Fix and improve __maps__fixup_overlap_and_insert
Ian Rogers
reviewercover Fix and improve __maps__fixup_overlap_and_insert
Fix latent unlikely bugs in __maps__fixup_overlap_and_insert.
Improve __maps__fixup_overlap_and_insert's performance 21x in the case of overlapping mmaps. [email protected] reported slowness opening perf.data files from chromium where the files contained a large number of overlapping mappings. Improve this case primarily by avoiding unnecessary sorting.
Unscientific timing data processing a perf.data file with overlapping mmap events from chromium:
Before:
real 0m9.856s user 0m9.637s sys 0m0.204s
After:
real 0m0.675s user 0m0.454s sys 0m0.196s
Tested with address/leak sanitizer, invariant checks and validating the before and after output are identical.
Ian Rogers (3):
perf maps: Fix use after free in __maps__fixup_overlap_and_insert perf maps: Reduce sorting for overlapping mappings perf maps: Add/use a sorted insert for fixup overlap and insert
tools/perf/util/maps.c | 113 +++++++++++++++++++++++++++++++++-------- 1 file changed, 92 insertions(+), 21 deletions(-)
Signature · 1 line
2.45.0.rc1.225.g2a3ae87e7f-goog
Ian Rogers
reviewerPATCH 1/3 perf maps: Fix use after free in __maps__fixup_overlap_and_insert
In the case 'before' and 'after' are broken out from pos, maps_by_address may be changed by __maps__insert, as such it needs re-reading.
Don't ignore the return value from __maps_insert.
--- tools/perf/util/maps.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)
tools/perf/util/maps.c +5 -4
@@ -741,7 +741,6 @@ static unsigned int first_ending_after(struct maps *maps, const struct map *map)*/ static int __maps__fixup_overlap_and_insert (struct maps *maps,struct map *new) { -struct map **maps_by_address;int err =0 ;FILE *fp =debug_file ();
@@ -749,12 +748,12 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)if (!maps__maps_by_address_sorted (maps))__maps__sort_by_address (maps); - maps_by_address =maps__maps_by_address (maps);/* * Iterate through entries where the end of the existing entry is * greater-than the new map's start. */ for (unsigned int i =first_ending_after (maps, new); i <maps__nr_maps (maps); ) { +struct map **maps_by_address =maps__maps_by_address (maps);struct map *pos = maps_by_address[i];struct map *before =NULL , *after =NULL ;
@@ -821,8 +820,10 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)/* Maps are still ordered, go to next one. */ i++;if (after) { -__maps__insert (maps, after); + err =__maps__insert (maps, after);map__put (after); +if (err) + goto out_err;if (!maps__maps_by_address_sorted (maps)) {/* * Sorting broken so invariants don't
@@ -851,7 +852,7 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)check_invariants (maps); }/* Add the map. */ -__maps__insert (maps, new); + err =__maps__insert (maps, new); out_err:return err; }
Signature · 1 line
2.45.0.rc1.225.g2a3ae87e7f-goog
Ian Rogers
reviewerPATCH 2/3 perf maps: Reduce sorting for overlapping mappings
In reply to Ian RogersWhen an 'after' map is generated the 'new' map must be before it so terminate iterating and don't resort. If the entry 'pos' is entirely overlapped by the 'new' mapping then don't remove and insert the mapping, just replace - again to remove sorting.
For a perf report on a perf.data file containing overlapping mappings the time numbers are:
Before:
real 0m9.856s user 0m9.637s sys 0m0.204s
After:
real 0m5.894s user 0m5.650s sys 0m0.231s
--- tools/perf/util/maps.c | 55 +++++++++++++++++++++++++++--------------- 1 file changed, 36 insertions(+), 19 deletions(-)
tools/perf/util/maps.c +36 -19
@@ -744,7 +744,6 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)int err =0 ;FILE *fp =debug_file (); -sort_again:if (!maps__maps_by_address_sorted (maps))__maps__sort_by_address (maps);
@@ -820,36 +819,54 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)/* Maps are still ordered, go to next one. */ i++;if (after) { - err =__maps__insert (maps, after); -map__put (after); -if (err) - goto out_err; -if (!maps__maps_by_address_sorted (maps)) { -/* -* Sorting broken so invariants don't -* hold, sort and go again. -*/ - goto sort_again; - }/* -* Maps are still ordered, skip after and go to -* next one (terminate loop). +* 'before' and 'after' mean 'new' split the +* 'pos' mapping and therefore there are no +* later mappings. */ - i++; + err =__maps__insert (maps, new); +if (!err) + err =__maps__insert (maps, after); +map__put (after); +check_invariants (maps); +return err; } +check_invariants (maps); }else if (after) { +/* +* 'after' means 'new' split 'pos' and there are no +* later mappings. +*/ map__put (maps_by_address[i]); - maps_by_address[i] = after; -/* Maps are ordered, go to next one. */ - i++; + maps_by_address[i] =map__get (new); + err =__maps__insert (maps, after); +map__put (after); +check_invariants (maps); +return err; }else { +struct map *next =NULL ; + +if (i +1 <maps__nr_maps (maps)) + next = maps_by_address[i +1 ]; + +if (!next ||map__start (next) >=map__end (new)) { +/* +* Replace existing mapping and end knowing +* there aren't later overlapping or any +* mappings. +*/ +map__put (maps_by_address[i]); + maps_by_address[i] =map__get (new); +check_invariants (maps); +return err; + }__maps__remove (maps, pos); +check_invariants (maps);/* * Maps are ordered but no need to increase `i` as the * later maps were moved down. */ } -check_invariants (maps); }/* Add the map. */ err =__maps__insert (maps, new);
Signature · 1 line
2.45.0.rc1.225.g2a3ae87e7f-goog
Ian Rogers
reviewerPATCH 3/3 perf maps: Add/use a sorted insert for fixup overlap and insert
In reply to Ian RogersData may have lots of overlapping mmaps. The regular insert adds at the end and relies on a later sort. For data with overlapping mappings the sort will happen during a subsequent maps__find or __maps__fixup_overlap_and_insert, there's never a period where the inserted maps buffer up and a single sort happens. To avoid back to back sorts, maintain the sort order when fixing up and inserting. Previously the first_ending_after search was O(log n) where n is the size of maps, and the insert was O(1) but because of the continuous sorting was becoming O(n*log(n)). With maintaining sort order, the insert now becomes O(n) for a memmove.
For a perf report on a perf.data file containing overlapping mappings the time numbers are:
Before:
real 0m5.894s user 0m5.650s sys 0m0.231s
After:
real 0m0.675s user 0m0.454s sys 0m0.196s
--- tools/perf/util/maps.c | 65 ++++++++++++++++++++++++++++++++++++++---- 1 file changed, 59 insertions(+), 6 deletions(-)
tools/perf/util/maps.c +59 -6
@@ -735,6 +735,60 @@ static unsigned int first_ending_after(struct maps *maps, const struct map *map)return first; } +static int __maps__insert_sorted (struct maps *maps,unsigned int first_after_index, +struct map *new1,struct map *new2) +{ +struct map **maps_by_address =maps__maps_by_address (maps); +struct map **maps_by_name =maps__maps_by_name (maps); +unsigned int nr_maps =maps__nr_maps (maps); +unsigned int nr_allocate =RC_CHK_ACCESS (maps)->nr_maps_allocated; +unsigned int to_add = new2 ?2 :1 ; + +assert (maps__maps_by_address_sorted (maps)); +assert (first_after_index == nr_maps || +map__end (new1) <=map__start (maps_by_address[first_after_index])); +assert (!new2 ||map__end (new1) <=map__start (new2)); +assert (first_after_index == nr_maps || !new2 || +map__end (new2) <=map__start (maps_by_address[first_after_index])); + +if (nr_maps + to_add > nr_allocate) { + nr_allocate = !nr_allocate ?32 : nr_allocate *2 ; + + maps_by_address =realloc (maps_by_address, nr_allocate *sizeof (new1)); +if (!maps_by_address) +return -ENOMEM ; + +maps__set_maps_by_address (maps, maps_by_address); +if (maps_by_name) { + maps_by_name =realloc (maps_by_name, nr_allocate *sizeof (new1)); +if (!maps_by_name) { +/* +* If by name fails, just disable by name and it will +* recompute next time it is required. +*/ +__maps__free_maps_by_name (maps); + } +maps__set_maps_by_name (maps, maps_by_name); + } +RC_CHK_ACCESS (maps)->nr_maps_allocated = nr_allocate; + } +memmove (&maps_by_address[first_after_index+to_add], + &maps_by_address[first_after_index], + (nr_maps - first_after_index) *sizeof (new1)); + maps_by_address[first_after_index] =map__get (new1); +if (maps_by_name) + maps_by_name[nr_maps] =map__get (new1); +if (new2) { + maps_by_address[first_after_index +1 ] =map__get (new2); +if (maps_by_name) + maps_by_name[nr_maps +1 ] =map__get (new2); + } +RC_CHK_ACCESS (maps)->nr_maps = nr_maps + to_add; +maps__set_maps_by_name_sorted (maps, false); +check_invariants (maps); +return 0 ; +} +/* * Adds new to maps, if new overlaps existing entries then the existing maps are * adjusted or removed so that new fits without overlapping any entries.
@@ -743,6 +797,7 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new) {int err =0 ;FILE *fp =debug_file (); +unsigned int i;if (!maps__maps_by_address_sorted (maps))__maps__sort_by_address (maps);
@@ -751,7 +806,7 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)* Iterate through entries where the end of the existing entry is * greater-than the new map's start. */ -for (unsigned int i =first_ending_after (maps, new); i <maps__nr_maps (maps); ) { +for (i =first_ending_after (maps, new); i <maps__nr_maps (maps); ) {struct map **maps_by_address =maps__maps_by_address (maps);struct map *pos = maps_by_address[i];struct map *before =NULL , *after =NULL ;
@@ -824,9 +879,7 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)* 'pos' mapping and therefore there are no * later mappings. */ - err =__maps__insert (maps, new); -if (!err) - err =__maps__insert (maps, after); + err =__maps__insert_sorted (maps, i, new, after);map__put (after);check_invariants (maps);return err;
@@ -839,7 +892,7 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)*/ map__put (maps_by_address[i]); maps_by_address[i] =map__get (new); - err =__maps__insert (maps, after); + err =__maps__insert_sorted (maps, i +1 , after,NULL );map__put (after);check_invariants (maps);return err;
@@ -869,7 +922,7 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new) } }/* Add the map. */ - err =__maps__insert (maps, new); + err =__maps__insert_sorted (maps, i, new,NULL ); out_err:return err; }
Signature · 1 line
2.45.0.rc1.225.g2a3ae87e7f-goog
Ian Rogers
reviewerQuoted 22 lines · Ian Rogers
Fix latent unlikely bugs in
__maps__fixup_overlap_and_insert.Improve __maps__fixup_overlap_and_insert's performance 21x in the case of overlapping mmaps. [email protected] reported slowness opening perf.data files from chromium where the files contained a large number of overlapping mappings. Improve this case primarily by avoiding unnecessary sorting.
Unscientific timing data processing a perf.data file with overlapping mmap events from chromium:
Before:
real 0m9.856s user 0m9.637s sys 0m0.204sAfter:
real 0m0.675s user 0m0.454s sys 0m0.196s
Quoted 7 lines
Tested with address/leak sanitizer, invariant checks and validating the before and after output are identical.
Ian Rogers (3):
perf maps: Fix use after free in__maps__fixup_overlap_and_insertperf maps: Reduce sorting for overlapping mappings perf maps: Add/use a sorted insert for fixup overlap and insert
Ping. Thanks,
Ian
Quoted 6 lines
tools/perf/util/maps.c | 113 +++++++++++++++++++++++++++++++++-------- 1 file changed, 92 insertions(+), 21 deletions(-)--
2.45.0.rc1.225.g2a3ae87e7f-goog
Quoted 24 lines · Ian Rogers
Fix latent unlikely bugs in
__maps__fixup_overlap_and_insert.Improve __maps__fixup_overlap_and_insert's performance 21x in the case of overlapping mmaps. [email protected] reported slowness opening perf.data files from chromium where the files contained a large number of overlapping mappings. Improve this case primarily by avoiding unnecessary sorting.
Unscientific timing data processing a perf.data file with overlapping mmap events from chromium:
Before:
real 0m9.856s user 0m9.637s sys 0m0.204sAfter:
real 0m0.675s user 0m0.454s sys 0m0.196sTested with address/leak sanitizer, invariant checks and validating the before and after output are identical.
Quoted 8 lines
Ian Rogers (3):
perf maps: Fix use after free in__maps__fixup_overlap_and_insertperf maps: Reduce sorting for overlapping mappings perf maps: Add/use a sorted insert for fixup overlap and inserttools/perf/util/maps.c | 113 +++++++++++++++++++++++++++++++++-------- 1 file changed, 92 insertions(+), 21 deletions(-)
I'm wondering if there is any point in the non sorted insert any more?
Maps could be always either sorted by name or sorted by address and insert() is always a sorted/fixup-overlaps insert depending on the sort style of the list.
Ian Rogers
reviewerQuoted 37 lines · James Clark
Quoted 24 lines · Ian Rogers
Fix latent unlikely bugs in
__maps__fixup_overlap_and_insert.Improve __maps__fixup_overlap_and_insert's performance 21x in the case of overlapping mmaps. [email protected] reported slowness opening perf.data files from chromium where the files contained a large number of overlapping mappings. Improve this case primarily by avoiding unnecessary sorting.
Unscientific timing data processing a perf.data file with overlapping mmap events from chromium:
Before:
real 0m9.856s user 0m9.637s sys 0m0.204sAfter:
real 0m0.675s user 0m0.454s sys 0m0.196sTested with address/leak sanitizer, invariant checks and validating the before and after output are identical.
Quoted 8 lines
Ian Rogers (3):
perf maps: Fix use after free in__maps__fixup_overlap_and_insertperf maps: Reduce sorting for overlapping mappings perf maps: Add/use a sorted insert for fixup overlap and inserttools/perf/util/maps.c | 113 +++++++++++++++++++++++++++++++++-------- 1 file changed, 92 insertions(+), 21 deletions(-)
Quoted 7 lines
Reviewed-by James Clark <[email protected]>I'm wondering if there is any point in the non sorted insert any more?
Maps could be always either sorted by name or sorted by address and
insert()is always a sorted/fixup-overlaps insert depending on the sort style of the list.
One of the motivations for the sorted array, instead of an rbtree, was to simplify reference count checking. We're in much better shape with that these days, I think the worst "memory leak" is the dsos holding onto a dso and its symbols indefinitely, instead of removing older unused dsos. It'd be hard to go back to an rb tree with reference counting.
For the rbtree insert it was O(log N), the sorted insert these changes add is O(N) and the regular insert without sorting is O(1). A common case from scanning /proc/pid/maps is that when map entries are inserted they are inserted in order. The O(1) insert checks that and maintains that the maps are sorted.
https://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git/tree/tools/perf/util/maps.c?h=perf-tools-next#n469
For the synthesized maps we should be able to insert N map entries in O(N). The sorted insert would be similar as the memmoves would always copy 0 elements. So I can see an argument for keeping the array always sorted. For the perf.data files we commonly process synthesized mmaps dominate. In the case for this patch, JIT data dominated, with frequent overlapping mapping changes. I guess the biggest hurdle to just always keeping things sorted is the mental hurdle of ignoring the worst insert complexity, which should never apply given how the maps are commonly built.
Thanks,
Ian
Quoted 9 lines · Ian Rogers
Fix latent unlikely bugs in
__maps__fixup_overlap_and_insert.Improve __maps__fixup_overlap_and_insert's performance 21x in the case of overlapping mmaps. [email protected] reported slowness opening perf.data files from chromium where the files contained a large number of overlapping mappings. Improve this case primarily by avoiding unnecessary sorting.
[...]
Applied to perf-tools-next, thanks!
Best regards,
Namhyung