mm/vmalloc: use rb_tree instead of list for vread() lookups
Subsystems (2)
PATCH v2 1/1 mm/vmalloc: use rb_tree instead of list for vread() lookups
vread() has been linearly searching vmap_area_list for looking up vmalloc areas to read from. These same areas are also tracked by a rb_tree (vmap_area_root) which offers logarithmic lookup.
This patch modifies vread() to use the rb_tree structure instead of the list and the speedup for heavy /proc/kcore readers can be pretty significant. Below are the wall clock measurements of a Python application that leverages the drgn debugging library to read and interpret data read from /proc/kcore.
Before the patch:
-----
$ time sudo sdb -e 'dbuf | head 3000 | wc' (unsigned long)3000
real 0m22.446s
user 0m2.321s
sys 0m20.690s
-----
With the patch:
-----
$ time sudo sdb -e 'dbuf | head 3000 | wc' (unsigned long)3000
real 0m2.104s
user 0m2.043s
sys 0m0.921s
-----
---
Changed in v2:
- Use __find_vmap_area() for initial lookup but keep iteration via va->list.
mm/vmalloc.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
mm/vmalloc.c +4 -1
@@ -2860,7 +2860,10 @@ long vread(char *buf, char *addr, unsigned long count) count = -(unsigned long ) addr;spin_lock (&vmap_area_lock); -list_for_each_entry (va, &vmap_area_list, list) { + va =__find_vmap_area ((unsigned long )addr); +if (!va) + goto finished; +list_for_each_entry_from (va, &vmap_area_list, list) {if (!count)break ;
Signature · 1 line
2.17.1
Quoted 59 lines
vread()has been linearly searchingvmap_area_listfor looking up vmalloc areas to read from. These same areas are also tracked by arb_tree(vmap_area_root) which offers logarithmic lookup.This patch modifies
vread()to use therb_treestructure instead of the list and the speedup for heavy/proc/kcorereaders can be pretty significant. Below are the wall clock measurements of a Python application that leverages the drgn debugging library to read and interpret data read from/proc/kcore.Before the patch:
-----$ time sudo sdb -e 'dbuf | head 3000 | wc' (unsigned long)3000real 0m22.446s
user 0m2.321s
sys 0m20.690s
-----With the patch:
-----$ time sudo sdb -e 'dbuf | head 3000 | wc' (unsigned long)3000real 0m2.104s
user 0m2.043s
sys 0m0.921s
-----Signed-off-by Serapheim Dimitropoulos <[email protected]>---
Changed in v2:- Use
__find_vmap_area()for initial lookup but keep iteration viava->list.mm/vmalloc.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)mm/vmalloc.c +4
-1@@ -2860,7 +2860,10 @@ long vread(char *buf, char *addr, unsigned long count) count = -(unsigned long ) addr;spin_lock (&vmap_area_lock); -list_for_each_entry (va, &vmap_area_list, list) { + va =__find_vmap_area ((unsigned long )addr); +if (!va) + goto finished; +list_for_each_entry_from (va, &vmap_area_list, list) {if (!count)break ;Signature · 1 line
2.17.1
Much better :)
--
Vlad Rezki