Add ecmult_gen, ecmult_const and ecmult to benchmark #662

jonasnick · 2019-09-05T19:03:58Z

I was trying to determine the impact of ecmult_gen in schnorrsig signing and noticed that there is no way to bench this right now. The new benchmarks look like this:

$ ./bench_ecmult
ecmult_gen: min 20.9us / avg 21.2us / max 21.7us
ecmult_const: min 63.9us / avg 64.3us / max 64.8us
ecmult 1: min 49.4us / avg 49.7us / max 50.3us
ecmult 1g: min 39.8us / avg 40.0us / max 40.3us
ecmult 2g: min 27.2us / avg 27.3us / max 27.8us
ecmult_multi 1g: min 39.8us / avg 40.0us / max 40.2us
ecmult_multi 2g: min 27.2us / avg 27.4us / max 27.7us
ecmult_multi 3g: min 22.8us / avg 22.9us / max 23.1us
ecmult_multi 4g: min 20.6us / avg 20.8us / max 21.1us
ecmult_multi 5g: min 19.3us / avg 19.5us / max 19.7us

(Turns out ecmult_gen is 37% of the 55.8us that schnorrsig sign takes)

real-or-random

Approach ACK

elichai · 2019-09-27T20:28:36Z

src/bench_ecmult.c

+/* Hashes x into [0, POINTS) twice and store the result in offset1 and offset2. */
+static void hash_into_offset(bench_data* data, size_t x) {
+    data->offset1 = (x * 0x537b7f6f + 0x8f66a481) % POINTS;
+    data->offset2 = (x * 0x7f6f537b + 0x6a1a8f49) % POINTS;


where's this thing from?
I see it's also in bench_ecmult_setup but can't find the source of this

I think is this is just a standard integer hash function (akin to https://en.wikipedia.org/wiki/Universal_hashing#Hashing_integers) whose parameters sipa selected. But looking at it now it may be give poor results because POINTS is not prime.

I don't think points needs to be prime, so long as the multipliers are coprime to it.

fwiw they're all coprime to POINTS.

real-or-random

Looks good otherwise

real-or-random · 2019-10-28T14:09:08Z

src/bench_ecmult.c

+    bench_ecmult_teardown_helper(data, &data->offset1, &data->offset2, NULL);
+}
+
+static void bench_ecmult_1g(void* arg) {


Suggested change

static void bench_ecmult_1g(void* arg) {

static void bench_ecmult_0g(void* arg) {

real-or-random · 2019-10-28T14:09:34Z

src/bench_ecmult.c

+    }
+}
+
+static void bench_ecmult_1g_teardown(void* arg) {


Suggested change

static void bench_ecmult_1g_teardown(void* arg) {

static void bench_ecmult_0g_teardown(void* arg) {

and so on..

real-or-random · 2019-10-28T14:10:21Z

src/bench_ecmult.c

+    bench_ecmult_teardown_helper(data, NULL, NULL, &data->offset1);
+}
+
+static void bench_ecmult_2g(void* arg) {


jonasnick · 2019-10-28T16:59:21Z

@real-or-random I don't think so (unless we change it for ecmult_multi) but it took me a while to figure this out again. I added a commit that printfs an explanation. It's not pretty, but less confusing.

real-or-random · 2019-10-28T17:21:30Z

@real-or-random I don't think so (unless we change it for ecmult_multi) but it took me a while to figure this out again. I added a commit that printfs and explanation. It's not pretty, but less confusing.

Certainly pretty enough for a benchmark. Can you squash?

jonasnick · 2019-10-28T18:39:24Z

squashed

src/bench_ecmult.c

real-or-random · 2019-10-29T08:38:40Z

ACK a419f3d I read the code and tried it

jonasnick · 2020-04-30T14:06:23Z

Rebased and added a --help command instead of printing an explanation of the output every time this is run.

real-or-random · 2020-05-01T15:08:09Z

src/bench_ecmult.c

        } else {
            fprintf(stderr, "%s: unrecognized argument '%s'.\n", argv[0], argv[1]);
-            fprintf(stderr, "Use 'pippenger_wnaf', 'strauss_wnaf', 'simple' or no argument to benchmark a combined algorithm.\n");
+            fprintf(stderr, "Use '--help', 'pippenger_wnaf', 'strauss_wnaf', 'simple' or no argument to benchmark a combined algorithm.\n");


simply call help here?

Ah sorry, my suggestion was just to replace the entire else branch with help(), not to replace the --help by help. Feel free to change back, or to ignore.

real-or-random · 2020-05-01T15:23:05Z

src/bench_ecmult.c

            data.ecmult_multi = secp256k1_ecmult_multi_var;
-            secp256k1_scratch_space_destroy(data.ctx, data.scratch);
-            data.scratch = NULL;


That line got lost, and this was what made the difference between simple and combined? I see you merged this into "simple combined" in the printfs. Was this change on purpose?

Good catch. I also forgot what the simple algorithm actually does. I fixed this and added an explanation to the help().

jonasnick · 2021-05-28T17:42:55Z

@real-or-random care to reACK :) ?

real-or-random · 2021-05-31T20:37:14Z

ACK ee83dbf

jonasnick · 2021-05-31T20:56:19Z

Rebased to let CI do its thing

sipa · 2021-05-31T21:18:26Z

While we're at it: are no-g variants not interesting?

jonasnick · 2021-06-02T15:29:02Z

Hm with-g seems more interesting (schnorrsig batch verify, taproot tweaks, pedersen commitments,...) but we could add a flag or something for no-g (preferably in a separate PR imho).

real-or-random · 2021-06-05T15:58:44Z

ACK 8f879c2

elichai · 2021-06-06T11:38:46Z

tACK 8f879c2

nit, It is a little confusing that the first 5 benchmarks are unrelated to the flag being passed (Maybe we want to move the "Using XXX algorithm" print to just before the relevant benchmarks?)

real-or-random · 2021-06-06T11:57:03Z

nit, It is a little confusing that the first 5 benchmarks are unrelated to the flag being passed (Maybe we want to move the "Using XXX algorithm" print to just before the relevant benchmarks?)

Indeed but this can be done in a separate PR, too.

Let me take the opportunity to merge this PR now that we have two ACKs.

real-or-random reviewed Sep 5, 2019

View reviewed changes

elichai reviewed Sep 27, 2019

View reviewed changes

real-or-random reviewed Oct 28, 2019

View reviewed changes

jonasnick force-pushed the ecmult-bench branch from 5b3e10e to 6a5673f Compare October 28, 2019 18:39

real-or-random reviewed Oct 29, 2019

View reviewed changes

src/bench_ecmult.c Outdated Show resolved Hide resolved

jonasnick force-pushed the ecmult-bench branch 2 times, most recently from 5ebe192 to 82a5911 Compare April 30, 2020 14:05

real-or-random reviewed May 1, 2020

View reviewed changes

jonasnick force-pushed the ecmult-bench branch from 82a5911 to ee83dbf Compare May 1, 2020 20:45

jonasnick force-pushed the ecmult-bench branch 2 times, most recently from c6867e1 to 8f879c2 Compare May 31, 2021 20:45

jonasnick added 3 commits May 31, 2021 20:46

Clean up ecmult_bench to make space for more benchmarks

593e6ba

Add ecmult_gen, ecmult_const and ecmult to benchmark

2fe1b50

Fix array size in bench_ecmult

8f879c2

real-or-random merged commit 7973576 into bitcoin-core:master Jun 6, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add ecmult_gen, ecmult_const and ecmult to benchmark #662

Add ecmult_gen, ecmult_const and ecmult to benchmark #662

jonasnick commented Sep 5, 2019

real-or-random left a comment

elichai Sep 27, 2019

jonasnick Sep 28, 2019

gmaxwell Oct 15, 2019

jonasnick Oct 15, 2019

real-or-random left a comment

real-or-random Oct 28, 2019

real-or-random Oct 28, 2019

real-or-random Oct 28, 2019

jonasnick commented Oct 28, 2019 •

edited

Loading

real-or-random commented Oct 28, 2019

jonasnick commented Oct 28, 2019

real-or-random commented Oct 29, 2019

jonasnick commented Apr 30, 2020

real-or-random May 1, 2020 •

edited

Loading

jonasnick May 1, 2020

real-or-random May 31, 2021

jonasnick May 31, 2021

real-or-random May 1, 2020

jonasnick May 1, 2020

jonasnick commented May 28, 2021

real-or-random commented May 31, 2021

jonasnick commented May 31, 2021

sipa commented May 31, 2021

jonasnick commented Jun 2, 2021

real-or-random commented Jun 5, 2021

elichai commented Jun 6, 2021 •

edited

Loading

real-or-random commented Jun 6, 2021

	static void bench_ecmult_1g(void* arg) {
	static void bench_ecmult_0g(void* arg) {

	static void bench_ecmult_1g_teardown(void* arg) {
	static void bench_ecmult_0g_teardown(void* arg) {

Add ecmult_gen, ecmult_const and ecmult to benchmark #662

Add ecmult_gen, ecmult_const and ecmult to benchmark #662

Conversation

jonasnick commented Sep 5, 2019

real-or-random left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

real-or-random left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jonasnick commented Oct 28, 2019 • edited Loading

real-or-random commented Oct 28, 2019

jonasnick commented Oct 28, 2019

real-or-random commented Oct 29, 2019

jonasnick commented Apr 30, 2020

real-or-random May 1, 2020 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jonasnick commented May 28, 2021

real-or-random commented May 31, 2021

jonasnick commented May 31, 2021

sipa commented May 31, 2021

jonasnick commented Jun 2, 2021

real-or-random commented Jun 5, 2021

elichai commented Jun 6, 2021 • edited Loading

real-or-random commented Jun 6, 2021

jonasnick commented Oct 28, 2019 •

edited

Loading

real-or-random May 1, 2020 •

edited

Loading

elichai commented Jun 6, 2021 •

edited

Loading