio tree and interval tree tests Patch 1 implements support for interval trees, on top of the augmented rbtree API. It also adds synthetic tests to compare the performance of interval trees vs prio trees. Short answers is that interval trees are slightly faster (~25%) on insert/erase, and much faster (~2.4 - 3x) on search. It is debatable how realistic the synthetic test is, and I have not made such measurements yet, but my impression is that interval trees would still come out faster. Patch 2 uses a preprocessor template to make the interval tree generic, and uses it as a replacement for the vma prio_tree. Patch 3 takes the other prio_tree user, kmemleak, and converts it to use a basic rbtree. We don't actually need the augmented rbtree support here because the intervals are always non-overlapping. Patch 4 removes the now-unused prio tree library. Patch 5 proposes an additional optimization to rb_erase_augmented, now providing it as an inline function so that the augmented callbacks can be inlined in. This provides an additional 5-10% performance improvement for the interval tree insert/erase benchmark. There is a maintainance cost as it exposes augmented rbtree users to some of the rbtree library internals; however I think this cost shouldn't be too high as I expect the augmented rbtree will always have much less users than the base rbtree. I should probably add a quick summary of why I think it makes sense to replace prio trees with augmented rbtree based interval trees now. One of the drivers is that we need augmented rbtrees for Rik's vma gap finding code, and once you have them, it just makes sense to use them for interval trees as well, as this is the simpler and more well known algorithm. prio trees, in comparison, seem *too* clever: they impose an additional 'heap' constraint on the tree, which they use to guarantee a faster worst-case complexity of O(k+log N) for stabbing queries in a well-balanced prio tree, vs O(k*log N) for interval trees (where k=number of matches, N=number of intervals). Now this sounds great, but in practice prio trees don't realize this theorical benefit. First, the additional constraint makes them harder to update, so that the kernel implementation has to simplify things by balancing them like a radix tree, which is not always ideal. Second, the fact that there are both index and heap properties makes both tree manipulation and search more complex, which results in a higher multiplicative time constant. As it turns out, the simple interval tree algorithm ends up running faster than the more clever prio tree. This patch: Add two test modules: - prio_tree_test measures the performance of lib/prio_tree.c, both for insertion/removal and for stabbing searches - interval_tree_test measures the performance of a library of equivalent functionality, built using the augmented rbtree support. In order to support the second test module, lib/interval_tree.c is introduced. It is kept separate from the interval_tree_test main file for two reasons: first we don't want to provide an unfair advantage over prio_tree_test by having everything in a single compilation unit, and second there is the possibility that the interval tree functionality could get some non-test users in kernel over time. Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Hillf Danton <dhillf@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-10-09 07:31:23 +08:00
if (a <= b) {
nodes[i].start = a;
nodes[i].last = b;
} else {
nodes[i].start = b;
nodes[i].last = a;
}
}
for (i = 0; i < SEARCHES; i++)
random32: rename random32 to prandom This renames all random32 functions to have 'prandom_' prefix as follows: void prandom_seed(u32 seed); /* rename from srandom32() */ u32 prandom_u32(void); /* rename from random32() */ void prandom_seed_state(struct rnd_state *state, u64 seed); /* rename from prandom32_seed() */ u32 prandom_u32_state(struct rnd_state *state); /* rename from prandom32() */ The purpose of this renaming is to prevent some kernel developers from assuming that prandom32() and random32() might imply that only prandom32() was the one using a pseudo-random number generator by prandom32's "p", and the result may be a very embarassing security exposure. This concern was expressed by Theodore Ts'o. And furthermore, I'm going to introduce new functions for getting the requested number of pseudo-random bytes. If I continue to use both prandom32 and random32 prefixes for these functions, the confusion is getting worse. As a result of this renaming, "prandom_" is the common prefix for pseudo-random number library. Currently, srandom32() and random32() are preserved because it is difficult to rename too many users at once. Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Robert Love <robert.w.love@intel.com> Cc: Michel Lespinasse <walken@google.com> Cc: Valdis Kletnieks <valdis.kletnieks@vt.edu> Cc: David Laight <david.laight@aculab.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Artem Bityutskiy <dedekind1@gmail.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-12-18 08:04:23 +08:00
queries[i] = prandom_u32_state(&rnd);
rbtree: add prio tree and interval tree tests Patch 1 implements support for interval trees, on top of the augmented rbtree API. It also adds synthetic tests to compare the performance of interval trees vs prio trees. Short answers is that interval trees are slightly faster (~25%) on insert/erase, and much faster (~2.4 - 3x) on search. It is debatable how realistic the synthetic test is, and I have not made such measurements yet, but my impression is that interval trees would still come out faster. Patch 2 uses a preprocessor template to make the interval tree generic, and uses it as a replacement for the vma prio_tree. Patch 3 takes the other prio_tree user, kmemleak, and converts it to use a basic rbtree. We don't actually need the augmented rbtree support here because the intervals are always non-overlapping. Patch 4 removes the now-unused prio tree library. Patch 5 proposes an additional optimization to rb_erase_augmented, now providing it as an inline function so that the augmented callbacks can be inlined in. This provides an additional 5-10% performance improvement for the interval tree insert/erase benchmark. There is a maintainance cost as it exposes augmented rbtree users to some of the rbtree library internals; however I think this cost shouldn't be too high as I expect the augmented rbtree will always have much less users than the base rbtree. I should probably add a quick summary of why I think it makes sense to replace prio trees with augmented rbtree based interval trees now. One of the drivers is that we need augmented rbtrees for Rik's vma gap finding code, and once you have them, it just makes sense to use them for interval trees as well, as this is the simpler and more well known algorithm. prio trees, in comparison, seem *too* clever: they impose an additional 'heap' constraint on the tree, which they use to guarantee a faster worst-case complexity of O(k+log N) for stabbing queries in a well-balanced prio tree, vs O(k*log N) for interval trees (where k=number of matches, N=number of intervals). Now this sounds great, but in practice prio trees don't realize this theorical benefit. First, the additional constraint makes them harder to update, so that the kernel implementation has to simplify things by balancing them like a radix tree, which is not always ideal. Second, the fact that there are both index and heap properties makes both tree manipulation and search more complex, which results in a higher multiplicative time constant. As it turns out, the simple interval tree algorithm ends up running faster than the more clever prio tree. This patch: Add two test modules: - prio_tree_test measures the performance of lib/prio_tree.c, both for insertion/removal and for stabbing searches - interval_tree_test measures the performance of a library of equivalent functionality, built using the augmented rbtree support. In order to support the second test module, lib/interval_tree.c is introduced. It is kept separate from the interval_tree_test main file for two reasons: first we don't want to provide an unfair advantage over prio_tree_test by having everything in a single compilation unit, and second there is the possibility that the interval tree functionality could get some non-test users in kernel over time. Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Hillf Danton <dhillf@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-10-09 07:31:23 +08:00
}
static int interval_tree_test_init(void)
{
int i, j;
unsigned long results;
cycles_t time1, time2, time;
printk(KERN_ALERT "interval tree insert/remove");
random32: rename random32 to prandom This renames all random32 functions to have 'prandom_' prefix as follows: void prandom_seed(u32 seed); /* rename from srandom32() */ u32 prandom_u32(void); /* rename from random32() */ void prandom_seed_state(struct rnd_state *state, u64 seed); /* rename from prandom32_seed() */ u32 prandom_u32_state(struct rnd_state *state); /* rename from prandom32() */ The purpose of this renaming is to prevent some kernel developers from assuming that prandom32() and random32() might imply that only prandom32() was the one using a pseudo-random number generator by prandom32's "p", and the result may be a very embarassing security exposure. This concern was expressed by Theodore Ts'o. And furthermore, I'm going to introduce new functions for getting the requested number of pseudo-random bytes. If I continue to use both prandom32 and random32 prefixes for these functions, the confusion is getting worse. As a result of this renaming, "prandom_" is the common prefix for pseudo-random number library. Currently, srandom32() and random32() are preserved because it is difficult to rename too many users at once. Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Robert Love <robert.w.love@intel.com> Cc: Michel Lespinasse <walken@google.com> Cc: Valdis Kletnieks <valdis.kletnieks@vt.edu> Cc: David Laight <david.laight@aculab.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Artem Bityutskiy <dedekind1@gmail.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-12-18 08:04:23 +08:00
prandom_seed_state(&rnd, 3141592653589793238ULL);
rbtree: add prio tree and interval tree tests Patch 1 implements support for interval trees, on top of the augmented rbtree API. It also adds synthetic tests to compare the performance of interval trees vs prio trees. Short answers is that interval trees are slightly faster (~25%) on insert/erase, and much faster (~2.4 - 3x) on search. It is debatable how realistic the synthetic test is, and I have not made such measurements yet, but my impression is that interval trees would still come out faster. Patch 2 uses a preprocessor template to make the interval tree generic, and uses it as a replacement for the vma prio_tree. Patch 3 takes the other prio_tree user, kmemleak, and converts it to use a basic rbtree. We don't actually need the augmented rbtree support here because the intervals are always non-overlapping. Patch 4 removes the now-unused prio tree library. Patch 5 proposes an additional optimization to rb_erase_augmented, now providing it as an inline function so that the augmented callbacks can be inlined in. This provides an additional 5-10% performance improvement for the interval tree insert/erase benchmark. There is a maintainance cost as it exposes augmented rbtree users to some of the rbtree library internals; however I think this cost shouldn't be too high as I expect the augmented rbtree will always have much less users than the base rbtree. I should probably add a quick summary of why I think it makes sense to replace prio trees with augmented rbtree based interval trees now. One of the drivers is that we need augmented rbtrees for Rik's vma gap finding code, and once you have them, it just makes sense to use them for interval trees as well, as this is the simpler and more well known algorithm. prio trees, in comparison, seem *too* clever: they impose an additional 'heap' constraint on the tree, which they use to guarantee a faster worst-case complexity of O(k+log N) for stabbing queries in a well-balanced prio tree, vs O(k*log N) for interval trees (where k=number of matches, N=number of intervals). Now this sounds great, but in practice prio trees don't realize this theorical benefit. First, the additional constraint makes them harder to update, so that the kernel implementation has to simplify things by balancing them like a radix tree, which is not always ideal. Second, the fact that there are both index and heap properties makes both tree manipulation and search more complex, which results in a higher multiplicative time constant. As it turns out, the simple interval tree algorithm ends up running faster than the more clever prio tree. This patch: Add two test modules: - prio_tree_test measures the performance of lib/prio_tree.c, both for insertion/removal and for stabbing searches - interval_tree_test measures the performance of a library of equivalent functionality, built using the augmented rbtree support. In order to support the second test module, lib/interval_tree.c is introduced. It is kept separate from the interval_tree_test main file for two reasons: first we don't want to provide an unfair advantage over prio_tree_test by having everything in a single compilation unit, and second there is the possibility that the interval tree functionality could get some non-test users in kernel over time. Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Hillf Danton <dhillf@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-10-09 07:31:23 +08:00
init();
time1 = get_cycles();
for (i = 0; i < PERF_LOOPS; i++) {
for (j = 0; j < NODES; j++)
interval_tree_insert(nodes + j, &root);
for (j = 0; j < NODES; j++)
interval_tree_remove(nodes + j, &root);
}
time2 = get_cycles();
time = time2 - time1;
time = div_u64(time, PERF_LOOPS);
printk(" -> %llu cycles\n", (unsigned long long)time);
printk(KERN_ALERT "interval tree search");
for (j = 0; j < NODES; j++)
interval_tree_insert(nodes + j, &root);
time1 = get_cycles();
results = 0;
for (i = 0; i < SEARCH_LOOPS; i++)
for (j = 0; j < SEARCHES; j++)
results += search(queries[j], &root);
time2 = get_cycles();
time = time2 - time1;
time = div_u64(time, SEARCH_LOOPS);
results = div_u64(results, SEARCH_LOOPS);
printk(" -> %llu cycles (%lu results)\n",
(unsigned long long)time, results);
return -EAGAIN; /* Fail will directly unload the module */
}
static void interval_tree_test_exit(void)
{
printk(KERN_ALERT "test exit\n");
}
module_init(interval_tree_test_init)
module_exit(interval_tree_test_exit)
MODULE_LICENSE("GPL");
MODULE_AUTHOR("Michel Lespinasse");
MODULE_DESCRIPTION("Interval Tree test");