Re: [PATCH 5/8] commit-graph: check all leading directories in changed path Bloom filters

["René Scharfe" <l.s.r@web.de>]

Am 15.06.20 um 22:14 schrieb SZEDER Gábor via GitGitGadget:
> From: =?UTF-8?q?SZEDER=20G=C3=A1bor?= <szeder.dev@gmail.com>
>
> The file 'dir/subdir/file' can only be modified if its leading
> directories 'dir' and 'dir/subdir' are modified as well.
>
> So when checking modified path Bloom filters looking for commits
> modifying a path with multiple path components, then check not only
> the full path in the Bloom filters, but all its leading directories as
> well.  Take care to check these paths in "deepest first" order,
> because it's the full path that is least likely to be modified, and
> the Bloom filter queries can short circuit sooner.
>
> This can significantly reduce the average false positive rate, by
> about an order of magnitude or three(!), and can further speed up
> pathspec-limited revision walks.  The table below compares the average
> false positive rate and runtime of
>
>   git rev-list HEAD -- "$path"
>
> before and after this change for 5000+ randomly* selected paths from
> each repository:
>
>                     Average false           Average        Average
>                     positive rate           runtime        runtime
>                   before     after     before     after   difference
>   ------------------------------------------------------------------
>   git             3.220%   0.7853%     0.0558s   0.0387s   -30.6%
>   linux           2.453%   0.0296%     0.1046s   0.0766s   -26.8%
>   tensorflow      2.536%   0.6977%     0.0594s   0.0420s   -29.2%

Nice!

>
> *Path selection was done with the following pipeline:
>
> 	git ls-tree -r --name-only HEAD | sort -R | head -n 5000
>
> The improvements in runtime are much smaller than the improvements in
> average false positive rate, as we are clearly reaching diminishing
> returns here.  However, all these timings depend on that accessing
> tree objects is reasonably fast (warm caches).  If we had a partial
> clone and the tree objects had to be fetched from a promisor remote,
> e.g.:
>
>   $ git clone --filter=tree:0 --bare file://.../webkit.git webkit.notrees.git
>   $ git -C webkit.git -c core.modifiedPathBloomFilters=1 \
>         commit-graph write --reachable
>   $ cp webkit.git/objects/info/commit-graph webkit.notrees.git/objects/info/
>   $ git -C webkit.notrees.git -c core.modifiedPathBloomFilters=1 \
>         rev-list HEAD -- "$path"
>
> then checking all leading path component can reduce the runtime from
> over an hour to a few seconds (and this is with the clone and the
> promisor on the same machine).
>
> This adjusts the tracing values in t4216-log-bloom.sh, which provides a
> concrete way to notice the improvement.
>
> Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
> Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
> ---
>  revision.c           | 35 ++++++++++++++++++++++++++---------
>  revision.h           |  6 ++++--
>  t/t4216-log-bloom.sh |  2 +-
>  3 files changed, 31 insertions(+), 12 deletions(-)
>
> diff --git a/revision.c b/revision.c
> index c644c660917..027ae3982b4 100644
> --- a/revision.c
> +++ b/revision.c
> @@ -670,9 +670,10 @@ static void prepare_to_use_bloom_filter(struct rev_info *revs)
>  {
>  	struct pathspec_item *pi;
>  	char *path_alloc = NULL;
> -	const char *path;
> +	const char *path, *p;
>  	int last_index;
> -	int len;
> +	size_t len;
> +	int path_component_nr = 0, j;
>
>  	if (!revs->commits)
>  		return;
> @@ -705,8 +706,22 @@ static void prepare_to_use_bloom_filter(struct rev_info *revs)
>
>  	len = strlen(path);
>
> -	revs->bloom_key = xmalloc(sizeof(struct bloom_key));
> -	fill_bloom_key(path, len, revs->bloom_key, revs->bloom_filter_settings);
> +	p = path;
> +	do {
> +		p = strchrnul(p + 1, '/');
> +		path_component_nr++;
> +	} while (p - path < len);

Hmm, that "+ 1" makes me a bit nervous.  Can we be sure that path is not
an empty string?

And shouldn't we use is_dir_sep() or find_last_dir_sep() instead of
hard-coding a slash?

> +
> +	revs->bloom_keys_nr = path_component_nr;
> +	ALLOC_ARRAY(revs->bloom_keys, revs->bloom_keys_nr);
> +
> +	p = path;
> +	for (j = 0; j < revs->bloom_keys_nr; j++) {
> +		p = strchrnul(p + 1, '/');

Same here, of course.

Also note that this puts shorter sub-strings first.


> +
> +		fill_bloom_key(path, p - path, &revs->bloom_keys[j],
> +			       revs->bloom_filter_settings);
> +	}
>
>  	if (trace2_is_enabled() && !bloom_filter_atexit_registered) {
>  		atexit(trace2_bloom_filter_statistics_atexit);
> @@ -720,7 +735,7 @@ static int check_maybe_different_in_bloom_filter(struct rev_info *revs,
>  						 struct commit *commit)
>  {
>  	struct bloom_filter *filter;
> -	int result;
> +	int result = 1, j;
>
>  	if (!revs->repo->objects->commit_graph)
>  		return -1;
> @@ -740,9 +755,11 @@ static int check_maybe_different_in_bloom_filter(struct rev_info *revs,
>  		return -1;
>  	}
>
> -	result = bloom_filter_contains(filter,
> -				       revs->bloom_key,
> -				       revs->bloom_filter_settings);
> +	for (j = 0; result && j < revs->bloom_keys_nr; j++) {
> +		result = bloom_filter_contains(filter,
> +					       &revs->bloom_keys[j],
> +					       revs->bloom_filter_settings);
> +	}

This checks shorter sub-strings first, contradicting the "deepest first"
strategy mentioned in the commit message.

This can easily be fixed by inverting the traversal of one of the loops,
of course.  Or perhaps do something like this?

	revs->bloom_keys = NULL;
	revs->bloom_keys_nr = 0;
	strbuf_add(&path, pi->match, pi->len);
	strbuf_trim_trailing_dir_sep(&path);
	for (;;) {
		const char *sep;
		ALLOC_GROW(revs->bloom_keys, revs->bloom_keys_nr + 1, alloc);
		fill_bloom_key(path.buf, path.len,
			       &revs->bloom_keys[revs->bloom_keys_nr++],
			       revs->bloom_filter_settings);
		sep = find_last_dir_sep(path.buf);
		if (!sep)
			break;
		strbuf_setlen(&path, sep - path.buf);
	}
	strbuf_release(&path);

The find_last_dir_sep() calls scan the first part of the string over and
over, which is a bit silly.  A strbuf_trim_trailing_path_component()
could start at the end and scan backwards if that turns out to be an
actual problem.

ALLOC_GROW wastes memory on revs->bloom_keys, and reallocating instead
of allocating the right size from the start has a cost as well, but I'd
expect this to be dwarfed by the actual revision walk.

>
>  	if (result)
>  		count_bloom_filter_maybe++;
> @@ -782,7 +799,7 @@ static int rev_compare_tree(struct rev_info *revs,
>  			return REV_TREE_SAME;
>  	}
>
> -	if (revs->bloom_key && !nth_parent) {
> +	if (revs->bloom_keys_nr && !nth_parent) {
>  		bloom_ret = check_maybe_different_in_bloom_filter(revs, commit);
>
>  		if (bloom_ret == 0)
> diff --git a/revision.h b/revision.h
> index 7c026fe41fc..abbfb4ab59a 100644
> --- a/revision.h
> +++ b/revision.h
> @@ -295,8 +295,10 @@ struct rev_info {
>  	struct topo_walk_info *topo_walk_info;
>
>  	/* Commit graph bloom filter fields */
> -	/* The bloom filter key for the pathspec */
> -	struct bloom_key *bloom_key;
> +	/* The bloom filter key(s) for the pathspec */
> +	struct bloom_key *bloom_keys;
> +	int bloom_keys_nr;
> +
>  	/*
>  	 * The bloom filter settings used to generate the key.
>  	 * This is loaded from the commit-graph being used.
> diff --git a/t/t4216-log-bloom.sh b/t/t4216-log-bloom.sh
> index c7011f33e2c..c13b97d3bda 100755
> --- a/t/t4216-log-bloom.sh
> +++ b/t/t4216-log-bloom.sh
> @@ -142,7 +142,7 @@ test_expect_success 'setup - add commit-graph to the chain with Bloom filters' '
>
>  test_bloom_filters_used_when_some_filters_are_missing () {
>  	log_args=$1
> -	bloom_trace_prefix="statistics:{\"filter_not_present\":3,\"zero_length_filter\":0,\"maybe\":8,\"definitely_not\":6"
> +	bloom_trace_prefix="statistics:{\"filter_not_present\":3,\"zero_length_filter\":0,\"maybe\":6,\"definitely_not\":8"
>  	setup "$log_args" &&
>  	grep -q "$bloom_trace_prefix" "$TRASH_DIRECTORY/trace.perf" &&
>  	test_cmp log_wo_bloom log_w_bloom
>